PROFESSOR OF GERMANIC PHILOLOGY, EMERITUS, IN THE JOHNS HOPKINS UNIVERSITY, BALTIMORE, MARYLAND

PRESENTED BY A GROUP OF HIS PUPILS AND FRIENDS ON THE OCCASION OF HISnbsp;SEVENTY-FIFTH BIRTHDAY,nbsp;FEBRUARY 4, 1930

Franklin Edgerton, New Haven, Connecticut. Ewald Eiserhardt, Rochester, New York.

Urban T. Holmes, Jr., Chapel Hill, North Carolina. J. Preston Hoskins, Princeton, New Jersey.nbsp;William Guild Howard, Cambridge, Massachusetts.nbsp;Sanki Ichikawa, Tokyo, Japan.

Elizabeth F. Johnson, Rock Hill, South Carolina. Friedrich 0. Kegel, Bethlehem, Pennsylvania.nbsp;Robert J. Kellogg, Ottawa, Kansas.

Ernst H. Mensel, Northampton, Massachusetts. Ernst C. P. Metzenthin, Chapel Hill, North Carolina.nbsp;Leopold L. Meyer, Houston, Texas.

N. nbsp;nbsp;nbsp;V. Martinus Nijhoff, 's-Gravenhage, Holland.nbsp;John Rathbone Oliver, Baltimore, Maryland.

Allen W. Porterfield, Morgantown, West Virginia. Louise Pound, Lincoln, Nebraska.

D. nbsp;nbsp;nbsp;B. Shumway, Philadelphia, Pennsylvania.nbsp;Eduard Sievers, Leipzig, Germany.

E. nbsp;nbsp;nbsp;H. Sturtevant, New Haven, Connecticut.nbsp;Archer Taylor, Chicago, Illinois.

Die Oeffentliche Bibliothek der Universitat in Basel, Switzerland. Benedictine High School, Cleveland, Ohio.

Borsenverein der Deutschen Buchhandler, Leipzig, Germany. Bryn Mawr College Library, Bryn Mawr, Pennsylvania.

Dartmouth College Library, Hanover, New Hampshire. Universitatsbibliothek, Freiburg i. Br., Germany.

George Washington University Library, Washington, D. C. Goteborgs Stadsbibliotek, Goteborg, Sweden.nbsp;Universitatsbibliothek, Gottingen, Germany.

The Grosvenor Library, Buffalo, New York. Universitatsbibliothek, Halle a. d. S., Germany.

Staats- und Universitatsbibliothek, Hamburg, Germany. Harvard College Library, Cambridge, Massachusetts.

Holy Cross College Library, Worcester, Massachusetts. University of Illinois Library, Urbana, Illinois.

The Johns Hopkins University Library, Baltimore, Maryland. Universitats- und Stadtbibliothek, K�ln, Germany,

Lehigh University Library, Bethlehem, Pennsylvania. Universitets-Biblioteket, Lund, Sweden.

Bookstore, Mount Holyoke College, South Hadley, Massachusetts Bayerische Staatsbibliothek, M�nchen, Germany.

Library, University of North Carolina, Chapel Hill, North Carolina. Northwestern University Library, Evanston, Illinois.

Library, University of Pennsylvania, Philadelphia, Pennsylvania. University of Pittsburgh Library, Pittsburgh, Pennsylvania.nbsp;Princeton University Library, Princeton, New Jersey.

Rutgers University Library, New Brunswick, New Jersey Stanford University Library, Stanford University, California.nbsp;Taylor Institution, Library, Oxford, England.

State College of Washington Library, Pullman, Washington. Washington University Library, St. Louis, Missouri.

West Virginia University Library, Morgantown, West Virginia. University of Washington Library, Seattle, Washington.nbsp;Library of the University of Wisconsin, Madison, Wisconsin.nbsp;Yale University Library, New Haven, Connecticut.

TO PROFESSOR COLLITZ.

The scholar gentleman as a knight of old Valiant and loyal, chivalrous and bold.

Recks not of self or fame, but takes the field In the name of truth to make blind error yield.

As olden heroes by the play of swords He wins in reason�s war by force of words.

Such a scholar knight has charmed my mind. Not less urbane than learned, keen and kind;

Simple, lucid, and trenchant, with careful ease. He wins at once by logic and the art to please.

In the new world and homeland accounted great, He wears his knightly plume immaculate.

TO PROFESSOR COLLITZ.

The glory of gray hair and youthful heart Is yours and yours the glory that you wearnbsp;Your crown of scholarship and learning�s lovenbsp;So gracefully �� lord master of your art.

Then there are traits that place a man apart From lesser men, � of these you have full store �nbsp;For personality is something more,

Some feet leave earth the richer where they tread. Some minds that move along life�s loftier ways,

TABLE OF CONTENTS.

TRUMAN MICHELSON: LINGUISTIC MISCELLANY .... GEORGE MELVILLE BOLLING: A MATTER OF SEMANTICS .

TENNEY FRANK: ON THE NAME OF LUCRETIUS CARUS . JAMES TAFT HATFIELD: AENEID, II, 557nbsp;nbsp;nbsp;nbsp;.......

XII

BIOGRAPHICAL SKETCH

Hermann Collitz was born on Sunday, February 4th, 1855, at Bleckede, a Hanoverian hamlet, a little over a thousand years old,nbsp;situated on the Elbe, about midway between Hamburg andnbsp;Magdeburg. His ancestors on the paternal side had been for several generations burghers engaged in various lines of business. Hisnbsp;mother whose maiden name was Friederike Schafer, came fromnbsp;the nearby ancient village of Barskamp. Her father is identicalnbsp;with the �Hauswirt Schafer� mentioned in J. K. Wachters Sta-tistik der im K�nigreiche Hannover vorhandenen heidnischen Denkmaler, Hannover 1841, for the reason that his estate included anbsp;piece of ground on which at that time was situated one of the largest �H�nengraber� (cairns) then in existence

At the age of seven, Hermann was sent to a private school which had just been opened in his native town. V/as it because henbsp;had always heard High and Low German spoken alongside ofnbsp;each other at his home or for any other reason that from thenbsp;outset he manifested a special interest in the study of languages ?nbsp;In this private school two classes were combined in one schoolroom. While young Hermann had been assigned to the beginners�nbsp;class who were expected to attend to caligraphy, he was found tonbsp;be listening all the time to the instruction given to the older pupils in Latin and French. Upon examination it was found thatnbsp;he had duly profited from the instruction in languages and hencenbsp;was allowed to join the upper class. A few years later, Englishnbsp;was added to his curriculum and when thirteen years old, he wasnbsp;permitted to join a select class in Greek. Thus he was well prepared

1 This �giant�s grave� we are sorry to say is no longer in existence, the stone blocks having been meanwhile used for erecting a church in the village of Stie-pelse on the Elbe.

to enter, at the age of fourteen, the Lyceum of the Johanneum at L�neburg. The latter was a municipal Academy which could tracenbsp;its origin to a school of Benedictines; at the time of the Reformation it became a municipal institution for training in thenbsp;humanities and as such took its place among the leading schoolsnbsp;of this kind in Northern Germany. In the humanistic divisionnbsp;of the Johanneum stress was laid chiefly, in accordance with anbsp;long tradition, on the study of Greek, Latin and Mathematics.nbsp;In the spring of 1875, the young scholar passed the final exami-sations (Maturitatsexamen) and thereupon entered the University of Gottingen in the autumn. The interval between thenbsp;�Abiturientenexamen� and the Immatriculation at the University he employed in acquainting himself with the elements ofnbsp;Sanskrit.

At Gottingen, he took up the study of Classical Philology under E. von Leutsch, Herm. Sauppe and Curt Wachsmuth.nbsp;It must suffice here to mention particularly Sauppe�s lectures on Latin Grammar and on Greek and Latin Epigraphy,nbsp;and Wachsmuth�s Introduction to the study of Ancient History. Besides, he pursued courses in Philosophy and the Germanic Languages. In Philosophy, his teachers were H. Lotze andnbsp;J. Baumann, in Germanic Philology, Adalbert Bezzenberger andnbsp;W. Miiller. The appointment, in 1876, of August Pick, to a professorship of Comparative Philology at the University of Gottingen was an event which had a decided influence on his studies.nbsp;He not only derived much benefit from Pick�s courses on Greeknbsp;and Latin grammar, but also joined Pick�s �Grammatische Ge-sellschaft�. In the lectures of the newly appointed professor onnbsp;Greek and Latin grammar, these languages were given a background in comparative Indo-European Philology. The �Grammatische Gesellschaft� which met at Pick�s house served to supplement these lectures and to give his students practice in independentnbsp;research. In this connection should be mentioned also Bezzen-berger�s courses on the Zend-Avesta and in Lithuanian, and anbsp;course given by Benfey on the Hymns of the Rigveda.

After three years at the Georgia Augusta, he spent the summer semester of 1878 at the University of Berlin in order to studynbsp;Comparative Philology and Sanskrit with Joh. Schmidt andnbsp;Albr. Weber, German Philology with M�llenhoff and Scherer,nbsp;Celtic with H. Zimmer, and Slavic languages with V. Jagic.

Among his fellow-students were F. Bechtel, Maurice Bloomfield, Paul DeuBen, Felix Hartmann, Georg Mahlow and other well-known scholars. In the next semester he returned to Gottingen,nbsp;chiefly in order to take his Doctor�s degree. The subject of hisnbsp;dissertation was the origin of the Indo-Iranian palatal series, embodying a discovery he had made before going to Berlin. Thisnbsp;dissertation is to be found in Bezzenberger�s Beitrage, vol. III.nbsp;On the recommendation of his Gottingen teachers he received anbsp;stipendium intended for future Pri vat dozenten which enabled himnbsp;to give his time for several years entirely to a continuation of hisnbsp;studies and to research work. This was the time when Comparativenbsp;Philology was undergoing a thorough change as compared withnbsp;the views held formerly by leading scholars. So the time seemednbsp;favorable for replacing Schleicher�s Compendium by a work morenbsp;in accordance with current views. This situation accounts for thenbsp;plan conceived at this time by Fick, Bezzenberger, Bechtel andnbsp;Collitz to unite in supplying the need for such a work. It was tonbsp;consist of four volumes, one volume to be contributed by each ofnbsp;the four scholars, namely: i. Phonology 2. Word-Formation 3. Inflection 4. Syntax. The whole was meant as a set of companionnbsp;volumes to Tick�s Vergleichendes W�rterbuch and to be publishednbsp;by the same publisher. This plan, unfortunately, was abandonednbsp;after a few years, partly because the project of collaborating innbsp;bringing out the Greek dialect inscriptions seemed even morenbsp;urgent, and partly because soon afterwards Brugmann�s Grund-rip made the need of such a work less imperative. That the worknbsp;of the Gottingen scholars would have received a form somewhatnbsp;different from Brugmann�s Grundrifi may be inferred, e. g., fromnbsp;Bechtel�s important book HauptproUeme der indogermanischennbsp;Lautlehre seit Schleicher (Gottingen 1892).

In 1879 Dr. Collitz returned to Berlin, where he resumed his studies in Indo-European linguistics with some of his formernbsp;teachers. The plan of collecting the Greek dialect inscriptions hadnbsp;meanwhile made substantial progress and Dr. Collitz was entrusted with the editorship. The first number appeared in 1883 andnbsp;the first volume was finished in the next year.

Another task undertaken in Berlin was the editing of a dictionary of the Waldeck Low German dialect left in manuscript by the late Karl Bauer. In 1883 he left Berlin in order to accept annbsp;appointment at the University Library of Halle where a vacancy

had been created owing to the resignation of Karl Verner who had been called to the University of Copenhagen. Collitz hadnbsp;every reason to enjoy the work since it included the writing ofnbsp;the new catalogues (begun by Verner) for General and Comparative Linguistics and for Philosophy. In 1885 he was granted thenbsp;venia docendi at the University of Halle for Sanskrit and Comparative Philology. In the same year, he received a call as Associate Professor of German, to the newly founded Bryn Mawrnbsp;College, near Philadelphia. Among his early colleagues were Missnbsp;Carey Thomas (Professor of English and Dean), Woodrow Wilsonnbsp;(Professor of History), Paul Shorey (Professor of Latin), andnbsp;Washburn Hopkins (Professor of Sanskrit and Greek). After thenbsp;latter had been called to Yale University, Collitz�s title wasnbsp;changed to that of Professor of German and Comparative Philology.

In connection with the graduate courses he gave at Bryn Mawr College, his studies centered more and more around comparativenbsp;Germanic Philology. This probably was the reason why he wasnbsp;invited, along with Professor Sievers, to represent Germanicnbsp;Philology at the Congress of Arts and Sciences held in connectionnbsp;with the Universal Exposition at St. Louis in 1904. For severalnbsp;years he served as von Jagernann�s successor in editing the Germanic division of Modern Language Notes.

Much as he felt attached to Bryn Mawr College he did not hesitate to accept a call, extended to him in 1907, to fill the newly established chair in Germanic Philology at the Johns Hopkinsnbsp;University. This university, as is well known, had adopted at itsnbsp;foundation the German methods of university study with regardnbsp;to seminaries, the requirements for a Doctor�s degree, and in layingnbsp;stress, in a general, way, on research work. It could claim, moreover, to be foremost in this country especially in the variousnbsp;branches of linguistic studies. This tendency apparently wasnbsp;responsible for the endeavor to foster the study of Germanicnbsp;Philology. As a matter of fact, the relations between the Germanicnbsp;department and other philological departments of the Universitynbsp;soon became quite intimate.

Germanic Philology, as Professor Collitz conceived it, means on the one hand the comparative study of the Germanic languagesnbsp;in connection with the cognate Indo-European languages, abovenbsp;all Sanskrit, Greek and Latin, on the other hand, a comparison

of the Early Germanic languages with each other with a view to the reconstruction of the common ancestor, i. e. the Primitivenbsp;Germanic. Naturally, however, the instruction of graduate students had to take the form chiefly of grammatical courses combined with the interpretation of texts in the various old Germanicnbsp;languages, such as Gothic, OldPIigh German, Old Norse, Old Frisiannbsp;and Old Saxon. The study of Gothic generally was combined withnbsp;that of the elements of Comparative Philology. Considering thatnbsp;Low German resembles English much more closely than High German, the study of Low German in its various periods seemed tonbsp;deserve special attention. Professor Collitz therefore made it a pointnbsp;to interpret, in each semester, to his students either portions ofnbsp;the Heliand or one of the Middle Low German dramas or extractsnbsp;from Klaus Groth and Fritz Reuter. The Johns Hopkins University probably was the only university in the United Statesnbsp;where such courses in Low German and Frisian were offered withnbsp;some regularity. Other courses given by Professor Collitz werenbsp;General Phonetics, seminary courses on OPIG. glosses, on Hartmann�s Iwein and other MHG. authors, and on MHG. �Urkunden�.nbsp;After the retirement (in 1920) of Professor Henry Wood, provisionnbsp;had to be made for the courses formerly given by him. Professornbsp;Collitz took over the graduate course in Faust (Partii) which hadnbsp;been on his repertory for advanced students at Bryn Mawr College.

In 1911, he attended, as a delegate of the Johns Hopkins University, the centennial celebration of the University of Christiania (Oslo). This occasion furnished an opportunit57 of visiting Denmark, Sweden and Norway, and to make or renew the personalnbsp;acquaintance with many Scandinavian scholars, such as Hj. Falk,nbsp;A. Torp; O. Jespersen, H. M�ller, Holger Pedersen, W. Thomsen;nbsp;E. Wadstein; 0. A. Danielsson, P. Persson, E. Hj. Psilander andnbsp;others. Needless to add that the museums and libraries of thenbsp;various Scandinavian capitals and Universities were a sourcenbsp;to him of both enjoyment and instruction, and that the Codexnbsp;Argenteus in Upsala was not overlooked.

In the following year there appeared his monograph on the origin of the Germanic weak preterite (Das schwache Prdtentumnbsp;und seine Vorgeschichte) as the first volume of the series Hesperia.nbsp;In 1916, the University of Chicago conferred on him the honorary degree of L.H.D.

Professors Leonard Bloomfield, G. M. Bolling and Edgar H. Stur-tevant for the organization of the American Linguistic Society. After having attended the organization meeting held in New Yorknbsp;in December 1924, he was elected the first President of thisnbsp;Society for 1925. That year the Society met in Chicago togethernbsp;with the Modern Language Association of which Professor Collitznbsp;also had been elected President.

Among the festivities which marked the celebration, in 1926, of the semicentennial anniversary of the Johns Hopkins Universitynbsp;was a special meeting of the Germanic department to which thenbsp;former students were invited. Mr. Collitz was privileged to welcome the guests in the name of the department and to read thenbsp;principal paper on the various meanings of the word �Dialect�.nbsp;At the commencement exercises of the next year an oil portraitnbsp;of him, to which friends, colleagues, and former students hadnbsp;contributed, was presented to the University.

At the age of seventy two, after having served for forty one years as a college and university Professor, he felt entitled to asknbsp;to be relieved from active duties and retired with the title ofnbsp;Professor Emeritus of Germanic Philology.

The 24th of November 1928 marked the fiftieth anniversary of the day when the degree of Ph. D. was conferred upon him bynbsp;the University of Gottingen. On this occasion the Philosophicalnbsp;Faculty of the Georgia Augusta sent Professor Collitz its congratulations together with a beautifully executed new diploma.

BIBLIOGRAPHY

OF PROFESSOR COLLITZ� WRITINGS*).

*) In quoting AnzfdA.nbsp;nbsp;nbsp;nbsp;=

AfslavPh. nbsp;nbsp;nbsp;=

ATPh. nbsp;nbsp;nbsp;=

APA. nbsp;nbsp;nbsp;=

BB. nbsp;nbsp;nbsp;=

CP. nbsp;nbsp;nbsp;=

DLZ. nbsp;nbsp;nbsp;=

IF. nbsp;nbsp;nbsp;=

JAOS. nbsp;nbsp;nbsp;=

JEGP. nbsp;nbsp;nbsp;=

JHU. Circ. nbsp;nbsp;nbsp;=

KZ. nbsp;nbsp;nbsp;=

Lang. nbsp;nbsp;nbsp;=

MLN. nbsp;nbsp;nbsp;=

Ndd. Korr.bl. nbsp;nbsp;nbsp;=

PBB. nbsp;nbsp;nbsp;=

PMLA. nbsp;nbsp;nbsp;=

Scand. Studies = ZfdPh.nbsp;nbsp;nbsp;nbsp;=

current periodicals the following abbreviations have been used: Anzeiger f�r Deutsches Altertum und Deutsche Literatur.nbsp;Archiv f�r Slavische Philologie.

Beitrage zur Kunde der indogermanischen Sprachen, heraus-gegeben von A. Bezzenberger.

Beitrage zur Geschichte der deutschen Sprache und Literatur (Paul u. Braunes Beitrage, z. Z. hrsg. von Ed. Sievers).nbsp;Publications of the Modern Language Association.nbsp;Scandinavian Studies and Notes.

Etymologien. i. Gr. T�fpqa, Lat. favilla, febris. 2. Gr. glyog, Lat. frigus = *sngos; BB. 3, 321�323.

Skr. car-, cira-m, Gr. rsX�dm, TtdXai; BB. 5, loi�102. Rev. of Brugmann-Osthoff, Morphologische Untersuchun-gen, vol. I; AnzfdA. 5, 318�348.

Rev. of G. Curtius: Das Verbum der Griechischen Sprache, vol. II; DLZ. 1880, no. 13.

1881. nbsp;nbsp;nbsp;Rev. of Braune, Gotische Grammatik;nbsp;nbsp;nbsp;nbsp;12, 480�482.

1882. nbsp;nbsp;nbsp;Rev. of Sievers, Phonetik, 2. Auflage; DLZ. 1882, nr. 6.nbsp;Rev. of Gust. Meyer, Griechische Grammatik; BB. 7,

Ueber eine besondere Art Vedischer Composita ; Abhand-lungen des 5. Internationalen Orientalistenkongresses zu Berlin, Vol. II, 2, 287�298.

Homerisch ^u-s, �v-s und Vedisch ayu-s; KZ. 27, 183�189. Rev. of Ziemer, Junggranimatische Studi�n; DLZ. 1882,nbsp;col. 1567.

Der Germanische Ablaut in seinem Verhaltnis zum Indo-germanischen Vocalismus (I. Teil); ZfdPh. 15, i�10. Zur Einteilung der Niederdeutschen Mundarten; Ndd.nbsp;Korr.bl. 7, 81�82.

nack, F. Bechtel, A. Bezzenberger, F. BlaB, H. Collitz, W. Deecke, A. Fick, H. van Gelder, O. Hoffmann, R. Meister, P. M�llensiefen, W. Prellwitz. Herausgegeben von

(1. Band, Heft i�4, 1883�84. �� 11. Bd., Heft i�6, 1885�99. �� IH. Bd., hrsg. von C. F. und F. Bechtel.

I. nbsp;nbsp;nbsp;Halfte, Heft i�5, 1888�99, 2. Halfte, Heft i�5,nbsp;1898�^1905. �� IV. Bd., Wortregister und Nachtrage.nbsp;Heft I, bearbeitet von R. Meister, 1886; Heft 2, hrsg.

von H. C. und F. Bechtel, 1901; Heft 3 und 4, hrsg. von H. C. und 0. Hoffmann, 1910�^15.)

Rev. of Hoffory, Professor Sievers und die Sprachphysio-logie; ihid., col. 1613 seq.

im Altindisclien und im Griechischen. Teil I, Die Casus des Singulars; BB. 10, i�71.

(Accepted as �Habilitationsschrift� by the Philosophical Faculty of the University of Halle, March 1885.)

Die Verwandtschaftsverhaltnisse der Griechischen Dialekte. Gottingen, Vandenhoeck amp; Ruprecht.

germanischen Ablautes; BB. ii, 203�242 (also publ. separately, Gottingen, Vandenhoeck amp; Ruprecht) andnbsp;(Nachtrag) 304�307.

Ueber das Vergleichende Studium der Niederdeutschen Mundarten; Ndd. Korr.hl. ii, 23�-32.

Sprachen; AJP. 9, 42�^57. (Reprinted in BB. 17, 227�244. Translated under the title �The Origin ofnbsp;the Weak Preterit�; PMLA. 3, 196�209.)

Rev. of Jellinek�s Beitragezur Erklarung der Germanischen Flexion; AnzfdA. 17, 275�282.

Rev. of Jackson �The Avestan Alphabet and its Transcription�; ihid. 12, 489�-492.

Die drei indischen Wurzeln ksi und ihre Verwanten im Griechischen; BB. 18 (1892), 201�226.

For additional comment on these roots and on the existence in the I.-Eur. parent speech of a palatal s- sound in addition to the dentalnbsp;sibilant s see the author�s articles on Lat. saeculum (1921, Festschriftnbsp;for Bezzenberger) and on Saturnus {1930, Oriental Studies ed. by Jalnbsp;Dastur C. Pavry).

Zur Bildung des Instrumentals der wair-Stamme im Alt-indischen; ihid. 18, 231�241.

�Frisian Language and Literature�; Johnson�s Cyclopaedia, Vol. Ill amp; IV (1894).

Rev. of Paul�s GrundriB der germanischen Philologie (vol. I, no. i); MLN. 8, col. 99�-106 amp; 160�169.

read before the Oriental Club of Philadelphia, 1888/g^. �� Boston, Ginn amp; Co.), 177�201.

Articles �Low German� and �Plattdeutsch�; Johnson�s Cyclopaedia, Vol. V (1894) amp; VI (1895).

The Etymology of aga and /xatp', APhA. Proceedings, Special Session, 1894, XXXIX�^XL.

i8gg. Sammlung der Griechischen Dialekt-Inschriften, 2. Bd. u. 3. Bd., I. Halfte (hrsg. v. H. Collitz und F. Bechtel).nbsp;Gottingen, Vandenhoeck amp; Ruprecht.

Die Niederdeutsche Mundart im F�rstentum Waldeck (Sonderabdruck aus Bauers Waldeckischem W�rter-buche). Norden, D. Soltau, i8gg.

The Vedic word ndvedas. JAOS. Vol. XX, Pt. 2, 225�228. igo�. The Home of the Heliand; PMLA. 16, 123�^140.nbsp;igo3. Waldeckisches W�rterbuch nebst Dialektproben, gesammeltnbsp;von K. Bauer. Im Auftrage des Vereins f�r niederdeutsche Sprachforschung bearbeitet und herausgegeben vonnbsp;H. Collitz; pp. XVI 106 320; Norden u. Leipzignbsp;igo2; D. Soltau�s Verlag.

�German Language�; New International Encyclopsedia (New York: Dodd, Mead amp; Co.) in vol. 7.nbsp;igo4. Zum Awesta-Alphabet; Verhandlungen des XIII. Internationalen Orientalisten-Kongresses, Leiden igo4, 107nbsp;�108.

Inhalts�bersicht: i. Die t-Deklination im Rigveda 8i; II. Die �-Deldination im Lateinischen 82; Exkurs: Zu den i'�-Stammen dernbsp;latein. V. Deklination 83; III. Die �-Deklination in den �brigen euro-paischen Sprachen 85; IV. Parallelismus der a- und der �-Deklinationnbsp;im Altindischen 89; V. Die a- und �-Deklination im Altiranischen 96;nbsp;VI. Zur Vorgeschichte der I-Deklination 98; VII. Ergebnisse f�r dienbsp;d-Deklination 104; VIII. Abgeleitete Stamme auf -e-ya- im Altindischennbsp;109; IX. Die �-Deklination in den europaischen Sprachen iio.

Note. When this article was printed in BB., it should have been stated that it had been written for the volume T�pai;. Abhandlungennbsp;zur idg. Sprachgeschichte�, published in honor of Prof. August Picknbsp;in 1903, but owing to adverse circumstances had been received too latenbsp;for publication in that volume.

Sammlung der Griechischen Dialekt-Inschriften, hrsg. von H. Collitz und F. Bechtel. III. Band, 2. Halfte. Gottingen,nbsp;Vandenhoeck amp; Ruprecht.

igo6 Problems in Comparative Germanic Philology; Congress of Arts and Science, Universal Exposition, St. Louis

12 (Boston; Houghton, Mifflin amp; Co.), vol. 3 (Language,nbsp;Literature, Art), 286�302.

1907 Segimer oder; Germanische Namen in Keltischem Gewande; JEGP. 6, 253�-306.

Rev. of Gall�e, Vorstudien zu einem Altniederdeutschen W�rterbuche; ibid. 6, 472�477.

1909 nbsp;nbsp;nbsp;Rev. of Pick, Vorgriechische Ortsnamen; CP. 4, 206�209.nbsp;Rev. of Streitberg, Die Gotische Bibel L; MLN. 24,

nation in Anglo-Saxon; JEGP. ii, 25�26 Zum Hildebrandsliede; PBB. 36, 366�373.

�Missingsch�; Pestschrift, Chr. Walther gewidmet von dem Verein fiir niederdeutsche Sprachforschung (= Jahr-buch des Vereins f. ndd. Spr. XXXVII, Heft i,nbsp;no�113).

1912 Das Schwache Prateritum und seine Vorgeschichte (Hesperia. Schriften zu�' Germanischen Philologie i.) pp. XVI 256; Gottingen, Vandenhoeck amp; Ruprecht (Baltimore,nbsp;The Johns Hopkins Press).

In the preface a review is given of recent contributions to Germanic Philology in the United States. An appendix contains a discussion ofnbsp;the Latin perfect tense and the Greek passive aorist in their relationnbsp;to the weak preterit. A theory advanced here concerning the origin ofnbsp;the Greek passive aorist has recently found an able advocate in Adolfnbsp;Walter in the Festschrift for W. Streitberg (Stand u. Aufgaben dernbsp;Sprachwissenschaft, Heidelberg 1924) p. 351 f.

Rev. of Petzet amp; Glauning, Deutsche Schrifttafein des IX. bis XVI. Jahrhunderts, i. Abteilung; MLN. 27,nbsp;115�117.

I. �3. Abteilung (SchluB der Sammlung). Gottingen,nbsp;Vandenhoeck amp; Ruprecht.

Rev. of Sipma, Phonology and Grammar of Modern West Frisian; MLN. 30, 215�217.

Zu den Mittelhochdeutschen kurzen Praterita gie, fie, lie', MLN. 32, 207�215, 449�458.

The Etymology of the word Degen] JHUCirc. no. 296, 887. The Greek noun n^o/Mi;] ibid. 889.

Note. An abstract only of a more comprehensive and more detailed treatise which remains to be published. In endeavoring to reconstructnbsp;the Primitive Germanic vocalism on the basis of the various Old Germanicnbsp;languages the author arrived at the result that the common foundationnbsp;is seen almost unaltered in Gothic. The innovations generally ascribednbsp;to Gothic were inherited from Primitive Germanic, whereas numerousnbsp;and fundamental changes took place afterwards in the West Germanic-Scandinavian period.

Rev. of Meillet�s Caract�res G�n�raux des Langues Ger-maniques; AJP. 39, 408�418.

Rev. of Gaidoz�s Deux �rudits Gallois, John Rhys et L. Reynolds; ibid. 205�209.

The Systematization of Vowel-Sounds, or: Bell�s Vowel-System modified and simplified; JHU.Circ. no. 316, 588�589.

Rev. of Joh. Enschede�s Die Hochdeutschen Schriften aus dem 15. bis zum 19. Jahrhundert; MLN. 34,nbsp;492�496.

1920. nbsp;nbsp;nbsp;The Causes of Phonetic Change; JHUCirc. no. 325, 30�32.nbsp;Anglo-Saxon rcefnan] ibid. 52�54.

1922. Germanische Wortdeutungen i. Gotisch inn, inna MLN. 37, 215�217. � 2. Gotisch duginnan �beginnen�; ibid.nbsp;274�297.

Wodan, Hermes und P�shan; Festskrift t�lagnad Hugo Pipping, 547�587. Helsingfors; Skrifter utg. av Svenskanbsp;Litteratursdllskafet i Finland, CLXXV.

Das Wort Keizerquot;, Germania, Sievers-Festschrift zum 75. Geburtstage, 115�128.

World Languages (Presidential address. Joint meeting of Modern Language Association and Linguistic Society ofnbsp;America); PMLA. 41, 43�55; Lang. II, i�13.

An address delivered in 1933 on the occasion of the centenary of the 2nd edition of the first vol. of J. Grimm�s Grammar. It includes anbsp;mnemonic help for remembering the working of the Law.

Rev. of K. Lokotsch, Etymologisches W�rterbuch der Europaischen W�rter orientalischen Ursprungs; MLN.nbsp;42, 412�414.

no. 3. 593�601.

A supplement to the author�s monograph on the weak preterit published in 1912. In his present opinion this preterit must be regarded as a complexnbsp;formation, due to the combination of certain forms of the I.-Eur. perfectnbsp;middle with a peculiar type of the I.-Eur. simple aorist.

Antediluvian Kings and Patriarchs in the Light of Comparative Mythology; JAOS. 48 (1928) p. 341.

(Syllabus of a Paper read at the Washington Meeting of the Amer. Oriental Society. Not yet published.)

1930. K�nig Yima und Saturn; Oriental Studies, edited by Jal Dastur C, Pavry, London [To be published shortly].

Contents: I. Einleitung; II. Yima in den Gathas; III. Yima im J�ngeren Avesta; IV. Yimo xsa�t� (Djemsid) und ved. ksaitah', V. Yimonbsp;xsa�l� und Saturnus.

(Editor;) Sammlung der Griechischen Dialekt-Inschriften; 4 vols., Gottingen, Vandenhoeck amp; Ruprecht, 1884�1915 (vol. 3nbsp;jointly with F. Bechtel, and vol. 4 with O. Hoffmann).nbsp;(Editor:) Hesperia. Schriften zur Germanischen Philologie; nos.nbsp;I�18, Gottingen, Vandenhoeck amp; Ruprecht, and Baltimore,nbsp;The Johns Hopkins Press, 1912�^1929.

(Associate Editor:) Modern Language Notes; vols. 17�28, Baltimore, 1902�1913.

(Co-operating Editor:) Journal of English and Germanic Philology; vols. 8�28, The University of Illinois, 1909�-1929.

(Co-operating Editor:) American Journal of Philology; vols. 41�50, Baltimore, 1920�-1929.

NEUTER PRONOUNS REFERRING TO WORDS OF DIFFERENT GENDER OR NUMBER i

Nine times in the Iguvinian Tables a relative pronominal form, forse (forsi, forsei) refers to an antecedent that is masculinenbsp;singular or neuter plural, and opinions differ as to the nature ofnbsp;forse itself. Biicheler, Umbrica 46, 192, 215, saw in rse an encliticnbsp;particle *de, with which he compared the final element of Latinnbsp;quamde^. Brugmann, Berichte der Sachsischen Akademie 1893,nbsp;135 f., considered fo an indeclinable relative adverb strengthenednbsp;by an enclitic *di. But forse appears to be related to the conjunction ftire (from *qtwd-�) in the same way as first to firenbsp;(from *quid-i). Other forms also of the relative show a finalnbsp;particle *1 in Umbrian; but there is no clear trace of a particlenbsp;*di. As far as form goes, there is every reason to consider forse,nbsp;like fiire, the neuter singular of the pronoun with an enclitic *1.nbsp;Consequently Bugge, Kuhn�s Zeitschrift fur Vergleichende Sfrach-forschung 3. 35, suggested that it was a stereotyped neuter usednbsp;for all genders and both numbers. He has been followed by severalnbsp;scholars I agree with them in the main, except that I considernbsp;the idiom, not a peculiar development of Umbrian^, but a remnantnbsp;of a very ancient usage.

In four of the eight passages forse is equivalent to Latin is qui, and the clause describes a person whose identity is unknown.

1 I am under obligations to my colleagues. Professors Edgerton, Prokosch, and Willard, and to my daughter, Mrs. F. W. Hopkins, for assistance in securingnbsp;material on languages with which I have little familiarity.

^ See Br�al, Les Tables Eugubines 42: �C�est un commencement d�appauvris-sement de la d�clinaison qui annonce ce qui s'est pass� en italien pour le pronom che et en franjais pour qne.� Cf. ib. 355.

The sentences are so similar to one another that it will be enough to cite 6 b 63:

In 6a 9 porsei refers once to angluto �ah angulo� and once to anglome �ad angulum�, which word is probably masculine. In 6 anbsp;19 and 6 b 40 the antecedent is uaso(r) �vasa�, which is neuternbsp;plural; and tudero �finis�, the antecedent in 6 a 15, is also a neuternbsp;plural.

The clauses of the first type, in which the antecedent of the relative is an unidentified pei'son, find a close parallel in certainnbsp;phrases in Latin; for example;

Plautus, Casina 451: erit hodie tecum quod amas, �today your sweetheart will be with you�.

Cu. 170: qui homo quod amat videt, �whoever sees his sweetheart�. Epidicus 653: tibi quidem quod ames domi praestost, �there�snbsp;somebody at home for you to love�.

Poenulus 327: ecquid amare videor? : : damnum, �do I seem to be in love with anybody? : : with loss�.

Cu. 38: ama quidluhamp;t, �make love to anybody you please� Terence, Andria 464: nam quod peperisset iussit tolli, �for he toldnbsp;us to rear the baby�.

longaevosque senes ac fessas aequore matres et quicquid tecum invalidum metuensque pericli est,

�aged men and sea-weary women and whoever in your company is a weakling and afraid of danger�.

Such sentences as these do not stand alone in Latin; there is a marked tendency to use neuter pronouns and pronominalnbsp;adjectives to refer to persons who are not fully identified.nbsp;Catullus, 9. II; quid me laetius est beatiusve? �who is happiernbsp;than I ?�

Cicero, Ad Atticum 9. 16. 3: Dolabella tuo nihil scito mihi esse iucundius, �be assured that nobody is dearer to me than yournbsp;Dolabella�.

Pliny, Epistulae i. 221: nihil est enim illo gravius, sanctius, doctius, �nobody is more dignified, more reliable, or betternbsp;educated than he�.

� For other examples with amo, see Lodge, Lexicon Plautinum i. 112. Festschrift CoUitz.nbsp;nbsp;nbsp;nbsp;2

Cicero, Tusculanae Dis-putationes i. 7. 13: quid enim tam pugnat quam non modo miserum sed omnino quicquam esse quinbsp;non sit ? �for what is so inconsistent as for one who does notnbsp;exist to be unhappy or even to be anything?�

Livy I. 53. II: in se ipsum postremo saeviturum, si alia desint, �that he would finally turn his fury upon himself, if therenbsp;were no others�.

We may fairly conclude from a comparison of Umbrian and Latin that Primitive Italic employed neuter pronouns andnbsp;pronominal adjectives to refer to persons in an indefinite way.nbsp;That the usage may be carried back to Italo-Celtic times seemsnbsp;to follow from a notice in Pedersen�s Vergleichende Grammatik dernbsp;Keltischen Sprachm 2. 64 f., to the effect that Old Irish sometimesnbsp;employs the neuter of similar words to refer to persons. Amongnbsp;his examples are nechtar n-di �one of the two� and cechtar n-atharnbsp;�both of us�.

In the Germanic languages of all periods neuter pronouns very frequently refer to persons, although examples are relativelynbsp;few in the Gothic Bible on account of the imitation of Greeknbsp;syntax which prevails there. I shall illustrate principally withnbsp;Old English.

Gothic. Mark 6. 3: niu ]gt;ata ist sa timrja ? �isn�t this the carpenter ?� Old High German. Tatian 81. 2: ih him iz, �ich bin es�.

Old English. Beowulf 237: hwcet syndon ge . . . ? who are ye . . . ?� Alfred, Boethius 36. 22 (Sedgefield): and he hine het secgannbsp;hwcet his geferan wseron, �and he ordered him to say whonbsp;his companions were�.

Aelfric, Homilies i. 14. 4 (Thorpe): hu mihte Adam tocnawan hwcet he wsere ? �how could Adam know who he was ? �nbsp;Aelfric, Genesis 27. 32: hwcet eart pu? �who art thou?�

Bede 4. 156. 1139 (Schipper): ac ic cu�lice wat ge hwcet ku eart ge for hwon ku gnornast, �scio enim certissime qui es et quarenbsp;maeres�.

Orosius 17. 26 (Sweet): and lgt;cet wseron eall Finnas, 'and they were all Finns�.

In one usage the employment of the neuter is extended to all adjectives in the Germanic languages; the neuter is used tonbsp;combine masculine and feminine nouns, as is well illustratednbsp;in the early versions of Luke i.'6. As we are at present interestednbsp;merely in the pronouns, I shall quote only the Gothic version innbsp;full.

Gothic, wesunuh kan garaihta ba in andwairpja gudis, gaggandona in allaim anabusnim jah garaihteim fraujins unwaha, �andnbsp;they were both righteous before God, walking in all thenbsp;commandments and ordinances of the Lord blameless�.nbsp;Norse, pau voro retlot bgepe.

Old High German. Tatian: siu warun rehtiu beidu fora gote. Otfrid; uuarun siu bethiu gote filu drudiu.

In Slavic the neuter singular of the pronoun is used to refer to a masculine or feminine noun of either number which has notnbsp;been previously mentioned (although this limitation is notnbsp;consistently obsei'ved, and has no importance for our investigation). A few examples will suffice.

Russian, kto eto takoy ? �who is that ?� eto moi deti, �these are my children�.

The interrogative pronoun, kto (Church Slavonic kuto) almost certainly represents a phrase of this type; for ku- is Sanskrit has,nbsp;and -to is Sanskrit tad. The relative pronoun also is used in thenbsp;neuter singular to refer to a masculine singular or plural. Thenbsp;construction is rare in Old Church Slavonic, but more commonnbsp;later, especially in Old Czech. Two examples from Church Slavonicnbsp;follow.

Codex Suprasliensis 17. 5; bogu suojemy, jeze u nebesexu, 'to our God who is in the heavens�.

In the other Indo-European languages neuter pronouns and pronominal words do not often refer to persons; but there arenbsp;a few scattering examples such as the following.

o�r�Q y' 'Argstdrji; evgv y.gdmv 'Ayafi�fivaiv, ai^Kp�iegov, fSaaJ.evg r� dyaamp;�g y.Qareg�Q r� atxfirjxrjq,

�this is Atreus�s son, Agamemnon, ruler of many lands, both a good king and a mighty spearsman�.

Aristophanes, Equites 854; nehroitMlai ml xvQOTiamp;lm' tovro � elg sv �ari avyHsxvfog, �dealers in honey and in cheese; fornbsp;these have conspired together�.

Plato Apology 41: �av nbsp;nbsp;nbsp;ri slvai, jurj��v oneg, �if they seem

Sophocles, Trachiniae 1107; mvrd jui^��v �, 'even if I am nobody�. Sanskrit. Chdndogya Upanisad 6. 16. 2; tat satyam, sa atmanbsp;tat tvam asi, �that is truth, that is the self, that thou art�.nbsp;Vajasaneyi Samhitd 32. i:

tad evagnis, tad adityas, tad vayus, tad ii candramah, tad eva sukram, tad brahma, tad apah, sa prajapatih,

�that is Agni, that is the sun, that is the wind, that is the moon, that is the bright one, that is Brahman, that is thenbsp;waters, that is Prajapati�.

�there is nothing higher than thou, Indra, nothing greater, slayer of Vrtra, nothing even like thee�.

Except for three Umbrian examples, we have so far confined our attention to the use of a neuter pronoun referring to persons.nbsp;It is evident, however, that Umbrian parse referring to a masculinenbsp;singular or a neuter plural antecedent cannot be separated fromnbsp;parse �is qui�. In Germanic and Slavic there is no distinction innbsp;this matter between personal nouns and others, and elsewherenbsp;also neuter singular pronouns sometimes refer to impersonalnbsp;masculine and feminine nouns or to neuters plural. Particularlynbsp;important is the use of a neuter singular pronoun to govern anbsp;partitive genitive.

Latin. Plautus, Mostellaria 627!.: quasi quadraginta minas; ne sane id multum censeas, �about forty minae; don�t thinknbsp;that much�.

Cicero, Pro Sulla 17. 49: cum honos agebatur familiae vestrae amplissimus, hoc est, consulatus parentis tui, �when high

The stanza occurs also in Rig Veda 4, 30. 1, where the usual masculine is employed: na kir .^ . uttarah, etc.

office for your household, that is, your father�s consulship, was at stake�.

Caesar, De Bello Gallico 3. 16: navium quod ubique fuerat, �what ships they had had anywhere�.

Livy 3. 17. 5; quicquid patrum plebisque est, �all the patricians and plebeians�.

Gothic. Ephesians i. 18: hwa ist wens? �what is the hope?� {wens is feminine).

Skeireins 7. 14 f.: bigitan was pizei hlaibe ib tainjons fullos, patei aflifnoda, �there was found of these loaves twelvenbsp;baskets full, that was left�.

Old English. Crist 694: hwcet sindon pa gimmas buton god sylfa? �what then are gems but God himself?� {gimmas is femininenbsp;plural).

89; hwcBt is tgt;eos wundrung ? �what is this excitement ?� {wundrung is feminine).

Alfred, Cura Pastoralis 21, p. 154 (Sweet): hwcet ryhtlices, �aliquid iusti�.

Orosius 226. 12 (Sweet); siJhan masst call forwearS pcet on kseni lande waes, ge monna, ge nietena, ge wildeora for hsem stence,nbsp;�afterwards most all that was in that land perished, bothnbsp;of men and of cattle and of wild things, on account of thenbsp;smell�.

Greek. Odyssey i. 159: tovroioiv ju�v ravta ju�kei, xldagig xai aoi�i], �their interest is this, the lyre and song�.

Herodotos 4. 23; xagnov cpoQel xvajucp �ami . . . ., rovto �neav y�vrjtai Ti�nen' . . . ., �it bears fruit like a bean; when this grows ripe�nbsp;{xagn�g is masculine).

Sammlung Griechischer Dialektinschriften 5653. a. 5: es Tj)r xqiobov �ig � and rovr� . . ., �to the fork of the road, six (boundarynbsp;stones); from there� {rglo�og is feminine).

Euripides, Helena 1687: evysvecjrdtrji; yva)}J.rjg, o noXXah; �v yvvai^iv ovK evi, �of a noble disposition, which in many women doesnbsp;not exist�.

Iliad 21. 360: xi noL eQvdoQ gt;cal dqcoyrfQ; �what is warfare and defense to me?� (cf. Aristophanes, Lysistrata 514, etc.: xlnbsp;ds aoi xavxa).

Avestan. Yasht 5. 77: avavat da�vayasnanam, �so many of the devil-worshipers �.

Yasna 34. 12: kat vasi, kat va st�t� kat va yasnahya? �what do you want; what of praise or what of sacrifice ?�

Sanskrit. Atharva Veda 7. 97. 2: sam indra no manasa nesa gobhih . . . sam bhrahmanam devahitaih yad asti, �bring usnbsp;together, Indra, with mind, with cattle, . . . with whatevernbsp;of incantations is god-established�.

Kathaka Samhita 8. 14 = Taittirlya Samhita i. 5. 3. 2 = Mai-trdyanl Samhita i. 7. i: yat te, �whatever of thee (Agni)�.

Keilschrijturkunden aus Boghazk�i 14. i. i. 48 (= G�tze, Maddu-wattaS p. 12): [KA]RAS-za-kan ku-i-e-es te-pa-u-es i[s-par]-te-ir a-pa-a-at-ma.-kan hu-u-ma-an a[r-ha ha]-as-pi-ir-pit, �and what few escaped from (his) army, all those theynbsp;likewise de [stroyed] �.

Keilschrifttexte aus Boghazk�i 4. 10. i. 50 f.: nu ka-a-sa a-pi-e-da-ni me-mi-ni LI. IM DINGIR. MES tu-li-ya hal-zi-ya-an-te-es nu us-kan-du is-ta-ma-as-kan-du-ya na-a^ ku-ut-ru-

* See Hrozny, Die Sprache dev Hethiter 1343, 1423, I43i; Friedrich, Zeitsclwift dev Deutschen Morgenlandischen Gesellschaft 76. 164, Zeitschrift fiir Assyriologie,nbsp;N. F. 2. 290�296.

e-es a-sa-an-du/and, behold! on this occasion the one thousand gods are called to council; let them see and hear, let themnbsp;be witnesses�.

KUB I. I. 2. 66�68 (= G�tze, Hattusilis p. 20): [n]u-za ki-e K�R. K�R. MES dan-na-at-ta . . . EGIR-pa a-se-sa-nu-[n]u-un TXdi-at EGIR-pa Ha-at-tu-sa-an i-ya-nu-un, �andnbsp;these devastated lands I caused to be inhabited again; andnbsp;I made them subject to Hattusas again�.

KBo. 5. I. 2. 38 f. (= Sommer and Ehelolf, Boghazk�i-Studien 10, p. 8*): VII ANSU. K�R. RA K�R VII GUD KIR VIInbsp;MU SEN K�R VII dug pu-ul-lu-ri-ya na.-at IS.TU LALnbsp;su-u-wa-an, �seven horses of asphalt, seven cattle of asphalt,nbsp;seven birds of asphalt, seven pulluriya vessels; these fillednbsp;with honey�.

KBo. 3. 6. 2. 52�53 (= G�tze. Hatt. p. 22): nu-mu-kan PA. AN. SeS.YA ku-it KI.KAL.BA[D] ANgU.KU[R.R]A.nbsp;[MES] KUR Ha-at-ti S�-i e-es-ta, na-an am-mu-ug da-par-ha, �what infantry and cavalry was, subjectnbsp;to my brother, under my control, that I commanded�.

Neuter plural pronominal forms are used in precisely the same way as the neuter singular, and Friedrich [ZA NF 2. 295 f.)nbsp;thought that the confusion of the two numbers was a secondarynbsp;development, due to the fact that certain neuter nouns are alike,nbsp;or nearly so, in singular and plural. The similarity of the Hittitenbsp;usage illustrated above to what we have found in the Indo-European languages makes it much more likely that the use ofnbsp;the neuter singular to refer to plural antecedents, personal ornbsp;not, was an inherited construction.

This collective use of the neuter pronoun is obviously akin to Umbrian parse with a neuter plural antecedent, to Greek tovtonbsp;with two masculine plural antecedents (Ar., Eq. 854), to Latinnbsp;id referring to quadraginta minas (PL, Most. 627 f.) to Gothicnbsp;patei with a plural antecedent, to Old English hwat and Pcet withnbsp;plural nouns, to Russian eto with the plural, and particularlynbsp;to the neuter singular pronoun governing the partitive genitivenbsp;(e. g. hoc litterarum) which we have observed in Latin, Greek,nbsp;Germanic, Avestan, and Sanskrit. The collective use of neuternbsp;pronouns may, then, confidently be ascribed to Pre-Indo-Europeannbsp;as well as to Indo-European. The precise history of the othernbsp;idioms which we have noticed (the use of a neuter pronoun to

refer to a person in an indefinite way, or to refer to a masculine or feminine singular antecedent, and the use of a neuter pluralnbsp;to refer to persons of different sex) cannot so easily be determined,nbsp;although there is a fair probability that they all date from Indo-European times. I do not maintain, however, that every usagenbsp;here illustrated was inherited by the language which exhibits it.nbsp;A secondary spread of the neuter construction is possible at anynbsp;period.

DIALECTIC PHONETICS IN THE VEDA: EVIDENCE FROM THE VEDIC VARIANTS

I. It is sufficiently familiar to all Indologists that the horizontal classification of the Indo-Aryan languages based on the relativenbsp;dates of their literatures � Vedic, Sanskrit, Pali-Prakrit, thenbsp;modern vernaculars � is anything but rigorously scientific;nbsp;or, at any rate, that the terms �early� and �late� cannot benbsp;universally and lightly applied to the structural features of thosenbsp;dialects on the basis of such relative dating of their respectivenbsp;literatures. On the one hand, we know that dialects recorded at latenbsp;periods, and showing in general unquestionably late phonologynbsp;and morphology, nevertheless reveal not a few features which arenbsp;undeniably �early� �� earlier than the corresponding featuresnbsp;of some older dialects. This is most familiarly illustrated by thenbsp;well-known features of Pali and Prakrit which are clearly pre-Sanskritic, or even pre-Vedic, in character. A by no meansnbsp;exhaustive list is found in Wackernagel, Altindische Grammatik,nbsp;I. I, p. XIX f.; and see in general Pischel, Grammatik der Prakrit-Sprachen, � 6; Geiger, Pali (in Grundrip der indo-arischen Philo-logie), p. I. Paul Tedesco, in a brilliant study on verbs meaningnbsp;�give� and �take� in Indo-Aryan, which has not attracted thenbsp;attention it deserves, has shown convincingly that even modernnbsp;Indo-Aryan vernaculars contain linguistic features directlynbsp;inherited from Indo-European, which nevertheless do not appearnbsp;anywhere in the older recorded literature, Vedic, Sanskrit, ornbsp;Prakrit

1 JAOS. 43. 362!.; the prehistoric perfect passive participle of IE. root *d�, �give�, namely *d9'.�, Vedic and Sanskrit *ditd (not found), appears in phonetically regular development in the modern vernaculars (Hindustani diya, Naipali

2. nbsp;nbsp;nbsp;The converse of this is the equally familiar fact that thenbsp;oldest literature we have in India, that of the Veda, reveals notnbsp;a few traces of the existence of dialects (whether geographicalnbsp;or social in basis is a question which need not concern us now)nbsp;that were, as regards their phonology, essentially in the stagenbsp;of Pali-Prakrit. For examples I may refer again to Wackernagel,nbsp;op. cit., p. XVIII f., and the literature there quoted in footnote 2.

3. nbsp;nbsp;nbsp;I think, however, that the extent to which such dialecticnbsp;phonology appears in the Veda has never been fully appreciated.nbsp;A flood of new light will be thrown on this subject, as on almostnbsp;every other aspect of Vedic grammar, by the corpus of Vedicnbsp;Variants (as revealed primarily by the Vedic Concordance), whichnbsp;was begun by the late Maurice Bloomfield, and several volumesnbsp;of which (prepared by him and the present writer) are now practically ready for print The volume on Phonetics shows, amongnbsp;other things, a large mass of phonetic variations in the repeatednbsp;mantras of Vedic literature which are of the same character .asnbsp;the phonetic changes characteristic of the Prakrit languages.nbsp;Obviously these have an important bearing on the question ofnbsp;dialects in ancient India. They seem, indeed, to show the prevalence of dialectic tendencies in Vedic phonology to an extentnbsp;hitherto unsuspected.

4. nbsp;nbsp;nbsp;The phonetic variants are not all uniform in character. Onlynbsp;to a limited extent do they concern what may be called purelynbsp;phonetic variations, as in the following:

pikah (KSA. pigah) ksvinka nllasirsni te �ryamne TS. KSA. The name of this bird is otherwise always writtennbsp;pika; the KSA. form piga is isolated, and it is hardly possiblenbsp;to doubt that it represents a dialectic pronunciation of thenbsp;word pika, in which we recognize the familiar Prakritic changenbsp;of surd mute between vowels to sonant. No real lexical changenbsp;can conceivably be involved. On the other hand, innbsp;puman enam tanuta ut krij.atti RV.: puman enad vayaty udnbsp;gry,atti AV.,nbsp;diy�, etc., cf. Bengali, Assamese diyd, Marathi etc.), whereas Vedic, Sanskrit,nbsp;and all Prakrit dialects have secondary and analogical substitutes, generallynbsp;based on the present stem (Skt. dattd, Pkt. dinna). If Tedesco is right, and I amnbsp;fully convinced that he is, this old IE. form must have existed underground,nbsp;so to speak, during fully two thousand years of recorded speech in India.

^ For a brief account of the plan of this work and its present status, see Language, 5, pp. 129 ff.

the AV. form is, indeed, primarily a phonetic corruption for krnatti, but simulates a form of the root grath, granth, �tie�,nbsp;which clearly floated before the mind of the redactor. And innbsp;other, much more numerous cases, both forms are lexicallynbsp;defensible; we have what are really two different words, standingnbsp;in a certain phonetic relation to each other. Thus:

Here jigatn�, �swift�, is as possible as cikitn�, presumably �intelligent� (tho not quoted in the lexica). Nevertheless, the phonetic bearings of even the most intelligible of lexical variants of thisnbsp;sort are obvious, and obviously important for our problem. Theynbsp;occur in such large numbers that they furnish strong supportnbsp;for the purely phonetic variants of the type pika: piga, helpingnbsp;to demonstrate the strong tendency for surds to shift to sonants.

5. It must further be emphasized that in no case, perhaps, do the variants point to change exclusively in one direction. Thatnbsp;is, the changes from surd to sonant, for instance, are flanked bynbsp;a perhaps equal number of cases in which an older sonant isnbsp;replaced by a surd. These are not to be regarded as cancellingnbsp;each other. Both groups are really evidence for the same tendency,nbsp;namely, the change of surds to sonants � not the reverse. Broadlynbsp;speaking, and in so far as it is purely phonetic in nature, thenbsp;shift of sonant to surd is to be classed as what we call hyper-Sanskritism; that is, a leaning backward to avoid dialectic pronunciation. It is clear that the high speech avoided these Pra-kritisms. A quite natural result of this shyness was that it wasnbsp;sometimes overdone; just as, in our own time and land, a half-educated person may refer to the well-known capitalist as^'Mr.nbsp;John D. Rockefellow�, because he has painfully learned thatnbsp;the American dialectic word feller is vulgar and that in correctnbsp;English one should say fellow instead. So, for instance, a wordnbsp;which contained an original sonant may be anomalously writtennbsp;with a surd, as:

The MS. form can only intend the word devi, �goddess�, unless we suppose an otherwise unknown laletic (nursery) word for

For Prakritic change of sonant to surd in certain obscure dialects see Pischel, of. cit. �� 190�191. It is not impossible that some such dialect maynbsp;have had some influence on Vedic speech; but it seems to me that the abovenbsp;explanation of these Vedic forms is more plausible.

quot;mother,� which is implausible. Cf. TA. i. 5. i matrk kascana vidyate; the comm, simply says that matrk means madrk, �likenbsp;me�, and no other interpretation is at all conceivable. Yet bothnbsp;editions of TA., apparently with all known mss., read matrk.

6. nbsp;nbsp;nbsp;In the rest of this brief report I shall limit myself, in general,nbsp;to a few examples of the various types of what may reasonablynbsp;be called purely phonetic variants. The list is, of course, far fromnbsp;exhaustive, even for them; and let it be remembered that theynbsp;are supported in most cases by a very considerable number ofnbsp;variants which can be more or less defended lexically. The totalnbsp;mass of the materials is highly impressive; I must postponenbsp;their full presentation for another occasion. Exact referencesnbsp;are not given since they can easily be got in every case from thenbsp;Vedic Concordance.

7. nbsp;nbsp;nbsp;The change from surd to sonant has already been illustrated.nbsp;Dialectic in a broad sense, whether accurately to be called Pra-kritic or not, is the shift of voiced aspirates to h, as in the adjectivenbsp;kakuhd, quot;tall�, exclusively found in the RV., tho its originalnbsp;kakubhd occurs in Yajurvedic texts (cf. Grassmann, s. v. kakuhd;nbsp;Wackernagel, p. 251). This very word occurs in both its formsnbsp;in one variant formula, see the Concordance under kakuhhamnbsp;r�fam etc.; and as an instance of a similar case, .not noted previously I believe, in which h stands for dh, I may quote:

grdhrah �itikaksi vardhranasas te divydh (KSA. vdrhinasas te �ditydh) TS. KSA.

Von Schroeder emends the KSA. (7. 10) reading to vdrdhri^, but I consider this emendation unjustified; there is no reason tonbsp;doubt that KSA. presents a genuine phonetic variant.

8. nbsp;nbsp;nbsp;It is well known that the Prakrit languages frequentlynbsp;replace dentals by linguals; and, contrariwise, some of themnbsp;show the reverse tendency. See Pischel, �� 218�225. We find innbsp;the Veda such variants as:

This very word occurs in two other variants, the RV. showing each time the form avatd, for which later texts substitute avatdnbsp;(not known in RV.). The theory of Bartholomae (IF. 3. 179), approved by Wackernagel, p. 167, that avatd is derived from thenbsp;adverb avdr, ignores the RV. avatd and is, in my opinion, mostnbsp;unlikely. Tho relatively not numerous (perhaps two dozen,nbsp;including all sorts), the Vedic variants under this head are

specially interesting because genuine lexical influences are scanty; they are for the most part purely phonetic in character.

9. nbsp;nbsp;nbsp;There are a number of cases of the familiar Prakritic changenbsp;of d to � before y (Wackernagel, p. 163), as:

sahasriyo dyotatdm (TS. TB. dipyatam; MS. Viyo jyotatam) VS. TS. MS. KS. SB. TB.

�Off, the bow-string as it were from the bow!quot; The HG. variant, in the same passage, would mean literally: �Off, heaven as itnbsp;were of (or, from) a bowman!� But of course dydm is simply anbsp;would-be learned or mincing pronunciation of jydm, perpetratednbsp;by a half-educated person.

10. nbsp;nbsp;nbsp;Similarly the change of y to �, and that of j to y, are bothnbsp;Prakritic in character (Wackernagel, pp. 208 f., 163; Pischel,nbsp;�� 252, 236). ^

The �correct� form of this proper name is saujdmi. Cf. AV. 19. 34. 2 a jdgrtsyas tripancdslh (so mss.), for which read ydhnbsp;(or yds ca) krtyds etc., see note in Whitney ad loc.

11. nbsp;nbsp;nbsp;The change of ks to kh (Pischel, �� 317, 319) is not common,nbsp;but the following appears to be a certain instance:

idhmasyeva praksdyatah (SS. prakhydyatah) TB. TA. SS. ApS. The form praksdyatah is none too clear; the TB. comm, andnbsp;Caland on ApS. 4. ii. 5 take it as an unparalleled formationnbsp;from root ksi, �wasting away, being consumed�. It is morenbsp;probably derived from root ksd, �burn�, and means thenquot; burningnbsp;up, burning away�. In any case it seems clear that prakhydyatahnbsp;is merely a phonetic corruption of the other form; see Hille-brandt�s critical note on S�. 4. 12. 10, p. 249.

12. nbsp;nbsp;nbsp;There are a number of instances of the familiar change ofnbsp;ks, ts, and ps to ch (Wackernagel, p. 158; Pischel, ��3171!.,nbsp;327, 328). It happens that lexical considerations play a part innbsp;most of them, but the following is a sample of a purely phoneticnbsp;example, unusually complicated in character:

The form rchara occurs in AV. lo. 9. 23. The word means �fetlocks� (of an animal); its original form is unknown. Either MS. or VS. (or both) must be hyper-Sanskritistic back-formationsnbsp;from the �Prakritic� form with ch. Note also the Prakriticnbsp;variation of r: a in the first syllable (� 20, below).

13. nbsp;nbsp;nbsp;On m for v (and the reverse), which may be Prakritic,nbsp;cf. Bloomfield, JAOS. 13, p. XCVIIff.; Wackernagel, p. 197;nbsp;Pischel, � 251. The cases are few but interesting:

14. nbsp;nbsp;nbsp;The familiar cases of shift between h and v are certainlynbsp;dialectic, whether properly to be called Prakritic or not; seenbsp;Wackernagel, p. 183 f., Pischel, �� 201, 300. Much uncertaintynbsp;exists as to the writing of h and v thruout Sanskrit literarynbsp;tradition, for reasons pointed out by Wackernagel. The variantsnbsp;are numerous, and fall into three groups of nearly equal size, innbsp;two of which b and v seem respectively to be original, while innbsp;the third there is no way of deciding with confidence. They arenbsp;almost all purely phonetic; few genuine lexical changes arenbsp;included. School custom may have played a role here, alongsidenbsp;of true dialect; thus the Vajasaneyin or White Yajurveda schoolnbsp;shows a tendency to prefer v (of course not rigorously carriednbsp;thru). I will content myself with quoting three variants containingnbsp;the word fadbUa or Hlsa (also %insa), �fetter�. It is alwaysnbsp;spelled with b in the RV., which is not encouraging to the viewnbsp;which would connect it with Lat. vincio (Wackernagel, p. 183),nbsp;a view for which the form 'gt;vin�a may seem to lend support (butnbsp;see � 18 below):

15. nbsp;nbsp;nbsp;Like the b: v variants, those between r and I are perhapsnbsp;hardly to be called Prakritic in a strict or exclusive sense, sincenbsp;they appear familiarly thruout the history of Sanskrit; I constantly encroaches on r. Yet at bottom they are of the same nature.nbsp;There are Prakrit dialects which change every r to I (Pischel,nbsp;� 256), and the reverse change is not unknown tho only sporadic

in the dialects recorded (id. � 259); cf. Wackernagel, p. 215 f. The variants are numeroiis and almost wholly phonetic innbsp;character, with scant lexical bearings. A single example mustnbsp;suffice:

The r-form of this word is not recorded outside of RV., which is the more remarkable because of its obvious connexion with thenbsp;common noun �rl. For this the suffix ~ra is no doubt responsible,nbsp;first by dissimilation of one of the liquids to I, and then againnbsp;by assimilation of the remaining r to 1.

16. nbsp;nbsp;nbsp;The thoro confusion of the sibilants in the middle Indienbsp;dialects (Wackernagel, p. 226, � 197 e; Pischel, ��227�229)nbsp;must certainly be connected with the extensive occurrence ofnbsp;the like confusion in the Vedic texts. One or two out of manynbsp;examples:

And other cases of sru = sm, �flow�, so that BR. even postulate a root �ru, �flow�! Here s is original and � secondary; thenbsp;contrary is the case in:

sukesu me (AV. sukesu t�) harimanam RV. AV. TB. ApS. Since different Prakrit dialects change all sibilants to both snbsp;and s, both these changes may be directly dialectic and notnbsp;hyper-Sanskritic. Less numerous, but still not rare, are interchanges between s and the other sibilants:

17. nbsp;nbsp;nbsp;Especially in Jaina Maharastri (Jacobi, Ausgewahlte Er-zahlungen in Mdhdrdstri, p. XXII, � 7), but also to some extentnbsp;in other Prakrit dialects (Pischel, � 353), and in Pali (Geiger,nbsp;� 73). y serves as a separator or �Hiatustilger� between otherwise juxtaposed vowels, where it has no etymological justification.nbsp;Traces of such a tendency seem to be found to a rather surprisingnbsp;extent in the Veda, altho to be sure they are always complicatednbsp;by other considerations; that is, the variant forms are grammatically or lexically explicable. Thus in the following examplenbsp;the evidently secondary reading of TS. is, of course, a correctnbsp;dative case-form; but it is certainly not very intelligent, andnbsp;seems to involve a tendency to avoid hiatus betw�en vowels bynbsp;use of the glide-sound y:

There are similar cases in which alternative presence and absence of V and of the nasals may be similarly interpreted (cf. Pischel,nbsp;� 353: Geiger, � 73).

18. In the Pali-Prakrit languages no syllable can be �over-long�, that is, none can consist of more than two morae. No long vowel can be followed by more than one consonant, and thenbsp;nasalization (anusvdra, anundsika) counts as a consonant; or,nbsp;to put it otherwise, a nasalized vowel counts as a long vowel.nbsp;Where such combinations originally occurred, either the twonbsp;consonants are simplified to one, or the long vowel is shortened.nbsp;In the final outcome, after numerous analogical developments,nbsp;it boils down to this, that Pali and Prakrit feel as phoneticallynbsp;equivalent the following three cases: (i) long vowel plus singlenbsp;consonant, (2) short vowel plus double consonant, (3) nasalizednbsp;vowel plus single consonant �. These three come to interchangenbsp;at random, with no fixed relation to etymological differencesnbsp;of origin: Pali-Prakrit siha, Pali sdrambha = Sanskrit sihha,nbsp;samrambha; Pali-Prakrit damsana, sumka = dar�ana, sulka (intermediate forms *dassana or *ddsana, *sukka or *s�ka); Pali-Prakrit nidda, n�dda = n�da; Prakrit vdsa = varsa; Pali sdsapa =nbsp;sarsapa, etc.

The beginnings of this phonetic confusion seem to be present in certain Vedic variants. For long vowel vs. nasalized vowel,nbsp;cf. the variant quoted above, � 14:

Here there is no lexical variation; we have forms of one and the same word. Whichever may be the older form (cf. � 14), it isnbsp;clear that they are phonetic equivalents.

As for variations between long (double) consonant and long vowel with single consonant, I know of no variants which cannbsp;be called �purely� phonetic. That is, in every case known tonbsp;me, both forms of the variant are capable of independent lexicalnbsp;or morphological explanation. And yet, if we confront such

^ Or, we may phrase it thus: Long vowel, nasalized vowel, and long consonant are interchangeable. A concise and excellent statement of the conditions isnbsp;found in Geiger, Pali, �� 5, 6. In Pischel the facts are more scattered, and statednbsp;less clearly and less happily; see his �� 90, 62�65, 74, 76. I do not believe thatnbsp;word-accent has an3'thing to do with it (cf. Pischel, � 90).

variants as the following with each other, it seems to me hard to doubt that they are bound together by a common phoneticnbsp;character, which makes its influence distinctly felt in spite ofnbsp;the individually distinct explanations which suggest themselvesnbsp;easily enough under each one;

ague samrdd ise raye (ApS. rayyai) . , . A�. ApS.: ise raye (ApS. rayyai) . . .VS. MS. SB. TB. AS. Ap�. Of course, rayyainbsp;(and also rayya, once in the RV. itself) is connected with thenbsp;stem rayi, parallel to rai whence comes raye. I do not questionnbsp;the propriety of this connexion, even tho the stem rayi is usuallynbsp;masculine and these are distinctively feminine forms; other case-forms of rayi, such as rayih, are also occasionally feminine. Ifnbsp;these variants were isolated, they would perhaps prove nothingnbsp;for the phonetic tendency under consideration. But taken withnbsp;the rest of this group, they do seem to me to suggest an urgenbsp;in the direction of treating short vowel plus double consonantnbsp;as the phonetic equivalent of long vowel plus single consonant.

socasva (VS. also, rocasva) devavltamah (KS. Hittamah) RV. RVKh. VS. TS. MS. KS. SB. TA. Roots vl and vid; quot;god-rejoicing� and quot;god-finding�. Cf. next.

istam ca vltam (SB. AS. vittam) ca (SS. cabh�t) MS. SB. TB. AS. SS. Cf. preceding.

indrdya (MS. MS. add tva) susuttamam (VS. SB. sus�tamam) VS. MS. KS. SB. MS. Equivalent bases su-t (with euphonic tnbsp;after final short vowel of monosyllabic base) and sm.

apsarassu (HG. apsarasu ca) yo gandhah ApMB. HG. And others with the same interchange. Here noun-declension isnbsp;concerned; apsarasu is from an a-stem apsard, parallel andnbsp;secondary to apsaras. This, of course, does not exclude thenbsp;likelihood of phonetic considerations, any more than in the othernbsp;cases of this group.

19. There are distinct traces among the variants of the very familiar Prakritic tendency to assimilate adjoining consonants;nbsp;this is most marked with semi-vowels and liquids, which arenbsp;easily assimilated to other consonants (cf. Jacobi�s table ofnbsp;consonants in the order of their resistance to assimilation, Ausgew.nbsp;Erz. in Mdh., � 27, p. XXXI):

abhinne khilye (TB. khille) nidadhdti devayum RV. AV. TB. The TB. reading is well attested, and can certainly mean nothingnbsp;but khilye. (Comm, khillbh�te.)

20. A very numerous group of variants concerns interchanges between vocalic r and other vowels, a, i, and u] also r and consonantal r plus vowel; cf. Wackernagel, �� 9, 16, 19. It is well-known that Pali and most Prakrit dialects have no r; they mostnbsp;commonly substitute a, i, or u for older r. And the modernnbsp;Hindus pronounce r (in Sanskrit words) as ri in most partsnbsp;of India (in some regions, notably Maharastra, the vowel followingnbsp;the consonantal r has rather an M-coloring). I select a few instancesnbsp;which show this dialectal phonetics in the Veda most clearlynbsp;(cf. also � 12 above):

ava sma durhan�yatah (SV. durhrn�) RV. SV. Here the RV. has Prakritic vocalism, for which SV. (secondarily, of course)nbsp;has substituted the more �correct� or �hifalutin� form with r-tejo yasasvi sthaviram samiddham (SG. samrddham) SG. PG.nbsp;ApMB. HG. It is highly likely that samiddham is a Prakritismnbsp;for samrddham, �plenteous�, as suggested by Kirste on HG.nbsp;1.4.6, and approved by Winternitz, ApMB. Introduction, p. XXIII;nbsp;�inflamed� [root idh) is incongruous in this context (despite thenbsp;et5unological meaning of tejas). Yet three texts have the dialecticnbsp;form, and only one the �correct� samrddham (and even ^G hasnbsp;a V. 1. samiddham, in two mss., probably owing to a scribe�snbsp;recollection of the variant form as found in the other texts).

puramdaro gotrahhid (MS. %hrd; TB. maghavdn) vajrabahuh VS. MS. KS. TB. Here we have a curious and most interestingnbsp;hyper-Sanskritism in MS. All the terms refer to Indra, one ofnbsp;whose regular epithets is gotra-hhid, �splitter of the cow-pensnbsp;(clouds or mountains?)�. All the mss. of MS. agree here on thenbsp;form gotrabhrd, which is otherwise unrecorded. It can only be anbsp;pedantic substitution for gotrabhid, apparently under the sciolisticnbsp;assumption that i stood for r; at any rate, the alteration cannbsp;scarcely be understood except in the light of this dialectal phonetics. For goira-bhrd is hardly capable of intelligent interpretation.nbsp;Perhaps it was felt as meaning �supporting (bhr) the clan (gotra)� �,nbsp;but if so, its incongruity, wedged in between puramdara, �cleavernbsp;of fortresses�, and vajrabahu, �armed with the club�, is obvious.

tvastrmantas (MS. M�. tvastri^; ApS. tvastu'^) tvd sapema VS. MS. KS. SB. K�. ApS. M�.quot;

The same Prakritic u is substituted for vocalic I in one variant: samvatsara rtubhih samviddnah (KS. %his cdklpd,nah, ApS.nbsp;%his cdkupdnah) MS. KS. ApS.

trsucyavaso (MS. tnsu�) juhvo nagneh RV. MS. The adverb trsu is the only word which can be involved here.

bhrmim (TB. hhrumim) dhamanto afa gd avrnata RV. TB. The mss. of TB. appear to be unanimous (if we -may trust thenbsp;two editions) in reading hhrumim, a form not otherwise recorded,nbsp;for bhrmim. � The reverse of this, r by hyper-Sanskritism fornbsp;ru, is found in the form prsvd, occurring several times in textsnbsp;of the Taittiriya school for regular frusvd; the somewhat similarnbsp;phenomena included under the name of sawprasdrana may, tonbsp;be sure, have a bearing on this:

21. nbsp;nbsp;nbsp;There are very many variants in which long diphthongsnbsp;(ai, au) interchange with the corresponding short diphthongsnbsp;(e, o). In most of these morphological matters are concerned.nbsp;But in a few, at least, it appears that the shift is purely phonetic,nbsp;and therefore belongs with the well-known universal law ofnbsp;Pali-Prakrit by which ai and au are everywhere reduced:

sukurird svaupa�d VS. TS. KS. SB.: sukarird svopasd MS. But several mss. of MS., and its pada-patha, read svaup^, which maynbsp;then be the correct reading.

22. nbsp;nbsp;nbsp;Allied to these are a number of interesting cases in whichnbsp;we seem clearly to have Prakritic reduction of aya to e, and avanbsp;to o (Bloomfield, AJP. 5. 27!!.; Wackernagel, p. 53!.):

namah kih�ildya ca hsayatj^dya (MS. kseyAya) ca VS. TS. MS. KS. ksena is not otherwise recorded.

to-to (MS. KS. MS. tava-tava, TS. ApS. to-te) rdyah VS. TS. MS. KS. SB. KS. ApS. MS. to for tava, gen. sg. of the 2d personalnbsp;pronoun.

23. nbsp;nbsp;nbsp;The insertion of an epenthetic vowel, generally i, betweennbsp;two consonants, usually a liquid and a sibilant or h, is likewisenbsp;to be regarded as a Prakritism, tho fairly wide-spread in Sanskritnbsp;(cf. Wackernagel, pp. 56 ff., especially p. 57, � 51 end; Pischel,nbsp;�� 140)- It is found in not a few Vedic variants, as in thisnbsp;sigmatic aorist from the root pr, �cross�:

tan nah parsad (MS. parisad) ati dvisah TS. MS. KS. TB. Note that the meter is against the inserted vowel. (Read parisad innbsp;both occurrences in MS.)

of variants showing the kind of shifts which are found in Middle Indie phonology. Granted that many, and under some rubricsnbsp;even most, of them may be explained lexically or morphologically : it would nevertheless be rash to deny that phonetic influencesnbsp;are concerned even in the latter, especially since they are supported by a very considerable number of purely phonetic cases.nbsp;The principle is, indeed, far from being a new discover}^ as hasnbsp;been indicated. But it has never been illustrated so extensivelynbsp;as it will be in theVedic Variants; and some of the rubrics herenbsp;included are presented for the first time in Vedic phonology.nbsp;The special character of the illustrations, namely the occurrencenbsp;of double forms of each variant, with and without Prakriticnbsp;phonology, makes them especially valuable and interesting. Thenbsp;cases of hyper-Sanskritism are perhaps the most interesting ofnbsp;all, as pointing to a rather definite consciousness, on the part ofnbsp;the handlers of the texts, of the antithesis between the phonologynbsp;of the high speech and that of the popular dialects.

I. Pali and Ardhamagadh�.

As is known Geiger considers Pali as a kind of Ardhamagadh�, a �Verkehrssprache� whose basis was Magadhi and which Buddhanbsp;himself spoke. I have protested against this in Language, IV,nbsp;pp. 104�105, without, however, presenting linguistic proof thatnbsp;the differences between Pali and Ardhamagadh� are too far-reaching to permit us to label Pali as a kind of Ardhamagadh�.nbsp;Some apparent similarities I have accounted for as being relics ofnbsp;a Magadhan original of which the Pali canon is a transformation.nbsp;It should be noted that Ardhamagadh� is so much later thannbsp;Pali that a good deal of material can not be relied upon, for thenbsp;changes may be secondary, and some sevca-tatsamas may havenbsp;crept in; also influence of Maharastri as the Prakrit literarynbsp;language par excellence is also to be reckoned with. I thereforenbsp;only list a few differences such as are quite convincing:- Amg.nbsp;nom. sing. masc. a-stems, -e (in verse -0 owing to Maharastri influence), P. -0; Amg. gen. pi. ra-vnam, P. rannam, raj�nam',nbsp;Amg. loc. sing, tamsi, etc., P. tamhi (cf. Sanskrit tasmin); Amg.nbsp;bemi (Skt. hravlmi), P. br�mi (Epic Skt. br�mi); Amg. dh�yd,nbsp;P. dhitd ; Amg. iha, P. idha ; Amg. ciyatta, P. catta (Skt. tyakta)\nbsp;Amg. giha, P. gaha (Skt. grha ; ghara usually replaces gaha in P.);nbsp;Amg. feccha-i-, P. pekkhati.

2. Asokan Petenika.

In my paper �Walleser on the home of Pali� {Language, IV: loi�105) I argued against the derivation of Petenika from *Prd-

tisthanika (a vrddhi derivative of Pratidhdna), laying emphasis on the early ablative Patithdna, the later locative Patithdne, andnbsp;Pa-i�tjidna (in a compound) which correspond to Sanskrit Pra-tisthdna. I overlooked some evidence from the middle of the thirdnbsp;century B.C. favoring my view, namely, prdtithdnasa bhichuno,nbsp;no. 12 of the Sanchi stupa, Tope i, p. 98 of E. I. 11, ed. Biihlernbsp;(see also pp. 104 and 387), which means �of the prdtithdna monk�,nbsp;that is, �of the monk from Pratithdna (Sanskrit Pratisthdna)�.nbsp;The fiaiOava of Ptolemy, also corroborates my view. For the derivation of Petenika, my old explanation that it stands for *Pdi-trayanika (I. F. 24: 52 f.) still holds valid.

3. Girnar traidasa.

Charpentier, JRAS. 1926, p. 139 derives traidasa from *trai-dasa without referring to the explanation given by Johansson, Shb. I, p. 136 (22 of the reprint) or my note JAOS. xxxi, p. 234;nbsp;and incidentally criticises Gray most sharply. Such archetypes asnbsp;^traidasa and *trayada�a (in spite of Hultzsch�s return to this,nbsp;CII, vol. I [new ed.], p. Ivii) will not satisfy the phonetic requirements of the dialects of the Asokan inscriptions. There still isnbsp;nothing for nor against Johansson�s theory that the archetypenbsp;is *tray�da�a from an earlier *trayazda�a, as far as the phoneticsnbsp;are concerned. I now hasten to observe that Turner, Bull S.O.S.nbsp;4, pp. 363, 364 goes back to Johansson�s explanation withoutnbsp;giving him any credit whatsoever.

4. Additions to Whitney�s Root-Book.

Whitney�s Root-Book of Sanskrit will ever impress any one who uses it as being truly marvellous. Minor additions have beennbsp;made from time to time, and probably will continue to be made.nbsp;In my �Linguistic Archaisms of the Ramayana� [JAOS. 25)nbsp;I added a few items; and in the present paper hope to add a fewnbsp;more, but it should be borne in mind that these are not the resultnbsp;of systematic reading: on the contrary they are the result ofnbsp;reading by snatches in leisure moments.

The following forms (not roots) are listed as occuring only in Classical Sanskrit; they occur in the Bombay edition of thenbsp;Ramayana of 1902: k%ic [samcukoca, v. i. 33), k�j [cuk�jatus,nbsp;vi. 90. 49), budh [buddhvd, v. i. 102, v. i. 180, v. 48. 39, v. 58. 12,nbsp;V. 67. 44, vi. 60. 55, vi. 64. 14), bhds [samabhi-hhdsisye, vii. 15.

25), vrdh {vardhitum, v. i. 92), vraj {vraje, v. 41. 9), sphut {pus-phota, vi. 77. 7, vi. 77. 16). I may add the infinitive gaditum (gad) listed by Whitney as occuring in Classical Sanskrit only, occursnbsp;in a manuscript of the Ramayana in the British Museum, but Inbsp;have unfortunately mislaid the exact reference.

The form liyati is given as occuring in the Upanishads and Classical Sanskrit only: observe, however, avaliyantl in thenbsp;above mentioned ed. of the Ramayana at vi. 114. 33.

Listed as occuring but once in the S�tras is jigah�re: but vija-gahire occurs in the Ramayana at vii. 31. 35; no middle voice is given for Classical Sanskrit, but the active jagaha is given asnbsp;occuring in Classical Sanskrit and that only.

The gerundive patitvd, listed as AV. B., occurs in the Bombay 1898 ed. of the Bhagavata Purana at v. 17. 6.

Under kup cukopa is listed as U.E. It is found in the Bhagavata Purana at iv. 4. 9, the Kathasaritsagara (ed. Nirnayasagara),nbsp;XV. I. 148, the Raghuvahsa (ed. Nirnayasagara), 3. 56.

The active perfect of krus is listed as being peculiar to Epic Sanskrit: cukrusus occurs in the Raghuvahsa at 8. 39, andnbsp;cukrosa in the Bhagavata Purana at ix. 16. 13.

Under smr sasmdra is given as E.: it is found in the Bhagavata Purana at ix. 16. 3 (see also JAOS. 25) and Raghuvahsa at 15- 45-

Under sah the middle perfect sehe is given as authorized by native grammarians but non-quotable; the third person pi. sehirenbsp;is found in the Raghuvansa at 12. 94; see also the Kumarasamb-hava (Nirnayasagara ed. 1893) at xii. 48.

The infinitive spardhitum is listed as AV. B., but it is found in the Naisadhiyacarita (ed. Bombay 1902) at 8. 30.

Under i pat the aorist apdti is listed as B.: it is found in the Kumarasambhava at xv. 29.

The perfect mamdda under mad is given as V. B.: see the Kumarasambhava at ix. 51.

The infinitive �ahsitum is listed as E.; it occurs in the Kumarasambhava at V. 51.

The aorist asevista (under sev) is listed as non-quotable though authorized by the native grammarians; the 3 pers. dual of thenbsp;middle of the aorist asevisdtdm occurs in the Kumarasambhavanbsp;at vii. 42. The commentator Mallinatha correctly explains thenbsp;form as sevater lun.

The gerund grastva is not in the Root-Book; it is found in the Vayu Purana (Anandasrama ed.) at 7. 72.

The gerund charditvd is given as authorized by the native grammarians but non-quotable. It occurs in the Vayu Purananbsp;at 61. 19.

5. The alleged Fox stem anemAdfci- �be cold�.

On page 49 of Bulletin 85, Bureau of American Ethnology, I have given a stem dnemAHci- be cold. Professor Leonard Bloomfield calls my attention to the fact that the alleged stem shouldnbsp;be divided dnem and aHcI. This is no doubt correct; cf. Foxnbsp;dnemi- �be in agony�; the stem of the second member is �-�, whichnbsp;takes post-verbal -�^tci- -t-: compare Fox nepA^'tci- �be cold, chilled�; slgA'^tci- �be frozen to death�; kepHenwi �it is frozen over�nbsp;{kepi- �enclose�, -en- inanimate copula, -wi pronominal ending);nbsp;Ojibwa ningVkatc �I am cold� (Jones� Ojibwa Texts, i. 62. 8),nbsp;gashkadin �it freezes over� (from Baraga); Cree so'kAtin �it isnbsp;frozen strongly� (restored from Watkins� sokutin). Professornbsp;Bloomfield cited a Menomini correspondent which I have unfortunately misplaced

6. Fox -/a-.

Years ago Jones construed -ta- as an inanimate copula; and I have followed him in this. Recently Bloomfield has given thenbsp;form as -eta-. Now, no doubt, -et�'-isthe living stem, but presumably historically the stem is -td-, not -eta-, as is shown by Foxnbsp;klwd!gwAtdwi �it lies about��, pemipahotdwi �it goes by� (of anbsp;train), wVcAtawi �it is hot�, nemAtdwi �it is erect�, dA^kAtaginbsp;�it is burned�. Decisive proof can only be given when the conditions in other Central Algonquian languages are known, excludingnbsp;Sauk and Kickapoo which are so close to Fox. If -eta- is historically correct then we must assume contraction of ac to a, andnbsp;contraction of oe to 0.

7. A non-Algonquian trait of Arapaho.

As I stated years ago Arapaho is a very divergent Algonquian language; and I still adhere to this view. The grammar is funda-

^ Correspondents occur in a number of other Algonquian languages, but unfortunately I have not been able to control the exact phonetics of the wordsnbsp;in other sources to make it worth while citing them, save Kickapoo kepAtenwi.

mentally Algonquian; from time to time I have been able to note some such traits as were previously unknown; the amount ofnbsp;vocabulary that can be shown to be Algonquian apparently isnbsp;but a small percentage of the total vocabulary: yet as the demonstrable phonetic shifts are very complicated and of an unexpectednbsp;character, it is always possible that more lexical material willnbsp;eventually fall in line So far it has been the lack of certain Algonquian grammatical features that has impressed any one, notnbsp;the presence of features which are un-Algonquian. I now proceednbsp;to list a feature which can not be duplicated so far as I know (andnbsp;I have had several years of experience with Algonquian languages)nbsp;in any other Algonquian language, and that is infixation wdthinnbsp;the primary stem. [For an infixation of this type note Teton Siouxnbsp;pahta �bind� pawahta �I bind�.] The syllable -en- is infixed afternbsp;the initial consonant to express customary or habitual actionnbsp;{ne- is prefixed to stems beginning with n-). So we have be'ning^

' A few notes on the symbols used in this paper: A is nearly like tt in sun but with a distinct a-tinge; e is open; � is long and close; i is open; � is long and close;nbsp;0 is very open; cu is open but long (like aw in law); ci is as in hat; a is as in bad;nbsp;� is long unrounded �] t and k are unaspirated; 6 is a surd interdental spirant; tcnbsp;is ch in church but unaspirated; q is nasal o, occuring when not rhetorical onlynbsp;after �; qgt; is nasal co, occuring only after h when not rhetorical; f is aspirated t,nbsp;occuring only when terminal; ft' is aspirated, but occurs only in the combinationnbsp;-ft'� (in which � is voiceless); ' medially before consonants denotes an aspiration,nbsp;hardly the surd velar spirant; � is a strong glottal stop. The complicated phoneticnbsp;shifts to which I have alluded are almost all conditioned by the quality of adjacent vowels, palatization being extremely common. Thus normal Algonquiannbsp;p appears as ft, but as tc before palatal vowels (whether these have been subsequently lost or not); m as ze; but b before I and i, and after e; t; (surd velar spirant)nbsp;is the normal correspondent to Fox 'eft. Ojibwa eft, Menomini and Cree sft (Centralnbsp;Algonquian eft according to Bloomfield) but appears as s before i; some othernbsp;shifts are: normal Algonquian tc becomes 6; Cree 'i, normal Algonquian s, appearsnbsp;as e; Cree /, normal Algonquian n, appears as 0 (which as a matter of fact Bloomfield without knowing the Arapaho correspondent, set up as the Central Algonquian prototype); w appears as � as does y, though ny- appears as y-; corresponding to Fox 'c we have s; a becomes co; u becomes t; 5 becomes t; which are onlynbsp;a few of the shifts which I have worked out. Thus it happens words which arenbsp;reall)! Algonquian may be enormously disguised, e. g.. Fox ni'pi^tct� �my tooth�,nbsp;n� pii^n^^ my teeth� correspond to Arapaho n�'tcid, n�'tcito^; Cree ispimi�k,nbsp;Ojibwa icpimiflg, Menomini ispamiah, Kickapoo i'^pemegi, Fox A^peniegi �above�nbsp;correspond regularly to Arapaho ixteebe^. � With the words in the text comparenbsp;Fox mi-n- �give�, pl^tci- �into�, mVei- �fuzzy�, -negwd- �arm�, -igwd- quot;face�,nbsp;-cin- �fall� (animate), ml si- �cacare�, notaw- �hear� (animate obj.). Furthernbsp;explanations would involve too much space.

�I give him, her� but hcot' bi'ng�quot; �I shall give him, her�; tceni'dik�do^ �I put him, her in�, tcen�'dik�'t�ncg �I put it in�nbsp;as contrasted with hco'f tciBik��tinco ��I shall put it in�;nbsp;bent'sinesaf �he, she has a fuzzy arm� , but hco'f hi'sinesUt' quot;he,nbsp;she will have a fuzzy arm� {tdbinesd'si^ �he broke his armnbsp;by falling�); beni's��f �he has a fuzzy face�, but ha'bi'sief �henbsp;had a fuzzy face�; ten�'tconiquot;tquot; �he, she is quiet� but t�!tomAk^'^^'nbsp;�be quiet!�, hoj'f t�'tmnif �he, she will be quiet�; henitcaco'tifnbsp;,,he gradually pours it out�, but hi'tmnoy �pour (sing.) it out!�;nbsp;be�nihif �he, she defecates�, but Acu'f bihincg �1 shall defecate�.nbsp;[For the prefixing of the element ne- observe nen�'detcwf �he,nbsp;she is afraid� but hco'f n�'detccof �he, she will be afraid�; neni'-tAng^ �I hear him, her� but hco'f nl't^ng^ �I shall hear him, her�.]nbsp;As intimated above, this trait is unique among Algonquiannbsp;languages; whether it is self-evolved within Arapaho or borrowednbsp;from some non-Algonquian stock is a question which I can notnbsp;answer at present. To an Indo-Europeanist such a question maynbsp;seem superfluous; it is not to an Americanist: such borrowingsnbsp;are well-recognized (e. g. the post-positions of Upper Chinook arenbsp;due to Shahaptian influence); it is only debatable as to whethernbsp;such borrowings can occur on a large scale: this, if I am not mistaken, is the chief point at issue between Boas and Sapir.

A MATTER OF SEMANTICS

I wish to treat of a group of words, the formal connections of which have been suggested by others^. The semantic relationsnbsp;do not seem to me to have been described satisfactorily, andnbsp;perhaps this is the reason that there is still difference of opinionnbsp;about the exact delimitation of the group. The matter maynbsp;also have a more general interest; for I shall assume a connectionnbsp;not considered under the sense of smell by Bechtel in his worknbsp;Ueber die Bezeichnungen der sinnlichen Wahrnehniungen in dennbsp;idg. Sprachen, Weimar, 1879.

To begin with the word whose connection is least recognized^. In his discussion of the suffix -aMog Debrunner� notes thatnbsp;it is associated above all with w-stems, and yet misses the prettiest example, a/neg�aX�og : afisQ�v�g. In classifying the meaningsnbsp;Liddell and Scott place �terrible to see� ahead of �terrible tonbsp;hear;� but the Homeric usage (and, bar a few imitations, thenbsp;words are confined to Homer) indicates that historically thenbsp;reverse is true. So also does the connection to be suggested withnbsp;ajuagayeco.

First the Homeric usage. The words are used of loud (and frightful) sounds emitted by human organs of speech: a/isQ�v�vnbsp;Poomv O 687 = 732; (rfiEQ�aX�a Idymv E 302 0 321 11 785 T 41nbsp;Y 285 382 443 ^ 81; afXEQ�aX�ov � �fi�rjaev 0 92 � 305 m 537;nbsp;a/xsQ�aX�ov d (d� fi�y�) mfua^ev Z 35 t 395: ansQ�aX�ov � mnoiOLvnbsp;EHETcXero T 399; vfjsg | a/uegdaliov xovd^rjaav dvcrdvTcov V7i �Ayaimv

^ In the following etymological dictionaries: Walde-Pokorny 2. 691; Walde^ 478�9; M�ller 432; Trautmann 271; cf. also the references given in note 8.

^ Proposed first by Fay, Am. Journ. Phil. 26. 173 (1905) and now approved by Muller.

^ 334 = n 2y'j. Of other noises: a/jicpi d� dc�/.ia \ a/jieQ�aXiov xovd^i^e (v. 1. xavdxdie) x 399 of the wailing of Odysseus� comrades,nbsp;Q 542 of Telemachus� sneeze; avzdg v%� \ airngbaliov xovd^it,Bnbsp;Tiobmv avxmv rs xal mncov B 466; and of the crash (xovafiZiv, xova-(ilCsiv) of arms, xalx�i; N 498 0 255,nbsp;nbsp;nbsp;nbsp;0 648, xvrjjutg O 593.

Separately may be noted the use of thunder Zsvg \ ansgbaX�a xrvTt�cov H 479.

In words that are reactions to such stimuli a shift of meaning can come about in one of two ways, (i) The emotional elementnbsp;may become so far predominant that the sense element becomesnbsp;indifferent, and the word is then associated with any stimulusnbsp;that produces such an emotion. (2) A speaker applies an epithetnbsp;because an object affects him through one of his senses, while anbsp;hearer, who is affected more powerfully through another sense,nbsp;takes the word in an unintended meaning and afterwards usesnbsp;it accordingly.

As a result of such processes most of the other Homeric examples are in a transitional stage, and their precise meaning cannotnbsp;be determined. Thus xaXxog auegbaMog may just as well benbsp;quot;dire clashing� as �dire gleaming,� cf. N 192 nag b� aganbsp;I afiegbaMm xsxdXvmo. Nor is it pinned down in either directionnbsp;by the context in:

�laa� of^gi^ov �yxog, | a/uegbaXeu)' /.l�ya b� d^cpl adxog jivxe bovgog dxcoxfj. Either meaning will suit in either passage; a preference willnbsp;depend in reality on whether one likes reiteration or contrast. Sonbsp;also in: A 579 afiegba^�oj b� X�ovxs bv tv ng�xrjai (i�saai 1 xavgovnbsp;�gvy/.ir]Xov �x�xxjv. Both slayers and slain may have their epithetsnbsp;either from the same or from different sense spheres�contrastnbsp;or parallelism again. Perhaps it is worth noting also that thenbsp;only part of the body to which these adjectives are applied isnbsp;that from which the voice comes. There are no oaas a/LisgbaX�ojnbsp;nor %�(jo�s ap.egbaX�ai; but the Gorgon�s head is bem] xe a/usgbvrjnbsp;xe E 742, and there is a ajxegbaUri xecpaX^ p, 91 on each ofnbsp;Scylla�s necks while her voice was commented upon uncom-plimentarily 86) a few lines before.

Uncertain similarly is the meaning in B 309 bgdxagt;v �nl vagt;xa bacpoivdg, \ ajjiegbaXtog and in Y 65, where a/jbegbaXi evgwevxa^ is

said of the dwelling of A�doneus. These may be cases in which the change has advanced considerably; and with them may benbsp;put:nbsp;nbsp;nbsp;nbsp;401 Pijar� aiyi�a �vaav�saaav \ afxEQ�aX�rjV and X 6og aiisq-

�aX�oQ d� ol afxcpl tceqI ar�)j�saaiv aoQxrjQ. Both have, to be sure, to do with arms, but arms of such a nature that an epithet derivednbsp;from the noise they make is hardly to be expected.

Clear evidence for the completed change is found in but three passages; one connecting with the use of armor, another like B 309nbsp;used of a serpent, and the third textually uncertain. They are:

O 6og TtrjXrj^ | OfiEQ�aX�ov XQoraxpoKyi, rivdaoEXo /xaQvafi�voio, X g5 (dQdxcav) a/i�Q�aX�ov d� d�doQxsv^, C I37 a/LiEQdaX�oQ � avxfiai lt;pdvrjnbsp;xexaKco/^avoQ aljurj In the scholia on the last passage XEvyaX�ognbsp;and dgyaX�og (Zenodotus) are cited as variant readings.

I see no possibility for doubting that the words were used originally of sounds. To put it otherwise, a/xegdvog must benbsp;regarded as a verbal adjective to * a/iegdai a verb used of loudnbsp;noises. On account of the etymological connection with Skt.nbsp;mrdndti, Lat. morde�, etc.�, we may say more definitely usednbsp;of crushing and crashing noises. We should expect it to havenbsp;been used of the sea and of thunder, which would account fornbsp;the connotation of terror that it acquired. Such is the range alsonbsp;of afiagayeco that I should regard as coming from the same basenbsp;but with different �root determinatives�: *snier-, *smer-d-,nbsp;*smcr-ag-. Compare B 210 a/mgaysi d� xs vdvxog, 0 igg dsiv^v xenbsp;Pgovxijv, ox''an ovgav�dev a/xagayriar], and note also its use of thenbsp;battle crash Hesiod, Theog. 679 (a/xagayiCco ib. 693). The underlying word (T/iogayos occurs in Horn. Epigr. 14. 9, while (a)iJiaQdaamnbsp;is said to be cited by Erotianus and by the Etymologicum Magnumnbsp;720. 58. We may nov/ note the rhyme with ndrayog �clatter, clash�nbsp;and its kindred �

Words used of thunder are frequently applied to sounds coming from the bowels. The association is evidenced as early as the

� Cf. Walde-Pokorny 2. 278�279; Boisacq 884; Walde 495; Muller 275�276; Falk-Torp 2. 1080�81.

� If the vocalism of naiayos has been influenced by that of a/xdgayost the former may derive from the base *peta-, meaning originally �noise made by anbsp;fall.�

Homeric h. Merc. 295 f. by the way (cf. Allen-Sikes, ad loc.) that the infant Hermes parodies the omen of Zevg vxpifiQS^�rrjQ.nbsp;Strepsiades too had evidently been struck by the parallelismnbsp;of PQovt-^ and tioq�'^ and is delighted (Arist., Clouds 394) withnbsp;the explanation of the similarity in which natayto (389) andnbsp;PQOvrdm (391) had been used. For a/xagay�co Hippocrates 658. 29nbsp;is cited, while ^qs^o} and its cognates will be considered below.nbsp;The same may be presumed for *a/LisQ�oj and we will then notenbsp;that *sm6rd- and *perd- were rh5ane words^�. That may help explainnbsp;why in the differentiation of *perd- and *pezd- the former camenbsp;to mean the loud sound; and why *afj,SQ�co dropped out of thenbsp;literary tradition so early and completely that even Aristophanesnbsp;knew nothing of it.

It gives also the semantic link between o/zegdaXtog and a group of Balto-Slavic words, that may be represented by Lith. smirdziu,nbsp;smirdeti �stink�, for which Trautmann reconstructs *smirdy5nbsp;as a startform. That is, in preference to the suggestions of anbsp;�beiBender Geruch� or a �reibender, kratzender Geruch� whichnbsp;do not seem adequate, I should derive the specialized meaningnbsp;of *smerd5 �stink� (recognized by Walde-Pokomy) from annbsp;earlier meaning �make a crashing noise�, the change being aidednbsp;by the existence of the rhyme word *perd5. Another line, of development from this earlier meaning ends in afiegdvog, a/usQ�aX�ognbsp;�frightful to hear�, �frightful�.

Latin merda lt; *sm�rdd may be exactly parallel to *p�rdd continued by Albanian porde^^, and differ from only in itsnbsp;vocalism, a restriction that is not surprising, cf. Hirt, Hdb. d.nbsp;gr. Laut- u. Formenlehre 343. To separate Gothic smarn�snbsp;�axvPaXa� seems difficult. Falk-Torp 2. 1086�87 connect it withnbsp;OHG smero, etc. thus bringing it into a group with which it contrasts semantically. I should prefer to consider the possibility thatnbsp;within Indo-European itself, under certain conditions at least,nbsp;-rdn- became -rn-; for a/j,sQdv�g could easily be a new formationnbsp;in Greek.

In the present both verbs seem to have varied between *-5 and *-y�. Greek nigSoiJLm may represent either, if the treatment of *-dy- parallels that of *-gy-after consonants as elsewhere; a y�-present would be the normal thing besidenbsp;the future anonaQS^Oofiai.

Jokl, IF 37. 96. For the change of meaning, contrast dialectic (Carinthian) Schass �noQamp;ij .

The change of meaning here assumed for Indo-European may be parallelled perhaps in Greek. No instances of (iq�fjioi and thenbsp;kindred nouns to indicate intestinal noises seem quotable; butnbsp;from the third century B. C. on Pq�/uoi; or (Sqamp;juog (the MSS arenbsp;said to vary frequently in the spelling) and the compound (^qo/ntt)-d�]g ((^Qwn�drjg) are well attested in the sense �stink,� �stinking.�nbsp;Phrynichus (133, Rutherford) seems to have been unable to findnbsp;warrant for this use in Attic. It seems likely that it started asnbsp;a euphemism for noqbrj and then gained a wider meaning.

ZUR DUENOSINSCHRIFT

I. Ueber die Geschichte der �Duenosforschung� hat neuerdings EmilGoldmannin einem besonderen Buche ^ so eingehendnbsp;berichte!, daB ich mich wohl jedes R�ckgreifens auf Vergangenesnbsp;enthalten darf, wenn ich versuche, im folgenden noch einigesnbsp;zur Erklarung der merkw�rdigen Inschrift nachzutragen, um dienbsp;es sich handelt. Denn so vieles auch im einzelnen dar�ber bereitsnbsp;richtig gesagt sein mag, eine einleuchtende Gesamtl�sung dernbsp;Aufgabe steht meiner Ueberzeugung nach doch noch aus. Zeugnisnbsp;daf�r ist mir, neben den Unertraglichkeiten in Stil und Gedanken-form, mit denen man uns auch hier �fter nicht verschont hat,nbsp;vor allem die Unsprechbarkeit samtlicher bisher vor-geschlagener Textformen, d. h. die Unm�glichkeit, eben diesenbsp;Textformen in dem ihnen beigelegten Sinne ohne Stimm-h e m m u n g frei so vorzutragen, daB dabei herauskommt, wasnbsp;sonst f�r jeden Satz verstandlicher menschlicher Rede notwen-diges Ingrediens ist. Ich meine damit: eine faBbare Personal-k u r V e , eine faBbare Stimmqualitat, eine faBbarenbsp;Satzmelodie, und dazu bei Verstexten auch noch einenbsp;faBbare Art der Taktf�llung^. DaB das so ist, erklart sichnbsp;leicht aus der Art, wie man beim Entziffern und Deuten vonnbsp;Inschriften �berhaupt vorzugehen pflegt. Denn ganz nat�rlichnbsp;halt man bei solcher Arbeit zunachst nach Buchstabengruppennbsp;Umschau, die einem einen Anhalt f�r die weitere Deutung zunbsp;geben scheinen. Von da aus baut man dann weiter, Steinchen an

K. Goldmann, Die Duenos-Inschrift, Heidelberg, Carl Winter, 1926; dazu kommen dann noch die weiteren Ausf�hrungen von R. G. Kent, Language 2nbsp;(1926), 207 ff,

2 Ueber alle diese Dinge habe ich kurz gehandelt in meinen �Zielen und Wegen der Schallanalyse� (= ZuW.), Heidelberg, Carl Winter, 1924 (Sonderdrnck ausnbsp;der Festschrift f�r W. Streitberg; Stand und Aufgaben der Sprachmssenschaft).

Steinchen reihend, und oft auch ohne viel zu fragen, ob der so entstehende, und manchmal sehr locker geschichtete, Haufe vonnbsp;Kleinst�cken in seiner Gesamtheit auch etwas ergibt, wasnbsp;man f�r ein m�gliches Sprach-, Sinnes- und Stilgebilde haltennbsp;kann. Da liegt eben eine L�cke in dem herk�mmlichen Verfahren,nbsp;die einmal ausgef�llt werden muB, damit eine neue, erganzendenbsp;Kontrolle einsetzen kann. Ich halte es demnach auch nur f�rnbsp;angemessen, beispielsweise auch bei der Behandlung der Duenos-inschrift das Arbeitsverfahren einmal direkt umzukehren, d. h.nbsp;an er ster Stelle nach der Gesamtform des Textes zu fragen,nbsp;und erst nachher zu untersuchen, wie sich diese Gesamtformnbsp;zwanglos in einzelne Glieder aufl�sen laBt. Die dabei zu be-waltigende Arbeit ist, nach einiger Vor�bung, gar nicht so schwer,nbsp;wie manche zu denken scheinen: man muB nur durch Prebieren lemen, seinen Text in lebendigem Ton so vorzutragen,nbsp;daB er ebenso �k 1 i n g t� wie andere menschliche Rede, die mannbsp;kennt und versteht. Wende ich dies Probeverfahren (zunachstnbsp;also noch ohne bewuBte R�cksicht auf den ja erst noch im ein-zelnen festzulegenden Sinn) auf die Duenosinschrift an, so ergibtnbsp;sich mir das Folgende als ein f�r mich klanglich anstoBfreiernbsp;Sprechtext:

B) nbsp;nbsp;nbsp;Nei t�'d endo co'smis / vi�rco si��'d,nbsp;as t�'d nois io'p�t / o'ite'si�a'i

C) nbsp;nbsp;nbsp;D u e'n o s m�d f�'ced, / e'n manom �e'inomnbsp;due'noi; n� m�'d / ma'los tat��d!

Darin mar kieren die Zeichen ', ', quot;, quot; einerseits die Ton-stellen (bzw. Hebungen, Nr. 3), andrerseits geben sie die verschiedenen Intonationsformen an, die f�r den hem-mungslosen Vort rag erforderlich sind, und zwar bezeichnen 'nbsp;und ' e i n f a c h e (d. h. durch die g a n z e Silbe hindurch-laufende) S t e i g - bzw. Fallt�ne, quot; und quot; aber �g e b r o -chene Steigfall- und Fallsteigt�ne, bei denennbsp;innerhalb der Silbe ein Wechsel der Tonrichtung statt-findet Ferner wolle man beachten, daB die Her�bernahme der

2 Weiteres dazu siehe z. B. Indogerm. Forschungen 43,158 ff. (Nr. 99). 45, 124 ff. (Nr. 10). Auch mag es nicht unangezeigt sein, einmal auf die vor Jahren an einemnbsp;freilich etwas versteckten Orte (KongreB f�r Aesthetik und allgemeine Kunst-wissenschaft . . . igiSi Bericht herausgegeben vom OrtsausschuB, Stuttgart,

Schreibungen virco, iopU und tat�d aus dem Original nur ortho-graphisch gerneint ist, daB man aber beim Vortrag virgo, iob�t (drei-silbig!) und dat�d sprecben muB. Die gesperrt gesetzten W�rter j�', qo'i und due'nos, due'noi verlangen ferner einen gewissen (imnbsp;Grunde antitbetiscben) Nacbdruck, aucb m�ssen sie so gesprocbennbsp;werden, daB man eine gegenseitige Beziebung zwiscben ibnennbsp;berausf�blt. Bei dem Zeicben � endlicb tritt stets der sog. Stimm-sprung ein, �ber den icb anderwarts ausf�brlicb gebandeltnbsp;babe Nacb dem Scblusse des mit � ausgezeicbneten Wortesnbsp;(oder des durcb - abgetrennten Wortst�ckes, das mit � eingeleitetnbsp;ist) kebrt dann die Stimme jedesmal zu der laufenden Grund-artikulationsweise zur�ck.

F. Enke. 1914, S. 461 f.) vorgetragene Beobachtung hinzuweisen, nach der das, was das gemeitie (d. h. nicht besonders akustiach-analytisch eingestellte) mensch-liche Ohr als 'tiefer� oder 'h�her' bzw. als �s t e i g e n d� und �f a 11 e n d� ein-scbatzt, nicht lediglich auf einer Verschiedenheit der Schwingungs-zahlen beruht, sondern auf einer Komplexwirkung von Verschiebung dernbsp;Schwingungszahl, Spannung und Klangfarbe, bei der dienbsp;Schwingungszahl durchaus nicht immer der ausschlaggebende Faktor ist. Gra-phische Aufnahmen, die einseitig nur die Schwingungszahlen aus dem Komplexnbsp;herauszuziehen vermogen, k�nnen daher auch nichts gegen die vom Menschennbsp;empfundene melodische Abstufung der Rede beweisen. Wohl aber habennbsp;wir f�r diese ein sehr bequemes Kontrollmittel in den durch lockeres Auflegennbsp;einer Fingerspitze auf die Vorderzunge zu verfolgenden Zungenbewe-g u n g e n beim Sprechen. Bei jedem Steigton speziell schnellt die Zungen-spitze innerhalb der betreffenden Silbe energisch nach dem Vordergaumennbsp;hinauf, beim F a 111 o n schiebt sie sich (da sie sich nicht wesentlich senkennbsp;kann) unter dem kontrollierenden Finger in etwas absteigendem Bogen nachnbsp;vorn zu hinweg, so den hinteren Resonanzraum der Mundh�hle vergr�Cerndnbsp;und dadurch die Klangfarbe verdunkelnd. Dieser einfache Versuch wird beinbsp;gen�gender Sorgfalt der Ausf �hrung niemand im Zweifel lassen; �Steigt�ne�nbsp;in dem von mir gemeinten Sinne sind eben zugleich 'Tone mit Steig-z u n g e�, 'F a 111 � n e� aber �Tone mit Zungenvorschiebung�nbsp;(das letztere war a. a. O. noch nicht richtig erkannt; ich glaubte damals nochnbsp;Zur�ckziehung der Zunge mit Herabdr�ckung der Hinterzunge beobachtet zunbsp;haben).

� S. meine Schrift �Zur englischen Lautgeschichte� (= Abhandlungen der Sachs. Akademie der Wissenschaften, phil.-hist. KI. Bd. 40, Nr. I) S. ii ff. 23 ff.nbsp;Weiteres dazu wird eine Abhandlung �A^eue Beitrage zur Lehre von der Kasus-intonation� bringen, die demnachst am gleichen Ort erscheinen wird, desgleichennbsp;ein weiterer Aufsatz Elnonensia, dessen Erscheinungsort noch nicht angegebennbsp;werden kann. � Der Stimmsprung bedeutet einen pl�tzlichen Wechsel zwischennbsp;den beiden Artikulationsarten der menschlichen Stimme, die ich (s. ZuW. S. 75 ff.)nbsp;als N und U (d. h. �ATormalstimme� und �Umlegstimme�) bezeichnet habe. Einenbsp;von diesen Stimmarten ist jeweilen die 'Grundstimme� des einzelnen Textes,nbsp;an deren Stelle dann durch den �Sprung� die andere Art als �Gegenstimme� tritt.nbsp;1st die Grundstimme N, so geht der �Sprung� zu U, ist die Grundstimme XJ,nbsp;so �springt� man zu N (Schemata N/U und U/N). Den ersteren Fall haben wirnbsp;bei der Duenosinschrift in dem Stiick B, den zweiten in den St�cken A und C.

3- Wenn man sich den Text der Inschrift nach diesen Vor-schriften frischweg vorspricht und dazu bei jedem der gesetzten Tonzeichen versuchsweise einmal Takt schlagt, so wird mannbsp;wohl alsbald bemerken, dad das Ganze in V e r s e n abgefaBtnbsp;ist. Bei naherer Priifung mit Hilfe der von mir eingef�hrtennbsp;sog. Taktf�llkurven (ZuW. 77 ff.) entpuppen sich diese Versenbsp;als regelrechte Saturnier, die wieder nach Ban und Abkunftnbsp;mit dem identisch sind, was ich auf germanischem Boden alsnbsp;�Sagverse� bezeichnet habe. Das Wesentliche der ganzennbsp;Versart ist die Anwendung des geraden (sei es steigenden, seinbsp;2.2

blatt aus (s. dazu vorlaufig Sievers, Zeitschr. f. Assyriologie Bd. 38 [NF. 4], S. 7).

4. Die drei Absatze A, B, C stammen nicht von einem und dem-selben Verfasser. Das folgt schon aus der Verschiedenheit ihrer Personalkurven (ZuW. 74). Die Personalkurven von Anbsp;und B geh�ren zwar beide dem GroBtypus I an, sind aber dochnbsp;unter sich verschieden. Die von A hat annahernd die Gestalt dernbsp;Figur 4 auf der Kurventafel S. 174 meiner Deutschen Sagvers-dichtungen des IX.�XI. Jahrhunderts (Heidelberg, Carl Winter,nbsp;1924), in B aber ist die Figur um 180 � gedreht, so daB dernbsp;keulenf�rmige Kopf nach oben, die Spitze aber nach unten steht;nbsp;die Kurve von C aber ist eine Unterform des Typus III, aus dernbsp;Figur 16 der Kurventafel S. 73 der ZuW. dadurch herzustellen,nbsp;daB man die beiden Spitzen des Bogens sich noch kreuzendnbsp;�berschneiden laBt. �� Dazu treten parallelgehende Verschieden-heiten der Stimmqualitat (ZuW. 90ff., ausf�hiiicher innbsp;meinen �Metrischen Studi�n� 4 [= Abhandlungen der Sachs. Aka-demie der Wiss., phil.-hist. Klasse Bd. 35, Leipzig 1918] S. 31 ff.

� Jjiese Deiden Bruchformeln� zeigen an, daB den Einstellungen auf die Hauptstimmarten 5 w und 4 kq Elemente der Einstellung f�r die Stimmart 2nbsp;eingemischt sind, Hier�ber werde ich erst spater im Zusammenhang handelnnbsp;konnen,

52 sichtlich eine gewollte Einheit, denn die Kleinheit der Wider-stande beim Uebergang von A zu B und von B zu C zeigt (nachnbsp;Metr. Studi�n 4, � 184. 193 usw.), dafi B als Fortsetzung an A,nbsp;und C als Fortsetzung an B �angearbeitet� ist, d. h. also, daB dienbsp;in der Inschrift vereinigten drei Spr�che einem gemeinsamennbsp;Zweck dienen sollten und f�r diesen von den drei Feuten ge-schaffen wurden, die an der zweckdienlichen Herstellung desnbsp;kleinen Kunstwerks (Abbildungen bei Goldmann Tafel I und II)nbsp;ein gleiches Interesse batten. Diese drei Manner werden wir unsnbsp;doch aber wohl in der Werkstatt vereinigt zu denken haben,nbsp;in der das aus drei Einzelt�pfchen zusammengesetzte �Drillings-gefaB� gearbeitet und mit den drei Spr�chen versehen wurde,nbsp;die es uns so wichtig machen.

5. Was mogen denn aber die hier gemutmaBten drei Leute mit ihrem Produkt und ihrer Zusammenarbeit an ihm gewollt haben ?nbsp;Ich glaube, das ist leicht zu sehen: denn das ist doch wohl bereitsnbsp;als sicher anzusehen, daB das DrillingsgefaB der Ausf�hrungnbsp;einer s�hnenden Liebeswirkung dienen sollte,nbsp;bei der im Hintergrunde ein G o 11 mitzuarbeiten hat: der Gottnbsp;eben, der ja allein imstande war, dem durch Menschenhandnbsp;gestalteten GefaB die gew�nschte Einwirkung auf den Empfangernbsp;zu sichern. DaB es ohne einen solchen lenkenden Gott nichtnbsp;immer glatt gehen w�rde, dessen sind die braven T�pfer sicher:nbsp;darum soil er eben auch dazu m i t h e 1 f e n , daB die ganzenbsp;Sache in guter Bahn verlaufe. Und sie sind auch darinnbsp;ehrliche und wohlgesinnte Leute, daB sie w�nschen, nur einnbsp;braver Liebhaber moge sich der gew�nschten Wirkung desnbsp;GefaBes zu erfreuen haben: �ein Braver* hat�, so sagt janbsp;direkt C, �dies GefaB dazu gemacht, daB es einem Bravennbsp;zu Handen komme: kein B � s e r soil es (sc. zum Zweck dernbsp;Einwirkung auf die Umworbene) verschenken d�rfen.� Undnbsp;daf�r, daB das GefaB nicht in unrechte Hande falie, soil auch dernbsp;Gott mit sorgen helfen, wenn das GefaB einmal die Werkstattnbsp;verlaBt, in der es entstanden ist: �d a r a u f moge der Gott sehen,nbsp;wem er mich zuf�hrt� (d. h. �in wessen Hande er mich geratennbsp;laBt�), wie es in A heiBt. Das ist ein ganz einfacher, und auchnbsp;ein wohl verstandlicher Gedankengang.

� Die Intonation zeigt, daB in duenos und diienoi unm�glich Eigennamen gesucht werden d�rfen; denn diese beiden Wortformen entbehren des melo-dischen Auszeichnungstones, der auch imLat. die Eigennamen von den gew�hn-lichen Appellativen scheidet.

6. nbsp;nbsp;nbsp;Ich glaube demnach den Text nun etwa folgendenuaBennbsp;umschreiben zu k�nnen:

B) nbsp;nbsp;nbsp;Wenn dir ein Magdlein nicht hold ist, so heiBt er [namlich der Gott] dich,nbsp;von uns [namlich den drei T�pfchen des GesamtgefaBes] Gebrauch machen,nbsp;um Frieden f�r euch zu gewinnen.

C) nbsp;nbsp;nbsp;Ein Braver hat mich gemacht, daB ich einem Braven zu Handennbsp;komme; kein B�ser soil mich verschenken [d�rfen] I

A) Eo visat deus, c u i me mittat. � B) Ni in te comis virgo siet, ast t� [sc. deus ille] nobis jubet uti pacari [= ad pacandum] vobis. � C) B o n u s me fecitnbsp;in manum ire [= ut in manum eat] b o n o: ne me malus dato!

7. nbsp;nbsp;nbsp;Das klingt ja nun wohl im ganzen ziemlich glatt und ein-leuchtend, aber man wird sich doch fragen m�ssen, ob es auchnbsp;m�glich ist, diesen Sinn sprachlich aus dem heraus zu gewinnen,nbsp;was dasteht. Ich muB also dar�ber noch einiges anf�gen.

8. nbsp;nbsp;nbsp;Die beiden Eingangsworte k�nnen, wie in Nr. 2nbsp;bereits angegeben wurde, stimmfrei nur als �j�'-veisat gesprochennbsp;werden, d. h. mit Stimmsprung und starkem Nachdruck auf demnbsp;steigend-fallenden j�quot;. Auch muB eine psychisch-syntaktischenbsp;Bindung zwischen j� und veisat bestehen (bzw. zum Ausdrucknbsp;gebracht werden, vgl. schon Nr. 2), die das j�quot; stimmlich sonbsp;stark hervorzuheben gestattet, daB das Wort veisat ohne AnstoBnbsp;dahinter in die Senkung treten kann. Beides gewinnt man sofort,nbsp;wenn man nur das j� als eine durch sog. �Akzentumsprung� ent-standene satzphonetische Dublette zu dem gew�hnlichen eonbsp;�dorthin� auffaBt, also der Stelle den schon oben gegebenen Sinnnbsp;�do'rthin (= da'rauf) schaue der Gott� beilegt, an den sich dannnbsp;auch das folgende nach Sinn und Klang gut anschlieBt (auchnbsp;der so oft beanstandete Konjunktiv mittat kommt damit zunbsp;seinem Recht). Auf germanischem Boden ist der hier gemeintenbsp;Akzentumsprung ja z. B. aus dem Altnordischen allbekannt, undnbsp;auch im Altenglischen ist er weit verbreitet, wenn er auch nochnbsp;nicht in den Grammatiken gehucht ist (vgl. aber z. B. Sievers,nbsp;Metr. Studi�n 4, � 102). Aus dem Lateinischen selbstnbsp;sowie aus dem U m b r i s c h e n (s. v. Planta 2, 732 f.) kennenbsp;ich zwar derlei Formen nicht, wohl aber hat das O s k i s c h enbsp;gerade bei dem Pronomen, das dem lat. is entspricht, ganznbsp;analoge Formen entwickelt, wie joc nsf. apn., jonc asm., juscnbsp;npm. zu lat. ea, eum, e�s (statt ii); desgleichen heiBt es marru-zinisch jafc apf. = lat. eas. Man vergleiche z. B. aus der ebenfalls

in Saturniern laufenden Inschrift der Tabula Bantina (v. Planta 2, 494 ff.) die Verse

in saa'epis j o n c fo'rtis / meddis mo�ltaum �he'r�st 12 j o'n c suaepis he'r�st meddis / mo'ltaum quot;li'cit�d 17 (vgl. 26)nbsp;po'izad li'gud j use �ce�nstur 20

Man wird also doch wohl auch f�r das alteste Latein die Existenz solcher Dubletten f�r m�glich halten d�rfen: sie werden nurnbsp;spater in �blicher Weise durch Ausgleichung wieder beseitigtnbsp;worden sein. �

9. Die annoch bestehenden Schwierigkeiten im Absatz B l�sen sich, wie mir scheint, in befriedigender Weise, wenn man aus dernbsp;Buchstabenfolge astednoisiopetoitesiai zunachst das St�ck iopetnbsp;als (dreisilbig zu sprechendes) altlat. io' h�t ^ = lat. jubet heraus-l�st, als dessen Subjekt dann nat�rlich der deivos der ersten Zeilenbsp;fungieren muB. Das jubei fordert dann weiter einen abhangigennbsp;Infinitiv hinter sich, und der steht da, wenn man das folgendenbsp;oitesiai unzertrennt laBt. Denn dies oitesiai enthalt (daran wirdnbsp;man wohl nicht zu zweifeln brauchen) gewiB nur eine Parallelenbsp;zu dem spateren Typus von amdrier, vid�rier, opper�rier, �rier,nbsp;ferrier (Sommer, Handbuch ^ 594 f.; Stolz-Schmalz, Lat. Gramm.�nbsp;328), nur noch ohne das nachtragliche passivisch-deponentialenbsp;Anhangsel -r, und insofern allerdings etwas auffallig, als esnbsp;meines Wissens der bisher erste Beleg f�r den ganz vollen Typusnbsp;*�tener (statt des nachher gebrauchlich gewordenen �tier) beinbsp;einem Verbum der 3. Konjugation ist (vgl. indessen doch dasnbsp;zitierte ferrier). Von dem Infinitiv oitesiai hangt sodann weiternbsp;das vorhergehende (auch an richtiger Satzstelle stehende) noisnbsp;�nobis� ab, dessen Pluralitat sich, wie man weiB, aus der Dreizahlnbsp;der vereinigten T�pfchen erklart. Etwas schwieriger bleibt syn-taktisch das folgende pdkdri vois, mag man nun pdkdri aktivischnbsp;oder passivisch fassen. Es muB aber doch wohl auf alle Falienbsp;ein Ziel infinitiv sein, im Sinne von etwa �ad pacandum�, wienbsp;schon oben angedeutet wurde (s. auch unten Nr. 22 �ber einom).nbsp;Der Dativ vois �f�r euch� d�rfte syntaktisch wieder leichter zunbsp;verstehen sein. Auch die k�rzere Form pdkdri gegen�ber dem

emphatischen, zweihebigen o'ite'sia'i mit Volldiphthong am SchluB (vgl. dazu auch unten Nr. i6 �ber umbr. heriei: heri)nbsp;begreift sich wohl aus den besonderen Betonungsverhaltnissennbsp;der Stelle. Zur Endung -esiai vergleiche man �berdies noch dasnbsp;altindische -cidhydi (wie in iyddhyai usw., Delbr�ck, AUindischesnbsp;Verbum � 211) und ahnliches.

10. nbsp;nbsp;nbsp;Nun aber �der� viel beanstandete �Rhotazismus� innbsp;pdkdri gegen oitesiai mit unverandertem s? Ich glaube, um dennbsp;braucht man sich nur so lange Sorge zu machen, als man es f�rnbsp;gestattet halten wird, sich auch beim Latein nicht um die Ein-wirkung der verschiedenen Intonationsarten auf die Ausgestaltungnbsp;der Wortk�rper zu k�mmern. Denn alles hier bisher Zweifelhaftenbsp;wird alsbald deutlich, wenn man nur das Verhaltnis der in Fragenbsp;kommenden s und r zu den Steig- und Fallt�nen der einzelnennbsp;Belegstellen ins Auge faBt. So k�nnen wir gleich als Ausgangs-punkt f�r weitere Betrachtung feststellen, daB unser o�e'siainbsp;auf der Silbe e's einen Steigton tragt, die Form pdkd'ri aber aufnbsp;dem d'r einen Fallton. Ebenso unbestreitbar ist, daB die sekun-daren r aus s fr�herer wie spaterer Zeit (wenigstens soweit ichnbsp;das habe verfolgen k�nnen) stets auf einen Fallton folgen, und um-gekehrt die unveranderten s der �vorrhotazistischen Zeit� ebensonbsp;wie die Dauer-s der stekenden �Ausnahmsw�rter� (wie cdsa, miser,nbsp;caesari�s, ndsus u. dgl., Sommer 191; Stolz-Schmalz � 141) aufnbsp;einen Steigton

11. nbsp;nbsp;nbsp;DaB dem Wechsel von s\r ein alterer Wechsel von s\ znbsp;zugrunde liegt, ist allgemein anerkannt, und dieser Wechsel trittnbsp;ja auch, und zwar in gleicher Verteilung nachnbsp;Steig- und Fallton, im Oskischen direkt zutage,nbsp;sobald man nur dessen metrische Texte richtig als Verse spricht.nbsp;So stehen z. B. in der Inschrift der Tabula Bantina (v.nbsp;Planta 2, 494 ff., Nr. 17), die ja s und z auch graphisch scheidet,nbsp;nebeneinander �steigtonige s� in

PiVn p4 e i s u m bra'teis / anti ca'de�s �a'mn�d 17,6 e' i s � c-en quot;zi'cul�d 16nbsp;censa'mur �'s u f in �e'ituam 19nbsp;in eizeic vi'ncter / �suf co'menei la�ma'tir 21 (s. u.)

* Znr bequemen Veranschaulichung des Verhaltnisses kann auch. der bekamite Ausspruch des Pomponius dienen: �R litteram invenit, ut pro Vale'siis Vale'riinbsp;essent, et pro F� sits F� rii, der (wie ahnliche andere Angaben) nat�rlich nichtsnbsp;anderes zu besagen braucht, als daB man damals angefangen habe, der falltonigennbsp;Dublettform den Eingang in die otfizielle Schreibung zu gestatten.

i'z�c eizeic si'celei / co'mono �ni-�h�'pid 7 si'om dat e'iza'sc / idle ta'ngineis quot;de'icnm 9nbsp;ta'dait e'z u m , nep �fe'facid 10nbsp;po'd pis dat e'i zac �e�gmad lonbsp;i'z�c co'mono �ni-�h�'pid 14nbsp;po'n censtur Ba'nsa� / ta'utam cen�sa'zet 19nbsp;po'izad l�'gud jusc �ce'nstur 19nbsp;ce'nsau'm �ange t u'z e t 20

i'n eizeic vi'ncter / �suf co'menei la^ma'dr 21 (s, o.) in e'i(tuo) s�'vom / pa'ei ei zeis fu'st 22nbsp;sua'e pis op e' i z o i s com �a'tr�d 23nbsp;manim a'serum e'i z a z u n c �e'g m a z u m 24,

12. nbsp;nbsp;nbsp;Die Inschrift des Cippus Abellanus (v. Plantanbsp;2, 513 ff., Nr. 177), die �berhaupt viel mehr mit Steigt�nennbsp;arbeitet als die der Tabula Bantina, hat allerdings, dank dernbsp;abweichenden pers�nlichen Sprechweise ihres Verfassers� nurnbsp;unverandertes s:

de'ketasiui Nu'vlanui 127, 5 inim te'erum pud up e'i s � d / sa'karakl�d �i'st 13nbsp;pru'f t u s e t re'ht�d �a'mn�d 16nbsp;mu'inikum mu'inikei / t�'re� �f �'s i d 19

inim e'i s e i s sakara'kleis inim �t�'reis 19 f. tr�'barak�ka't t u s e t 39

Ich betone ausdr�cklich, daB man hier an keiner Stelle ohne Stimmst�rung ein stimmhaftes z statt des geschriebenen s sprechennbsp;kann: wir haben also sicher auch nur wieder den Niederschlagnbsp;eines rein lautgesetzlichen Zustandes vor uns: der etwaige Ein-wand, man habe gesprochenes z im oskischen Originalalphabetnbsp;graphisch nicht ausdr�cken k�nnen, hat also hier keine beweisendenbsp;Kraft, da �berhaupt keine Belege f�r klanglich notwendiges 2nbsp;vorliegen.

13. nbsp;nbsp;nbsp;Ganz analog stehen die Dinge auch im Altumbri-s c h e n f�r den Inlaut. Das kann man speziell wieder an dennbsp;ebenfalls metrischen Texten des altumbrischen Teiles der I g u -vinischenTafeln(v. Planta 2, 557 ff.) ersehen. Die einzigenbsp;Abweichung vom oskischen Branch besteht darin, daB das er-

� Ueber solche pers�nliche, nicht dialektische Differenzen s. Sievers, Zur engUschen Lauigeschichte S. i8. 86 FuSnote. � Ich scheide hier and im folgenden,nbsp;um das Druckbild nicht zu �berladen, bei den Texten in altoskischer Schriftnbsp;nicht zwischen i, u und i, �.

weichte z auch schon im Alt umbrischen zu r weiterentwickelt ist. Man vergleiche etwa

ante'ntu 2'

sv�'pu �'s u m - e k ��'s u n u i ^ ik_ g. 14, 38. 2�. 20. 21. 3, I. 14nbsp;str�'ii9la pe'tenata / i's e k �arve'it� 4, 4nbsp;s V i's e V � fe'rt� �p�'ne,nbsp;e'tr� s V i's e V � / vi'nu �fe'rt�,nbsp;te'rti� s v i's e v � / �'tur �fe'rt� 2^, 14 ff.nbsp;i's u n t kre'matru �pru'sektu 2**, 28; ahnl.nbsp;se'sten t a's i ar� / u'rn a s i�a'r� 3, 2nbsp;pl�'n a s i �'r ur'ii a's i �r 5nbsp;nbsp;nbsp;nbsp;2

a'rfert�Te e'r u pe�pu'r kurent 5�, 5; ahnl. 5 nbsp;nbsp;nbsp;26. 29

erek �'sun�s-ku �ve'purus 5�, ii; ahnl. ere 5^,4 �'r a-hunt quot;ve'a i'

e're r� �'sune k��ra'ja^i 5�i 41 ahnl. 5b 24. 26. 29 �'nnk hapinar u / e�rn's tit�, quot;ze'r�f 1nbsp;nbsp;nbsp;nbsp;33

se'stentasi a r u / u'rnasi�a'ru 3,2 (vgl. oben a; ahnl. 2* e't ape fra'ter / ^e'rsna't�r quot;f�'rent 5^,22nbsp;p�ne pu'rtinfus ka'��tu, / p�'fe apr�f �fa'k u r e n t i'�, 33nbsp;p�'re ulu be'n u r e n t / �pru'si kurent re'hte / . . . 5nbsp;nbsp;nbsp;nbsp;25;

14. Im A u s 1 a u t hat sich bekanntlich, auch schon im A 11 -umbrischen, eine dreifach verschiedene Behandlung des urspr�nglichen -s herausgebildet. Auch diese richtet sich nachnbsp;der Klangbesonderheit der einzelnen Belegstellen, und diese Be-sonderheiten hangen ihrerseits wieder von den deklamatorischennbsp;Gewohnheiten oder Neigungen der einzelnen Verfasserpers�nlich-keiten ab. Das -s bleibt namlich entweder a) erhalten, oder esnbsp;ist b) geschwunden, oder c) durch r ersetzt. In dem Wechsel dernbsp;verschiedenen Wortausgange herrscht aber, nach der Ueberliefe-^bing gemessen, keineswegs das sinnlose Durcheinander, das mannbsp;allgemein anzunehmen beliebt, sondern: a) das erhaltene -s

Das e dieses Wortes ist, wie die Klangverhaltnisse zeigen, sicher lang: �su gehort also (gegen v. Hanta) doch zu osk. eis- usw. Die n e u umbrischen Formennbsp;essu und Fisstu in dem einen Satze

ti'om e s s u b�'e / pera'cr� pi�ha'cl� e'tr� o'criper F i's s i � 6*, 43

haben sichtlich sekundare Gemination im Verband mit Vokalverk�rzung, denn auch der Name F�s- nebst seinen Ableitungen verlangt sonst stets die Aussprachenbsp;mit -Is- (Belege bei v. Planta 2, 736).

58 ist an vorhergehenden einfachen(d.h. die ganze Silbe durch-laufenden, oben Nr. 2) Steigton(') gebunden; � b) dernbsp;S c h w u n d ist nur nach e i n f a c h e m (d. h. wieder die ganzenbsp;Silbe durchlaufendem) tiefemFallton (') eingetreten; �nbsp;c) das r nur bei hakenf�rmigem, tief lierabgehendemnbsp;Steigfallton (^) innerhalb derSilbe, die es beschlieBt

15. nbsp;nbsp;nbsp;DaB es sich bei dieser Regelung in der Tat mindestens zumnbsp;guten Teil nur urn Stilistisch-Deklamatorisches handelt, kannnbsp;man besonders deutlich daraus ersehen, daB auch stehendenbsp;Formeln je nach dem klanglichen Zusammenhang, in demnbsp;sie erscheinen, intonatorisch wie lautlich verschieden behandeltnbsp;werden k�nnen. So geht z. B. die oft wiederkehrende Formelnbsp;quot;arefes �a'fv�s (am VersschluB) stets auf steigtoniges awes' odernbsp;arvls' aus (letzteres bei besonderer Hochlage der letzten Silbe),nbsp;aber das erste Wort heiBt entweder arep�s' (oder arip�s') mitnbsp;schlieBendem Steigton, oder verkiirzt mit tief herabgehendemnbsp;Fallton am SchluB arep^, oder endlich areper^ mit -r undnbsp;steigend-fallendem Hakenton. Jede der drei Formen paBt klang-lich an die Stelle, wo sie geschrieben steht: keine aber kannnbsp;ohne Stimmst�rung an eine Stelle geschoben werden, wo wir einenbsp;der anderen Formen lesen. Man vergleicht z. B. die Verse

a) nbsp;nbsp;nbsp;s�'vum kut�f quot;pc'sni�m� / a'rep�s' �a'rv�s' i 6.nbsp;fe'it� kut�f quot;pe'rsn� m� / a'rip�i' �a'rv�s' 7.

l ta'jez �pe'sn�'m� / a'�ep�' �a'rv�s' 26 � p�'ni f�tu. �pe'�aja �f�'t�;

16. nbsp;nbsp;nbsp;Weitere Beispiele aus dem Gebiet der Deklination anzu-f�hren er�brigt sich wohl, da die Regel ausnahmslos zu verlaufennbsp;scheint und man sich die Belege m�helos aus den ausf�hrlichennbsp;Listen von v. Planta zusammenstellen kann. Aber aus dem annbsp;sich begreiflicherweise viel sparlicher vertretenen Verbal-gebiet sei wenigstens die dem lat. sive�-sive entsprechende Formelnbsp;(eigentlich = lat. vel�s�veils, oder allenfalls auch indikativischnbsp;vis�vis) hergesetzt, bei der �brigens neben der Intonation im

DaB aber die k�rzeren Formen mit bloBem l in der Endung mindestens konjunktivisch sein k�nnen, zeigt die deutliclie 3. Sing. Conj. sv� pis hert... 4, 26.

engeren Sinne auch noch Stimmlage, Tempo und Dynamik maB-gebend eingewirkt haben. Sie erscheint in folgenden Gestalten:

fe'itu. kut�f �pe'rsn�m� / a'�ip�s �a'rvis iquot;, va'tuva fe'rine �fe'it�nbsp;her�s' v�'nu hcri' p�'ni, / vi'kri-per quot;F�'si� .nbsp;va'tuva fe'rine �f�'t�,

heri' vi'nu herj' p�'ni, / a'rviu �uste'ntfl. i herj' v�'nu heri' p�'ni �f�t�.nbsp;va'tu� fe'rine �f�'t�, 6*, 57nbsp;a'rvio �f�'t�.

heri' v�'nu heri' p�'ni �f�'t� �'�, 46 fo'ndlir-� a'brof / tri'f �f�'t�.nbsp;he'rici' r�'f�, / he'riei' quot;p�'j� . . . 7�, 3nbsp;va'tu� fer'ine quot;fe'tu,nbsp;hcrie' vi'iiu herie' p�'ni �f�'t�

17. nbsp;nbsp;nbsp;Ebenso richtet sich im Umbrischen nach Steig- undnbsp;Fallton das Stehen oder Fehlen des auslautenden m {%), sowie desnbsp;-/ aus -ns, wie z. B. in dem Vers

Fu'ntl�r-�' tri/ a'pr�/' / r�'fr/�' �te �p�'jM' ... nbsp;nbsp;nbsp;24.

Im letzteren Falie kann es nur fraglich sein, ob direkt ein / ge-schwunden ist, oder ein s nach vorherigem Schwund des Nasals (unter Nasalierung des vorhergehenden Vokals). Die Entschei-dung ist nat�rlich auch f�r die chronologische Frage wichtig,nbsp;ich kann aber an dieser Stelle nicht naher darauf eingehen. Ichnbsp;begn�ge mich also damit, zu sagen, daB mir nach den begleitendennbsp;Umstanden die zweite Alternative wahrscheinlicher vorkommtnbsp;als die erste.

18. nbsp;nbsp;nbsp;Auch das Lateinische zeigt nun wieder ganz ahnlichenbsp;Erscheinungen. DaB auch dort das -m nur nach Fallton abfallt,nbsp;kann ein Bliek z. B. auf die Scipioneninschriften lehren, f�r dennbsp;Abfall des -s lassen sich auch genug inschriftliche Zeugnisse bei-bringen, ganz abgesehen von dem alten Paradebeispiel facilenbsp;omnihu' frinceps u. dgl. Ja selbst aus der spateren Literatur laBtnbsp;sich die einstige G�ltigkeit der Regel noch demonstrieren an dernbsp;Hand der Verschmelzungsformen mit es, est,

Ich habe daraufhin z. B. den ersten Akt des Plautinischen Amphitruo durch-gesehen. Dort heiBt es aber bei erhaltenem m ebenso konsequent im Neutrum mit steig tonigem Ausgang oppidum�st igi, exitum�st 219, actum�st 227, pugnd-tum�st 249, certum�st 2(gt;^. 339. 372, pessumum'st 314, conlubitum�st 343, futurum�stnbsp;374) datum�st 418. 538, quaerundum�st 423, factum�st 431, redeundum�st 527 (v'gl.

6o auch verb�rum�si 247) wie beim Maskulinumfall tonig -u st: ndtu'st 179,nbsp;additu�st 250, it�ru�st 263, n�Uu�st 293, supersHti�su�st 323, t�tdtu'st 352, prae-fectu'st 363, eru�st 381, ingressii'st 429, s�mindtu�st 482, soitu'st 506; desgleichennbsp;exercit�ru�s 324, cmentlhi�s 411, vaniloquo�s 379, Dagegen ist die Endung -usnbsp;ebenso konstant s t e i g tonig, wenn keine k�rzende Verschmelzung eintritt:nbsp;conspicdtus est 242, accept�rus est 296, validus est 299, onerandus est 328, enis estnbsp;362, solitus est 419, �loc�tus est 420, saevos est 541. Auch die -a der F e mi n i n-formen bleiben trotz der Verschmelzung (wie das ~^m� der Neutra) steigtonig:nbsp;vgl. d�ra�st 166, aurea�st 260, exorta�st 274, scita�si 288, nupta�st 364, sol�ia'st 412,nbsp;n�Ua�st 509, data�st 534. Dazu halte man das mannliche S�sia�s 427 und beliebigenbsp;andere Verschmelzungsformen mit Steigton, wie quid-, quisquam�st 271. 400,nbsp;Sosiae�st 332; superque�st 168, [con]simile'st 443. 447, quale�st 537!., necesse�stnbsp;501; tib�'st 350. 364. 402, ml'st 406, simil�st 446', ade��st i6g, imago�st 265, e��stnbsp;345, sign�'st 421, fact��st 505 (2), curdti��st 519; diu�st 302, 530. Steigtonig bleibennbsp;bei der Verschmelzung auch die us, die iricht dem Nom. Sing, von o-Stammennbsp;angeh�ren: onus�t 175, opus�t 445; satius�t 176, domus�t 362.

Da also tatsachlich nur bei den eben genannten Nom. M. Sing, bei der Verschmelzung ein Fallton eintritt, wahrend bei Nicht-verschmelzung auch deren ~us Steigton behalt, so bleibt wirklichnbsp;nichts anderes �brig, als von einer alten, d. h. vor der Verschmelzung liegenden Doppelheit steigend -us': fallend -u' auszugehen,nbsp;die diesmal zugunsten des falltonigen -u' ausgeglichen wurde,nbsp;wahrend sonst �berall die steigtonige Variante den Sieg davon-trug. Wie das steigtonige onus�t nebst Genossen zeigt, hat alsonbsp;Sommer, Krit. Erlautenmgen S. 92 ff. darin gegen Leo recht, daBnbsp;auch eine K�rzung von -us est zu -us�t �ber -us'st an sich ni � g -1 i c h ist: aber der mannliche Typus -u�st ist doch eben tatsachlich offenbar aus -u' est entstanden. � Uebrigens geitennbsp;die hier vorgef�hrten Intonationsregeln auch noch f�r die klas-sische Latinitat.

19. Es stimmt ferner wieder zur Hauptregel, daB in Fallen mit mehrsilbiger Senkung wie quibus me deceat Amph. 201, facinusnbsp;nequiter , lassus sum, vectus h�c sum 329, servos ne an liber 343,nbsp;servus sum 356, inmutdtus sum 456; oder vlribus superbi 212,nbsp;f�nibus exercit�s 215, vicimus vi 237, complexus cum 290, wonbsp;Satz- und Versmelodie steigtonigen Ausgang des ersten Wortesnbsp;verlangen, aus klanglichen Gr�nden auch dessen -s mitzusprechennbsp;ist, wahrend das s aus gleichem Grunde unterdr�ckt werden muB,nbsp;WO an entsprechender Stelle ein Fallton auftritt; opindtu' fui 186,nbsp;instr�ximu' legi�n�s 221, familidri' sis 354, sdnu' n�n est 402.nbsp;Es ist also wirklich nicht abzusehen, warum nicht auch z. B.nbsp;mentlris' nunc 344 und vocdre' 382 u. dgl. einfach lautgesetzlichenbsp;Intonationsdubletten sein sollen. Auch das falltonige pote's(t)nbsp;w�rde zur Regel stimmen: selbst das standige qui's: qui' k�nnte

schlieBlich hierher geh�ren, wie demi auch Formen wie ai-n Amph. 284. 344, pergi'-n 349. 539, scl'-n 356, sati'-n 509, abi'-n 518nbsp;tatsachlich. den zu erwartenden Fallton zeigen. Auf Zweifelhafteresnbsp;m�chte ich hier nicht eingehen. ~

20. nbsp;nbsp;nbsp;In der intonationsbedingten Erweichung eines innerennbsp;und dem intonationsbedingten A b f a 11 eines schlieBenden snbsp;gehen also die altitalischen Hauptsprachen (auch die Klein-sprachen liefern noch einiges einschlagende Material) in solchernbsp;Ausdehnung zusammen, daB man schwer glauben kann, es handlenbsp;sich hier nicht um bereits uritalische Prozesse, deren Wirkungennbsp;nur hernach durch Ausgleichungen mannigfacher Art wiedernbsp;un�bersichtlicher gemacht wurden. EinigermaBen abseits stehtnbsp;nur das Umbrische mit seinem bereits in altumbrischer Zeitnbsp;beginnenden auslautenden hakentonigen -f (oben Nr. 15), undnbsp;das muB denn doch wohl als eine auf sekundarem Into-nationswandel beruhende nachtragliche Abbiegung ausnbsp;dem alteren steigtonigen s' angesehen werden: der lange fallendenbsp;Ast des Hakentons hatte dann im Umbrischen ahnlich gewirkt,nbsp;wie der einfache Fallton im Vorumbrischen. F�r das Umbrischenbsp;hatten wir demgemaB neben dem �alteren� auch noch einennbsp;�j�ngeren� Rhotazismus anzusetzen.

21. nbsp;nbsp;nbsp;Damit kann ich denn endlich wieder zum Ausgangspunktnbsp;dieser langen Digression zur�ckkehren. Wenn namlich im Oski-schen je nach der Intonation inneres s und inneres z, im Umbrischen inneres s und inneres r nebeneinander stehen konnten, sonbsp;weiB ich nicht, warum ein solches Nebeneinander (das ja akzen-tisch gesch�tzt war) im Lateinischen unm�glich gewesen seinnbsp;soil. F�r meine Auffassung der Geschichte des �Rhotazismus�nbsp;vertragen sich also auch das oite'siai und das paka'ri unserernbsp;Duenosinschrift auf das beste miteinander. �

22. nbsp;nbsp;nbsp;Zu dem Abschnitt C der Inschrift sind nur noch ein paarnbsp;Worte zu sagen. Das o von manom gegen�ber gemeinlat. manumnbsp;mit altem u ist vom Intonationsstandpunkt aus gut verstandlich,nbsp;wieder als Produkt des Falltons. Das zwischen en und einoninbsp;formelgemaB akzentisch eingeklemmte Wort sinkt namlich bisnbsp;zum Schlusse des o, um dann in dem m wieder anzusteigen (dasnbsp;nun als steigtonig erhalten bleibt). In einom sehe ich dann syn-taktisch wieder einen (an den Supinumgebrauch erinnernden)nbsp;Infinitiv des Zieles (wie oben in pakdri, Nr. 9), formell einennbsp;akkusativisch gebildeten Infinitiv der Wurzel ei �gehen�, der

genau dem germanischen Bildungstypus zumal von Verbis wie d�-n, gd-n, std-n usw. entsprechen w�rde, der ja seinerseits auchnbsp;von einem Suffix -no- ausgeht. Neben den oskisch-umbrischennbsp;ebenfalls akkusativischen Infinitiven auf -tim scheint mir dienbsp;Annahme besonders unbedenklich, unser ei-no-m k�nne ein Restnbsp;einer einst weiter verbreiteten Gattung von Infinitiven sein, dernbsp;aus der Zeit vor der groBen Neuregelung des lateinischen Infinitiv-wesens stammt.

Ich m�chte also glauben und boffen, daB die im Vorstehenden verfochtene Deutung der Duenosinschrift auch der sprachlichennbsp;Formkritik gegen�ber werde standhalten k�nnen.

We know so little about the poet Lucretius that we are justified in resenting misinformation which is handed down without hesitation from volume to volume. Many years ago Marx^ madenbsp;the statement that the poet�s cognomen Cams was not foundnbsp;among Romans of the Republican period, and that while rathernbsp;frequent in the imperial period it belonged regularly to men ofnbsp;low station, that is, to slaves, freedmen and immigrants. He alsonbsp;noticed that Cams was frequently found in the Celtic provinces,nbsp;and therefore concluded that Lucretius was probably a Celt ofnbsp;freedman standing.

Now that several indices of the Corpus have appeared so that any one might have discovered that Marx� statements werenbsp;erroneous it is a pity that the standard reference books continuenbsp;to repeat and even to exaggerate these guesses. Mewaldt, whonbsp;has recently written the article on Lucretius for the great Pauly-Wissowa Realencyclopadie ,(XIH, 1659 ff., hated 1927) says:nbsp;�So wird es bei der These von Marx verbleiben mussen.�� Hosius,nbsp;in his revision of Schanz (I p. 273, dated 1927) also accepts thenbsp;statements of Marx, as did Kroll in his revision of Teuffel-Schwabenbsp;W' 473)- Finally, the article on Cams in the great Thesaurusnbsp;Linguae Latinae says ante aetaiem imp. non nisi serv. vel lib.,nbsp;cf. Marx.

Now let us examine the facts in detail. The first edition of vol. I of the Corpus (containing the republican Latin inscriptions)nbsp;gave one instance of Carus (no. 769) and that belonged to anbsp;slave. However the inscription had no right to a place in thatnbsp;volume because its date is 32 A. D. So much for the slave namenbsp;of Cams in republican times. We have no instance of it. The

64 cognomen is, to be sure, rare in republican times in any class,nbsp;but that has little significance since cognomina came into usenbsp;late except in the highest nobility and republican inscriptionsnbsp;are relatively few. The fact is that nine-tenths of the well-knownnbsp;cognomina are not vouched for at all in republican records.nbsp;Marx�s statement regarding the rarity of this cognomen amongnbsp;citizens of the republic was, therefore, hardly worth repeating.nbsp;The fact is that, leaving a.side the case of Lucretius, we havenbsp;found it in republican times only in Livy 39, 55, T. Aebutiusnbsp;Cams, a senator (where it is doubtful because the name seemsnbsp;to reappear as Aebutius Parus in 42, 4, according to the Viennanbsp;ms.), in an old inscription (vetustis litteris) of Aquila, in centralnbsp;Italy (C. I. L. 1796, Sex. Vibius Sex. lt;fgt; Cams), and in thenbsp;case of a Spanish chieftain (Appian, Hisp. 45). From this evidencenbsp;we can only say that there was a Celtic name which Appiannbsp;transliterated into Karos, that the Latin name Cams was innbsp;good standing in central Italy before there is any question ofnbsp;Celtic nomenclature there, and that perhaps a senator of thenbsp;Catonian epoch bore the cognomen.

During the empire, whence most of our inscriptions come, the cognomen is found in all strata of society. The Emperornbsp;Cams is late, but before his day there were L. Aemilius Carusnbsp;(consul about 50 A. D.), D. Junius Carus (a senator), C. Popiliusnbsp;Carus (consul about 148), Salvius Carus (a senator) and Seiusnbsp;Carus (of senatorial rank). The index of nomina of C. I. L. VI,nbsp;where the cognomen is found about fifty times, will add to thenbsp;above several centurians and praetorians of respectable standingnbsp;(2676; 3639; 32 520, VI, 14; 32 536, II, 5: 32 995; 37 184. I. 22).nbsp;Among the forty that cannot be classified only two are distinctlynbsp;proved to be freedmen (20 304, 16 823), though several of thenbsp;others appear to be related to freedmen, but are clearly one ornbsp;two removes from the class. And as it is a well known fact thatnbsp;while freedmen were required to keep their servile cognomina,nbsp;the children of freedmen were very apt to be given respectablenbsp;cognomina so that they might have a fair chance at civic andnbsp;social opportunity. The occurrence of the name Cams in thisnbsp;latter group does not support the Marxian hypothesis.

Those who are not intimate with the inscriptions might readily draw the incorrect inference from the fact that only 25% of thenbsp;Cari of the index are demonstrably of a respectable social class

But the fact is that that proportion is exceedingly high. One has only to compare the use of such servile names as Amandus,nbsp;Ampliatus, Auctus (cf. the Thes. L. L.), Hilarus, Donatus,nbsp;Optatus, Successus, or the numberless Greek servile names tonbsp;realize how estimable the name Cams is. A good place to testnbsp;the tone and quality of names is the second part of the sixthnbsp;volume of C. I. L. which contains the thousands of inscriptionsnbsp;of the columbaria where the ashes of slaves and freedmen werenbsp;interred in the early empire. In examining the first two thousandnbsp;inscriptions in that volume we find the favorite slave namesnbsp;recurring by scores while Cams occurs only once (VI 5263)nbsp;and then in the instance of a son of an imperial freedman. Innbsp;short no one who is familiar with Latin inscriptions could possiblynbsp;have accepted the Marxian statement. So far is Cams from beingnbsp;a cognomen of servile association that it is one of the mostnbsp;respectable of names outside of the distinctly senatorial group.nbsp;The fact that the majority of the people who bear it are of humblenbsp;station is, of course, due to the fact that more than 90% of ournbsp;epitaphs naturally come from that class.

The reference to Celtic origin has little more value^. There is no doubt that there was a Celtic name that resembled Caros andnbsp;that it occurs on Celtic headstones. But the first occurrence ofnbsp;the narne in Italy is too early to attribute to Celtic origins.nbsp;Furthermore, the frequent use of it as a designation of childrennbsp;in Italy shows that it was felt as a kosename at Rome in manynbsp;instances. Cams in fact is found as a name all over the Latinnbsp;part of the empire and is just about as frequent on the inscriptionsnbsp;of Rome and Italy as in Gaul and Spain. There is not the slightestnbsp;ground for assuming that Lucretius had any Celtic connections.

It must be because of the Marxian hypothesis that several scholars, especially countrymen of Marx, have assumed thatnbsp;Lucretius was of humble origin and immediately tried to provenbsp;that he was unusually deferential in his address to Memmius.nbsp;Sellar and Munro, to be sure, felt differently, and Merrill ^ wellnbsp;says Such expressions as volgus abhorret, and impia pectoranbsp;volgi show the intellectual aristocrat; but the austere sermon at

^ Brieger has already said so, in Bursians Jahresbericht, 1901, 159 ff. The cognomen is surprisingly rare in Cisalpine Gaul whence Marx suggests thatnbsp;Lucretius might have come.

3 Merrill's edition, p. 14. See also Giri, Bull. Phil. Class, rgio�ii and Tol-kiehn, in Woch. Klass. Phil. 1904, 362.

the opening of the second book on the variety of political ambition is evidence of a freedom of criticism and a liberty of thoughtnbsp;and expression which could hardly be found in a person of lownbsp;social standing at the time�. Once we are rid of the prejudicesnbsp;created by the conjectures of Marx I think that no reader ofnbsp;Lucretius will fail to see that the poet speaks like a free citizennbsp;conscious of an honorable position in society and that he addressesnbsp;Memmius as an equal.

AENEID, II, 557

That the body of Priam should come to lie on the sea-shore, has doubtless seemed incongruous to others than the writer.nbsp;During the progress of this study, I was pleased (Vivant qui antenbsp;nos nostra dixerunt!) to find a short note by J. Mahly in thenbsp;Zeitschrift fur die �sterreichischen Gymnasien, 1887 (38, 415):nbsp;�Aber wer hat denn das �ingens corpus� des K�nigs an das Gestadenbsp;geschleppt? Verstandlicher und natiirlicher ware �limine�.�

With the city in flames, the gods themselves taking part in its overthrow, Ulysses and Phoenix guarding the spoils and captives,nbsp;and, at the very close of the book, the Greeks holding all thenbsp;gates, there was no likelihood that anyone would be inclined tonbsp;transport the king�s body over the wide plain and through thenbsp;ford of the deep-eddying Xanthus to the Hellespont, some fivenbsp;miles away.

In the cases of Patroclus and Hector, the possession of the hero�s corpse became, to be sure, the object of much activity onnbsp;the part of the warriors: Hector wished to drag away the nakednbsp;body of Patroclus from the plain, in order to give it to the dogs ofnbsp;Troy (II. 17, 127), while the Greeks finally brought it to theirnbsp;camp, and rendered it most imposing funeral honors; Achillesnbsp;was unrelenting in the persistent indignities which he wreakednbsp;upon the dead body of Hector (II. 24, 15), but, in general, thenbsp;Homeric heroes were content that the bodies of their foes shouldnbsp;be devoured by vultures where they fell. This place, in the casenbsp;of Priam, was a sequestered limen of his palace.

This word, from its original meaning, �dividing-line,� �sill,� has a variety of derived, transferred uses in the Aeneid, but fornbsp;the purposes of our passage may be held to signify �place ofnbsp;entering in,� comparable to the 7iQ(hxf}ai �vgriaiv of II. 22, 66.

As in the case of litus, the word makes a convenient and perfect dactyl in the declined cases, the forms limine, limina; litore, litoranbsp;occurring unnumbered times in the fifth foot; litore here is somewhat more frequent than limine, as might be expected in a seagoing epic. In this position, litore (�place where�), without anbsp;preposition, occurs somewhat oftener than the prepositionalnbsp;phrase, in litore; whereas in limine, in this sense, occurs aboutnbsp;eight times as often as the simple limine. The clearest cases fornbsp;limine, without preposition, in the fifth foot, and indicatingnbsp;�place where,� are:

sedens niveo candentis limine Phoebi (8, 720). Illustrations could be multiplied as to the indifferent use ornbsp;omission of the preposition in in such constructions, e. g.:nbsp;nudus in ignota, Palinure, iacebis harena (5, 871)

A scribe, accustomed to litore in the fifth foot, would very easily substitute it in place of limine, though I believe that thenbsp;current reading litore in our line, 2, 557, is found in all standardnbsp;manuscripts and editions. Precisely this easy substitution is shownnbsp;in line 8, 555: Palatinus, y��-, c, and % read limina; Mediceus,nbsp;Romanus, and read litora.

The strongest argument for the proposed amendment lies in the fact that the whole action connected with the death of Priamnbsp;revolves continually about the word limen. Aeneas, with hisnbsp;valiant companions, while fighting in burning Troy, is drawn,nbsp;by the vehemence of the conflict there, to the palace of Priam.nbsp;The entrance is occupied by attacking Greeks, who have formednbsp;a testudo (2, 441):

caede Neoptolemum geminosque in limine Atridas. Priam, driven to fury by the destruction of the city (507 f.):nbsp;urbis uti captae casum convolsaque viditnbsp;limina tectorum et medium in penetralibus hostem,nbsp;rushes, in spite of his age, against the invader. Pyrrhus butchersnbsp;him, after slaying his son Polites.

In view of all these considerations, there seem to be convincing reasons for reading, at the close of this scene:��

THE GERMANIC VOWEL SHIFT AND THE ORIGIN OF MUTATION

I. THE INDO-EUROPEAN VOWEL SYSTEM. The foundations of our present understanding of Indo-European phonology have been obtained in two great strides. In 1822 Jacob Grimmnbsp;correlated the Germanic Consonant System with that of thenbsp;other Indo-European languages by recognizing the primary factorsnbsp;of the Germanic Consonant Shift. More than half a century later,nbsp;in 1878, Hermann Collitz, through his discovery of the Palatalnbsp;Law gave the proof that the Indo-European vowel system wasnbsp;closely akin to that of early Greek and Latin, while the Indo-Iranian vowels, which had been supposed to represent the primitive Indo-European condition, were shown to be a secondarynbsp;development. Neither Grimm nor Collitz, it is true, stood entirelynbsp;alone in their discoveries; Rask, Johannes Schmidt and othersnbsp;have furnished meritorious contributions to the new evidence,nbsp;but in the main the names Grimm and Collitz will always standnbsp;out as those of the most important pioneers in our �reconstruction� of the Indo-European consonants and vowels.

In regard to both sound systems, the Germanic languages, especially in their earlier periods, present a network of variationsnbsp;which seem kaleidoscopic at first consideration, but on closernbsp;analysis follow more or less clearly distinguishable directions ofnbsp;the type for which Sapir has coined the fortunate term �Drift.�nbsp;The drift in the Germanic Consonant System is especially obvious,nbsp;and the writer has attempted to formulate it in various articles,nbsp;especially in MPh. XV and XVI. The details of the transitionnbsp;of the Indo-European vowels to the various Germanic Vowelnbsp;Systems have also been investigated at various times, and Collitz�s

the most important attempt to clarify that involved problem. The writer cannot fully accept certain aspects of Collitz�s viewsnbsp;on the development of the Germanic vowels, especially hisnbsp;contention that the primitive Germanic and the Gothic vowelnbsp;system were essentially identical, but he is nevertheless indebtednbsp;to him for invaluable stimulation and information on this asnbsp;well as many other linguistic problems.

2. nbsp;nbsp;nbsp;THE GERMANIC VOWEL SHIFT. A mere tabulationnbsp;of those facts that are accepted by the majority of linguists innbsp;the Germanic field reveals for the development of the Germanicnbsp;vowels a consistent continuity that is no less amazing than thatnbsp;of the Germanic consonants. In fact, the assumption of a vowelnbsp;shift in the Germanic languages is quite as valid as the assumptionnbsp;of the well established consonant shift, and it might be best tonbsp;use the term �sound shift� as comprising both groups. To statenbsp;the external facts of the drift in concrete terms is a relatively easynbsp;matter; to explain its meaning or cause is difficult and may atnbsp;the present stage of linguistic understanding be even impossible.

The term �Germanic Vowel Shift� as I am using it in this paper should be interpreted as comprising the spontaneous vowelnbsp;changes of the pre-historic period before the formation of thenbsp;several Germanic languages that is, roughly speaking, of pre-Christian times (Collitz�s Proto-Germanic and Primitive Germanic). In principle, therefore, it excludes changes that dependnbsp;on neighbouring sounds, but this is not and should not be, carriednbsp;through with complete consistency. Also, theoretically the termnbsp;should not take into account later changes in the individualnbsp;Germanic languages. But we are justified in including certainnbsp;indications of a continuation of the trend into historical times.

3. nbsp;nbsp;nbsp;THE VOWEL TRIANGLE. In the following analysis ofnbsp;the Germanic vowel drift it is expedient to make use of thatnbsp;standardized diagram that we are accustomed to term the �Vowelnbsp;Triangle�. This may seem objectionable in view of Russell�snbsp;recent opposition to that concept (cp. especially The Vowel,nbsp;Columbus, O., 1928), and a brief discussion of his objections isnbsp;quite unavoidable for the purpose of the present paper. I am,nbsp;of course, aware of the fact that in spite of such over-^convention-alized diagrams as we find, for instance, in Vi�tor-Rippmann�snbsp;Elements of Phonetics the term �triangle� is not to be taken innbsp;its geometrical sense but merely indicates the acceptance of

intersecting spheres of tongue articulations in which the distances between the quot;points of articulation� are wider for �high vowels�nbsp;than for �low vowels.� The use of a triangle in representing thisnbsp;theory merely takes into account the practical fact that evennbsp;within a given language (say, German) the contrasts betweennbsp;the various low-vowels [a, a, se] are generally less marked thannbsp;those between front and back mid-vowels and between front andnbsp;back high-vowels, and that for certain practical purposes it isnbsp;expedient, but hardly more than that, to sum up all low vow^elsnbsp;in the convenient although rather vague symbol [a]. In thatnbsp;sense, there is really no contradiction between the triangle andnbsp;Bell-Sweet�s rectangular scheme. The latter merely goes to thenbsp;other extreme of standardization, assigning to the low vowelsnbsp;the same latitude that is given to the high vowels. The writernbsp;has attempted to compromise between these extremes in hisnbsp;Sounds and History of the German Language (p. 38) by using anbsp;spherical diagram of this shape:

The top curve in this diagram is supposed to run approximately parallel to the roof of the mouth, while the bottom line roughlynbsp;symbolizes the tongue in its lowest position. Evidently thisnbsp;diagram does not strive for physiological or geometrical accuracynbsp;any more than the triangle, or, for that matter, the rectanglenbsp;of the English school, but it does give recognition to the beliefnbsp;in the �xistence of �high-vowels,� �front-vowels,� etc. Nor doesnbsp;it seem to me that Russell�s admirable investigations have shakennbsp;that assumption in the least.

I wish to go on record with an expression of the highest appreciation of the fundamental scientific value of Russell�s experimental work. His x-ray photographs of vowel articulations are

beyond comparison the clearest and most reliable that have ever been published; in fact, most similar attempts dwindle intonbsp;insignificance before Russell�s attainments. But I am unablenbsp;to reconcile his photographs (and his tables, as well) with hisnbsp;conclusions. He characterizes the vowel triangle as a �fallacy�nbsp;and goes so far as to say (p. 149): �It is only by the biggestnbsp;stretch of the imagination that we can make out anything whichnbsp;at all resembles points of tongue arching which would remotely

To arrive at tangible results I made minutely accurate tracings of Russell�s outlines of tongue and hard palate (p. 257 ff. andnbsp;p. no ff.) and super-imposed and concentrated them into onenbsp;diagram by means of transparent tracing paper (the transparent diagram which Russell himself attaches as a help innbsp;comparing his pictures with the �traditional sweet terminalogy�nbsp;{sic, p. 257) is quite meaningless for the purpose). In that waynbsp;I obtained a composite outline, in which the �points of articulation� are located in the following way;

I was amazed to find that this careful tabulation of Russell�s own results did not substantiate his skepticism. More than that:nbsp;the �points of articulation� in this diagram that is, the points ofnbsp;highest tongue elevation, are in principle identical with thosenbsp;contained in the newest diagram of the International Phoneticnbsp;Association, in which the four �fundamental� points [i, u, a, a]nbsp;were also obtained by X-ray photography {Lautzeichen und ihrenbsp;Anwendung, Berlin, 1928, p. 20 ff.). The relatively greater altitudenbsp;of the latter diagram possibly indicates an extreme articulation,nbsp;for the purpose of the experiment, but I am by no means sure onnbsp;that point; several other interpretations are possible. Russell�snbsp;outlines show the most surprising resemblance with those containednbsp;in Bremer�s Deutsche Lautlehre, published as early as 1893 (p. 160 f.nbsp;and table II), gained not by photography but by extremely

74 skillful and conscientious mechanical measurements. Of course,nbsp;they cannot claim the same reliability as Russell�s results, butnbsp;they are thoroughly confirmed by them. Bremer�s diagrams arenbsp;also substantially in agreement with several other of Russell�snbsp;statements, notably the assumption that �the point of archingnbsp;for u is not much farther back than is that for the front vowels��nbsp;(The Vowel, p. i6o). Indeed, a part of his results almost seemsnbsp;another illustration of a remark of Sievers� in his lectures onnbsp;Phonetics which I took down in shorthand as a student innbsp;1905: �Die Experimentalphonetik hat bis jetzt nichts gefunden,nbsp;was die besten Phonetiker nicht ohnehin wuBten.� Doubtless,nbsp;in view of the great progress in experimental phonetics duringnbsp;these twenty-five years, Sievers himself would nowadays notnbsp;maintain this view in full, but in the present case it seemsnbsp;only slightly exaggerated as far as the �points of articulationquot;nbsp;are concerned. Here, indeed, Russell has in the main confirmednbsp;the previous statements of the �best phoneticians,�� notablynbsp;Bremer, and his quotations from Vi�tor, Jespersen, Meyer andnbsp;others do not prove the opposite for this part of his contentions.

True, the opposition to �points of articulation�� is only a part of Russell�s results. His discoveries in regard to the movementsnbsp;of the epiglottis, velum, and back of the tongue are to a largenbsp;extent new contributions of the greatest value, but I cannotnbsp;admit that these factors eliminate the concept of points ofnbsp;articulation. All of these elements of articulation constitute annbsp;interlocking complex, although up to a certain degree each elementnbsp;may vary independently from the others. The point of articulationnbsp;is, of course, not the cause of any given vowel articulation butnbsp;merely its most characteristic symptom. Taken alone, it meansnbsp;nothing. It is obviously true, as Russell points out, that anynbsp;vowel can be articulated with a virtually flat tongue. It is alsonbsp;true, I might add, that variations of the angle of the jaws arenbsp;by no means indispensable; many men are in the habit of speakingnbsp;while holding a pipe firmly between their teeth, thus insuringnbsp;an invariable angle of the jaws. All of this does not in the leastnbsp;controvert the fact that in average articulation each vowel familynbsp;ranges within a given sphere of tongue elevation, which in anbsp;diagram such as that of Bremer or of the IPA may quite legitimately be indicated as a point of articulation. That these spheresnbsp;of articulation are apt to intersect is no proof to the contrary.

(Compare Sounds and History of the German Language, p. 38 f.) I consider it therefore justified and even necessary to retain innbsp;the present analysis the concept and general scheme of the vowelnbsp;triangle, all the more since in the investigation of earlier periodsnbsp;of the language it is obviously impossible in most cases to definenbsp;the tongue positions with the accuracy of a contemporary X-raynbsp;picture. The general trend of vowel changes is quite certainnbsp;nevertheless.

4. nbsp;nbsp;nbsp;VOWEL QUANTITY AND VOWEL SHIFT. There existsnbsp;a striking parallelism between the quantity of Indo-Europeannbsp;and Germanic vowels and the Germanic Vowel Shift, all the morenbsp;striking since it presents a sharp contrast to similar groups ofnbsp;phonetic changes in other Indo-European languages. As will benbsp;shown below, in the Germanic languages long and short vowelsnbsp;move in opposite directions, thus increasing the divergencynbsp;between the two series. Thus IE a became Gc. � (HG. w) whilenbsp;IE 6 became Gc. d. Nowhere else is such a contrasting treatmentnbsp;to be found. In Indo-Iranian, both �, 5 and �, o are lowered tonbsp;a, d. In the Romance Languages the deciding factor is not quantitynbsp;but position in closed or open syllable. Slavic at first glance seemsnbsp;to offer a partial analogy since here as in Germanic (and Lituanian)nbsp;� and d on the one hand and 0 and d on the other fall together.nbsp;This fact is frequently considered a sort of connecting link betweennbsp;Germanic and Slavic (thus Kretschmer, [Einfilhrung in) die indo-germanische Sprachwissenschaft, p. 54). But, as a matter of fact,nbsp;the apparent similarity in this respect between Germanic andnbsp;Slavic leads to opposite results. In Slavic � is lowered to d (Lat.nbsp;do-num; SI. dd-ti) while d is raised to 0 (L. ar-dre: SI. or-ati). It cannbsp;safely be said that this vowel change was merely the first step,nbsp;or an accompanying and completing factor, in the obliterationnbsp;of the difference between long and short vowels; in Old Bulgariannbsp;a is long while 0 is short, but in the later development this distinction disappears. �� Vowel fronting or narrowing, which is suchnbsp;a common process in Greek dialects, in Celtic, and in Romancenbsp;and Germanic Languages that developed on Celtic soil, is eithernbsp;restricted to long vowels or at least far more general with them;nbsp;but no contrary movement toward vowel widening or relaxing,nbsp;such as we shall find in Germanic, occurs in those languages on anbsp;significant scale.

of the Germanic long vowels are well understood and its general direction is easily established. Its graphic representation bynbsp;means of the vowel triangle (see below) shows a remarkablenbsp;uniformity comparable in consistency and simplicity to thenbsp;Germanic consonant shift. If we arbitrarily arrange the vowelsnbsp;in the diagram in such a way that the direction of the currentnbsp;of breath follows our customary direction of writing, that is,nbsp;from left to right (this is Sievers� preferred arrangement), wenbsp;may state symbolically that the Germanic vowel trend followsnbsp;a quot;clock-wise� direction:

\./

This uniformity is too constant to be accepted as accidental, but what is its significance? Two definitions of the trend offernbsp;themselves, one physiological and one acoustic; they do notnbsp;exclude but rather support and confirm one another. Physiologically, it appears that the articulation of Germanic long vowelsnbsp;moves farther and farther away from the resting position of thenbsp;tongue (Ruhelage). I am aware of the fact that this position isnbsp;difficult to define and varies both individually and nationally.nbsp;Still, we are on fairly safe ground if we say that in fausa thenbsp;mouth is closed or nearly closed and the tongue is well forward,nbsp;its point being at or near the lower teeth so that the f-positionnbsp;is approximated. Every Germanic long-vowel change means anbsp;step farther away from this basis. If we examine the seriesnbsp;�gt;��gt;dgt;�gt;� we find the following: the tongue is at firstnbsp;lowered as far as possible with corresponding widening of thenbsp;angle of the jaws (IE � gt; Gc. w NWGc. a); when the lowestnbsp;point has been reached, the articulating activity is furthernbsp;increased by withdrawing the tongue, leading to its physicallynbsp;unavoidable elevation towards the soft palate {d gt; �). The latternbsp;aspect of the process becomes still more evident in the historicalnbsp;continuation of this change in a part of the Germanic languages:nbsp;Lat. frater, Goth, hr�par, OHG hruoder, NHG Bruder. The factsnbsp;as such are incontestable; the Germanic consonant shift seemsnbsp;to represent an analogous increase of articulatory activity.

However, this plain statement of actual physiological facts does not explain the Germanic vowel shift but merely describes it. Innbsp;the case of consonants, the physical contact (tongue-roof ornbsp;lip-lip) produces a distinct muscular feeling which is capablenbsp;of increase or decrease according to the degree of expiratorynbsp;pressure or muscular tension. The articulation of vowels is notnbsp;accompanied by any muscular feeling of similar distinctness andnbsp;we are hardly justified in attributing the change, say, from d to d.nbsp;merely to a tendency towards greater articulating energy, althoughnbsp;in point of fact the tongue is withdrawn, the angle of the jawsnbsp;is widened and incidentally the lips are more or less rounded.

A partial explanation of the phenomenon is offered by Sievers, GrundzUge der Phonetik^, p. 279: �Kurze und lange Vokalenbsp;schlagen bekanntlich bei derartigen Verschiebungen haufig ent-gegengesetzte Wege ein. . . . Hief�r liegt der Grund wohl in demnbsp;auch sonst vielfach zur Anw'endung kommenden Gesetz, daB dienbsp;Artikulationen eines Lautes um so energischer und sicherer voll-zogen werden, je starker derselbe zum BewuBtsein kommt, d. h.nbsp;je gr�Ber seine Starke und Dauer ist. Dies erklart beim langennbsp;Vokal sowohl eine Steigerung in der spezifischen Zungenartiku-lation (nach Stellung und Spannung) als der Rundung, fallsnbsp;solche vorhanden ist. Beim kurzen Vokal dagegen, der nur einennbsp;momentanen Zungenschlag erfordert, wird leicht das eigentlichenbsp;MaB der Entfernung von der Ruhelage wie der Spannung nichtnbsp;erreicht, d. h. es wird eine Wandlung von Vokalen mit starkerernbsp;spezifischer Artikulation zu Tauten von mehr neutraler Artiku-lation angebahnt, sowohl was Zungen- und Lippenstellung alsnbsp;w^as Spannung betrifft.� This is doubtless correct as far as it goes,nbsp;but it does not explain the fact that Germanic vowels follow thisnbsp;tendency with such unalterable consistency while other languages,nbsp;as has been pointed out above (and numerous instances could benbsp;added) carry it out only casually or even proceed in the oppositenbsp;direction.

An explanation may be found in the same principle that has actuated so many other changes in the Germanic languages,nbsp;namely, the striving for contrasts, (cp. author, �Inflectionalnbsp;Contrasts in Germanic, JEGPh. XX, 468 //.) Sievers� definitionnbsp;of the underlying cause is abstractly true, but tbe effect followednbsp;only where the contrast between long and short vowels, asnbsp;subsequent linguistic history shows, was sharply outlined in

78 speech consciousness, that is, principally in the Germanic languages. Variation in the degree of speech consciousness had beennbsp;a chief starting point for contrasts in quantity; it led graduallynbsp;to an increase of muscle tension and secondarily to such progressive changes in the position of the tongue as have been described.

But incidentally an acoustic aspect of the problem is suggested by the fact that each step of the long-vowel shift implies a loweringnbsp;of the vowel pitch. While personally inclined to interpret thisnbsp;consistent lowering of the pitch of long vowels as merely annbsp;incidental physical effect of the progressive variation of thenbsp;resonance chambers, I must admit that it is by no means impossible to consider it the primary cause. It might be argued thatnbsp;the greater carrying power of low-pitch vowels led to changesnbsp;of articulation in that direction in the case of long and thereforenbsp;relatively emphatic vowels, while short vowels followed thenbsp;opposite trend. But I am skeptical as to this interpretation.

6. THE SHORT VOWELS are less clearly understood as far as the historical facts are concerned. So much is certain: o (and onbsp;in syllables with the Germanic accent) became a; u became o,nbsp;especially in West-Germanic and Norse, under certain conditions;nbsp;� became � under more or less analogous conditions, and � became inbsp;under the opposite conditions. The exact relation between Gc.nbsp;� and � must be considered a moot question but there is no clearnbsp;evidence of any Germanic change from o to �. � A conditionalnbsp;change of a to � takes place in historical times, too late to appearnbsp;in (literary) Gothic. For this, see section 7.

Diagram B on page 76 schematizes the trend of these changes, broken lines indicating conditional changes. It is obvious thatnbsp;the direction of this contrary-clock-wise development is just thenbsp;opposite of the process observed with long vowels, so the explanation should be simple since we would merely have to assumenbsp;muscular relaxation for the short vowels as against muscularnbsp;tension for the long vowels. But we are somewhat hamperednbsp;by the uncertainty of the historical facts. Since the problemnbsp;has been restated so admirably by Collitz (cp. � i) I maynbsp;restrict myself to the briefest remarks. These three points arenbsp;in doubt:

I. Did Gc. 0 and � change to � and i in Primitive Germanic ever5rwhere except before h and r (and in West Germanic beforenbsp;i and u in the following syllable) ?

2. nbsp;nbsp;nbsp;Did IE � and i change to o and � under the same conditionsnbsp;in which Gc. o and � were preserved, namely, before h and r andnbsp;in West-Germanic before a in the following syllable ?

3. nbsp;nbsp;nbsp;Was the Germanic vowel that developed before IE m, n, I, r,nbsp;perhaps 0 rather than u, as Loewe assumes, and did this 0 changenbsp;to ft under the same conditions as those in which IE ft wasnbsp;preserved ?

If these questions are answered in the affirmative. Diagram B is wrong and we must assume this development instead:

At first, IE ft is lowered to Prim. Gc, o just as IE 0 had been lowered to a. Later, the two mid-vowels (� and 0) are raised tonbsp;high-vowels unless prevented by /t or r (Collitz); still later, thesenbsp;two high-vowels, ft and � (IE as well as Gc.), are lowered undernbsp;the influence of a following low-vowel, while the mid-vowels arenbsp;raised before high-vowels. In that case we should have to assumenbsp;alternating stages of different phonetic tendencies in Germanic.nbsp;But in view of its great uniformity in other respects, e. g. in thenbsp;development of the long vowels, such an assumption of contradictory drifts could only be justified by definite historical evidence.nbsp;But the facts in the questions formulated above are ambiguousnbsp;and can be used equally well for the affirmative as for the negative.

Therefore clearness must be sought elsewhere than through historically accessible material, and I believe that we have herenbsp;a striking instance of the methodical value of the concept ofnbsp;phonetic tendency, in the sense of the assertion of Vendryesnbsp;{M�l. ling., p. 116): �Une loi phon�tique ne peut done �trenbsp;reconnue valable que si^est d�accord avec les principes quinbsp;r�gissent le syst�me articulatoire de la langue au moment ounbsp;elle agit ... La notion de tendance phon�tique est plus exactenbsp;theor�tiquement, et pratiquement plus f�conde que celle de loinbsp;phon�tique. Elle seule permet de d�terminer avec pr�cision lanbsp;cause des changements phon�tiques et d�interpreter scientifique-ment ceux m�mes qui paraissent le plus rebelles a toute disciplinenbsp;scientifique.�

Perhaps the following is in keeping with the facts: Apparently (and for this Russell�s experiments have given new evidence) thenbsp;articulating spheres of i and e are not only contiguous but evennbsp;overlap, so that e is not infrequently pronounced with a highernbsp;tongue elevation than i. Germanic e and i virtually constitutenbsp;one phonetic concept (one phoneme). Its articulation was deter-

8o mined by various factors. When unaccented it inclined towardsnbsp;the resting position, and i resulted. In accented (non-nasal)nbsp;syllables its development depended on the vowel of the followingnbsp;syllable. If that was low (Gc. a, ce) the mid-vowel variety of thenbsp;phoneme prevailed regardless of its origin from IE i or e. Thusnbsp;IE *ni-zdom, *wiros became Gc. *nesta-, *wera-, and e wasnbsp;preserved in the present stems of the fourth and fifth classes ofnbsp;strong verbs (neman, etan). Under all other circumstanc�s inbsp;resulted: It remained, for instance, in the preterite plural of thenbsp;first class, and it resulted in the 2nd and 3rd sing. pres, of thenbsp;fourth and fifth classes: OHG ziguni, nimit, gibit. Certain detailsnbsp;of this general trend require comment. First of all, the negativenbsp;effect of a low-vowel in the following syllable did not overcomenbsp;the underlying trend towards i when the latter was supportednbsp;by association with other grammatical forms, therefore thenbsp;�a-Umlaut� did take place in isolated words like *nesta-, *wera-,nbsp;but the high-vowel variety of the phoneme was protected bynbsp;analogy with the other verb forms in the past participles of thenbsp;first class of strong verbs: gizigan, gistigan.

The effect of u in the following syllable is not easy to judge. I am at present inclined to believe that in this respect Old Highnbsp;German presents the original Germanic condition, insofar asnbsp;there the e-i phoneme yields i if followed by u (gihti, nimuj.nbsp;Gothic does not help us in that respect, since it gives the preference to i regardless of the following vowel, and conditions innbsp;Norse and Old English appear to be secondary.

Thus the interchange between � and % in Germanic, aside from Gothic, is conditional in the sense that it really rests on a spontaneous trend just as much as the change of 0 to d, but it is eithernbsp;supported or counteracted by association with the vowel of thenbsp;following syllable. I dare not decide whether this association isnbsp;chiefly psychological or chiefly physiological, nor has this oldnbsp;controversy any bearing on the general problem. The changenbsp;of w to � is analogous in character, but it is simpler insofar asnbsp;there is no corresponding change of o to �. We can say: u followsnbsp;the general trend of short vowels, that is, the tongue elevationnbsp;is lowered (the muscles of the tongue are relaxed) so that 0nbsp;results, but this change, too, is conditional. It takes place onlynbsp;under conditions that favor it, namely, under the associative influence of a following low-vowel: OHG gibotan.

7. nbsp;nbsp;nbsp;THE ORIGIN OF MUTATION. The conditional changesnbsp;� gt; � and � gt; � followed the general Germanic trend and belongnbsp;to the Primitive Germanic period, although in Gothic the interference of new phonetic laws has largely obscured them. Theynbsp;implied the establishment of a new linguistic device of greatnbsp;importance, namely, the influence of inflectional vowels on thenbsp;quality of stem vowels. At first it functioned only as a contributingnbsp;factor that either favored or hampered a phonetic tendency ofnbsp;somewhat unstable character, but it affected the inflection,nbsp;especially of the verb, so strongly that it could hardly remain innbsp;this tentative condition. It was bound to disappear or to spread.nbsp;Gothic, with its strong leveling tendency, discarded it by standardizing � and %. In West Germanic and Norse, however, the newnbsp;device became one of the most important morphological factors.

First of all, probably in Primitive Germanic times, although not apparent in Gothic, the variation between � (both IE ft andnbsp;Gc. % from I, r, m, n,) and o, and between � (IE �) and t, spread tonbsp;IE �. That was natural, although against the basic phonetic trend,nbsp;which, however, exerted a restricting influence, as stated above.

The leveling tendency of Wulfila�s Gothic did not accept this change as a distinguishing element in its inflection; but it spreadnbsp;widely in the other Germanic Dialects. It was first transferrednbsp;to Germanic a, affecting it in the direction of the basic trend asnbsp;diagram B indicates. Due to the great number of nouns withnbsp;the stem vowel a, this change became a frequent plural characteristic and brought about corresponding changes of other stemnbsp;vowels that had no longer any connection with the originalnbsp;direction of the Germanic Vo-wel Shift. It is not impossible thatnbsp;in Old English and Norse the mingling with speakers of Celticnbsp;and Finnic languages may have favored the adoption and spreadnbsp;of the new device since the Celtic �consonant colors� and thenbsp;Ugro-Finnic �vowel harmony� in their results resemble vowelnbsp;mutation, although their phonetic basis is different. But thenbsp;origin of mutation is doubtless to be sought in the Germanicnbsp;Vowel Shift.

8. nbsp;nbsp;nbsp;LATER DEVELOPMENTS do not belong to the scope ofnbsp;this paper, which is limited to general Germanic conditions,nbsp;nor was the influence of consonants taken into consideration innbsp;the discussion of the Vowel Shift. I believe that in strictly Germanic vowel development it represents only a negative element.

Nasal combinations prevent the change of % to � under the influence of a following a, but require � instead of e regardlessnbsp;of the following vowel. This means that a nasal vowel is notnbsp;subject to the influence of the following vowel, the nasal combination forming a more effective syllable division than othernbsp;consonants or consonant groups. Moreover, nasalization, perhapsnbsp;on account of the lower pitch of nasal vowels, tends to preservenbsp;the low pitch of u. �� Influences exerted by h, r, and othernbsp;consonants are not Primitive Germanic but independent phenomena in the several Germanic languages which have nothingnbsp;to do with the vowel drift.

SALIC LITUS

I. The term litus, lidus, letus, ledus, laetus in the Salic Law ^ designates a person of the semi-servile class intermediate betweennbsp;serf and free Frank: adscriptus glehae, Horiger. In the ill-preservednbsp;Frankish glosses of the Law our word appears as letu, lexm, ledo,nbsp;lito (4 MSS), leciim, ISi�, lamp;us (2 MSS) From the Salic Lawnbsp;litus w'as taken into the other laws and into documentary use.

Grimm, 1. c., recognized litus as the latinized form of a Frankish word, which he set up as *lita-', this he connected, by vowel-variation, with OHG laz and cognates, of the same meaning;nbsp;this, in turn, with the adjective �^lata- �slack, lazy�� (Go latanbsp;�okn�re� etc.), and thus with ^l�t� �I let�� (Go Man, etc.). Asidenbsp;from the romantic view of the meaning, � the master is keen,nbsp;the servant slack, � we know today that an alternant withnbsp;PGic i (PIE e) has no place in the alternation of PGic �:�:a.

Von Kralik, Neues Archiv der Gesellschaft fur altere deutsche Geschichte 38 (1913), 430 developed a new etymology; he supposesnbsp;that the i, e of the Merovingian scribes represents a PGic �,nbsp;early Frankish [e:] or [s;], and connects our word with fourth-century Latin laetus �semi-servile Germanic colonist on Gallicnbsp;soil�, and with OHG laz �litus� and cognates (whose vowel isnbsp;now known to be long); all three represent a PGic *Ma- in thenbsp;sense of a past participle of the verb *M� �I let�: ^Ma- is �con-

^ Ed. Hessels, Lex Salica (1880), with index; cf. Schramm, Sprachliches xur Lex Salica (Marburger Beitrage zur romanischen Philologie, 3 [1911]), withnbsp;stemma of MSS on p. 2. On the meaning of our term, see Grimm, Deutschenbsp;Rechtsaltertmner, I* (1899), 424 (305); Schroder-Kiinszberg, Deutsche Rechts-geschichte^ (1922), 239 with references.

^ See Grimm in Merkel�s ed. (1850), xxx. Hessels� Index (not his text) gives one of the two last-named occurrences as letur.

84 quered person who has been left on his land or allowed to settlenbsp;on ours�. That a type ¹l�ta- had the meaning of a past participlenbsp;is proved by the parallel formation Go fral�ts �apeleutheros�, OHGnbsp;frlldz �libertus�, etc.

This led Meyer-Liibke, Rontanisches etymologisches W�rter-buck ^ (Sammlung romanischer Elementar- und Handbiicher, III^ 3 [1924]), no. 4994, to derive OFr liege �bound by feudal tie�nbsp;from a type PGic ¹letika- �vassal� This ¹l�tika- Gamillschegg,nbsp;Etymologisches Worterbuch der franz�sischen Sprache (1926 ff.),nbsp;s. V. lige, explicitly connects with our litus, letus, which he identifies, ignoring the difference of meaning, with the words fornbsp;�libertus�, citing the Go word as �let�. The forms litus andnbsp;OFr lige he explains, regardless of geography, as due to �HU�, anbsp;late Go variant of �let�.

We thus obtain a whole series of words in the three related meanings �libertus�, �inquilinus�, and �adscriptus glebae� ��nbsp;a confusion of forms and meanings which needs ordering.

2. The Gic type ¹-leta- as second member in words for �libertus� occurs in Go fral�ts (i Cor 7, 22) and possibly in the variantnbsp;reading fralatz in Lex Baiwariorum (ed. Schwind, MGH Legum 4�nbsp;I, 5, 2 [1926]), 338. Another compound appears in the variants,nbsp;friilaz, etc., 1. c., fern, frilaza, 358, and in glosses, libertus:nbsp;frilaz, etc., Steinmeyer-Sievers, Die althochdeutschen Glossen I,nbsp;760, 41; II, 122,16; 611,28; III, 426,41; 645,26. In OE thenbsp;word is an w-stem: frioleta, freolceta, etc.. Sweet, Oldest Englishnbsp;Texts (EETS 83), 75; 105; 115; Wright-Wiilcker, Anglo-Saxonnbsp;and Old English Vocabularies I, in, 22; Wright, A Second Volumenbsp;of Vocabularies (1873), 51. Another compound is OHG liberti'.nbsp;hantlazza, Steinmeyer-Sievers II, 139, 37.

This queer formation is paralleled, as von Kralik 431 points out, by OHG superstitem: aftarlaz, St-S I, 318,3. It may be anbsp;personalization of the neuter action noun in Go aflet, fral�tnbsp;�aphesis�, in the manner of NHG Bedienung, Go hliftus �thief�,nbsp;OB sluga �servant�; cf. Brugmann, Grundrifl ^ II, i (1906),nbsp;599; 610 and Brugmann-Delbr�ck, GrundripV^ III (1893), no.nbsp;As Grimm remarks, it would not do to confuse �libertus� withnbsp;�litus�; accordingly, von Kralik sees in his ¹leta- �litus� onlynbsp;a parallel formation with Go fral�ts �libertus�. However, queer

With the erroneous statement (due to misreading of von Kralik 435; repeated by Gamillschegg) that terras laeticae occurs in the Lex Baiwariorum,

as is a second member ¹-llta(n)- in passive meaning, an uncompounded ¹leta- would be queerer still. Except for a possible confusion in German (cf. below, �3), the above words for �libertus� are in all likelihood entirely distinct from the terms litus and laz.

3. nbsp;nbsp;nbsp;Latin laetus in the second half of the fourth century designatesnbsp;semi-seiwile Germanic (Batavian, Frankish, Suebian, Alemannic)nbsp;colonists in Gaul; laetus in Gaul is what inquilinus, at this time,nbsp;is in the other provinces

Kluge, Urgermanisch (Paul, Grundrili der germanischen Philologies, 2 [1913]),33takestorepresentaGicwordwithPGic�; the ae would represent, then, a stage intermediate between the �nbsp;reflected in Suehi (Caesar), Segimerus (Tacitus) and the historicalnbsp;German a. Bremer, PBB ii (1886), i has, indeed, shown thatnbsp;PGic � long remained in Frankish; names in -merus do not appearnbsp;with a before 500 A. D. A Frankish [e:] is reflected in OFr biere,nbsp;Meyer-L�bke no. 1038; cf. von Kralik 434. This interpretationnbsp;leads, of course, to the etymology worked out by von Kralik. However, it is nothing more than a possibility. Latin aeior the earlynbsp;Frankish form of PGic � occurs nowhere; Ammianus, who hasnbsp;laetus (ed. Clark, 1910) 16, ii, 4; 20, 8, 13; 21, 13, 16, writes Teuto-meres 15, 3, 10. Pliny�s� glaesum �amber� (for which Tacitus hasnbsp;glesum) leaves us at a loss: has it PGic � (M�llenhoff, ZfdA 23nbsp;[1879], 23; Bezzenberger, KZ 44 [1911], 291) or ai (Osthoff, MU 4nbsp;[1881], 145) or a (Bremer, PBB ii [1886], 16) ? A fourth-centurynbsp;Roman may have represented the quality of Gic short e by hisnbsp;ae; cf. Zosimus� spelling with Gk short e. The consonant maynbsp;have been Gic cf. Kluge, 1. c., 31. To name extreme possibilities, laetus may represent a Gic ¹Uta- or a Gic ¹lepu-.

4. nbsp;nbsp;nbsp;The type ¹l�ta- is in the WGic vernaculars the equivalentnbsp;of litus. In OF it occurs but once, in the oldest laws, as lest, andnbsp;is glossed peowa (Liebermann, Die Gesetze der Angelsachsen, I

For occurrences and meaning see Sch�nfeld in Pauly-Wissowa, Realenzyhl.^, s. V.; of the older discussions Boeking, Notita dignitatum II (1853), 1044 is thenbsp;fullest. One notes especially Theodosiani libri xvi (edd. Mommsen et Meyer,nbsp;1905) 13, II, 10 tevrae laeticae, and Zosimus (ed. Mendelssohn, 1887) 2, 54, inbsp;�is Leto�s, ethnos Galatikdn. � Du Cange�s mention, s. v. leti, of Vita S. Gildaenbsp;Sapientis relates only to the name Letavia] cf. the text (ed. Mommsen, MGHnbsp;Auct. Antiqu. 13 [1898]) 96, 12. � Hardly our word is Jordanes� (ed. Mommsen,nbsp;MGH Auct. Antiqu. 5, i [1882]) io8, 3 liticiani (later MSS litigiani, litiani)nbsp;among a list of Gic tribal names; cf. below, � 8.

Br�ch, Der Einfluj} der germanischen Spvachen auf das Vulgarlatein (Samm-lung romanischer Elementar- und Handbiicher V, i [1913]) 16 is in error when ke attributes the word to Caesar.

[1903], 4gt; 26) �. In OFris we have the plurals letar, Icthar (on the th see van Helten, Altostfriesische Grammatik [1890], 90), letannbsp;and the compounds let(h)ma, letsloxhta, letslachte, letslaga (Richthofen, s. vv.). Van Helten, Verhandelingen der K. Akademie vannbsp;Wetenschappen Amsterdam, afd. Letterkunde, n. r. 9 (1907), 104,nbsp;points out that West Frisian spellings with ee assure a long vowel.nbsp;The MLG (Schiller-Liibben, Mittelnicderdeutsches W�rterbuch, s. v.nbsp;lot �) and MDu forms (Verwijs en Verdam, Middelnederlandschnbsp;woordenboek, s. vv. laet and late-, a wealth of examples) do notnbsp;show the vowel quantity. The later forms here, as in HG, shownbsp;�-inflection. In OHG we have, as a variant of a gloss cited � 2nbsp;above, libertini'. laza (St-S III, 645, 26); this must be due to anbsp;confusion of the two words. But a gloss on the Lex Ribuaria,nbsp;litus-. laz (St-S II, 354, 30) is unmistakable. In Latin documentsnbsp;we have Hucbald (MGH Scriptores 2 [1829]) 361, 50 sunt quinbsp;lassi dicuntur', Nithard (ibid.) 668, 46 qui lazzi illorum linguanbsp;dicuntur; 669, 3 frilingis lazzibiisque; 377, 45 (spurious but old)nbsp;multis lazzis; Constitutioncs et acta publica (ed. Weiland, MGHnbsp;Legum 4� 4, I [1893]) 88, 14 (1035 A. D.) illis qui dicuntur lazinbsp;(with variants laz, lass, lasz); Diplomata regum 4 (ed. Bresslau,nbsp;MGH [1909]), 296, 16 (1035 A. D.) illis qui dicuntur laszi. Ournbsp;word has little life in HG; for the later usage see Lexer, Mittel-hochdeutsches W�rterbuch s. v. lazze; Haltaus, Glossarium Ger-manicum (1758) 1195; Grimm, Deutsches W�rterbuch 6, 212. Thenbsp;modern Lasse in learned works is artificial. It survives in opprobrious terms; so probably prostituta: huorra lazza, St-S I, 344, 41.nbsp;In modern dialects it is subject to popular etymology (Latznbsp;�latchet, bib�) which even the dictionary-makers seem to accept.nbsp;Alsatian (Martin-Lienhardt, W�rterbuch der elsassischen Mund-arten i [1899], 633; Stoeber, Die dcutschcn Mundarten 3 [1856],nbsp;483 recognized the origin) Jul, dumm wi e Latz; Hanslatz (1575nbsp;A. D.), etc.; Swabian (Fischer, Schwabisches W�rterbuch 4 [1914],nbsp;1019) in field names Latz, Burenlatz, Ldtzen, as an opprobriousnbsp;term, and in the name of the Shrovetide and Pentecostal clownnbsp;or dummy, Latzmann (cf. Sartori, Sitte und Branch [Handbiichernbsp;zur Volkskunde, 7�-8], 3 [1914], 231; Birlinger, VolkstUmliches

� The passage in Palgrave, Rise and Progress II (1832), ccxxxiv (Liebermann I, 400) cited by Richthofen, Altjriesisches W�rterbuch (1840), s. v. let, containsnbsp;a different word; cf. Liebermann III (1916), 237 and here, below, � 6.

Von Kralik 432 too readily accepts the popular etymology (Sachsenspiegel) there cited.

aus Schwaben, 2 [1862], 114; references due to Archer Taylor); in the passages from older literature note especially the farfetched popular etymology from the Zimmer Chronicle in explanation of Latzwein �lov/ grade of wine�; Swiss (Staub-Tobler,nbsp;Schweizerisches Idiotikon, 3 [1895], 1546) Laz �Dime�, Latznbsp;�awkward fellow�; Latsch, Latsch �clumsy, stupid person� (col.nbsp;1529) may be a contamination of our word. The [ts] of the modernnbsp;dialects probably does not point to old tt, but rather takes ranknbsp;among the examples of [ts] from old final i!; cf. Swiss suts �SchuB�nbsp;guts �GuB� (Streiff, Die Laute der Glarner Mundarten [Beitrdgenbsp;zur schweizerdeutschen Grammatik, ed. Bachmann, 8; 1915], 78;nbsp;see Behaghel, Geschichte der deutschen Sprache � [Paul, Gnindrip, 3;nbsp;1928], 420, with references).

5. Grimm, we have seen (� i), took Salic litus and variants to represent a Frankishnbsp;nbsp;nbsp;nbsp;von Kralik sets up Frankish

*leta-. Van Helten, PBB 25 (1900), 425 explains the spelling lidus (ledus) as due to Romance weakening of intervocalic t,,,nbsp;citing a few graphs of Latin words with d for t. Graff, Althoch-deutscher Sprachschatz 2 (1836), 191 seems to be alone in suggestingnbsp;a prototype with OHG d (i. e., PGic Igt;). Actually, however, thenbsp;MSS of Lex Salica do not often write d for Latin t] the few instances are suspiciously concentrated on a fev/ words. The onlynbsp;word comparable in this respect with litus, is fritus �king�snbsp;peace�, from Gicnbsp;nbsp;nbsp;nbsp;In the list below, the Latin words

with d for t occur once each, unless otherwise stated. The vowel variations of litus and fritus are here ignored.

8 9nbsp;Bnbsp;Fnbsp;Gnbsp;H

Leiden

Herold

Emendati

Latin words with d for t

dode, poledrum

strada

deladore, emancada, mederit, repedit 2, sudem, uinidorenbsp;tribudarium

dido (quot;digito�) 4, mederit, legadario

tribudarium

poledrum

tribudario, medere

toscada (�toxicata�)

medere

poledrum

sudem 2, poUedrum 2, capridum sudenn (�sutem�), puledrum 2nbsp;-8 sudem 2, puledrum 2

fridus

2 8nbsp;6

6 6nbsp;6nbsp;6nbsp;6nbsp;5nbsp;8

litus

8 6nbsp;6nbsp;5nbsp;5

4 4nbsp;4nbsp;2

(I)

1 o

2 onbsp;o

Of inverse writings I find only: MS 2 retat (reddat), MS 7 peiica. Special factors are involved in MSS i, 2, 6, 9 quatru-,nbsp;quatro-; MS 5 and Herold spato, spathum, espaiauerit; MSSnbsp;I, 2, 3, 4, 6, B, F, G, H, Emm cinitum, etc. (cinaedum?). Confusion in final position is another matter: MS 2 aliquit; MSS 3nbsp;(twice), 4 (twice), 7 aput; and MS 4 tor quid; MSS 8 (twice),nbsp;Leiden capud.

The spelling letus appears also in the glosses of the lost MSS d�Este (Hessels, col. xxii) and Thou (Merkel�s ed., xcvii).

The divergence between Utus and fritus may be due to the fact that the latter was in living use in German asnbsp;fridu, in Romance (names) as *jredu-, while the former wasnbsp;known only from law writings; even our MSS i and 2nbsp;once mistake the word and write lex. The Carolingian reform (Herold, Emendati) normalizes to lidus, fredus; the semi-reformed Leiden matches its two archaic Utus with two similarnbsp;fretus.

The later tradition of Utus (see Du Cange, s. v.) derives from the Salic Law. In the laws d-spellings prevail, with t in MSSnbsp;that otherwise also show Merovingian irregularities. Forms withnbsp;t occur in the Saxon laws (ed. Richthofen, MGH Legum P 5nbsp;[1889]), beside jrido, fredo', Lex Ribuaria (ed. Sohm, ibid.), withnbsp;^^-variants and misunderstandings Ubertum, licitum, beside fretus,nbsp;fredus) Roman Law of the Burgundians (ed. Bluhme, MGHnbsp;Legum P 3 [1863]), with one misunderstanding, et, beside Frede-gisclus, Fredemundus] Lex Frisionum (ed. Richthofen, ibid, andnbsp;Patetta, Mem. r. Accad. d. Sc. di Torino, ser. 2, t. 43 [1893]),nbsp;beside freda) Alemannic Laws (ed. Lehmann, MGH Legum 4�nbsp;5, I [1888]) Pactus, fragment 2 Utus, fragment 5 lita in one MS,nbsp;where the rest have lida and miswritings (French?) lisa, lesa,nbsp;beside fridus, fredus, except once (8th century MS) freto. Bothnbsp;lidus and fredus have d in Lex Francorum Chamavorum (ed. Sohm,nbsp;MHG Legum P 5) and Leges Langobardorum (edd. Bluhme etnbsp;Boretius, MGH Legum P 4[1868]). The Formulae (ed. Zeumernbsp;MGH Legum 4� 5 [1886]) have the derivative letimonium, liti-munium, with lidemonio in one MS. The archaic freta keeps up:nbsp;Lex Baiwariorum (ed. Schwind, MGH Legum 4� i, 5, 2 [1926])nbsp;in the archaic MSS; Lex Romana Raeiica Curiensis (ed. Zeumer,nbsp;MGH Legum f� 5 [1889]).

Karolingia (ed. M�hlbacher, MGH [igo�] �) 113; iig; Diplomata Regum {MGH, 1 and 2 ed. Sickel [1884.1893]; 3 and 4 ed. Bresslaunbsp;[1903. 1909]) passim. Spellings with d are rare: Dipl. Kar. 177nbsp;lidos-, Dipl. Reg. i, 401,39; 529,25; 2, 38,41. A late form isnbsp;liddo with �-inflection, in the spurious documents Dipl. Kar. 401;nbsp;Dipl. Reg. I, 294, I; it is genuine after the year 1000: Dipl. Reg.nbsp;3, 626, I; 4, 168, 31 Intermediate forms are litones, littones,nbsp;Dipl. Reg. 3, 10, 24; 4, 218, 17. The �-inflection and probablynbsp;the double writing of the consonant are due to vernacular HGnbsp;lazze. These documents consistently write jreda.

Litus and its variants represent a purely graphic learned tradition derived from the Salic Law. The spellings of the Salicnbsp;Law are inconclusive; what they tell is in favoi of P as the consonant of the underlying Frankish word; unfortunately theynbsp;give us no counter-example of a word with Frankish t.

We have seen that von Kralik takes the vowel of litus to represent Gic �; he blames the Merovingian scribes for the spellings with i. Certainly Briich, ZsfrP 38 (1917), 701 errs when henbsp;says that litus is quot;so gut wie immer mit i geschiiebenquot;; yet thenbsp;orthography of the Salic Law does not warrant von Kralik�snbsp;conclusion. In stressed syllables of Latin Words fluctuation ofnbsp;i and e is almost entirely confined to certain forms where specialnbsp;factors come into play: a type ficit of perfects with suffix zeronbsp;(namely, fecit, fregit, venit, presit and compounds), with i writtennbsp;for �; and a type reteneat of compounds with e written for i onnbsp;the model of the simple word (namely, exteterit, perdederit, ven-dederini, reteneat). Outside of this, i for Latin � occurs only innbsp;the few cases listed below, and c for Latin � only once (MS 2nbsp;deuidant), for Latin i twice (MSS 2 and 9 fedem)^^. The tablenbsp;below ignores the fluctuation of t and d.

Note especially that MS 2, which does fairly often (34 times in all) write i for Latin �, has only one litus and no fritus.

� I owe the references in this volume to C. A. Williams of the University of Illinois.

Special cases are (omnes) cinitum, cinidum, etc. (cinaedum?); MS 7 sipem; the various spellings, inium, etc., in all the MSS, for aeneuin �kettle�. I do notnbsp;count the river-name Ligerim, Ligere �Loire� or �Lys, Leye� (see Hessels 633),nbsp;Written variously with i and e. MS i has praeserit once, MS 6 thrice, beside oncenbsp;praesit.

8 9

Leiden

Herold

Emendati

-90

9 3nbsp;8

9 2

�

7-8

Latin words with i for �. tricinus, nonaginus, pli-bium, uioissimusnbsp;accidant, succidant, di-hilem, tricinus, quin-ginus, iudicihw, ma-nire, niinsis, pleuium,nbsp;secricius, stispinsus,nbsp;tiiro, dibeo 7

plebium

plebium, regiiiritur dibilem, minsts, plebium,nbsp;riteinnbsp;plebium

dibilium (pointiirg to dibilis ?)

The difference between the vowel writings of litus and fridus may, again, be due to the living character of the latter word.nbsp;The occasional laetus is doubtless due to the familiar if irrelevantnbsp;word-picture of laetus �glad�.

In the later laws the norm is lidus with i, fredus with e. MSS of the older type for Lex Alamannomm show variants with thenbsp;reverse vowels; letus, frido. The charters consistently write litus)nbsp;exceptions are letos Dipl. Reg. i, 413, 33; 2, 127, 36; 3, 133, 30;nbsp;4, 18, 34; laetos I, 66, 34. For *fripu- they write freda.

In Gic names the Merovingian scribes write *-mera~ as -merus, later as -mams, never with i; cf. Bremer, 1. c.

In so far as the spellings of litus tell us anything, they suggest a *lipu- parallel with *fnpu-. In both words the fluctuationnbsp;between i and e may date back to the Gic prototype and therenbsp;be due to the familiar alternation (OHG jihu, fehu; sciff, sceff)]nbsp;in ^'friPii- the living usage of Romance names later favored the e.

6. A Gic *lipu-, *lepu-, though swamped by *l�ta-, has left some traces. ON lidar (�Sat eru fylgSrmenn�) is set up as annbsp;�-stem (Cleasby-Vigfusson, Fritzner) and may well represent anbsp;type PGic *galidjan- �one who travels along� yet a technical

term *lipu- may have been merged with it and may also have contributed to the formation of lid n. �retinue� (*galidja-).nbsp;ME lith is doubtless from the Norse (NED), though the phrasenbsp;land and lith (reinterpreted in ne lathes ne landes, Liebermann

1, nbsp;nbsp;nbsp;400; cf. above, footnote 6) is reminiscent of OFris umbe londnbsp;ne umbe lethar. In German we have the standing expressionnbsp;der tiejal unde sine lide (Notker, ed. Piper [1883], 2, 116, 10;nbsp;619,8), des tiefels lide (occurrence, MSD^ [1892] �, 166; 2, 259;nbsp;Lexer, Mhd. Wb. 1, 1938); strangely, of Christ�s followers, Notker

2, nbsp;nbsp;nbsp;187, 31 This is scarcely *lipu- �limb�, though it may havenbsp;been so reinterpreted; in the same way E limb of Satan maynbsp;be a reformation of an older ^lith of Satan. Finally, MHG lidel�n,nbsp;tidl�n (Benecke-M�ller-Zarncke, Mhd. Wb. i, 1042; Lexer i, 1940nbsp;and Nachtrage, 301; Schmeller, Bayer. Wb.^ i [1872], 1442),nbsp;taken by Heyne (Grimm, D. Wb., s. v. Liedlohn) as a derivativenbsp;of OHG lldan �travel�, is specifically payment to the humblernbsp;sort of servants: Liedlohn soil man vor alien Schulden zahlen',nbsp;Verdienter Liedlohn schreit m Gott im Himmel (Bachtold-Staubli,nbsp;Handw�rterbuch des deutschen Aberglaubens [1929], s. v. Dienst-bote).

The adjective type '^liLuga-, *lepaga-, MHG lidec, ledec, NHG ledig is not exactly synonymous with frei (cf. especially, thenbsp;modern use of lediglich); as Heyne, 1. c. s. v. points out, itnbsp;reflects a term of social organization whose import is lost tonbsp;us. The meaning �free� appears also in the sister tongues, e. g.nbsp;ON: lauss ok lidugr, G los und ledig. It survived best in HGnbsp;because it was there taken into elevated language (MHG poetry)nbsp;us a synonym of vri.

7- OFr lige, liege, borrowed in Prov litge. It ligio, E liege, latinized to ligius, and hellenized to Uzios (Du Cange, s. v. ligius),nbsp;occurs typically in ome liges �vassal� (Godefroy, s. v.) andnbsp;derivatives. However, we have some divergent uses: lige seigneurnbsp;quot;liege lord�; le roy vertueux, franc et liege; property is givennbsp;lige; quitte et lige �free�; esligier, eslegier �to free�

Graff�s reference, Ahd. Sprachschatz 2, 191, to litten in the Vienna Genesis, though confirmed by Massmann, Deutsche Gedichte (BNL I, 3 [1837]) 2, 305,nbsp;verse 5673, is to be canceled: the MS has luten, Dollmayr, Die Sprache der Wienernbsp;Genesis (QF 94 [1903]), 2, for liuten �people�; cf. p, 72, line 2 and Hoffmann,nbsp;Pundgruben 2 (1837), 79, 32. � The Vienna Notker has Hdir at 3, ro2,6.

To the occurrences in Godefroy T. A. Jenkins kindly adds the following: Guernes, Vie St. Thomas 1854. 1856 huem liges-, 2517 lige seigneur] Ambroise,

The distribution of i and ie does not agree with that in any native type (Briich, ZsfrP 38 [1907], 701)

The older etymologies are given by Du Cange, s. v. ligius. Diez, Worterhuch^ (1887) decides that G ledig is the source. Thisnbsp;has been widely accepted (Nyrop, Grammaire historique i[i889],nbsp;9; Mackel, Franz�sische Studi�n 6 [1887], 82; Korting, Lateinisch-romanisches W�rterhuch^ [1907]. no. 5506; and others) but questioned on account of the discrepancy of meaning. Similarlynbsp;G. Paris, Romania 12 (1883), 382 derived esligier as ex- plus Gnbsp;*ledig�n, against Tobler, Jahrbuch f. rom. u. engl. Lit. 8 (1867),nbsp;342, who took it to be ex-litigare. Tobler�s explanation fails tonbsp;account for the e-varaint, eslegier; one suspects that the elitigarenbsp;of medieval law (Du Cange, s. v.) is a learned re-interpretation:nbsp;Capitula (ed. Boretius, MGH Legmn 4�, 2, i [1883]) 337, i resnbsp;litigosa nuUathenus potest dare neque vendere antequam elitigetur.nbsp;Note the spellings Dipl. Imper. (ed. Pertz, MGH 1872) 107, 20nbsp;evindecatum atque elidicatum', 108,11 evindicatas adque elidiatas)nbsp;similarly Formulae (ed. Zeumer, MGH), see Index, p. 753.

Baist, ZsfrP 28 (1904), 112 derives lige, following an old etymology, from Gic *leoda- �people (collective)�, *leudi- pi.nbsp;�people�, which appears in medieval Latin writings as leudis,nbsp;leodis', eligier he derives from leudis �wergeld�. This fails tonbsp;account for the Fr forms with f; nor does it seem that leodisnbsp;in Latin usage ever went farther than �retainer�.

Thomas [Diet, g�n.^ [1920], s. v.) connects lige with litus (*li-ticu-), as do Meyer-Liibke and Gamillschegg, cited at the beginning of this article. But litus, letus is a merely graphic and learned tradition which could not lead to such phonetic development as is implied in lige. There is no trace of an adjectivenbsp;in -icu- or of a verb derived from litus. Moreover, a noble ornbsp;even a freeman who entered into the feudal relation would notnbsp;call himself litus �adscriptus glebae�; the two relations werenbsp;entirely different.

In spite of Baist, 1. c., ledig and lige-ligius have some uses in common. Beside erledigen: esligier-elidigare, we have los und

Guerre sainte, ligece'. Vie de Guillaume le Mareschall 2oig homes Hges) Chrestien, Erec 3868 home ligequot;, Charetle 1731 chose lige.

However, it is perhaps not so clear-cut as Godefroy�s examples lead one to suppose; thus Aiol et Mirabel (ed. Foerster, 1876 ff.), beside the normal Picardnbsp;(Foerster, p. 495) en liege, verse 10 210, has also hon liges, verse 7590. (Referencenbsp;due to T. A. Jenkins).

ledig: quitte et lige. Du Cange, s. v. ledighman and Haltaus 1214 give US, from the 13th century, such turns as these: ligius homo,nbsp;quod Teutonice dicitur ledighman] homines legii dicti ledigman]nbsp;homo suus absolutus, quod vulgo ledighman nuncupatur] ent adjutornbsp;noster, quod ledichman dicitur in vulgari, contra quemlibet hominemnbsp;in hoc mundo (cf. semper liberi vasalli erimus). With ledigez eigen,nbsp;ledigez guot (Lexer, Haltaus) cf. OFr chose lige, sa lige mansion.nbsp;The OFr word is a loan from an antecedent of the documentednbsp;MHG ledec, lidec] probably both the native*and the Fr wordnbsp;changed their meanings before the time of our records; somenbsp;semantic common ground is left.

8. Our records are of technical terms; we have not the spoken forms from which these terms were specialized.

There seems to have existed a PGic or early general Gic term (spread by loans at the time of the new settlements?) *lelu-,nbsp;*lihi-, with the derived adjective *lilyuga-, *lepaga- and verbnbsp;*uz-lipug�mi, *uz-lepag�mi. The meaning of these words seemsnbsp;to have been close to that of *frija- �free�, but in some waynbsp;uiore restricted 1�.

It is barely possible that Jordanes� liticiani (footnote 4, above) IS an early trace of the adjective, in a specialized technical use,nbsp;3-s a synonym of laeti. The laeti of the fourth century may similarlynbsp;reflect our *lehu-, but it is ambiguous, since it may just as wellnbsp;reflect Gic *leta-.

The earliest relatively certain deposit is the letus, liUis of the Salic Law. If Gic *lepu-, *lilgt;u- was close to �free person� (innbsp;some way more restricted than *frija-), then the technical termnbsp;*lihi-, litus must at the time of its specialization have stressednbsp;the advantageous phase of adscriptio glebae, perhaps in contrastnbsp;with the fate of other conquered persons or groups, who werenbsp;reduced to thraldom, servitudo. Whether ON lidar �retainers�nbsp;has ^lipu- at least as a component, is uncertain. In WGic speechnbsp;lipu- was soon crowded out by Htta-, an entirely differentnbsp;word, of unknown origin, � perhaps at first the name of somenbsp;conquered tribe. Our litus is a �Latin� (that is, of course, annbsp;artificial, graphic) word, the prescribed equivalent, in Latin-writing, of the vernacular lot, laz. Relics of *lipu- are OHG lidenbsp;�followers (of Satan)� and MHG lide-lon �wages�.

Ultimate connection with *lipu- �limb� or with *llp� quot;I travel� is a matter beyond our ken.

specialized use, the adjective ^lepaga-, *lipuga- has perhaps been somewhat generalized in meaning by the time of our records.nbsp;Yet in ON lidugr and even more in the WGic languages it showsnbsp;traces of specialization in the legal sphere. In the language ofnbsp;MHG poetry, however, it became an elevated synonym of vri)nbsp;with this, its career was made. Of the servile connotation innbsp;which the underlying noun had once been specialized it showsnbsp;no trace; its French loan lige, liege has on the contrary beennbsp;carried into the noble sphere of the feudal relation, again withnbsp;traces of older meaning.

The verb doubtless lurks behind the elidiare, elitigare �free from legal claims� of medieval law. In German it has remained closenbsp;to the adjective: erledigen is merely the verb of ledig', but thenbsp;French loan-word elegier, eligier does not share the specializationnbsp;of liege, lige.

These words are different from *l�ta-, synonym and successful rival of *lepu-, *lipu-, and from derivatives of *l�t5 �I let�,nbsp;such as Go fralets �libertus�.

DER GENITIV PLURAL AUF -� IM GOTISCHEN

Das Gotische zeigt im Gegensatz zu allen anderen Sprachen der indogermanischen Sprachgruppe im Genitiv Plural samt-licher Stamme beim Masculinum und Neutrum (die �- und ein-Stamme nat�rlich ausgenommen) die eigent�mliche Endung -�:nbsp;dag�-waurd�, hairdj�-kunj�, gast�, suniw�, a�an�-hairtan�, br�pr�,nbsp;nasjand�, mann�. Ja diese Endung kommt auch beim Femininumnbsp;der z-Stamme (q�n�) und der weniger zahlreichen Stamme auf -unbsp;(kandiw�), auf -r (sehr wahrscheinlich, obwohl nicht belegt:nbsp;*swistr�, *dauhtr�) und der konsonantischen (baurg�) vor. Imnbsp;groBen und ganzen ist j edoch ein deutlicher Unterschied zwischennbsp;dem Plural der Maskulina-Neutra und dem der Feminina verhanden; die letzteren haben bis auf die erwahnten Ausnahmennbsp;-� (gib�-frdistubnj�, laisein�, qin�n�, managein�), die ersteren -�.nbsp;Das -� beim Femininum findet sich nur bei den Stammen,nbsp;WO das Maskulinum stark vertreten ist; also wohl Ausgleichnbsp;innerhalb der betreffenden Deklinationsklassen. Bei den 5-, j�-,nbsp;w5- und ��-Stammen, die �berhaupt keine Maskulina haben,nbsp;ist nur -� verhandenDas Femininum der �-Stamme hat gleich-falls nur -�. Wie fest im BewuBtsein sonst bestimmte Endungennbsp;f�r Mask.-Neut. und Fem. der a- und �-Klassen liegen, zeigt sichnbsp;auch beim Dat. Plur. attam, hairtam�qin�m.

Die Ansetzung einer indogermanischen Endung -�m, die eine Ablautform^ zu -dm (ai. (ved.) vrkam, d�vdm, gr. ^vyMV, lat. demn,nbsp;fabrum, inschriftl. Romanom, osk. Nuvlan�m) sein soil, entbehrtnbsp;jedes Haltes. Auch innerhalb des Germanischen ist, trotz dernbsp;Behauptung R. Kogels, PBB 14, 114, daB altsachs. kinda,

friunda, Hrodbertinga � u. a. Genitive auf -em seien, keine Spur aufzudecken. Also niuB mit einer spezifisch gotischen Neuerungnbsp;gerechnet werden. Van Helten, PBB ly, 570 //. vermutet dahernbsp;eine qualitative Angleichung an das -e- der Endung -eso beimnbsp;Gen. Sg. Diese Auffassung scheint mir unannehmbar, erstensnbsp;weil das Gotische jedes kurze � (auBer vor r, h, und hv) in � verwandelt batte und deshalb eine qualitative Angleichung vonnbsp;vornherein unm�glich ist, und zweitens ist die Annahme einernbsp;Beeinflussung eines Endungsvokals durch den Vokal einer Mittel-silbe immerhin bedenklich. Unter diesen Umstanden wird es wohlnbsp;nicht ungerechtfertigt sein, einen neuen Versuch zu einer Er-klarung zu machen

AuBer beim Gen. Plur. kommt das lange -� im Gotischen in den folgenden Fallen vor: beim Instr. der Pronomina }�, hv�\nbsp;beim Dat. der Indefinita hvamm�h, hvarjamm�h, ainumm�him',nbsp;bei den Adverbia n�, hvadr�, hidr�, jaindr�, swar�, siml�, hisunjan�nbsp;[hv�h �jedenfalls�) und den Konjunktionen sw�, unt�, Hnd�. Wasnbsp;den Ursprung der Endung bei den Pronominibus betrifft, gehennbsp;die Meinungen auseinander. In ]gt;� und hv� sehen die meistennbsp;Forscher einen alten [nominalen ? Instrumental (vgl. ved. vrkd,nbsp;ai. Adv. �paged] daneben ther. rfj-�e �hier�, lak. nrj-noxa)^, wasnbsp;wohl das Richtige trifft�, zumal die syntaktische Funktion sichnbsp;mit der Bildung deckt. Bei hvamm�-h herrscht noch vorlaufignbsp;keine Uebereinstimmung. Paul, PBB 2, 239 ff., M�ller ebd. 7, 490,nbsp;Bremer, ebd. ii, 35 f., Kluge, Urgermanisch halten es f�r einennbsp;Ablativ; J. Schmidt, Festgrufi an B�htlingk, S. 3, Collitz, BB. 17,nbsp;13 ff. f�r einen Dativ; Streitberg, Urgerm. Gra'm'm. S. 269, Bethgenbsp;bei Dieter, Laut- und Fomienlehre, S. 554, u. a. f�r einen Instrumental. Es ist sogar bei vielen �berhaupt noch die Frage, ob got.nbsp;hva�mma [daga usw.] und hva'mm�-h auf dieselbe Urform zur�ckge-hen. Streitberg, Got. Elementarbuch �. S. 171 bemerkt folgendes

� S. Schl�ter, Untersuchungen zuy Gesch. d. aHsc�chs. Sprache, S. 108; Collitz, Bezz. BeitrSge 17, 13.

In der Zeit zwischen der i. und 2. Korrektur bin ich zufallig auf einen wichtigen Aufsatz von Brugmann [IF 33, 272�284) gestoBen. Seine Beweis-f�hrung deckt sich z. T. �berraschend mit der meinigen, aber das Endergeb-nis ist verscliieden, da er den Gen. PI. auf c aus einer denominativen Adjek-tivbildung auf -�ia- herleitet.

�berden Dativ: �Der got. Dativ ist ein Mischkasus. Er entspricht formell dem idg. Instrumental (a- und maskul. �-Stamme), demnbsp;Lokativ (fem. i-; u- und kons. Stamme aller Genera) und demnbsp;Dativ (�-Stamme), niemals jedoch der idg. Ablativbildung�. Dasnbsp;mag wohl zum groBen Teil richtig sein, schlieBt aber nicht aus,nbsp;daB durch syntaktische Verschiebung ein ablativischer Sinn auchnbsp;im Dativ enthalten sein kann, oder daB sogar Adverbia kasuellenbsp;Funktion �bernehmen. Es ist ja allgemein bekannt, daB dasnbsp;Baltisch-Slavische den Ablativ als Gen. benutzt. WackernagePnbsp;{Mel. de Saussure, 125 ff.) hat fiir den Gen. Sg. der o-Stammenbsp;(gall. Gen. Segomari, ir. Gen. maicc) im Italo-keltischen AnschluBnbsp;an indo-iranische Adverbia auf -I hergestellt, indem er Wen-dungen wie multi facio mit ai. sami kr verband. Aehnlich laBtnbsp;Sommer � die Form peregr� neben peregri (Lokativ) durch Adverbianbsp;auf -� beeinfluBt sein. Solche formelle und syntaktische Verschie-bungen sind also in den indogermanischen Sprachen nichts AuBer-gew�hnliches. Ob nun auch in den gotischen Adverbia auf -�nbsp;wie siml�, swar�, bisunjan�, hidr�, jaindr� und hvadr� ein Ablativnbsp;auf -�d � (vgl. lat. facilum�d �facillim�, ai. pagcdt gt; *posk�d)nbsp;Oder ein Lokativ auf -dm � stecke, ist auch immer noch einenbsp;Streitfrage. Eng mit den Adverbien auf -� hangen die auf -�nbsp;zusammen: galeik�, sintein�, usdand�, andaugj�, allandj�, !gt;iubj�,nbsp;sniumund�, unw�nigg�, ussmd�, sundr�, hvapr�, jainPr� usw.,nbsp;Worin die meisten Forscher einen idg. Ablativ auf -od sehen (vgl.nbsp;got. [Prap.] undar� = ai. adhardd\\z.\. Gnaiv�d). Um die Erhaltungnbsp;der Lange zu erklaren, haben einige, wie z. B. Jellinek, zu dernbsp;Hypothese gegriffen, daB das -d erst spat abgefallen sei und dennbsp;vorausgehenden Vokal vor Verk�rzung gesch�tzt habe. Dagegennbsp;Streitberg und Collitz Wie dem auch sein mag, jedenfallsnbsp;sehen die Endungen -� und -� bei Formen wie swar�, siml�, viel-leicht auch bisunjan�, piubj�, sniumund� wie erstarrte Kasus vonnbsp;Substantiven aus. Sie k�nnen allerdings auch �bertragen sein.nbsp;Besonders bei W�rtern wie swar� und siml� ist mit einer Ueber-tragung des -� aus den Adv. hv� und p� zu rechnen. Ob man nunnbsp;Bie Erhaltung einer urspr�nglichen idg. Endung oder eine Neu-

bildung annimmt, muB man doch f�r die o-Stamme, denn bei diesen allein sind die betreffenden Endimgen in erster Linie zunbsp;suchen, einen vorgermanischen Kasus auf -� und -� voraussetzen.nbsp;Bei dem Kasus (ob Instr. oder Abl.) auf -� war ein Anhalt an demnbsp;Instr. des Pronomens P�, hv�, wie auch im Westgermanischennbsp;(ahd. tagu, wordu � diu) gegeben. Im Altindischen besteht auchnbsp;ein Zusammenhang zwischen vrkena-tena. Solche alte BeeinfluS'nbsp;sungen lassen sich auch sonst nachweisen: Ai. gen. sg. masc.nbsp;d�vdsya nach tdsya] instr. sg. fem. s�nayd-tdya; im Griech. gen. sg.nbsp;�eolo-rolo, nom. pl. OeoL-toi, gen. pl. ^mgaMv-tdayv, loc. pl. deo�ai-rolai; im Latein. nom. pl. lupi nach isi�, gen. pl. bell�rum-ist�rum, mensarum-istdrum, loc. pl. bell�s-istis, abl. sg. bell�(d)nbsp;nach ist�d.

Ich nehme nun an, daB auch im Gotischen ein Instr.-Abl. *dag�, *waurd� einst bestand Adverbia wie siml�, swar� k�nnennbsp;wohl erstarrte Reste davon sein (siehe j edoch oben). Das Alt-hochdeutsche und Altsachsische haben den (Abl.-) Instrumentalnbsp;auf -� (vgl. got. galeik� usw.) bei Substantiven der o- (und i-)-Deklination erhalten. Wie im Gotischen sind auch im Althoch-deutschen Instrumentalformen zu Adverbien geworden. Vgl.nbsp;ahd. allu werku �summopere'; diu dingu, disu dingu �hactenus;nbsp;diu mezu 'quemadmodum� u. a. m. (Vgl. auch hiutu �heute�, hiuriinbsp;�heuer�). Im Altnordischen kommt der alte Instrumental aufnbsp;-5 nur noch beimDativ des Adjektivs (z. B. blindu, -o; vgl. anorw.nbsp;hu) und vereinzelt bei den Substantiven� vor. Die andere Formnbsp;auf -� ist wie im Gotischen auch auf den Instr. des Pronomensnbsp;beschrankt {hv�, pv� [hat sein v vom Fragepronomen bezogen]).nbsp;Im Altenglischen fungiert der alte Lokativ auf -i als Instrumentalnbsp;(adverbial: thys geri �homo�, aengi thinga �quoquomodo�, gihuuelcinbsp;uuaega �quocunque�, sume daeli �partim�; hraecli �amiculo�,nbsp;gaebuli �aere alieno�) i�. Die Instrumentalform auf -� ist nur nochnbsp;beim Pronomen (hwon [h�], pon) erhalten. Das Altfriesische hatnbsp;den Instr. noch beim Pronomen; thiu, hwiu.

Sievers, PBB 8, 324 ff.: Dieter, Laut- u. Formenlehre S. 554, 666, Vgl. H. Ehrlich, Zur idg. Sprachgeschichte, S, 70.

auch als Gen. PI. aufgefaBt; ob mit Recht, wird sich schwer entscheiden lassen. Die Frage ist nur, ob das -� in diesen W�rternnbsp;zum alten (Abl.-) Instrumental gehore, oder ob es wirklich dienbsp;neue Genitivendung sei. Die enge Beziehung zwischen (Abl.-)nbsp;Instrumental und Gen. PI. ist auch sonst im Germanischennbsp;bezeugt. Im Altsachsischen stehen nebeneinander: �daru uuordunbsp;gibiodan (Hel. 3208) �� gib�d torohtero t�cno (Hel. 5943); that henbsp;them uueroldcuning sprdcono gesp�ni endi spahun uuordun (Hel.nbsp;2719). Im Althochdeutschen hat dieser Genitiv Plural in dennbsp;adverbialen Bestimmungen der Art und Weise stark um sichnbsp;gegriffen ^0, besonders bei worto: kurzero worto l�ren, wlsero wortonbsp;warn�n, gahero worto sprechen, frenkisgoro worto sag�n; pet�nonbsp;pittiu �prece posco�usw.; dann in Wendungen wie sul�cheronbsp;dato �auf solche Weise�, managero dingo 'in mannigfacher Weise�nbsp;giborganero werko �im verborgenen�. Im Angelsachsischen f�hrtnbsp;Behaghel^^ aus dem Beowulf 3209 an; swa s� secg secgendenbsp;wees Idbra spella. Dieser instrumentale Genitiv wird wohl nachnbsp;Behaghel durch Verschiebung der Gliederung aus dem partitivennbsp;hervorgegangen, muB aber alt sein, da er gleichzeitig im Angelsachsischen, und Althochdeutschen belegt ist. Ob das Gotischenbsp;ihn auch gekannt habe, laBt sich nicht nachweisen, wie so manchesnbsp;andere, dessen einstige Existenz aber �ber allen Zweifel erhabennbsp;ist. Es d�rfte dasselbe auch hier der Fall sein. Ein instr. Genitivnbsp;ist allerdings auch im Gotischen verhanden, besonders bei Verbennbsp;des F�llens^�: gr�dagans gas�pida piup� 7tsiv�ngt;rag �v�nkrjasv aya-'hc�v. Daneben der instr. Dativ^^; hvapr� pans mag hvas gas�pjannbsp;hlaibam n�hsv tovtovg �vv�)jaerat tig %0Qxdaai oQtwv. Vgl. gairnidanbsp;sad itan haurn� ensddpisi �yoQraaamp;fjrai �x rc�v xsQan'm'. �Danachnbsp;konnte sich bei Verben, die sonst mit dem Instr. verbundennbsp;wurden, auch ein Gen. einstellen� Zum Beispiel, neben demnbsp;Instrumental got. ni blandaip izwis h�ram; ags. heolfre geblonden;

Vgl. ai. s�masya jathdram prn�iham; daneben meistens der Instr.; som�na Jathdram dpipratdmandata. Ebenso bei lat. impl�re. F�r das Germ. vgl. Delbr�ck,nbsp;Synkretismus, S. 34.

Vgl. aucli den Gen. u. Instr. bei got. hr�kjan: daupeinim br�kjan �baptisma-tis uti� Sk. III, lo (vgl. ags. linenum hrceglum br�can); sonst der Gen.: ainis hlaihis br�kjan �x rov �v�g �qtov perexo/asv.

aisl. hl�le blanda erscheint der Genitiv: ags. nlda gehlonden, altsachs. haluwes giblandan So auch bei hladan, wasjan usw.nbsp;(das vollstandige Verzeichnis samtlicher hier in Betracht kommender Verba bei Delbr�ck, Synkretismus).

Bei einer Besprechung des Gen. Plur. im Gotischen darf die Form des Genitivs der ungeschlechtigen Personalpronominanbsp;(meina, peina, seina, ugkara, igqara, unsara, izwara) nicht auBernbsp;acht gelassen werden. Zwar gehen die Meinungen �ber die Endungnbsp;stark auseinander: Brugmann 2, 569, denkt an einen altennbsp;Ablativ; Janko IF. Anz. 15, 253 an einen Abl. oder Instr.; Bethgenbsp;bei Dieter, Laut- und FormenleJire S. 552 an einen Akk. PI. Neut.nbsp;oder Akk. Sing. Fem.; Walde, Auslautsgesetze S. 91 an einennbsp;alten Lokativ �bei mir� n� [vgl. lat. supern�]. Alle stimmennbsp;aber soweit �berein, daB sie in allen drei Numeri einen Kasusnbsp;der Possessivpronomina sehen. Dieser Kasus wird wohl im Singular zu suchen sein, war auch urspr�nglich kein Genitiv und wurdenbsp;schlieBlich auf den Plural �bertragen.

Man k�nnte auch zur Erklarung des Gen. Plur. auf -� vom Pronomen ausgehen, indem man die Form piz� durch Beein-flussung des Instrumentals p� entstehen und dann durch Ver-mittelung des Adjektivs auf das Substantivum �bertragen laBt,nbsp;was wohl die Beschrankung des -� auf den Gen. Plur. der Stamme,nbsp;die nur Maskulina und Neutra enthalten, leicht erklart, aber f�rnbsp;andere oben besprochenen Einzelheiten nicht gen�gt.

Der angeblich spurlos verschwundene Instrumental auf -� im Gotischen ist im Gen. Plur. der o-Stamme erhalten.

mentals mit �bernommen hat, ist es nicht immer sicher, ob wir es mit einem urspr�nglichen Dativ oder Instrumental zu tun haben.

GOTHIC SYNTACTICAL NOTES

The question here has reference to the construction of the particle ana and of the dative gastim.

Streitberg ^ construes ana as an adverb in direct connection with the infinitive usfilhan; i. e., du usfilhan ana = �to burynbsp;[strangers] in�.

To this interpretation G. W. S. Friedrichsen in his scholarly monograph The Gothic Version of the Gospels (London, 1926)nbsp;objects 2 on the ground that the verb (us)filhan elsewhere alwaysnbsp;governs the accusative, not the dative case.

But there is no reason for construing the dative gastim as the object of the infinitive usfilhan. The dative gastim, it seemsnbsp;to me, is best construed as having the same function as thenbsp;Greek dative rotg i�voig which it translates; i. e., rjy�qaaav r�vnbsp;^ygov Tov usQa/xecog slg ra(prjv rolg ^�voig, �They bought thenbsp;potter�s field for a burial for strangers� = �to bury strangers in�.nbsp;Gothic du usfilhan ana gastim.

Friedrichsen construes the particle ana [gastim] �in respect of [strangers]� as a preposition, apparently because he believesnbsp;the dative gastim to be the object of the verb usfilhan. But thenbsp;fact that we may assume the dative gastim to be a dative ofnbsp;interest exactly equivalent to the Greek dative xoig i�voignbsp;renders his contention highly improbable.

The very fact that Friedrichsen raises the question as to the rection of the Verb usfilhan tempts one to believe that he has here

� Cf. p. 140: nbsp;nbsp;nbsp;. .taking ana closely with gastim, and not, as Streitberg

(see W. B.), with du usfilhan: (us) filhan governs the accusative case, not the dative ...�

been led astray by the English translation �for burying strangers in�. But the infinitive usfilhan � is a verbal substantive and hasnbsp;no voice, corresponding exactly to the Greek verbal noun ralt;prjvnbsp;�burial� (active or passive); i. e., aig racprjv tolg ^evoig = dunbsp;usfilhan gastim �[the field to be used] for a burial for strangers� =nbsp;�for strangers to be buried in�.

Streitberg does not anywhere discuss the dative case gastim but there is no reason to believe that he considers it as the objectnbsp;(instrumental dative) of the infinitive usfilhan, as Friedrichsennbsp;implies.

Finally, as to the troublesome particle ana, which Friedrichsen falsely construes as a preposition. As an adverbial modifier of thenbsp;verbal substantive usfilhan the particle ana is used in its literalnbsp;sense of location, denoting the field where strangers are buried;nbsp;cf. Eng. �for burying strangers in�. This is a clear case of annbsp;addition (inflation) for the sake of explicitness. On the othernbsp;hand, if we construe ana with Friedrichsen as a preposition innbsp;connection ysti�h gastim (i. e., ana gastim = �in respect of strangers�) the inflation tends rather to confusion than to explicitness,nbsp;for ana in this sense occurs far less frequently than in its primarynbsp;sense of location.

Streitberg ^ mentions the fact that the Greek present tense (aside from the historical present) is sometimes represented bynbsp;the Gothic preterite and quotes one example: John XIV, 9,nbsp;swalaud melis mip izwis was; xoaovtov xqovov fie O' vfjimv eifitnbsp;�Have I been so long a time with you?�

Streitberg makes under this heading no attempt to explain the discrepancy between the Gothic and the Greek tense. It willnbsp;be seen from the following that the Gothic present tense sometimesnbsp;represents the same time function as the Greek present tense and

� For the reading du usfilhan Friedrichsen (loc. cit.) suggests du usjilha. There is no justification for this emendation; the reading of the text is clearlynbsp;du usfilhan. Apparently Friedrichsen has made this emendation to fit his interpretation of ana as a preposition, inasmuch as usfilh is a noun and as such couldnbsp;not properly have an adverbial modifier (cf. du usfilha ana �for a burial in�),nbsp;which is not true of the verbal substantive usfilhan (cf. du usfilhan ana �to benbsp;buried in�).

sometimes does not, a fact which Streitberg should have brought out under the category in question.

The example which Streitberg quotes, belongs to the category of the duralive perfect (i. e., time begun in the past and continuingnbsp;in the present) and therefore the Gothic preterite was does notnbsp;violate the function of the Greek present eliii. Since the verbalnbsp;action involves two points of time (the past and the present),nbsp;either one of these points of time may be utilized to expressnbsp;the durative nature of the verbal action; cf., e. g., the Greeknbsp;present tense ei/ti with the German (Luther) present �So langenbsp;hin ich bei euch?�, and the Gothic preterite tense was with thenbsp;English perfect �Have I heen so long with you?�

Further examples of this nature are: Luke XV, 29, swa filu jere skalkinoda pus;Toaavraenj dovlsvco aoi] quot;Thesemany yearsnbsp;do I serve thee� (Goth, skalkinoda = �have served and do nownbsp;serve�); Mark VIII, 2, unte ju dagans prins mip mis wesun^]nbsp;on T��] ^fi�Qai Toets uiQoajx�vovalv f.101; �Because they continuenbsp;with me now three days� (Goth, wesun = �have been and arenbsp;now�).

In certain cases, however, the Gothic preterite, used in the function of a durative perfect, violates the present time functionnbsp;of the Greek verb. In such cases no adverbial element of timenbsp;is expressed and there can, therefore, be no doubt that the Greeknbsp;present represents time without reference to action begun in thenbsp;past. The Gothic preterite, therefore, here represents an expansionnbsp;of the Greek present into a durative perfect.

John XIX, 4, ei witeip Patei in imma ni ainohun fairina bigat] ha yvmre on iv avtqj ovdsftiav alrlav e�glmco] �That yenbsp;may know that I find no crime in him� (Goth, ni bigat = �havenbsp;not found and do not now find�). For evQ'axm in this phrasenbsp;the Gothic scribe elsewhere regularly uses the present tensenbsp;bigita (cf. John XVIII, 38; XIX, 6).

� Cf. Luke XV, 31, pu seinteino mip mis [wast jah] is; av mivioTS /xtr i/xov Cl; �Thou art ever with me�. The interpolated preterite form wast is evidentlynbsp;based upon the Latin [tu mecum] fuisti [semper et es], but the Latin fuistinbsp;(= Grk. f() represents the principle involved.

� Textual corruptions (due to confusion with other passages or to the influence of the Latin) are, of course, excluded from my discussion; cf., e. g., John VIII,nbsp;45, rodida: Ie/oj; John VI, 37, ga[: SiStaaiv, etc. See Friedrichsen, op. oit., pp. 81,

JohnXIV, 31, eiufkunnai so manaseps patei ik frijoda attan meinana) �va yvw � a�ajxoQ on ayaTzm x�v nax�Qa', �That the worldnbsp;may know that I love the Father�� (Goth, frijoda � �I havenbsp;loved and do now love��).

Cases of this kind must be clearly distinguished from genuine perfect presents (i e., where the Goth, preterite representsnbsp;present time as a result of an action begun in the past); cf. Mat.nbsp;VI, 2, 16, andnemun mizdon seina; aneyovaiv xov fiiaddv avxmv;nbsp;�They have their reward�� (Goth, andnemun = �They have receivednbsp;and therefore now have��); cf. haband (Mat. VI, 5) = an�xovai.

The foregoing analysis shows that the Gothic preterite may represent the Greek present tense under three different conditions : i) when the Greek present functions as a durative perfect,nbsp;2) when the Gothic has expanded the Greek present into anbsp;durative perfect, and 3) when the Gothic uses a preterite (= perfect) present for the Greek present

The whole passage reads: du pammei arhaidja usdaudjands hi waurstwa sei inna uswaurkeip in mis in mahtai] elg o real noni�)nbsp;ayagt;viti�[iEvoQ naxa xxjv �v�qyeiav avxov xxjv ivsQyovjjLlvrjV iv �fiol ivnbsp;dwdfisf, �Whereunto I labor also, striving according to hisnbsp;working which worketh in me mightily��.

The question under consideration concerns the relative pronoun sei, which should, according to the Greek original, refer backnbsp;to waurstwa and therefore be in the neuter (patei, is ei andnbsp;not in the feminine gender (sei).

Streitberg� meets this apparent incongruence in gender by deriving the form sei from an adverb IE gt; Lat. s� (as innbsp;sl-c �so� lt; *sei-ke), in which case the relative form sei mustnbsp;be considered as genderless (cf. OHG so; ON sem, es: er, etc.)nbsp;and may therefore refer back to the neuter antecedent waurstwa.

Professor Curme on the other hand, maintains that the relative sei represents the regular feminine form, referring not

� In all other cases of the genuine preterite present the Gothic is in accord with the Greek; cf. Mark I, ii. galeikaida: evS�xrjaa; Luke XVI, 4, andpahta:nbsp;fyv(ov, John VIII, 52, ufkunpedum: tyv�xK/iev; John XVII, 7, tifkunpa: eyvoiv, etc.

to the preceding neuter wawstwa but to the following feminine mahtai. This contention he supports by the fact that in Gothic,nbsp;as well as in Greek, the relative often preceded its antecedent.

Professor Curme further contends that the Gothic scribe has entirely reconstructed the passage so that it reads (according tonbsp;Curme�s translation): �Whereunto I labor, striving with energynbsp;by the aid of that power (mahtai) which works within me�.

I agree with Professor Curme on both these points and that Streitberg�s recourse to an adverbial relative sei (see above) isnbsp;unnecessary.

Yet at the same time the difference in sense between the Greek original and the Gothic translation indicates not a deliberatenbsp;reconstruction of the original Greek on the part of the Gothicnbsp;scribe, as Curme implies, but simply a case of corrupted text

In reconstructing his relative clause (sei inna uswaurkeip) in place of the Greek participial construction (irjv �veQyov/x�vrjv)nbsp;the Gothic scribe apparently became confused as to the correctnbsp;antecedent of his relative and falsely construed it with mahtainbsp;instead of with waurstwa. This confusion may be attributednbsp;to the fact that in Gothic the relative often precedes its antecedent.nbsp;The omission of the equivalent for the Greek avrov (= Goth, fs)nbsp;is further evidence of confusion and of a corrupted text.

With this type of attraction in gender compare II Cor. V, 17 /abai hwo in Xristau niuja gaskafts; ei rig tv Xgiarcp, aaivr) miaig',nbsp;�If any man is in Christ, he is a new creature�.

It is here perfectly evident that the feminine pronoun hwo (instead of hwas = Gr. rig) is due to the influence of the followingnbsp;feminine substantive gaskafts

Now, since the Gothic relative often preceded its antecedent it is most reasonable to assume that in our passage in questionnbsp;the Gothic scribe falsely construed his relative with the followingnbsp;substantive mahtai instead of with the correct antecedent waurstwa, resulting in an attraction of gender similar to that existingnbsp;in IlCor. V, 17 (i. e.., hwo for hwas).

This is also Streitberg�s view [Got. Eleinentarb.^-^, � 346, i, Anm.; Got. Bibel^, p. 488) but he is unable to explain how the corruption came about.

Cf. likewise John XVII, 3, soh pan ist so aiweino libains �This is the eternal life�; where, however, the Gothic feminine soh corresponds to the Greek femininenbsp;nvTtj [Jj lunv 7) aitaviog

IV. }o samon frisaht ingaleikonda af wulpau in wulpu; xi]v avvrjv six�va ixsraixoQxpovusOa and d�i�)]Q sig d�^av, �We are transformednbsp;into the same image from glory to glory�, 11 Cor. Ill, i8.

At first blush one is tempted to view the Gothic accusative /o samon frisaht, used with the medio-passive ingaleikonda, asnbsp;an imitation of the Greek accusative avirjv etx�va andnbsp;not as a native Gothic construction.

However, when we consider the fact that the Gothic passive in-galeikonda is compounded with the prefix in-, it is quite possiblenbsp;to construe the accusative po samon frisaht as due to the influencenbsp;of this prefix in-, denoting the goal of the verbal action. Thenbsp;prefix in- (= Grk. /xsra-) here denotes a change from onenbsp;condition into another. As there is no essential difference betweennbsp;a verbal prefix and a preposition we might construe the accusativenbsp;Po samon frisaht as object of in- in close connection with thenbsp;verb whereas in our Modern Germanic languages a fullynbsp;developed prepositional construction must be used; cf. Eng. �Wenbsp;are transformed into the same image�. Germ. (Luther) �Wirnbsp;werden verklaret in dasselbige Bild�, etc.

It seems to me, therefore, that it is a moot question as to whether the Gothic scribe in preserving the Greek constructionnbsp;here did violence to his native Gothic idiom.

V. Galeik ist barnam sitandam in garunsim jah wopjandam an par an par is (ngoafoovovaiv rolg �r�qoig)', �It is like unto childrennbsp;sitting in the market places and calling out one to the other��,

Since the form an Par is nominative and yet stands in apposition with the dative barnam. We evidently have here a case of anacoluthon. This lack of strict grammatical agreement isnbsp;due to a constructie ad sensum. The reciprocal relation �one ��

So Gerhard H. Balg in his translation (1895) of Braune�s Got. Gramm.*, �Explanatory Notes� (p. 132), under this passage: �. . .this acc., with a pass,nbsp;vb., is an imitation of the corresponding Greek passage�.

** Cf. anakumbjan mip: mipanakumhjan with dative rection (Luke V, 29, wesun mip im anahumbjandans: Mark II, 15, mipanakumbidedun Jesua)', similarly bisitan: sitan hi with accusative rection (Luke I, 65, ana allaim agis paimnbsp;bisitandam ina: Mark III, 32, jah setun hi ina). Cf. Germ, quot;er durchfliegt die Luft�;nbsp;�er fliegt durch die Luft�.

the other� usually occurs in connection with a finite verb, anHr (nom.) standing in partitive apposition with the plural subjectnbsp;and anpar- (oblique case) standing as object of the verb; cf.,nbsp;e. g., anpar an parana munands sis auJiuman, �Each accountingnbsp;the other better than himself� (Ph. II, 3); timrjaip ainlvwarjizuhnbsp;anpar an par ana, �Build ye each other up� (Th. V, ii); sijuninbsp;anpar an par is lipus, �We are members one of the other�nbsp;(Eph. IV, 25).

In our passage in question the participial construction (si-tandam . . . wopjandam) is in sense equivalent to a relative clause, hence the nominative anpar is felt to be in partitive appositionnbsp;with the subject of a finite verb, just as in the cases quotednbsp;above; cf. galeik ist barnam poei sitand ... jah wopjand anparnbsp;anparis, �Like unto children who sit ... and cry out one tonbsp;the other�.

That the nominative anpar in the original passage with participial construction is good Gothic, is shown by the fact that in the original Greek we have nothing corresponding to anpar.nbsp;It is unfortunate for our discussion that in the parallel passagenbsp;(Luke VII, 32), where the participial construction (sitandam jahnbsp;wopjandam) occurs, the Gothic translator avoids the anpar �nbsp;anpar- construction (cf. wopjandam seina niisso, TiQoacpovovaiv

VI. Jah jagino in izwara, ei galaubjaip, unte ni was jainar; xai xaiQiO di vf-iaQ, �va moTevarjrs, ozi ovxnbsp;nbsp;nbsp;nbsp;�xel] �And I

am glad for your sakes that I was not there, to the intent that ye may believe�, John XI, 15.

The clause unte ni was jainar is the object of the verb fagino, hence we should expect in the place of the conjunction unte eithernbsp;patei or ci which are regularly used after verbs of �rejoicing�nbsp;and of �wondering�; cf., e. g., Luke X, 20, ni faginop, ei..:nbsp;firj xaigsre on; John XIV, 28 faginodedeip ei: �%aQrire dv on.

Why then did the Gothic scribe in our passage deviate from the normal usage of Patei or ei in favor of imte ?

The reason for this, I think, is obvious. The Greek on may either serve as a conjunction introducing a substantive clause

Cf. Streitberg, Got. Elemenlarh.^~^, � 354, 3. Besides patei and ei the dative form pammei is also sometimes found after faginon] cf. faginop mip mis pammeinbsp;bigat lamb mein] ovyxaQfin on evQov xb ttq�^bt�v /xov, Luke XV, 6.

(= Goth, patei: ei �that�) or represent a causal conjunction (= Goth, unte �for which reason, because�). The passage may,nbsp;therefore, be interpreted as meaning �1 rejoice because I wasnbsp;not there�. Had the Gothic scribe here used the regular particlenbsp;patei or ei instead of unte, the clause introduced by the particlenbsp;(pat) ei (i. e., (pat) ei ni was jainar) could have been construednbsp;as the object of the verb galauhjaip instead of the verb fagino;nbsp;i. e., ei galauhjaip (pat)ei ni was jainar �That ye may believenbsp;that I was not there�.

To avoid such a misinterpretation the Greek on is rendered by the particle unte which (unlike the particle Imtei ; ei) cannotnbsp;properly be construed with the verb galauhjaip but only withnbsp;the verb jagino; i. e., fagino . . . unte �I rejoice because (= that)�

As has been shown above, the Greek 8n may correspond either to Gothic patei:ei: Pei �that�, introducing a substantive clause,nbsp;or to Gothic unte �because, for�.

I have found one passage in which on account of this double meaning of the Greek conjunction on the Gothic scribe evidentlynbsp;used unte for Patei: ei: pei, namely Luke I, 58: jah hausidedun . .nbsp;unte gamikilida jrauja armahairtein seina; ual i)xovaa ... ortnbsp;�fxeyaXvvev uvqioq to eXeog avrov; �And they heard that the Lordnbsp;had magnified his mercyquot;.

There is no possibility of construing unte � on here (after a verb of �hearing� 1�) as having a causal sense, yet Streitbergnbsp;makes no mention of this misuse^quot; of the particle unte eithernbsp;in his Got. Elementarh.^~^ or in his Got. BibeP (Voc. under unte).

Since this use of unte == on in the sense of �that� introducing indirect discourse does not occur elsewhere, we may best explainnbsp;its occurrence here as due to the influence of the Latin qiUa,

In our modern versions of this passage we avoid the ambiguity in question by changing the order of the clauses so that the last clause immediately followsnbsp;the first; cf. Eng. �And I am glad for your sakes that I was not there, to thenbsp;intent that ye may believe�, Germ. (Luther) �Ich bin froh um euretwillen,nbsp;daB ich nicht da gewesen bin, auf daB ihr glaubet�.

For the usage of the conjunction patei: ei: pei after verbs of quot;hearing�, �seeing�, etc. compare Streitberg, op. cit., � 355.

Braune, however, in his Got. Gramm.^ (Voc. p. 185) gives besides denn and weil, the meaning dap, which would indicate that Braune recognizes the misusenbsp;of unte under discussion.

which also had the double sense of �because, inasmuch as�� and �that� (cf. OHG (h)uuanta, Tatian); cf. Luke I, 58 �et audieruntnbsp;. . . qida magnificavit Dominus misericordiam suam�.

VIII. nbsp;nbsp;nbsp;Jah qens meina framaldrozei in dagam seinaim; xal 17nbsp;yvvq jiiov jcQofiefirjxvla ev rah; rjfi�qaiQ avTfjg; �And my wife well

It will be noted that the Gothic here employs the comparative degree of the adjective (framaldrozei) to render the Greek positivenbsp;TiQoflsPrjnvta.

The same phrase is used in Luke II, 36 regarding the prophetess Anna �� framaldra dage managaize (nqofis^riKvla ev Tq/uegaignbsp;TioXlalg) �� but here in the positive degree corresponding to thenbsp;Greek. Why then did the Gothic scribe in the first passagenbsp;employ the comparative degree?

In the first passage the age of the wife of Zacharias is connected with the question of childbirth, so that we may construe thenbsp;comparative framaldrozei as meaning �older [than that she cannbsp;give birth to a child]� = �too old [to give birth to a child]�.nbsp;Such an idea is not present in connection with the age of thenbsp;prophetess Anna.

This usage of the comparative framaldrozei is obviously equivalent to the so-called absolute comparative; i. e., framaldrozei = �older [than women who can give birth to a child]�, in whichnbsp;a second party is implied. In the Greek this distinction betweennbsp;positive and comparative is not always observed, the idea beingnbsp;viewed (as in our passage) from the standpoint of only onenbsp;party 22;nbsp;nbsp;nbsp;nbsp;^f. Marfa Jacobis pis minnizins (rod futcQov),

IX. nbsp;nbsp;nbsp;Pairhinfeinandein armahairtein gndisunsaris, in Pammeinbsp;gaweisop unsara urruns us hauhipai; did anXdy%va �X�ovg rjfjiwv,nbsp;IV olg STUaxeipexai rjfxag dvaroXri �i vipovg] �Because of the tendernbsp;mercy of our God, whereby the dayspring from on high shall

It will be noted that there is a lack of agreement in gender between the relative pammei (= olg) and its antecedent arma-

Cf. ON ��ertu ok myklu vitrari ma�r, en J)� munir vilja . . quot;Thou art also a much wiser man than that thou shouldst wish � quot;too rvise a mannbsp;to wish�.

hairtein; instead of in Hfnmei we should expect in pizaiei fern, (in agreement with armahairtein).

The Greek phrase did anXdyyva flc'ous (� Lat. viscera miseri-cordiae �bowels of mercy�) has been recast by the Gothic scribe into pairh injeinandein armahairtein �durch erbarmendes Mit-leid� �because of pitying {= tender) mercy�, wherein a verbalnbsp;action (injeinandein) is implied. The relative pammei may thennbsp;be construed as a neuter singular referring to the act [of mercy]nbsp;implied in the present participle injeinandein; i. e., in pamniei =nbsp;�by which act [of mercy], whereby�.

In view of the Greek plural oig (which must refer back to ajiXdyyva) the Gothic scribe could not possibly have used Pammeinbsp;with reference to gudis unsaris.

Streitberg ^5 makes no reference to this type of apparent incongruence of gender (i. e., between relative and antecedent).

X. Jah gairnida sad itan haurne; ual inedv^si %OQxaa0rivai lx ramp;v xsoarmv; �And he would fain have eaten his fill of the husks�,

Regarding this phrase Friedrichsen says; �In German �er wollte sich satt essen� . . . Nothing could be more obviouslynbsp;idiomatic.�

Of course the German �er wollte sich satt essen� translates perfectly the idea expressed in the Gothic, but the Germannbsp;idiom does not correspond to the Gothic idiom. For an exactnbsp;parallel to the Gothic idiom Friedrichsen might have quotednbsp;the English �he would fain have eaten his jill (adj.) of�. Thenbsp;Gothic adjective sad (sap) is not a predicative adjective (masc.nbsp;sing., cf. sadana), as is the German satt, but a substantivizednbsp;neuter = �a sufficient amount of, enough�. We find severalnbsp;such examples of the substantivized neuter singular of the strongnbsp;adjective in Gothic cf. piup = to dyaOov, ubil = to novrjQov

Cf. Lat. �per viscera misericordiae Dei nostri, in quibus visitavit nos, oriens ex alto�.

Cf. Streitberg, Got. Elementarb.^��, � 236, quot;Inkongruenz der grammatischen Genera�.

It will be noted that the adjective mci denotes the idea of qiiantity and therefore easily lends itself to the collective idea, �nbsp;hence the substantive usage. This is of course a Common Germanicnbsp;idiom; cf. the Gothic substantive usage of sad with the substantive usage of the adjective for �enough� in all the Germanicnbsp;languages (Eng. �enough of the husks�. Germ, �der Trebernnbsp;genug�), similarly ON fdtt manna �few men�, mart manna �manj^nbsp;men�. The Gothic adjective sad in our passage is thus syntactically equivalent to the substantive filu with the partitivenbsp;genitive.

We find no cases of manag �much� plus the partitive genitive (parallel to ON mart manna), presumably because the substantivenbsp;filu is thus used in place of manag. But the substantive filu is,nbsp;of course, not capable of comparison and therefore in thenbsp;comparative degree �more� we find the neuter singular of thenbsp;adjective managizo substantivized in exactly the same way asnbsp;is sad in our passage; cf. nihai managizo^^ wairpip izwaraizosnbsp;garaihteins^ Pau pize bokarje; iav jurj TisQiaosvar] vjmv rj �ixaioam]nbsp;Ttlelov xvjv YQafijiax�mv, Matt, V, 20.

This latter passage illustrates the point in question, viz. that the reason for the substantivization of the neuter singular ad-sective (sad) is due to the idea of quantity � i. e., a collectivenbsp;substantive (cf. German genug, viel, mehr).

The Gothic datives Jacobau and Johanne correspond to the Greek accusatives ^Idncojov and �Iwawr/V, direct objects ofnbsp;inoi'rjaev (gawaurhta) in verse 14; i. e., �He made James andnbsp;John to be his disciples�.

These Gothic datives obviously represent a scribal error for the accusative. Directly following the passage in question wenbsp;read: jah gasatida im namna Bauanairgais: �And gave to themnbsp;the names Boanerges�. Similarly in the preceding verse (16),nbsp;jah gasatida Seimona namo Paitrus; �And gave to Simon thenbsp;name Peter�.

The Gothic datives Jacobau: Johanne, therefore, obviously represent a later contamination with the dative constructionnbsp;after gasatida (cf. im: Seiniona hence Jacobmk' Johanne). Thisnbsp;anacoluthon was further favored by the fact that the verbnbsp;gawaurhta, which governs these two names (Jacobau: Johanne),nbsp;is so far separated from them.

XII. Ei sahwamh izei usqimip izwis, puggkei} hunsla saljan guda; iva Tiag o o.noKXslvaQ v/xag dd^rj Xat qs lav nQoacpEQeiv repnbsp;deep ; �That whosoever killeth you shall think that he offerethnbsp;a service to God�, John XVI, 2.

The Greek singular larpeiav �religious service� is here rendered by the Gothic plural�� hunsla �sacrifices�. There is no reasonnbsp;to assume that the form hunsla represents a dative (instrumental)nbsp;singular after the verb saljan �to sacrifice�, especially since thenbsp;verb saljan elsewhere requires an accusative rection, just asnbsp;does atbairan = npoaepeqeiv which is the Greek verb for saljannbsp;in our passage. The form hunsla undoubtedly represents annbsp;accusative plural.

But why should the Gothic scribe have translated the Greek singular Xarge-av by a plural hunslal There is no reason fornbsp;considering hunsla as a pluralis tantum ��, especially since thenbsp;singular form of this word is regularly used to translate thenbsp;Greek Qvaia �sacrifice�.

The plural form is, it seems to me, best explained as due to the plural idea inherent in the subject sahwazuh (= nag 6)nbsp;�whosoever� � �all who�.

The subject sahwazuh �whosoever� (= nag oj is a collective pronoun; although singular in form it is plural in sense, denoting

2* There is no reason to believe that Wulfila himself used the dative case for these two names; the contamination was most probably due to a visual errornbsp;on the part of a later scribe. Streitberg does not mention {Got, Bibel^, undernbsp;this passage, p. 173) this contamination nor is it discussed by Friedrichsennbsp;(op. cit.).

Cf. ICor. X, 20, ak patei saljand piudos; on a O�ti ra iQvr]; �The things which the Gentiles sacrifice�.

Nor does Streitberg include this plural form in his list of pluralia tantum {pot. Elementarb.^-^, � 238). To this list I should add haurja = avOqaxuiv �coal-fire�, John XVIII, 18.

a number of individuals who perform the same act. The Gothic scribe, contrary to the Greek, viewed the verbal action fromnbsp;the standpoint of a plural subject (constructie ad sensum))nbsp;consequently the object (XaxQclav) of the verb was pluralizednbsp;(hunsla) corresponding to the several acts on the part of thenbsp;subject.

The Greek viewpoint is singular, i. e., �Every one who killeth you shall think that he offereth his service to God�; the Gothicnbsp;viewpoint is plural, i. e., �All who kill you shall think that theynbsp;offer their services to God�. For this type of fluctuation betweennbsp;a singular and a plural object compare German �Sie nahmennbsp;den Hut ab� and English �They took off their hats�.

SCANDINAVIAN-POLISH RELATIONS IN THE LATE TENTH CENTURY

The contacts between Scandinavians and Slavs in Russian territory during the ninth and tenth centuries are treated in respectable detail by both Russian and Norse sources. The Russian Primarynbsp;Chronicle, composed about iiiband covering the period 850�iiio,nbsp;preserves a coherent tradition of the establishment of Scandinavian domination over the Slavic and Finnish tribes of Northern Russia by three totally legendary brothers, ostensibly ofnbsp;Swedish origin This tradition, covering the extension of Scandinavian overlordship down the Dnieper valley toward the Blacknbsp;Sea, reflects the foundation of Scandinavian trading posts andnbsp;colonies at strategic points along the Russian watercourses andnbsp;the gradual consolidation of the peaceful but disunited Slavicnbsp;tribes into a state dominated by a ruling class of immigrantnbsp;Scandinavian warrior-merchants. Their voyages of trade andnbsp;piracy had carried them as far as the Caspian, Byzantium, andnbsp;the northern coast of Asia Minor before 900. The activities ofnbsp;these Norse colonists are abundantly attested both by archaeological remains and by the testimony of contemporary orientalnbsp;observers Intermarriage with daughters of the numerically

1 JlaspeHTtescKaa JleTonnci. ed. E. F. Karski (Leningrad, 1926), 19�21. Ci. W. Thomson; Relations between Ancient Russia and Scandinavia (Oxford 1877),nbsp;passim; V. Klyuchevski: Kypc PyocKoS HcTopnn I (Moscow, 1904), 150�173;nbsp;L. Niederle; Slovansk� StarozitnosH I ^ (Prague 1925), 90�126, and for summarynbsp;of literature on Varangian question, ibid., 278�281; Manuel de l�Antiquit� Slave Inbsp;(Paris 1913), 198�207.

^ Cf. T, J. Arne: La Su�de et I�Orient (Uppsala 1914), 18�61, 220�231; J. Marquart; Osteuropaische und Ostasiatische StreifzUge (Leipzig 1903), 203�204,nbsp;330�353. F. Westberg: �K Anajinsy Boctouhux Hctohhhkob O Boctouho�nbsp;EBpone�,JK. M.H. II., N. S. XIII (Leningrad 1908), 364�404; XIV, i�52, D. A.nbsp;Khvolson: HasecTna lItH-,3(aoTa 0 Xoaapax, C.aaBaHax n Pyccax, (Leningrad,nbsp;1869). A. S. Cook: �Ibn Fadlan�s Account of Scandinavian Merchants on the

dominant Slavic masses rapidly deprived these energetic immigrants of their exclusively Scandinavian character. Especially after the conversion of Vladimir I (988), intensive contact withnbsp;Byzantine civilization oriented the Varangian princes more andnbsp;more toward the intellectually superior south, and restricted thenbsp;influence of their original northern habitat. The route from thenbsp;Baltic to the Black Sea via the Russian watercourses�, whichnbsp;remained open until blocked by Turco-Tatar nomads towardnbsp;the end of the twelfth century, insured notwithstanding, as longnbsp;as it was practicable, a constant influx of Scandinavian adventurers, some of whom enlisted in the service of the Russiannbsp;princes, while others went on southward to seek gainful employment on the Bosporus The Russian state thus became a recognized haven of refuge for princes whose sojourn in Scandinavianbsp;was, for one reason or another, fraught with momentary peril,nbsp;while intermarriage between the Russian and the Scandinaviannbsp;ruling families kept alive the sense of kinship uniting the Norsenbsp;princes with the house of Rurik.

The closest relations between Scandinavia and the early Russian state prevailed during the reign at Kiev of a friend and contemporary of St Olav: Yaroslav the Wise (978�1054), who wasnbsp;himself son-in-law of Olaf Skotkonungr and father-in-law ofnbsp;Harald Har�r��i�. For this period the Norse sources (e. g., Agrip,nbsp;Morkinskinna, Fagrskinna, Heimskringla, Eymundarpattf Hrings-sonar) supply abundant data as to contacts between Scandinavianbsp;and Russia. The various sagas dealing with Olaf Tryggvason alsonbsp;contain a considerable body of information (most of it legendary

Volga in 922,� Journ. Eng. and Germ. Philol. XXII (Trbana �923), 51�63. V. G. Vasilievski: 'PycoKo-BiiaaHTH�OKHa IIocjie,noBaHHa�, JleTounct ApKeorpa^naecKoSnbsp;KoMMHOcnH IX�X (Leningrad, 1893), pt. 2 M. Rostovtzeff: Iranians andGreeksnbsp;in South Russia (Oxford 1922), 210�222; �Les Origines de la Russie Ki�vienne,quot;nbsp;Revue des Etudes Slaves II (Paris 1922), 5�18).

^ ilasp. JleT., 7. Cf. A. Bugge: �Die nordeuropaisclien Verkehrswege im fr�hen Mittelalter�, Vierteljahrsschrift fiir Soz. u. Wirtschaftsgeschichte IV (Berlinnbsp;1905�1906), 245�252. A. Szelagowski: Najstarsze Drogi z Polski na Wsch�d (Cracow 1909), 9 ff.

^ Cf. V. G. Vasilievski: 'Baparo-PyccKaa n Baparo-AHranficKaa TtpysKnaa b KoHCTaHTHiionone XI b XII BeKos�, 3K. M. H. 11., CLXXVIII (March, 1875),nbsp;76�152.

� Cf. F. Braune: �Das historische RuBland im nordischen Schrifttum des X. bis XIV. Jhdts.�, Festschrift fur E. Mogh (Halle 1924), 150�196. S. H. Cross:nbsp;�Yaroslav the Wise in Norse Tradition�, Speculum IV (Cambridge, 1929),nbsp;177�197, and the sources there cited.

ii6 or distorted in transmission) concerning his sojourn at the courtnbsp;of Yaroslav�s father Vladimir I, and his entirely unhistoricalnbsp;connection with the latter�s conversion to Christianity. Thenbsp;Bjarnar Saga Hitdoelakaf-pa also preserves some remnant ofnbsp;tradition concerning the participation of a Norwegian adventurernbsp;in the conflicts of Vladimir with the nomads about 1008. It isnbsp;thus to be expected that the Norse historical compilations ofnbsp;the twelfth and thirteenth centuries might present correspondingnbsp;material concerning Scandinavian intercourse with the Westernnbsp;Slavs of the Baltic seaboard during substantially the same epoch,nbsp;in the course of which the trading centers established by Vikingnbsp;raiders along the Baltic maintained close contact with the Wendishnbsp;tribes then engaged in their perennial struggle against Saxonnbsp;expansion.

Among the Norse sources, the earliest (the Agrip and the anonymous Historia Norwegiae) contain but scanty references tonbsp;such contacts. The Fagrskinna, the two full redactions of Oddr�snbsp;Saga of Olaf Tryggvason, the J�msvtkingasaga, the Heimskringla,nbsp;and the Knytlingasaga expend considerable detail and extensivenbsp;romantic embellishment upon the relations between the Danesnbsp;and the Slavic tribes adjacent to the mouth of the Oder duringnbsp;the last half of the tenth century, especially in connection withnbsp;the foundation and history of the Danish colony at J�msborg,nbsp;on the island of Wollin. The Slavic tribes nearest this center werenbsp;obviously either Pomeranian or Polish. Before their subjectionnbsp;and absorption by German invaders, the Pomeranians nevernbsp;reached a sufficiently high intellectual level to permit the crystallization of any tradition concerning this early period. Westernnbsp;culture penetrated Poland itself only during the thirteenthnbsp;century after an interval of internal disruption in which thenbsp;memories of Mieszko I, the prince who first brought Poland intonbsp;the circle of European states, had fallen into almost completenbsp;oblivion. The Poles therefore possess no literary monument ofnbsp;historical nature in any way comparable with the Russian Primarynbsp;Chronicle, and for the second half of the tenth century, thenbsp;earliest Polish sources offer practically no information againstnbsp;w'hich the data of the Norse sagas may be checked. We are thusnbsp;limited to the evidence of Widukind, Thietmar of Merseburg,nbsp;and Adam of Bremen (together with Helmold and occasionallynbsp;Saxo Grammaticus) for material either contemporary or slightly

later and applicable as a corrective to the extant Norse texts, none of which antedates the last quarter of the twelfth century.nbsp;For the early life of Olaf Tryggvason prior to 986, Snorri Sturlusonnbsp;himself found little scaldic material to supplement oral traditionnbsp;and the scanty literary efforts of his precursors. By the timenbsp;the Norse historical sagas took shape as written literature,nbsp;Scandinavian contacts with Poland proper must also have beennbsp;severely restricted. After the death of Boleslav Krzywoustynbsp;(the Wr3nnouthed) in 1138, Poland was partitioned among hisnbsp;heirs, and the whole Baltic seaboard west of the Vistula was lostnbsp;to German conquest. From 1229 forward, the penetration of thenbsp;German military orders into East Prussia and Lithuania definitelynbsp;cut the Poles off from the sea. The coincidence of a period ofnbsp;civil strife in Norway, together with the gradual collapse of thenbsp;Polish principale, which did not recover until the coronation ofnbsp;Vladislav tokietek in 1320, provided an interval in which earliernbsp;information, unfixed by consignment to writing and uncontrollednbsp;by continuous contact between the nations concerned, couldnbsp;degenerate into a body of inaccurate legend wherein the remotenbsp;Icelandic compiler or sagnamadr could hardly be expected tonbsp;distinguish fact and romance. The Norse sources dealing withnbsp;Scandinavian-Wendish relations in the second half of the tenthnbsp;century thus merit the closest critical scrutiny before theirnbsp;content is in any wise historically acceptable.

While the Oder and the Vistula as trade routes never attained the importance of the Dnieper and the Volga, they still playednbsp;no mean part in the early mediaeval communication system. Thenbsp;finds of Arabic coins not only along the upper Oder in Silesia,nbsp;but also along the courses of the Warta and Notec, indicate thatnbsp;the Oder basin likewise fell within the range of the orientalnbsp;commercial influence which was so extensively felt in Northeasternnbsp;Russia and in Sweden in the course of the ninth and tenthnbsp;centuries �. During the Viking Age and early in the ninth century,nbsp;such Baltic trading points as Hedeby and Slesvig in Denmark,nbsp;or Birka and Sigtuna in Sweden, were the centers of Vikingnbsp;commerce. The raids of the Vikings along the south shore of thenbsp;Baltic resulted either in the capture of Wendish coastal towns

� Cf. Niederle III nbsp;nbsp;nbsp;369�374; R, Ekblom, in Arch. Slav. Philol. XXXIX

ii8 or the foundation of new Viking settlements adapted both fornbsp;trade with the interior and as intermediate stations for voyagesnbsp;toward destinations further eastward. Among these were Reric innbsp;Mecklenburg, Svarinshaugr (Schwerin), and Truso, on the easternnbsp;arm of the lower Vistula. For the contact of Scandinavians withnbsp;Polish and Pomeranian tribes, the most important center of thisnbsp;sort seems to have been a Danish settlement at J�msborg,nbsp;situated at the mouth of the Oder on the island of Wollin, apparently upon or near the site of the modern town bearing the latternbsp;name �. This colony, at least until its destruction by Magnus innbsp;1043, has the reputation of having been the outstanding Danishnbsp;settlement on the Baltic, and is correspondingly celebrated in thenbsp;Sagas �.

The related saga material is, in fact, so contradictory and so complicated with romantic accretions that one Swedish investigator has been led to deny the existence of J�msborg altogether 1�.nbsp;As will presently appear, however, regardless of the extensivenbsp;fiction surrounding this Danish strong-point, the evidence ofnbsp;Danish contact with at least one historical Polish prince throughnbsp;some Scandinavian outpost near the mouth of the Oder isnbsp;too well-established to justify such ruthless rejection of thenbsp;whole J�msborg tradition, though the archaeological remainsnbsp;gathered on the spot are remarkably scanty, as far as anynbsp;proof of lasting and organized Scandinavian occupancy is concerned.

In the earliest Norse historical compilations, J�msborg and the J�msvtkingar do not appear, apart from a brief mention ofnbsp;the settlement as Olaf Tryggvason�s winter-quarters on his return

� Cf. J. C. H. R. Steenstrup: Venderne og de Danske for Valdemar den Stores Tid (Copenhagen 1900), 25�59; Bugge, loc. cit., 238�239; A. Stubenrauch:nbsp;!igt;Untersuchungen im AnschluB an die Vineta-Frage�, Balt. Studi�n n. F. I�IInbsp;(Stettin 1897), 67�133; Niederle, op. cit. Ill, 150, and literature there cited.

� Heimskringla ed. F, J�nsson, III, 43, 44: �En er Magnus konungr kom til Vindlandz, p�, lagdi hann til J�msborgar ok vann pegar borgina, drap p�, mikitnbsp;folk ok brendi borgina ok landit vlda ut i fr�, ok gerdi par it mesta herverkiquot;,nbsp;corroborated by citation from Arn�rr Jarlaskdld, cf. F. J�nsson: Shjaldedigtningnbsp;A- 355, B 308. This destruction appears to have been only temporary, however,nbsp;as a ciuitas Vulinensis is again mentioned in 1140 (Niederle III, 149, n. 7). Hel-mold, however (Chron. Slav. ed. G. H. Pertz [Hannover r868] 15), refers tonbsp;J�msborg as long since destroyed: quot;quondam fuit nobilissima ciuitas lumneta . . .nbsp;omnes enim usque ad excidium eiusdem urbis paganicis studiis oberrarunt�.

L. Weibull: Kritiska Unders�kningar i Nordens Historia omkring Ar 1000 Lund 1911), 178�195.

from Russia in 986 The traditions of its foundation are highly conflicting. Sveinn Agason asserts that Harald Gormsson (Bla-tonn) fled before his son Sveinn Tjuguskegg into Sclavia, wherenbsp;he is said to have founded the city called J�msborg Thenbsp;Fagrskinna (second quarter of thirteenth century)^� contains thenbsp;statement that King Harald, while raiding in Vindland, causednbsp;the construction of a great fortress at a site called at J�mi, laternbsp;known as J�msborg, with the supplementary information thatnbsp;the J�msvtkingar seized a good part of the realm of Burizleifr,nbsp;who then ruled over Vindland A similar tradition is preservednbsp;in the KnytUngasaga, according to which Harald possessed anbsp;great domain (jarlsnki) in Vindland. He thus caused the construction of J�msborg, where he established a considerablenbsp;garrison, who were known as J�msvikingar^^. A parallel groupnbsp;of sources attributes the foundation of J�msborg not to Harald,nbsp;but to Sveinn Tjuguskegg�s foster-father Palnatoki. The J�ms-vikingasaga thus dates its establishment by this legendary heronbsp;as subsequent to Sveinn�s accession, and in the various redactionsnbsp;of this saga, the J�msvikingar appear not as Burizleifr�s enemies,nbsp;but as his allies, a role which they likewise play in the Arnamag-naean redaction of Oddr ^�. Snorri Sturluson omits any statementnbsp;as to the foundation of J�msborg, but Saxo Grammaticus attributes its establishment to Harald Obviously, if Harald was itsnbsp;founder, the establishment of J�msborg would fall somewherenbsp;between 950 and 986; if the foundation is attributed to Sveinnnbsp;Tjuguskegg, it must have occurred after the latter date, whichnbsp;appears intrinsically improbable. Modern authorities in general

Agrip, in Fornmanna S�gur X (Copenhagen, 1835), 392: �[�lafr] hafj)i ipuliga vetrseto sina i Veinlandi i borg peirri er heit J�msborg,� and Hisiorianbsp;Norwegies ed. G. Storm, p. 113.

M, G. SS. XXIX, 32: ��Qui [Haraldus] cursu celeri fugam arripiens . ., ad Sclauiam usque profugus commeauit ibique, pace impetrata, primus urbemnbsp;fundasse dicitur quae Hynnisburgh nuncupatur.�

Cf. F. J�nsson: Oldnor. og Oldisland. Litt. Hist. II (Copenhagen 1923), 611�620.

Fagrskinna, ed. P. A. Munch and C. R. Unger (Oslo 1847), 42: quot;Haraldr konungr herja^i d Vindland ok l�t gera borg mikla, pd er heitir at J�mi, ok ernbsp;su borg k�llu5 si^an J�msborg . . . J�msvikingar unnu mikit af riki B�rizleifsnbsp;konungs, er pd r�d fyrir Vindlandi.�

Forn. Sag. XI (Copenhagen 1828), 179: �Haf^i hann mikit jarlsriki i Vindlandi; hann l�t par gj�ra J�msborg, ok setti par herli�^ mikit. . . peir vorunbsp;kalla�'ir J�msvikingar.�

thus accept as historical the foundation of J�msborg by Harald Gormsson From these variants in the tradition, however, itnbsp;would appear that the compilers of the historical sagas knewnbsp;nothing positive of the foundation of J�msborg, and in thisnbsp;instance gave free rein once more to their habitual practice ofnbsp;legitimating some doubtful feature of their narrative by attachingnbsp;it to a known personage whose chronology and general characternbsp;harmonized at least ostensibly with the situation in question.nbsp;In view of the extensive and self-contradictory increments ofnbsp;the whole J�msviking tradition, Weibull is entirely justified innbsp;his contention that �the Norse J�msborg and the J�msvikingsnbsp;as the Danish and Icelandic traditions knew them never existed�nbsp;There is, however, one element of analogy connecting J�msborgnbsp;with Scandinavian colonization in Russia which, though notnbsp;noticed heretofore, is capable of casting some light on the actualnbsp;status of this stronghold. The Scandinavians who penetratednbsp;Russia actually founded no new towns there, but rather, throughnbsp;their superior energy and military ability, became in time thenbsp;ruling class both in the older cities on the Dnieper and in thenbsp;newer settlements like Novgorod and Rostov, which were themselves the product of Slavic expansion. In the case of J�msborg,nbsp;the commercial interest of the site cannot have been intrinsicallynbsp;as great as that of Novgorod (to cite an outstanding example innbsp;Russia), though the sheltered mouth of a major river connectingnbsp;with a prosperous hinterland was ipso facto of obvious importance.nbsp;There is, moreover, relatively satisfactory archaeological evidencenbsp;of a Wendish settlement on the site of Wollin during the tenthnbsp;century The Danish tradition as expressed by Saxo states:nbsp;�Haraldus, armis Sclauia potitus, apud lulinum, nobilissimumnbsp;illius prouinciae oppidum, competencia militum presidia col-locauit.� Without regard to other elements of the tradition, thisnbsp;text means nothing more than that Harald, on one of his raidsnbsp;in Vindland, set a garrison in the already existent Slavic town onnbsp;the site of Wollin, which is precisely what the Swedes in Russianbsp;appear to have done repeatedly. Such a garrison, provided withnbsp;suitable harbor facilities, then served as a base for further raids

Cf. J. W. Thompson: Feudal Germany (Chicago 1928), 534, 535, and notes; J. C. H. R. Steenstrup, op, oil., 34.

or trading expeditions into the interior, from which resulted the increased contact between Poles and Danes substantiated bynbsp;German sources for the period. The absence, however, of archaeological evidence as to any impressive and specifically Scandinaviannbsp;settlement on the site of Wollin also casts a shadow of doubt onnbsp;the fulsome description of J�msborg supplied by Adam of Bremen J�msborg is thus by no means to be conceived as anbsp;grandiose mediaeval mercantile port, but as a small Slavic townnbsp;dominated by a Danish garrison, who used it as a starting-pointnbsp;for their characteristic expeditions of combined trade andnbsp;piracy.

The statement in the A grip and the Historia Norwegics that Olaf Tryggvason, upon returning from Russia, made J�msborg his winter-quarters while raiding Flanders, England, Scotland, and Ireland, has alread}^ been cited. Neither in the Agripnbsp;nor in the latter is there any reference to a marriage of Olaf innbsp;Vindland. A tradition to this effect first appears in thamp;Fagrskinna,nbsp;where this wife is called Geila the Wend, and later identified asnbsp;daughter of Burizlafr, King of Vindland, and thus sister to bothnbsp;Gunnhildr, wife of SveinnTjuguskegg, and Astri�r, wife of Jarl Sig-valdi of J�msborg. {Fgsk., 56, 59). The Arnamagnaean redaction ofnbsp;Oddr contains a long account of the princess Geira, who, uponnbsp;Olaf�s arrival in Vindland, was exercizing the sovereignty in hernbsp;father Burizlafr�s absence {Forn. Sog. X, 233 ff.). In this text,nbsp;no reference is made to J�msborg in connection with Olaf�snbsp;arrival in Vindland, to which he was driven southward fromnbsp;Bornholm by storm. Geira, upon learning of his arrival fromnbsp;her chief councillor Dixin, despatches the latter to invite Olaf tonbsp;her court. At Dixin�s instigation, her marriage with Olaf is soonnbsp;consummated. The princess lives hut three years as Olaf�s consort,nbsp;and after her death he quits Vindland, according to Oddr (Arn.),nbsp;to undertake his second (apocryphal) voyage to Russia, butnbsp;according to Snorri {Hskr. I, 306), only to extend his raids to thenbsp;North Sea. This same redaction of Oddr also recurs later to Geiranbsp;as Olaf�s former wife in connection with his voyage to Vindlandnbsp;to recover the dowry of his third wife Pyri, Sveinn Tjuguskegg�s

ed. B. Schmeidler (Hannover 1917), 79, 80. Cf. C. Niebuhr: �Die Nachrichten von der Stadt Jumne�, Hansische Geschichtsblatter XXIII (Munich, 1917),nbsp;367�375, pointing out that Adam�s descriptions of both J�msborg-Iumneta andnbsp;Kiev show signs of bring derived from some fxfff^aaig on Constantinople.

122 sister It is, however, rather more than a coincidence that thenbsp;Stockholm redaction of Oddr omits all passages associating Olafnbsp;with Geira. In any case, Snorri adopts substantially the samenbsp;account provided by the Arnamagnaean redaction, not only as tonbsp;Olaf�s marriage to Geira {Hskr. I, 294�295, 297) and her deathnbsp;{ibid., 306), but also with regard to his recovery of Pyri�s dowrynbsp;from Burizlafr {ibid., 422�423, 429), with the addition thatnbsp;P3u:i herself suggests this errand will prove particularly easy onnbsp;account of Olaf�s long-standing friendship with the Wendish king.nbsp;Snorri rejects Olaf�s second voyage to Russia, but accepts thenbsp;erroneous association of Olaf Tryggvason with Otto II�s campaignnbsp;against Harald Gormsson in 974 2�. It is likewise significant thatnbsp;the traditional site of Olaf�s adventure with Geira was notnbsp;unanimously localized near J�msborg and the mouth of thenbsp;Oder, as may be gathered from the statement of Oddr (Arn.);nbsp;�Geira drottning rik�i bar er Germania heitir til vestrhaifu, ocnbsp;er t)angat bei�i betri landzcostr oc sva ly�rinn�� {Forn. S�g. X, 235).

Before attempting any conclusion as to the degree of historical verity attributable to the Geira tradition, it is essential to analyzenbsp;the various phases of the Norse material dealing with her supposednbsp;sister Gunnhildr. The first mention of the latter, as of Geira,nbsp;appears in the Fagrskinna (p. 42), and in Oddr According tonbsp;the former, when Jarl Sigvaldi learned of the death of Haraldnbsp;Gormsson he immediately made peace with Burizlafr of Vind-

Forn., S�g. X, 340: �En til Jjessa veitti Astridr konungi lid, d�ttir Biirizleifs konungs, oc hann var oc hinn mesti vin Olafs konungs, af J)vi er hann hafdinbsp;verit i Vindlandi, p�, er hann itti d�ttur hans.

As Olaf was born in 968, and returned from Russia only in 986�987, such participation is historically impossible. For the activity of B�rizl�,fr in thisnbsp;connection, v. infra p. 136.

Arn., Forn. S�g. X, 283 ff.; Stkh., in Kong Olaf Tryggves�ns Saga af Odd Snorreson, ed. P. A. Munch (Oslo, 1853), 29, 30.

The accounts of Harald�s death are varied and interesting. According to Fgsk. (p. 51), he died of an illness. The redactions of Oddr give no particulars.nbsp;The Hskr. represents him as attacked by Sveinn, and receiving in battle a woundnbsp;from which he died (I, 318, 319). Adam of Bremen, who claims to have derivednbsp;his information from Sveinn Ulfsson, the grandson of Sveinn Tjuguskegg,nbsp;recounts that the latter was supported against his father by the anti-Christiannbsp;party, and that Harald, beaten and severely wounded, �elapsus est ad ciuitatemnbsp;Sclauorum, quae lumna dicitur [= J�msborg]��, where �post aliquot dies exnbsp;eodem uulnere deficiens in Christi confessione migravit� {ed. cit. I, 27, 28). According to the J�msvikingasaga (Forn. S�g. XI, 64, Flatey. I, 161, 162), he isnbsp;assassinated by Palnatoki (cf. also Saxo Gram., ed. cit., 332, according to whomnbsp;Harald, beaten by Sveinn, flees to J�msborg for assistance, and upon landingnbsp;after a second battle with his son, is killed by Palnatoki).

land, with whom the J�msvikings are represented as waging constant war, and after marrying his daughter Astri�r, proposednbsp;to entice Sveinn Tjuguskegg to Vindland. Journeying to Seelandnbsp;for this purpose, he employed a successful ruse to lure Sveinnnbsp;on board his ship, and forthwith set sail for J�msborg. On hisnbsp;arrival there, he informed Biirizlafr of Sveinn�s capture, and alsonbsp;proposed to release the Danish king upon bhe latter�s acceptancenbsp;of such terms as Sigvaldi should dictate. Burizlafr likewisenbsp;consented to conclude peace with Sveinn on conditions to be setnbsp;by the Jarl of J�msborg. These stipulations included the marriagenbsp;of Sveinn to Gunnhildr, whose dowry was to include the Wendishnbsp;territory which the Danes had seized, while Burizlafr was betrothed to Pyri, Sveinn�s sister, who later married Olaf Tryggvason.nbsp;In the Arn. redaction of Oddr, the account of the marriage ofnbsp;Sveinn and Gunnhildr is yoked up with the story of the unionnbsp;of Sigvaldi with Burizlafr�s third daughter Astri�r. Sigvaldi, innbsp;fact, demands her hand, but the Wendish king stipulates asnbsp;a preliminary that Sigvaldi shall first kidnap Sveinn. The abduction is then recounted with extensive elaboration. As soonnbsp;as Sveinn is in his power, Sigvaldi frankly explains the situationnbsp;to his captive, and suggests the advisability of a match betweennbsp;the Danish king and Gunnhildr. As Sveinn has no choice but tonbsp;assent, the Jarl then visits Burizlafr and puts forward his proposition covering Gunnhildr�s marriage to Sveinn, which is dulynbsp;celebrated. In this redaction, the Pyri-Burizlafr match does notnbsp;appear at all in this connection, but instead {Forn. Sag. X, 309),nbsp;Burizlafr makes independent application for Pyri�s hand considerably later. In the Stockholm redaction of Oddr, the situationnbsp;is very briefly described with totally different implications. Atnbsp;the time of Harald Gormsson�s death, Sigvaldi is already marriednbsp;to Astri�r. Burizlafr conspires with Sigvaldi to deprive Sveinnnbsp;of his throne. The Jarl thus captures Sveinn, releasing him onlynbsp;on payment of ransom and the celebration of his marriage to Gunnhildr. Here again, there is no relation between the marriage ofnbsp;Gunnhildr and the union of Pyri with Burizlafr. The J�msvtkinga-saga {Forn. Sag. XI, 99 ff.) adds the detail that, of the Wendishnbsp;princesses, Astri�r was the oldest and Geira the youngest. Thenbsp;basis of the narrative corresponds in the main to the Arn. redactionnbsp;of Oddr, but with the addition that Astri�r herself suggests as anbsp;prior condition to her marriage with Sigvaldi that he either secure

124 the refund of all previous payments made by Burizlafr to thenbsp;Danes or else kidnap Sveinn Tjuguskegg. The Jarl succeeds innbsp;capturing Sveinn, and then announces to the Danish king thatnbsp;out of personal friendship he has planned the marriage withnbsp;Gunnhildr. Sigvaldi now claims Astri�r, having fulfilled his partnbsp;of the contract, and also secures from Sveinn the desired refundnbsp;prior to the latter�s marriage with Gunnhildr. In the J�ms-vikingasaga, there is again no mention of any alliance betweennbsp;Pyri and Burizlafr. The Hskr. (I, 319, 320) is closest to the Arn.nbsp;redaction of Oddr, but represents Sigvaldi�s marriage to Astri�rnbsp;as having already taken place before the abduction of Sveinn.nbsp;The Danish king, fearing for his life, consents to a reconciliationnbsp;with Burizlafr, sealed by his own marriage to Gunnhildr and thenbsp;betrothal of the Wendish king to his sister Pyri

The only consistent elements emerging from this welter of contradictory detail are the capture of Sveinn by Sigvaldi andnbsp;the latter�s mediation in the arrangement of a marriage betweennbsp;the Danish king and a Wendish princess. The sources dealingnbsp;most extensively with the Gunnhildr-episode (both redactionsnbsp;of Oddr, J�msvtkingasaga, Hskr.) are curiously reticent as to thenbsp;chronological relation between the marriage of Geira and thenbsp;former, though Olaf�s return from Russia and the accession ofnbsp;Sveinn Tjuguskegg (and therefore the latter�s marriage withnbsp;Gunnhildr) are, according to the customary dating, separatednbsp;by only two years at the most. In fact, if the implications ofnbsp;the Geira episode are strictly followed, Olaf should have been innbsp;Vindland when the abduction and marriage of Sveinn occurred.nbsp;In view, however, of the faulty nexus between the two sets ofnbsp;traditions, the conclusion is indicated that Geira, if she existed,nbsp;was actually not related to Gunnhildr and Burizlafr at all. Thenbsp;Arn. redaction of Oddr {sufra, p. 122) classes her as a Wendishnbsp;princess in Western Germany. Olaf�s early association with thenbsp;J�msvikings is mentioned in but two sources {Agrip, Forn. S�g.nbsp;X, 392, and Hist. Norm., p. 113), and is never picked up againnbsp;in later works save in a brief note in Oddr (Arn.) mentioning anbsp;tradition that Olaf once besieged J�msborg [ibid., 238), which isnbsp;obviously in contradiction with the statement of the sources citednbsp;that he made that center his winter-quarters. It therefore appears

2� For the subsequent development of the J)yri-Burizlafr motive, v. infra, p. 129.

that a scanty tradition regarding Olaf�s temporary union with a Wendish princess has been connected by the saga-compilers withnbsp;the account of Gunnhildr�s marriage, and has then been extensivelynbsp;used to motivate Olaf�s later success in recovering Pyri�s dowrynbsp;in Vindland. The rejection of any relationship between Gunnhildrnbsp;and Geira is further justified by the fact that while there is scaldicnbsp;confirmation for Olaf�s raids in Vindland, there is none whatevernbsp;for any contact between Olaf and the J�msv'ikingar.

According to the Fagrskinna (p. 43), Gunnhildr bore Sveinn two sons, Knutr inn Riki and Harald. This tradition is repeatednbsp;in both redactions of Oddr (Stkh., loc. cit., 37, 38; Arn., Forn. Sog.nbsp;X, 313) and in the KnytUngasaga {Forn. Sog. XI, 183), as wellnbsp;as in the Hskr. (I, 319). The same sources unanimously add thatnbsp;Sveinn later married Sigri�r St�rra�a, the widow of Eirikrnbsp;Sigrsaeli of Sweden (died 993), who is stated to have been thenbsp;mother of Olaf Skotkonungr by her first husband The Hskr.nbsp;is the only source stating explicitly that Gunnhildr died beforenbsp;Sveinn�s marriage to Sigri�r.

It so happens that the marriage of Sveinn to a Wendish princess is attested by German sources considerably closer to the event.nbsp;Thus Thietmar of Merseburg, referring with some animus to thenbsp;Scandinavians, writes:

�Sed quia nullus ad comprehendendas aquilonaris regionis uarietates, quas natura pre caeteris mirabiles ibidem operatur, et crudeles populi istius executionesnbsp;sufficit, omitto et de geniminibus uiperarum, id est filiis Suenni persecutorisnbsp;[i. e., Sveinn Tjuguskegg] pauca edissero. Hos peperit ei Miseconis filia ducis,nbsp;soror Bolezlaui successoris eius et nati; quae a viro suimet diu depulsa nonnbsp;minimam cum caeteris perpessa est controuersiam.�

The generation of vipers mentioned is speedily identified by Thietmar as quot;praedicti fratres Harald et Cnut� The princenbsp;appearing as Miseco, the father of Bolezlavus, his own successor,nbsp;is Mieszko I, Prince of Poland (died 992) the father of Boleslavnbsp;Chrobry. Thietmar is thus in agreement with the body of Norsenbsp;tradition as to the parentage of Knut. Adam of Bremen, however,nbsp;WTites:

2� For a general account of Mieszko, cf. H. ZeiBberg: �Miseco I (Mieczyslaw), der erste christliche Beherrscher der Polen� Archiv fiir Kunde �sterreich. Ge-schichtsquellen XXXVIII 2 (Vienna 1867), 27 ff.

�Post mortem diu optatam Herici [Eirikr Sigrsaeli], Sueinn ab exilio regressus optinuit regnum patrum suorum anno depulsionis uel peregrinationis XIIII.nbsp;Et accepit uxorem Herici relictam, matrem Olaph [Olaf Skotkonungr], quaenbsp;peperit ei Chnud.�

As Weibull has remarked, it would hardly seem likely that Adam of Bremen should have erred in a matter of recent genealogynbsp;where the grandson of Sveinn Tjuguskegg was his immediatenbsp;source, though his statement that fourteen years elapsed betweennbsp;the deaths of Harald Gormsson (986) and of Eirikr Sigrsaeli (993)nbsp;shows that in dealing with foreign subjects he was by no meansnbsp;exempt from the capacity to err Saxo Grammaticus likewisenbsp;mentions Sveinn�s marriage to Eirikr�s widow, whom he namesnbsp;Syritha, and characterizes as having borne to Sveinn his sonnbsp;Knut And finally, the Cnutonis Res Gestae relate that Knut,nbsp;after succeeding to the throne, went with his brother Harald tonbsp;Sclavonia, whence he brought back their mother who w'as sojourning there.

The evidence of the Norse sources, along with Thietmar and, by implication, the Cnutonis Res Gestae, thus indicates Knut'snbsp;mother to have been of Slavic origin. Adam testifies, however, innbsp;contradiction with Norse tradition, that Knut and Olaf werenbsp;sons of the same mother. If the mother of Knut was a Slavicnbsp;princess, and if, at the same time, Knut and Olaf Skotkonungrnbsp;were half-brothers by the same mother, the widow of Eirikrnbsp;Sigrsaeli cannot have been a Swedish lady. In fact, not onlynbsp;must Eirikr�s widow in this case have been Slavic, but she andnbsp;Sveinn�s Slavic wife (i. e., Gunnhildr) must have been one andnbsp;the same person.

It is only by analysis of the traditions surrounding Sigri�r St�rra�a that this radical correction of the sources can be justified,nbsp;Sigri�r first appears in the Morkinskinna casually mentionednbsp;as the mother of Olaf Skotkonungr and his sister Astri�r. In the

� II, 39; ed. Schmeidler, 99. He later (II, 73) refers to Knut and Olaf as �germani fratres�.

Unders�kningar, 109. That Adam was less infallible in matters of recens genealogy than Weibull�s remark would signify is also indicated by Adam'tnbsp;statement that Sveinn Tjuguskegg�s daughter AstrOr (her Christian name wasnbsp;Margaret) was married to Richard II of Normandy (II, 54, ed. cit., 114), whennbsp;she was really the wife of his son Robert I.

same connection, she is characterized in the Fagrskinna (p. 59) as daughter of Skoglar-Tosti and mother of Olaf Skotkonungr, instrumental in influencing her second husband, Sveinn Tjuguskegg,nbsp;to avenge the slight put upon her by Olaf Tryggvason through hisnbsp;unceremonious breaking of their engagement. She is later referrednbsp;to (ibid., no) as the mother of AstriSr by Sveinn Tjuguskeggnbsp;and as the previous wife of Eirikr Sigrsaeli. The statement ofnbsp;Adam of Bremen (II, 39) that Sveinn married Eirikr�s widownbsp;(whose name he does not mention) has already been noted,nbsp;together with the related passage in Saxo Grammaticus, whonbsp;gives Eirikr�s widow the name Syritha. Saxo is thus the firstnbsp;author to use the name Sigri�r or its approximation. Detailsnbsp;of Sigri�r�s marriage to Eirikr first appear in Oddr (Stkh., p. 7),nbsp;where it is stated that, according to common gossip, Eirikrnbsp;divorced her on account of her temper, but that she knew henbsp;had made a previous vow to Odin, according to which he hadnbsp;but a few years to live, which was dangerous for her in view ofnbsp;the Swedish custom that a queen should not survive her deceasednbsp;husband. This same redaction (p. 29) later reports that, innbsp;accordance with his vow to Odin, Eirikr committed suicide whennbsp;ten years had elapsed after his successful battle with Styrbjorn,nbsp;in connection with which his vow had been made. The Arn.nbsp;redaction {Forn. S�g., X, 220) first amplifies the correspondingnbsp;data to the immolation of a surviving consort over a deceasednbsp;king by the statement that Eirikr had only ten more years to live,nbsp;and then, in a later passage {ibid., 283) repeats the terms of thenbsp;vow, adding that at the time of Eirikr�s divorce of Sigri�r henbsp;made her queen in Gautland. Snorri Sturluson apparently undertook some elimination of the prolific romantic elements of thenbsp;tradition as it existed to this point in his own day, since he confinesnbsp;himself to the following statements: (i) Sigri�r, daughter ofnbsp;Skoglar-Tosti, was married to Eirikr Sigrsaeli (I, 245), by wEomnbsp;she had a son Olaf [Skotkonungr]; (2) Eirikr died of illness atnbsp;Uppsala (ibid.) ten years after Styrbjorn�s death (993�94);nbsp;(3) upon the arrival of Harald grenzki in Sweden (994), Sigri�rnbsp;was a widow, and had considerable estates in that countrynbsp;(I. 338). He thus rejects the whole tale of Eirikr�s vow and of hisnbsp;divorce from Sigri�r.

The next series of traditions associated with Sigri�r concerns her relations with Olaf Tryggvason. She does not appear in this

connection at all prior to the Fagrskinna which, as indicated above (p. 127), notes her hostility to Olaf on account of their broken engagement, but without detailed motivation {ed. cit.,p. 59). The episodenbsp;of the imitation gold ring offered as a gift from Olaf Tryggvasonnbsp;to Sigri�r appears, however, in both redactions of Oddr, whichnbsp;likewise report Olaf�s subsequent meeting with her and thenbsp;scornful slap dealt her with his glove when she refused to benbsp;converted. Both redactions also mention this episode as the basisnbsp;for her pronounced hostility to Olaf Snorri adopts substantiallynbsp;the same version. This enmity of Sigri�r toward Olaf Tryggvasonnbsp;is connected with his personal history only at one point, viz., innbsp;the events preceding the battle of Sv�l�, with the story of whichnbsp;has been interwoven a considerable amount of romantic embellishment connected both with Sigri�r and with Pyri, the sisternbsp;of Sveinn Tjuguskegg.

The motives underlying the collision between Olaf Tryggvason and his enemies at Sv�l� are variously explained in the earliernbsp;texts. According to Adam of Bremen (II, 39, p. 100), Olaf,nbsp;hearing of an alliance between Sveinn Tjuguskegg and Olafnbsp;Skotkonungr, and believing Sveinn easy prey, declared war uponnbsp;him at the instigation of his wife There (quot;cuius instinctu bellumnbsp;Danis intulit,� ibid., 36, p. 98). Olaf Tryggvason is thus thenbsp;aggressor. In the Agrip, however, the quarrel arises from Sveinn�snbsp;retention of the items promised and stipulated as Pyri�s dowrynbsp;after her marriage to Olaf Tryggvason, who, with a view tonbsp;wiping out the insult entailed, projects an expedition againstnbsp;Denmark {Forn. S�g. X, 393). While troops are being gathered innbsp;Norway, Olaf sails over to Vindland, but his soldiery turn backnbsp;as soon as he is out of the country, so that he is obliged to dependnbsp;solely on Wendish support. Sveinn now allies himself with Olafnbsp;Skotkonungr and Eirikr Hakonarson, and the allies attack andnbsp;defeat Olaf Tryggvason �furir Siolandi�. In Theodric�s Hisiorianbsp;de Antiquitate Regum Norwagiensium Sveinn, Olaf Skotkonungr,nbsp;and Eirikr Hdkonarson are the aggressors, catch Olaf Tryggvasonnbsp;unprepared, and defeat him �iuxta insulam quae dicitur Svoldnbsp;et iacet prope Slauiamquot;. In this text, there is not the slightest

Mon. Historica Norvegiae ed. G. Storm (Oslo 1880), 23, 24. For the most recent and a very satisfactory analysis of the sources on the battle of Svold,nbsp;cf. J. Schreiner; �Olaf Tryggvasons Siste Kamp� in Festskrifi til H. Falk, (Oslonbsp;1927), PP- 54�-77-

reference to Pyri. The Historia Norwegiae gives an account similar to that of the Agrip, relating that, when Sveinn retainsnbsp;Seeland, the dowry of his sister �nomine Thyri, quam priusnbsp;dux quidam de Sclauia desponsauerat inuitam�, Olaf declaresnbsp;war upon him. The Norwegian king, as in the Agrip, endeavorsnbsp;to collect a force wherewith to attack the Danes, but the Norwegians are unwilling to cross the frontiers, so that Olaf Trygg-vason is obliged to seek aid among the Sclavi, quot;quos in piraticanbsp;fidissimos habuerat socios�, Sveinn, upon hearing of Olaf�snbsp;preparations, concludes an alliance with Olaf of Sweden andnbsp;Eirikr Hakonarson. The three take Olaf of Norway by surprisenbsp;as he is sailing past Seeland, where they attack and defeat him.

As far as the earlier Norse sources may be credited, the basic element of these narratives seems to consist in a territorial disputenbsp;between Sveinn and Olaf Tryggvason, whether or not connectednbsp;with Pyri�s dowry; it can no longer be determined with certaintynbsp;who V(^as the aggressor. The romantic increment of this episodenbsp;in the later texts is attached to the fundamental account asnbsp;supplied by the fairly congruent narratives of ike Agrip and thenbsp;Historia Norwegice, both of which include as points of departure:nbsp;Pyri�s betrothal to a Wendish prince prior to her marriage withnbsp;Olaf Tryggvason, Sveinn�s retention of her dowry, Olaf�s preparation for hostilities, his journey to Vindland, the alliance ofnbsp;Sveinn, Olaf Skotkonungr, and Eirikr Hakonarson, and the fatalnbsp;naval engagement. Subsequent elaborations, which first occurnbsp;in the redactions of Oddr, thence find their way, for the mostnbsp;part, into the Hskr., and are summarized in the Fagrskinna,nbsp;include: an extended account of Pyri�s marriage to Burizlafr, hernbsp;flight from Vindland, her marriage to Olaf of Norway, her incitement of Olaf to war against Sveinn, and her request that Olafnbsp;recover her property in Vindland; Olaf�s successful errand tonbsp;Burizlafr in this connection; and finally, the activity of Sigri�rnbsp;St�rra�a in urging Sveinn to avenge on Olaf Tryggvason thenbsp;disgrace he had visited on the Danish king by marrying pyrinbsp;without Sveinn�s permission.

The various development of these elements may best be seen from the summaries following:

(a) Oddr (Stkh.); pyri, though betrothed to Btirizldfr, stays at home in Denmark, unwilling to go to Vindland and consummate the marriage until forced

to do so by Sveinn, who is chiefly influenced by the protests of his wife Gunnhildt. Jjyri, even after the marriage, refuses to live with Burizlafr, who allows her tonbsp;depart. She flees to Falster, and sends a messenger to Olaf Tryggvason requestingnbsp;his protection. Olaf goes to Falster, where he marries Jryri without consultingnbsp;Sveinn Tjuguskegg, who resents this slight (pp. 36, 37). After Sveinn�s marriagenbsp;to Sigri�r St�rra�a, she keeps his resentment hot, and finally (p. 46) inducesnbsp;him to concoct a plot with Olaf Skotkonungr whereby the latter shall pretendnbsp;his readiness to be converted and plan a meeting with Olaf Tryggvason, bynbsp;which pretext the Norse king shall be enticed out of his kingdom and thus laidnbsp;open to attack. Olaf Tryggvason falls into the trap, and when Olaf Skotkonungrnbsp;does not show up at the rendezvous (p. 48), sails on to Vindland, where he recoversnbsp;J)yri�s property. Meanwhile Sveinn, Olaf Skotkonungr, and Eirilcr H�,konarsonnbsp;prepare to attack him.

(b) nbsp;nbsp;nbsp;Oddr (Arn.): Olaf Tryggvason�s attention is first called to J)yri by one ofnbsp;his companions (Forn. Sag. X, 293). Biirizlafr, however, asks successfully fornbsp;her hand in Denmark, and returns home to Vindland, whither ]Dyri is supposednbsp;to follow. She refuses to do so, however, until Burizldfr requests Gunnhildr tonbsp;intervene with Sveinn so that he may induce her to set out (p. 309). The marriagenbsp;is consummated, but jpyri refuses food for eleven days, and on the twelfth thenbsp;king allows her to depart. She flees to Falster, whence she sends messengers tonbsp;Aki, her foster-father at Olaf�s court (the same person who had previously spokennbsp;to Olaf concerning her). Olaf goes to meet her at Falster, and they are marriednbsp;(P- 313)- Sveinn is displeased at the news, and shortly after, Gunnhildr dies,nbsp;whereupon he marries Sigri�r. She subsequently (p. 333) incites him to hostilitynbsp;against Olaf Tryggvason, and devises a plot whereby Olaf Tryggvason, Olafnbsp;Skotkonungr, and Sveinn shall meet the following summer at the Brenneyar.nbsp;Eirikr HAkonarson and his brother Sveinn join with Olaf Skotkonungr in thenbsp;conspiracy. Olaf Tryggvason consents to the meeting, and meanwhile ])yri urgesnbsp;him to recover her property in Vindland (p. 337). Olaf Tryggvason sets out fornbsp;Vindland, where he celebrates a reunion with his old friends Dixin, Biirizldfr,nbsp;and Astri�r, and recovers the dowry of J)yri (pp. 339, 340). In both redactionsnbsp;of Oddr, Sigvaldi of J�msborg is Sveinn Tjuguskegg�s emissary to Olaf Skotkonungr.

(c) nbsp;nbsp;nbsp;Heimskringla (I, 419�430): Biirizlafr complains of the non-fulfillment ofnbsp;the contract whereby he is to marry Sveinn�s sister pyri. Sigvaldi goes to Denmarknbsp;and brings her back to Vindland, where she is married to B�rizlafr. She refusesnbsp;all sustenance for a week, and finally flees to the woods with her foster-father.nbsp;They escape thence to Denmark, but dare not stay there in fear of Sveinn�snbsp;anger. They thus move on to Norway, where Jryri addresses herself for protectionnbsp;directly to Olaf Tryggvason. Being a persuasive woman (�ordsnjgll�, says Snorri),nbsp;she immediately v/ins his sympathy to the extent that he proposes marriage.nbsp;Upon its celebration, Jjjnri begins to regret her lost belongings in Vindland, andnbsp;suggests that Olaf should have no difficulty in recovering them, as long as Bdriz-14fr is such an old friend. The king thus plans a voyage for this purpose, disregarding the counteradvice of his friends because joyri taunts him v/ith beingnbsp;afraid of her brother Sveinn. He thus arrives in Vindland, and accomplishesnbsp;his mission successfully. Sigri�r is meanwhile inciting Sveinn Tjuguskegg tonbsp;attack Olaf. Sveinn therefore crosses over to Sweden to conclude an alliancenbsp;with Olaf Skotkonungr and Eirikr Hdkonarson, so that they may attack Olafnbsp;while he is on his voyage to Vindland.

(d) nbsp;nbsp;nbsp;Fagrskinna (pp. 58�61); At Gunnhildr�s instigation, Sveinn forces jryrinbsp;to depart for Vindland and marry Burizldfr. After seven days, however, she escapesnbsp;to Denmark, and thence to Norway, where she becomes Olaf�s wife without

Sveinn�s consent. After leaving Vindland, which he had visited for the recovery of Jpyri�s dowry, Olaf sails toward Denmark, and in the neighborhood of Sv�l�-comes into collision with the combined fleets of Sveinn Tjuguskegg, Olaf Skot-konungr, and Eirikr. Olaf Skotkonungr was there to avenge Olaf Tryggvason�snbsp;insult to his mother Sigri�r, who had also incited Sveinn to avenge the affrontnbsp;Olaf had put upon him.

This body of tradition which, in view of its extensive variations, must be viewed entirely as fiction, arises logically from an effortnbsp;to motivate the hostility of Sveinn Tjuguskegg and his contemporaries toward Olaf Tryggvason. The impulse to the wholesalenbsp;introduction of feminine influences derives initially from thenbsp;supposition that the quarrel first rose over Pyri�s marriage portion,nbsp;so that she was the injured party likely to spur Olaf Tryggvasonnbsp;to aggression. In the versions which made Sveinn Tjuguskeggnbsp;the aggressor, it was then necessary to supply a similar motivation,nbsp;hence the general appearance of Sigri�r St�rra�a in this role.nbsp;In both redactions of Oddr, Gunnhildr exercizes great influencenbsp;over Sveinn, especially in connection with the marriage of Pyri.nbsp;On the other hand, since another parallel tradition existed according to which Gunnhildr v.'as Olaf Tryggvason�s sister-in-law, itnbsp;would be incongruous if she were to prove hostile to him. Thenbsp;necessity thus arose of postulating as second wife for Sveinnnbsp;Tjuguskegg the entirely mythical Sigri�r St�rra�a, around whomnbsp;other legends could be centered. The somewhat nebulous characternbsp;of the whole account is further indicated by the absence of precisenbsp;information as to Gunnhildr�s death; only Snorri (I, 419) remarksnbsp;specifically that she �took sick and died� but with verynbsp;indefinite indication of chronology, and the dubious explanationnbsp;quot;sem nu var a�r fra sagt.�

There remains one further item requiring interpretation in this connection, viz.. Scholium 25 to Adam of Bremen, according tonbsp;which Eirikr Sigrsaeli of Sweden made an alliance with �Boliz-laus�, king of Poland, against the Danes and married �either hisnbsp;daughter or his sister.quot; The Polish prince thus designated is

Cf. Oddr (Arn., Forn. S�g. X, 313): quot;En er Sveinn konungr spyrr petta, Ucar honum st�rilla, er petta var gert at lileyfi bans [i. e., marriage of Olaf Tryggvason and pyri]; oc litlu siparr andadiz Gunnhilldr drotning, er amp;tti Sveinnnbsp;konungr.�

Substantially the same conclusion is reached by Weibull (Unders�kningar �24, 125), who suggests an analogy with the Sigurd-legend, urging that Brynhildnbsp;is the prototype for the evolution of Sigridr.

Ed. Schmeidler, p. 95: �Hericus rex Sueonum cum potentissimo rege Polanorum Bolizlao fedus iniit. Bolizlaus filiam uel sororem Herico dedit. Cuius

132 undoubtedly Boleslav Chrobry, Mieszko�s son, who actually didnbsp;control most of Russia for a brief period (ioi8�1019), and whonbsp;likewise subjected both the Prussians and certain Slavic tribesnbsp;west of the Oder. Mieszko, however, had died in 992; Eirikrnbsp;Sigrsaeli died in 993�994. The period within which this marriagenbsp;could have been consummated and the otherwise totally unknownnbsp;SlaVO-Swedish attack on the Danes could have been executednbsp;is thus reduced to an impossibly narrow margin. While Schmeid-ler attributes this scholium to Adam himself {op. cit., XLII), thisnbsp;attribution by no means exempts it from the possibility of error,nbsp;particularly since Adam was apparently misinformed anywaynbsp;as to the date of Eirikr�s death (v. supra p. 127). Furthermore,nbsp;Boleslav in 992�994 was in no position to undertake a warnbsp;against the Danes. In the first place, he had no navy and as yetnbsp;no direct contact with the sea. In the second, he was supportingnbsp;the Saxons during 992 against a Slavic revolt *2, and shortlynbsp;thereafter was involved in a serious Russian offensive into Galicia The fact that Adam did not know whether it was Boleslav�snbsp;sister or his daughter whom Eirikr was supposed to have marriednbsp;shows that his information was dubious. As a matter of fact,nbsp;Boleslav was born in 967 (cf. Zeissberg, op. cit., 102) and thusnbsp;could have had no marrigeable female issue in 992�994. Thenbsp;conclusion is therefore justified that we are here confronted withnbsp;a very scanty trace of the original marriage of Gunnhildr tonbsp;Eirikr Sigrsaeli, which must actually have taken place some eightnbsp;years previous. Adam, already familiar with Sveinn�s marriagenbsp;to a daughter of Mieszko, appears to have received subsequentnbsp;information of the marriage of Eirikr to a princess of the samenbsp;family. Since Adam obviously did not know the name of thenbsp;princess concerned, and, unlike Thietmar, was unfamiliar withnbsp;the tradition of Mieszko�s daughter�s divorce, he naturallynbsp;concluded that two separate individuals were involved. Since thenbsp;gratia societatis Dani a Sclauis et Sueonibus iuxta impugnati sunt. Bolizlausnbsp;rex christianissimus cum Ottone tercio confederatus omnem Sclauoniam subiecitnbsp;et Ruzziam et Pruzzos, a quibus passus est sanctus Adalbertus, cuius reliquiasnbsp;tunc Bolizlaus transtulit in Poloniam.�

Jlasp. JleT., ed. cit., 122 ad 992; Ann. Hildesheim, M.G. SS., Ill, 6g ad 992: �Bolizlao vero, Misachonis filius, per se ipsum ad dominum regem uenirenbsp;nequaquam ualens � imminebat quippe iili grande contra Ruscianos bellum �nbsp;suos sibi satis fideliter milites in ministerium regis direxerat.�

evidence is overwhelmingly in favor of the Slavic origin of the mother of Knut the Great, it may be inferred that Gunnhildrnbsp;was first married to Eirikr Sigrsaeli, by whom apparently shenbsp;was divorced After Eirikr�s death, she married Sveinn Tjugu-skegg. During his campaigns in England, she returned to livenbsp;in Poland, and was brought home from there by Knut after hisnbsp;accession to the throne.

The various traditions surrounding Astri�r, referred to in the younger Norse sources generally as Burizlafr�s third daughter,nbsp;have already been briefly noted {supra p. 123). Aside from thenbsp;daughter whom Thietmar of Merseburg and Adam of Bremennbsp;mention as married to Sveinn Tjuguskegg, we have no furthernbsp;data as to Mieszko�s female issue. He seems to have been firstnbsp;married in 965�966 to Dobrava, daughter of the Czechish princenbsp;Boleslav I, and then, upon her death (977), to the former nunnbsp;Oda, daughter of the German margrave Thiedrich Dobravanbsp;is attested as the mother of Boleslav Chrobry '��. It is thus apparentnbsp;that any daughters of Mieszko who acquired Scandinaviannbsp;husbands in the eighties of the tenth century must have beennbsp;children of Dobrava and sisters of Boleslav In view, however,nbsp;of the romantic elaborations, particularly in the redactions ofnbsp;Oddr and of the J�msvikingasaga, which surround the conditionsnbsp;of Astri�r�s marriage to Sigvaldi, and the complete lack of evidencenbsp;elsewhere as to any very consistent and close relations betweennbsp;Poles and Danes, it is rather more likely that Astri�r was actuallynbsp;the daughter of some minor Pomeranian prince in the vicinity ofnbsp;Wollin, and was linked up with Burizlafr-Mieszko in the general

The phrase of Thietmar (VIII, 39) �a uiro suimei diu depulsa� was rightly-taken by J. M. Lappenberg [M. G. SS. Ill ad loc.) to refer to Eirikr, not to Sveinn, as F. Kurze (ed. cit., 216, n. 5) interprets it. Kurze also considers, onnbsp;the basis of the scholium of Adam, that this princess was previously married tonbsp;Eirikr; quot;quae antea Erici regis Sueciae uxor fuisse apud Adam II c. 33, schol. 25nbsp;et c. 36 traditur� (ibid., n. 4). Kurze was, however, unfamiliar with the Norsenbsp;sources, since he remarks of Mieszko�s daughter (ibid.): quot;nomen ei fuisse Sigridnbsp;Storr^da Lappenberg argumentis ignotis nisus dixit.quot;

^5 Ann. Pol., M. G. SS. XIX, 615, Ann. Cracov. Vet., ibid., 577, Thietmar IV, 55�56, pp. 94. 95-

The tradition preserved in Gallus (Mon. Pol. Hist. I, 399) that Mieszko, while a pagan, had seven wives, is rather to be regarded as analogous to thenbsp;story presented by the Russian Primary Chronicle (Jlanp. Jler., 79) regardingnbsp;Vladimir�s maintenance, before his conversion, of six hundred concubines, whonbsp;are simply intended to place him in the same class with Solomon and thus tonbsp;make the contrast between his pagan and his Christian life the more impressive.

134 process of combination of initially unrelated elements whichnbsp;seems to have taken place in the evolution of the sagas of thenbsp;J�msvikingar and of Olaf Tryggvason

The foregoing analysis of Norse material dealing with the relations between Scandinavians and Wends at the mouth ofnbsp;the Oder in the last decades of the tenth century suggests thatnbsp;the extent of these contacts was considerably more restrictednbsp;than a literal reading of these sources would indicate. It remainsnbsp;to confirm this conclusion by examining the details of Polishnbsp;history during this period with a view to determining whatnbsp;relations exist between the well-attested activities of the contemporary Polish princes and the implications of the Norsenbsp;texts.

Regardless of the statement of Oddr {Forn. S�g. X, 285) that Burizlafr was a vassal of the king of Denmark, there is no evidencenbsp;elsewhere that the Poles during the reign of Mieszko were particularly concerned with controlling the mouth of the Oder or thatnbsp;their northern boundary lay north of the Warta river. The regionnbsp;known as Pomerania, bounded on the west by the Oder, on thenbsp;south by the Warta and the Notec, and on the east by the Vistula,nbsp;was indeed inhabited by Slavic tribes closely related to the Poles,nbsp;but hardly identical with them. The Russian Primary Chroniclenbsp;characterizes them as a subdivision of the Vistulan Slavs, whilenbsp;Helmold, likewise writing in the twelfth century, clearly differentiates them from the Poles Though generally not so inclusive,nbsp;the name Pomerania also applied in the twelfth century to thenbsp;west bank of the lower Oder, while the border with Poland propernbsp;ran up the Notec through the frontier towns of Zantok, Uzda,nbsp;and Naklo ��. Some indication that the Polish outposts laynbsp;somewhat further north on the Oder than generally supposednbsp;might conceivably be drawn from V/idukind�s mention that

The Arn. redaction of Oddr (Forn. S�g. X, 285) thus mentions an otherwise unknown pyri as a fourth daughter of Burizlafr in addition to Astriot, Gunnhildr,nbsp;and Geira.

Jlasp. Jlex., 6: �Slovene prisedse sedo�a na Visle prozvasasja Ljachove i ot tech Ljachov prozvasasja Poljane, Ljachove, druzii Lutici, ini Mazovsane,nbsp;ini Pomorjane.� Helmold: Chron. Slavorum ed. G. F. Pertz (Hannover 1868)nbsp;I, 40, p. 88: �Otto . . , adiit peregrinationem ad gentem Sclauorum, qui dicunturnbsp;Pomerani et habitant inter Oderam et Poloniam.� Cf. Adam of Bremen IV, 13,nbsp;p. 241: �Trans Oddaram comperimus degere Pomeranos, deinde latissima Po-lanorum terra diffunditur.� Cf. also Gallus, Mon. Pol. Hist. I, 394.

Wichmann incited the Vuloini to attack Mieszko in 967 It is clear, however, that in the official Saxon conception Mieszko�snbsp;territory extended in 972 only to the Warta, since the Polishnbsp;prince is expressly characterized as �imperatori fidelem tribu-tumque usque inUurta solventem�. This passage is practicallynbsp;the only extant reliable indication of the geographical extent ofnbsp;Poland in the course of Mieszko�s principate. While it has sometimes been assumed that Pomerania was in some degree subjectnbsp;to Poland even in Mieszko�s time, the best evidence points to itsnbsp;first conquest by Boleslav Chrobry during the early years of hisnbsp;reign Mieszko, in fact, during his later years appears to havenbsp;been too constantly engaged in disputes with his Czechish neighbors or, as an ally of the Saxons, in campaigns against the pagannbsp;Slavs along the Elbe, to have been able to undertake any organizednbsp;expansion up the Oder tov/ard the sea.

The historical evidence of Mieszko�s relation of vassalage to the contemporary Saxon emperors obviously precludes attachingnbsp;any significance to Oddr�s statement that he was a vassal of thenbsp;Danish king. In 966, Mieszko had married Dobrava, sister ofnbsp;Boleslav II of Bohemia, and apparently through her influencenbsp;was converted to Christianity At this period, conversion wasnbsp;synonymous with political dependence, and we thus find him

Widukind: Res Gestae Sax. ed, K. A. Kchr (Hannover 1904) 11, 69 (pp. 120, 122): �[Wichmannus] egit cum Sclauis qui dicuntur Vuloini quomodo Misacamnbsp;amicum imperatoris bello lacesserent.� The identity of this tribe has been variouslynbsp;defined. L. Giesebrecht {Wendische GescMchten I [Berlin 1843] 189, 190) wishednbsp;to locate them further south in Brandenburg, but their name clearly connectsnbsp;them with the island of Wollin, so that they can be placed with relative certainlynbsp;on the lower Oder. So J. Steenstrup: Venderne og de Danshe (Copenhagen 1910),nbsp;33} 36', Niederle, op. cit., HI, 149, 150.

^2 Thietmar ed. F. Kurze I, 29 (p. 37). As Thietmar�s father supported Mieszko when the latter was attacked at the time by the Margrave Odo, his informationnbsp;would appear trustworthy.

W. Barthold: Geschichte von R�gen u. Pommern I (Hamburg, 1839) 337; A. Naruszewicz: History a Narodu Polskiego I (Cracow 1859) 156.

Cf. Helmold I, 15 (p. 36): quot;Eodem quoque tempore Bolizlaus, Polonorum cristianissimus rex, confederatus cum Ottone tertio, omnem Sclauiam, que estnbsp;ultra Odoram, tributis subiecit.� Cf. Adam of Bremen, schol. 25, supra) W. vonnbsp;Sommerfeld: Geschichte der Gevmanisierung des Hevzogtums Pommern oder Slaviennbsp;(Leipzig 1896) 16, dates this conquest in 995, presumably on the basis of Giesebrecht, op. cit., I, 231, 232. In any case, there was a bishopric at Colberg undernbsp;Boleslaw�s friend Reinbern about 1000 (Thietmar IV, 45; p. 90).

Thietmar IV, 55; pp. 94�95; Ann. Pol., M.G.SS. XIX, 614�615; Ann. Cracov. Vet., ibid. 577.

present as a vassal at Otto I�s court in Quedlinburg on Easter Day, 973, together with his Czechish brother-in-lawnbsp;To what extremes Mieszko�s subservience to Saxon authoritynbsp;extended may be gathered from Thietmar�s subsequent lament,nbsp;in connection with the presumption of Boleslav Chrobry, thatnbsp;the latter�s father (Mieszko), as long as Margrave Odo lived, nevernbsp;even dared appear before him in a fur-trimmed cloak, much lessnbsp;remain seated if the Margrave arose Mieszko�s loyalty to hisnbsp;suzerain, however, was not entirely invariable. Regardless of thenbsp;fact that in 973 he put his son as a hostage in the hands of Otto I,nbsp;he became involved the following year, just prior to Otto II�snbsp;Danish expedition, in a conspiracy against the Emperor withnbsp;Henry 11 of Bavaria, Bishop Abraham of Ereising, and Boleslavnbsp;of Bohemia Upon Otto�s return from Denmark, he undertooknbsp;in 975 a punitive expedition against Bohemia and apparently,nbsp;with less success, against Mieszko Hence, if the Norse sourcesnbsp;understood Burizlafr to be prince of Poland, their statements asnbsp;to his participation as a German ally in any operations againstnbsp;Harald in 974 belong to the realm of fable Erom the statementnbsp;of the Annales S. Trudperti that Otto�s expedition against Mieszkonbsp;was unsuccessful, the conclusion may be drawn that duringnbsp;Otto II�s reign, Mieszko�s relations with the imperial court werenbsp;in abeyance. In view of the Emperor�s difficulties elsewhere, firstnbsp;in Bohemia (977), then in Lorraine (978�980), next in Italynbsp;(980�983), and finally, the revolt of the Elbian Slavs in 982,nbsp;he was in no position to enforce obedience upon the Polish prince.

Ann. Allah. Maior., 2nd edit. ed. E. von Oefele (Hannover 1891), p. ii; Thietmar II, 31, p. 38.

Thietmar V, 10; p. 113: quot;Vivente egregio Hodone pater istius Bolezlaui Miseco domum qua eum esse sciebat crusinatus intrare vel eo assurgente numquamnbsp;praesumpsit sedere.�

Ann. Allah. Maior., p. ii; Lambert of Hersefeld, ed. O. Holder-Egger (Hannover 1894), 42.

Ann. S. Trudperti, M.G.SS. XVII, 280: �Otto imperator cum Polanis uincitur, et Miseco dux filius Bolizlai uictor existens Polemiorum gentem abnbsp;eius imperio seiunxit.quot;

The idea of introducing Bilrizldfr into the narrative of Harald Gormsson�s conflict with Otto II rose very naturally from the verses of the Vellekla (F. J�ns-son: Shaldedigtning A 117) which speak of Wends as forming a part of Otto�snbsp;army. Cf. Hshr. I, 300: �pAs me�' fylki Frisa For gunnvi^'urr sunnan (Kvaddinbsp;vigs) ok Ven�fa (V�,gs blakri^i) Frakka.� The use of Wendish allies from thenbsp;Slavic tribes west of the Oder is by no means unknown, as is shown by Henry II�snbsp;similar enlistment of the Luticians against Boleslav Chrobry in 1003 (Thietmar;nbsp;VI, 22; pp. 146, 147).

It has generally been assumed heretofore that Otto�s campaign of 979 against certain Slavic tribes outside the boundaries of thenbsp;realm was directed against the Poles but its results werenbsp;apparently negative. Mieszko thus took no active part in Germannbsp;affairs until after the death of Otto II, when he was disposed tonbsp;support Duke Henry of Bavaria, the first guardian of the youthfulnbsp;Otto III, in his efforts to seize the realm. The Polish prince wasnbsp;thus present at Henry�s court in Quedlinburg on Easter Daynbsp;984 After the diet of Rara and the conclusion of peace betweennbsp;Henry and the partisans of the five-year old Emperor, bothnbsp;Mieszko and Boleslav of Bohemia returned to their previousnbsp;feudatory relationship, and were present once more at Quedlinburg at Easter 986 In the course of the same year, Otto initiatednbsp;a new campaign against the Elbian Slavs, in which he was joinednbsp;by Mieszko, who brought him the present of a camel �'*. A furthernbsp;Saxon sally against the Elbian Slavs in 987 resulted in thenbsp;reconstruction of the border blockhouses along the river (�reno-uatis iuxta Albim castellis�, says Thietmar), showing that at thenbsp;moment the Saxon frontier followed the course of the Elbe.nbsp;During 990, Mieszko became involved in a conflict of uncertainnbsp;origin and issue with Boleslav of Bohemia which seems to havenbsp;caused considerable momentary commotion without modifyingnbsp;the territorial status of either prince ��. The next Easter (991),nbsp;Mieszko again appeared before Otto and Theophano, whose deathnbsp;occurred the following June �^. In the same year, Mieszko madenbsp;his final appearance as a German ally in Otto�s first successfulnbsp;attack upon Brandenburg ��. The Luticians, however, shortlynbsp;retook the city, with the result that Otto resumed operationsnbsp;against it in 992; Mieszko was prevented from appearing in personnbsp;in consequence of a threatened Russian attack, but sent a detach-

Giesebrecht: Wend.Gesch. I, 255: ZeiBberg, loc. cit., 87 n., based ox^Gesta Episcoporum Cameracensium, M. G. SS. VII, 442, 443.

Ann. Allah. Maior., p. 15, Ann. Hild. and Qued., M. G. SS. Ill, 67. Lambert (p. 46) substitutes Boemos for Sclauos, which is obviously an error, in view ofnbsp;Boleslav�s attendance at court the same spring.

ment commanded by his son Boleslav (v. supra p. 132). Mieszko himself died in June of this year

It thus appears that the Pyri episode connected in the Norse sources with Burizlafr cannot by any means be attached tonbsp;Mieszko I, since, in the first place, he had been a Christian fornbsp;some twenty-five years; and for the further reason that, sincenbsp;Pyri�s marriage to Olaf Tryggvason took place after the latter�snbsp;occupation of the Norwegian throne (hence in 995 or later),nbsp;Mieszko had already been dead for at least three years whennbsp;Pyri�s marriage to Burizlafr is supposed to have occurred. It hasnbsp;also been shown (v. supra^. 131) that the account of this marriagenbsp;is chiefly fiction, quite apart from the historical fact that Mieszko�snbsp;second wife Oda, whom he married after Dobrava�s death (977),nbsp;actually survived him'^�.

There is similarly no historical basis for assuming any more intimate contact between Poles and Scandinavians during thenbsp;early reign of Boleslav Chrobry. It has already been shownnbsp;(v. supra^. 132) that the report in Scholium 25 to Adam of Bremennbsp;concerning a joint Suedo-Slavic attack on the Danes connectednbsp;with an alleged marriage of some female relative of Boleslav tonbsp;Eirikr Sigrsaeli is, in all likelihood, entirely apocryphal. In 992,nbsp;immediately after his accession, Boleslav, was threatened bynbsp;a Russian incursion which occupied his attention for the next twonbsp;years. In 995, Boleslav once more joined Otto III, this time in annbsp;expedition against the Obodriti in Mecklenburg '^1, and shortlynbsp;thereafter initiated his campaigns into Pomerania, which culminated in 997 by his recovery of the relics of St. Wojciechnbsp;(Adalbert) from the Prussians ^2. We also find him seizing thenbsp;pretext of a revolt among the Elbian Slavs to advance his westernnbsp;frontier to the Bober. The death of Boleslav II of Bohemia innbsp;999 and the subsequent disorders provided him with an opportunity for the recapture of Cracow and the simultaneous annexationnbsp;of Silesia and a part of Moravia. By this expansion of Polishnbsp;sovereignty, however temporary it was likely to prove in consequence of the as yet deficient internal organization of Poland

Ann. Hild., M.G. SS. Ill, 91: quot;Rex Abodritos uastauit, urbes et oppida disiecit; occurrit in auxilium Bolizlau filius Misaco cum magno exercitu.�

itself, Boleslav automatically became an outstanding European figure. It was accordingly in looo that Otto III undertook hisnbsp;celebrated pilgrimage to Gnesen. The Polish prince cannot wellnbsp;have entertained any interest in petty Scandinavian quarrels innbsp;999�1000, so that, like the other components of the Pyri episode,nbsp;the tradition of Olaf Tryggvason�s gathering of assistance at thisnbsp;juncture among the Wends of the lower Oder must be rejectednbsp;without reserve.

In summary, then, the only historical elements in the extensive body of Norse tradition surrounding the contact of the J�msvik-ings with the Slavs at the mouth of the Oder and the relationsnbsp;of the latter with the Scandinavians in general toward the closenbsp;of the tenth century consist in the probable presence of a Danishnbsp;garrison at the Slavic town of Wollin and the marriage of anbsp;daughter of Mieszko I of Poland first to Eirikr Sigrsaeli and laternbsp;to Sveinn Tjuguskegg. As is shown by the marriage of Yaroslavnbsp;the Wise to Ingiger�r some twenty years later, close politicalnbsp;affiliation was not an essential of dynastic unions at this period.nbsp;From the evidence of Scandinavian trade with central andnbsp;eastern Europe via the Oder and the Vistula it appears thatnbsp;these marriages with Scandinavian princes must have originatednbsp;in a desire on the part of Mieszko and Boleslav to maintainnbsp;friendly relations with Scandinavian warrior-merchants passingnbsp;through the entrepots of Gnesen and Kruszwica. The romanticnbsp;evolution of the J�msvtkingasaga also appears to have favorednbsp;the concentration about the legendary Burizlafr of all survivingnbsp;recollections of contacts with the Wends of the Baltic seacoast,nbsp;particularly with those tribes residing between Holstein andnbsp;R�gen, practically all of whom had been reduced to submissionnbsp;by the Germans before the thirteenth century. The apparentlynbsp;enigmatic prevalence of the name Burizlafr-Burizleifr as anbsp;cognomen of Slavic princes, which, in spite of its Slavic aspect,nbsp;has no Slavic counterpart whatever is hardly to be explainednbsp;satisfactorily from any confusion of Mieszko I with his son, butnbsp;rather by the greater Scandinavian familiarity with Boleslavnbsp;Krzywousty (the Wrymouthed), who from II02 to 1129 was

R. Ekblom: �Dic Warager im Wcichselgebiet�, Arch. Slav. Philol. XXXIX (Berlin 1924), 210.

In the tenth century, the Polish form, if derived from burja, quot;storm�, and fhe suffix -Slav, should have been *Burzislaw, which never occurs.

uninterruptedly engaged in the subjection of Pomerania, and who, after the reduction of Stettin in 1121, allied himself in 1130nbsp;with Nicholas of Denmark (son of Sveinn Ulfsson, and thereforenbsp;great-grandson of Sveinn Tjuguskegg) for the conquest of Wollin.nbsp;This alliance was sealed by the marriage of Boleslav Krzywousty�snbsp;daughter Rikiza to the Danish king�s son From Boleslavnbsp;Krzywousty, the name appears to have become the characteristicnbsp;appellation for Wendish princes in the Norse sagas, while thenbsp;transformation of the first two syllables results from the supposednbsp;analogy with Jarizleifr-Jarisldfr, the Norse equivalent of Yaroslav who, by virtue of his historical relationships with St Olafnbsp;and Harald Har�ra�i, was the Slavic ruler most frequentlynbsp;mentioned in the whole body of Norse literature reflecting thenbsp;Russian adventures of these two princes.

Saxo Gram. XIII, p. 420. Knytlingasaga 89 {Forn. Sag, XI, 327): quot;Magn�s Nikuldsson fekk Rikizu dottur Biirizldfs Vindakonungs, J)eirra synir voru peirnbsp;Kniitr ok Nikulas.�

For the variation between -zldfr and -zleifr, cf. Oldfv. Aleifr, A. Noreen: AUisland. u. altnorweg. Gramm. 3rd edit. (Halle 1903), 39.

BLOTNAUT

Although many scholars have in recent times referred to the worship of cattle in ancient Scandinavia no one has collectednbsp;and evaluated the evidence.

Cattle are mentioned in the tales of origins, as in the story (i) of the origin of the world and the gods in the Younger Edda. ^nbsp;A dripping frost turned into the cow Audhumla, from whose teatsnbsp;ran four rivers of milk. These rivers nourished the giant Ymir fromnbsp;whose body earth and sky and sea were made. The cow licked thenbsp;frost-stones �and the first day a man�s hair came out of thenbsp;stones in the evening, the next day a man�s head, and the daynbsp;after that the whole man was there.�� This man was Buri, thenbsp;ancestor of the gods. (2) There are two slightly differing versionsnbsp;of the tale of Gefjon and the origin of the island of Zealand.nbsp;Gefjon, kin of the Msir, hitched to a plow her four giant-begottennbsp;sons, who, in the first version were oxen, or whom, in the second, ^nbsp;she changed to oxen. She was to have as much land as she turnednbsp;over in a day; she plowed so deep that she dragged a piece ofnbsp;land, now the island of Zealand, out into the sea.

^ E. g., J. de Vries, ZfdPh XXXV (1928), 281. Gudmund Sch�tte, Hjemligt Hedenskab i allmenfattelig FremsHllning {K0b. og Kris. 1919) 121 ff. E. H. Meyer,nbsp;Germanische Mythologie (Berlin 1891), Par. 141; see also index under the namesnbsp;of the various animals. Finnur J�nsson, quot;Gudenavne-Dyrenavne,� Afnf XXXVnbsp;(1918), 309 ff, Kaarle Krohn, Skandinavisk Mytologi (Helsingfors, 1922), 79.nbsp;Eugen Mogk in Hoop�s Reallexikon der germ. Altertumskunde (StraBburg, igii tonbsp;19), s. V. rind. Karl Helm, Altgerm. Religionsgeschichte (Heidelberg, 1913) I,nbsp;202�213. Hjalmar Falk, �Odensheite,� Vid.-Selsk. Skr. ll.H.-F. Kl, (Kristiania,nbsp;1924), No. 10.

� Gylfaginning, chap. I; �Ynglinga saga,� chap. V in Finnur J�nsson, Heims-kringla (K0b. 1893�1900), I. Axel Olrik, �Gefjon�, Danske Studier nbsp;nbsp;nbsp;if.

142

The instances of cattle used in ritual are particularly interesting. ^ (3) The oldest written record of cattle-worship among peoples from present Scandinavian territory concerns the Cimbri. �nbsp;They released captive Romans on parole, requiring them to takenbsp;oath on a bronze bull. � (4) The Viga-Glums saga relates as follows:nbsp;�The man who should take a temple-oath took in his hand anbsp;silver ring which had been reddened in the blood of a neat whichnbsp;had been sacrificed.� ^ (5) In the North German Nerthus ritualnbsp;cows drew the sacred image of the goddess about the country innbsp;a car. � (6) Oxen, or more likely, bulls, were sacrificed in case ofnbsp;famine. �Then the Swedes performed great sacrifices at Upsala.nbsp;The first autumn they offered oxen, and the crops were notnbsp;bettered either.� � The next year they triedh uman sacrifices andnbsp;in the following year they sacrificed the king. (7) In Sweden annbsp;old bull intended for sacrifice had been so pampered that henbsp;became violent and broke away into the woods and was a sourcenbsp;of danger.^� (8) Saxo relates that a certain Danish king who wasnbsp;harrying in Sweden made an offering with furuis hostiis, �blacknbsp;victims,� to secure divine favor, and that the Swedes perpetuated

^ One is tempted to regard the vivid pictures of five bulls carved on the rocks at Tegneby, Aspeberget, in Tanums Socken, Tanums Harad in Bohuslan as annbsp;indication of bull-worship in Sweden in the bronze age. The manner in whichnbsp;they are depicted suggests that they are symbols of fertility. We know, however,nbsp;so little of the purpose of these rock-pictures that one dares only to suggest thisnbsp;interpretation as a possibility. Reproduced in Bidrag till kdnnedom om G�teborgsnbsp;och Bohusldns fornminnen och historia, \lll (1906), 514. Also Oscar Montelius,nbsp;Vdr forntid (Stockholm, 1919), 138.

� �Caius Marius,� in Plutarch�s Lives, IX, 525 = Losb's Classical Library (London and New York), Greek Authors, Cl (1920). The Cimbri arrived in Italynbsp;about 100 B. C., having left Jutland some twenty years before.

� The function of the bull here resembles that of the sonarggltr, the boar. quot;In the evening there was making of vows. The sonarggltr was led forth; men laidnbsp;thereon their hands and made vows on the vowing-cup�. See Helgakvi�a Higr-vardssonar, prose following stanza 30. �King Hei�rekr had a big boar fed (i. e.nbsp;reared). He was as big as the biggest bulls and so handsome that each hair seemednbsp;to be of gold. The king laid one hand on the boar�s head and the other on hisnbsp;bristles and took this oath . . .� See Jon Helgason, Heidreks saga (Kab., 1924),nbsp;text if, 54 Text H, 54 adds that the king worshipped Freyr and gave himnbsp;the biggest boar he could get. The oaths were taken on Yule-eve. Text U, i2g *nbsp;tells in addition to this that the boar was to be sacrificed to Freyr for good cropsnbsp;early in February.

� Gudmundar p�rldksson, Islenzkar Forns�gur, gefnar lit af hinu islenzka B�kmentaf�lagi (Kaupmannah�fn, 1880�83) I, 76.

this offering in the form of an annual sacrifice. Saxo added concerning it: Fr�hlod Sueones meant. �The Swedes call it Frey-sacrifice.� The �black victims� were in all probability black cattle. (9) Porvar�r bought the carcass of a slaughtered bull andnbsp;offered it to an elf. One could give a living animal to a god.nbsp;(10) Porgils P�rdarson had once given a calf to Thor. Later Porgils,nbsp;who in the meantine had become a Christian, was sailing tonbsp;Greenland. He dreamed that Thor was trying to win him backnbsp;to heathenism and that Thor had finally said, �If you are notnbsp;going to be good to me give me my goods anyway.� When Porgilsnbsp;awoke he remembered the calf, which was now an ox, and threwnbsp;it overboard. (ii) �. . . before Porkell went away from Pvera henbsp;went to Frey�s temple and led thither an old ox and spoke thus:nbsp;�Frey�, said he, �has long been my confidence and received manynbsp;good gifts from me and repaid them well. Now I give you thisnbsp;ox in order that Glumr may leave Pvera no less against his will thannbsp;I go now. And do you let signs be seen, whether you accept or not.�nbsp;And the ox gave a start and cried out and fell down dead.�

Bulls were employed in the ritual of the duel. (12) The participants had taken their positition on the duelling ground.�Then was led up a big old bull. It was called a bl�tnaut. That should he killnbsp;who had the victory. It was sometimes one neat; sometimes eachnbsp;one who went on an island (i. e. fought a duel) had his own lednbsp;up.� After the victor, Egill Skallagrimsson, had killed his opponentnbsp;he �sprang up quickly and to that place where the bl�tnaut wasnbsp;standing, grasped with one hand its muzzde and with the othernbsp;a horn and turned the bull so that its feet stuck up and its necknbsp;broke.� 1� (13) Kormakr disabled Porvar^r in a duel and the

Gudmundur |)orl�,ksson, Islenzkar Forns�gur, Gefnar �t af hinu islenzka B�kmentaf�lagi (Kaupmannah�fn, 1880�83) I, 29.

Bl�tnaut, from biota, �to sacrifice, to worship,� and naut, �neat,� signifies (a) �a bovine animal destined for sacrifice,� or (b) �one which is the object ofnbsp;�worship.�

Finnur J�nsson, Egils saga Shallagrimssonar nebst den gr�szeren gedichten Egils^ (Halle, 1924), chap. LXV, sect. 20ff. The editor says this manner ofnbsp;killing a bull is of course fictitious. This deed was similar to but far lessnbsp;difficult than �bulldogging a steerquot;, a fairly common exhibition feat of Americannbsp;cowboys. �In this a steer is set galloping at full speed, the cowboy afternbsp;kim. A quick cow-pony will soon catch up with a steer, and, just at thenbsp;nioment that the pony forges beside the steer, the rider must throw himself

latter had to yield. �Kormakr saw where the neat stood and killed it.� (14) The outcome of the duel was not satisfactory and it wasnbsp;tried again, but the same man was victorious as before, andnbsp;�Kormakr killed the U�tnaut according to custom.� (15) Twonbsp;brothers were leaving home for the duelling ground where onenbsp;of them was to fight a berserkr. �Porbjprn asks his brother Gisli,nbsp;�Which of us, brother, shall fight the berserkr today, and whichnbsp;shall kill the calf?� Gisli answers, �I advise this, that you killnbsp;the calf, but Bjorn and I shall test each other�.�

There are accounts of cattle which were objects of worship. Some of these tales concern cows. (16) King Olafr Tryggvasonnbsp;once asked a peculiar old man who came to him at Eastertide atnbsp;Ogvaldsness why the ness and farm were so named. The old mannbsp;answered: �Ogvaldr was a king and a great warrior. He worshipped most a cow and had her with him wherever he went bynbsp;sea or land. It seemed beneficial always to drink her milk, . . .nbsp;He was buried in a mound a short way from this ness and innbsp;another mound near by the cow was laid, and there people setnbsp;up the hautastones which still stand.�(17) The people of Kingnbsp;Eysteinn of Upsala �had great faith in a cow, and they callednbsp;her Sibilia. She was so much worshipped (and as a result, enchanted) that men could not endure her bellowing. And for thatnbsp;reason the king was accustomed, when he was expecting a (hostile)nbsp;army, to have this same cow in front of the lines, and so muchnbsp;craft of the devil was in her that his enemies, as soon as theynbsp;heard her, became so wild that they fought each other and did notnbsp;look out for themeselves.� In Eystein�s message to his army henbsp;said. �We shall have with us the cow Sibilia, our god.� Ragnar�snbsp;from the galloping horse to the horns of the galloping steer, grasping thenbsp;horns in such a way as to turn the animal�s head, then throwing himselfnbsp;to the ground, by main force he must twist its neck and drag the steer overnbsp;to the ground, taking care not to get the horns pinned in his body whilenbsp;doing so.� (Francis Rolt-Wheeler, The Book of the Cowboy [Boston, 1921],nbsp;390.) The object is to turn the steer over and tie its feet, but sometimesnbsp;the animal�s neck gets broken in the process. There is no reason for doubtingnbsp;that an Icelander could do what was imputed to Egill.

On the sacrifice of cattle in Scandinavia in recent times see Nils Lid, quot;Norske Slakteskikkar,� fyrste luten, Vid.-Selsk. Sky. II. H.-F. Kl. (Oslo, 1923), No. 4,nbsp;128; Paul Heuergren, Husdjuren i nordisk folktro (�rebro, 1925), 249�52.

Arngrimr J�nsson�s Latin summary of the lost Skjoldunga saga describes this same vaccam diaholicam.

Bulls were worshipped. (i8) A hero wished to find the origin of an urarhorn, the quest of which had been imposed upon him.nbsp;He learned that King Haraldr had harried Bjarmaland, and thatnbsp;the inhabitants had taken an animal and worshipped it; theynbsp;called it an 4rr. It became enchanted and destroyed men andnbsp;animals and subjected everything under itself. When Kingnbsp;Haraldr came to capture the beast a stern woman approached himnbsp;in his sleep and told him how to kill the urr, but in return demanded the horn in the front of its head. When she got this she tooknbsp;it to the temple, where it still was at the time. (19) There is anbsp;Norwegian tale of the suspected worship of a bull that was atnbsp;the head of a herd. Porsteinn accused Harekr in the presence of

*0 Magnus Olsen, Vohunga saga oh Ragnars saga lodbr�kar (Keib. 1906�08), �32�33, 138, 144, 147, 148�49.

Axel Olrik, �Skjoldunga saga i Arngrim Jonssons Udtog,� Aarhzger f. nord. Oldk. og. Hist., 1894, 133.

Felix Liebrecht, Zur Volkskunde {Heilbronn, 1879), 72�-2, says: �Die . . . Kuh Sibilia hat ihr Analogon (auch im Namen) in der g�ttlichen Kuh Sabala,nbsp;die durch ihr Briillen dem Vasischtha hundert Konige verschaf ft, welche dasnbsp;Heer Visvamithras vernichten (Julius Braun, Naturgeschichte der Sage [M�nchen, 1864�5] II, 431 ff. Angelo de Gubernatis, Die Thieve in der indogerma-nischen Mythologie [Leipzig, 1874], 56!).� Liebrecht gives further references tonbsp;what he regards as parallels. The passage which he here has in mind is innbsp;Manmatha Nath Dutt, The Ramayana, translated into English Prose etc.nbsp;(Calcutta, 1891�4) II, 124�8. See also Ralph �. H. Griffith, The Rdmdyananbsp;of Vdlmiki translated into English Verse (London and Benares, 1870�4) I,nbsp;226�33.

The word Sabala or Cabala or Savala,amp;s it is variously transliterated, means �spotted�, or �Dappleskin� as Griffith translates. Its resemblance to Sibilia isnbsp;a matter of chance. The latter was probably made up from elements contained in the foreign names wich pleased this age, such as Marsibil, Blanda,nbsp;Maria, Cecilia, and has nothing to do with Latin sihylla. There is no manuscript authority for making the first vowel of Sibilia long, as is often done.

The horn of an aurochs, OE ur, Icel Urr, a species of wild cattle, coexistent W'ith the European bison in Northern Germany in the eleventh century andnbsp;persisting longer in Sweden, Poland and Lithuania. The last of the species, anbsp;cow, died in the Polish province of Masovia in 1627. See Otto Keller, Die antikenbsp;Tierwelt (Leipzig 1909�13) I, 34. The events described in (15) are placed quot;westnbsp;of the river Vina� (i. e. Dvina), in Bjarmaland, the Perm. Mogk in Paul�s Grundrifinbsp;d. germ. Philol.^ (StraBburg 1900�1909) II, 845�47 inferentially puts this saganbsp;in the fourteenth century. There were plenty of aurochsen at this time, but I donbsp;not know that they ranged as far north as the Perm.

King Olafr Tryggvason of worshipping a U�tnaut in secret. Harekr said there was little to this, but the king demanded to see the animalnbsp;and they went into the woods until they came to a large herd ofnbsp;cattle. quot;There was with them a bull so terribly big and ugly thatnbsp;the king thought he had never seen the like. The bull bellowednbsp;frightfully and acted most viciously. Harekr said, �Here is thenbsp;bull, my lord, and I am so fond of this neat that he is very affectionate to me.� T certainly see it,� said the king, �and henbsp;looks ugly to me�.�� The king directed Porsteinn to kill the bullnbsp;and seized Harek�s possessions and drove him from the country.nbsp;(20) A king�s sister was held captive by a giantess who intendednbsp;to make the princess her successor as priestess of a heathennbsp;temple. The hero Herrau�r and his man B�si found that thenbsp;giantess ate a two-year-old heifer for a meal. quot;There is a bull innbsp;the temple, enchanted and worshipped. He is bound with ironnbsp;bands; he is to cover the heifer and (thus) the poison is mixednbsp;with her, and all become betrolled who eat (of her flesh). She isnbsp;to be prepared for food for Hlei�r, the king�s sister, and shenbsp;(Hlei�r) will become just as much a troll as the temple-priestessnbsp;was before.�� The heroes kill the heifer, break the bull�s neck, slaynbsp;the old priestess and rescue the princess.

Pampered bulls kept for cult purposes became dangerous and gave rise to fantastic tales. (21) �It happened one day thatnbsp;Hjalm{)�r was playing at tables with the king�s daughter. Herv�rnbsp;asked what he was to do for his winter�s lodging. He said he hadnbsp;to hunt up a calf. �Whither do you have to seek it ?� says she. �Thenbsp;king will not tell me anything about that,� says he. �That is nonbsp;calf,� says she, �but rather an old bull in the nineties; he eatsnbsp;live-stock and kills men and horses. He is more cruel than anynbsp;other animal; he has killed all the men who have gone afternbsp;him and have asked my father for lodging for the winter. It isnbsp;the greatest U�tnaut. ... He is in a high and strong enclosure innbsp;the center of the land; he must not be turned loose, for then henbsp;breaks down fortifications and castles and does much harmnbsp;to men and beasts.�� Hjalmtj�r�s enchanted companion, H��r,nbsp;secretly takes upon himself the task of killing the bull, accomplishes this in a violent struggle and brings home the hide and

** Flateyjarb�h I, 261 f., and Fornmannas�gur (Kaupmannahofn, 1825�37) II, 131�33-

Otto Luitpold Jiriczek, Die B6sa saga in zwei Fassungen nebst Proben am den B�sa-rimur (StraBburg 1923), 26 ff.

(22) The Icelander who translated chap. XXXIII of the Book of Exodus in the latter half of the fourteenth century had onnbsp;his tongue�s end a suitable word for the golden calf worshippednbsp;by the Children of Israel. He called it a bl�tkdlfr.

Dangerous cattle are often supernatural in nature or origin. They may be cows with uncanny powers, as in the Ragnars saga.nbsp;(23) quot;The people of the town own two neat-cattle, and they arenbsp;heifers, and men had left because they could not stand theirnbsp;bellowing and sorcery. . . . They now take and turn loose thesenbsp;cattle which they believed in, and when the heifers are let loosenbsp;they run hard and bellow frightfully.� A hero with supernaturalnbsp;powers kills the heifers and contributes to the defeat of the peoplenbsp;of the town. The troublesome animals are more often bullsnbsp;and of supernatural origin. (24) P�rolfr bsegif�tr made trouble fornbsp;his neighbors in his lifetime and still worse trouble after hisnbsp;death. When nothing else would stop his spooking people dugnbsp;up his body and burnt it far from the dwellings of men, and thenbsp;country-side had rest. But a broken-legged cow that had beennbsp;turned loose on the range after preliminary recovery from itsnbsp;injury had been seen licking the ashes where P�rolfr had beennbsp;burned. Some people said they had seen her there with a dapple-gray animal. At any rate she was with calf when she came homenbsp;in the fall. She bore a heifer calf towards spring, but, strange tonbsp;say, soon after bore another calf, and this time with great difficulty because it was so big. When it bawled the sound was sonbsp;evil that an old woman with second sight urged that it be killed.nbsp;But the calf was so superior that the owner kept it for a bull,nbsp;and it was soon larger and stronger than other cattle of its age.nbsp;The summer after it was three years old it became violent and

'�Hj�,lmp�rs saga ok �lvis,� FAS III, 498. Reykjavik edition III, 383. A somewhat semilar story is in Jon Arnason, Islenzkar pj�ds�gur og cefintyri,nbsp;(Leipzig 1862�64) II, 363.

C. R. Unger, Stjorn, gatnmelnorsk Bibelhistorie fra Verdens Skabelse til det Babyloniske Fangenskab (Chr., 1862), 312 f.

In a version of the Hervarar saga (Heidreks saga) given by Rafn there is mention of quot;berserkir roaring like blotneyti.quot; See FAS I, 425. For Modern Icel.nbsp;blotneyti, �ugly bull,� see Vigfusson, Icel.-Eng. Diet., Oxford 1874, s. v. bl�tnaut.nbsp;The characteristic color of supernatural cattle in Iceland. See (25) infra.

killed its owner. The farm-hands pursued it into a swamp where it mysteriously disappeared. The reader of the saga isnbsp;supposed to know without being told that the bull is the son ofnbsp;P�rolfr and his reincarnation. (25) �lafr pai had a remarkablynbsp;fine bull, dapplegray with four horns, one of which stuck out innbsp;front of its forehead. When the bull was eighteen years old thenbsp;front horn fell off and Olafr had him killed. The next night Olafrnbsp;dreamed that a large angry looking woman came to him. She said,nbsp;�You have had my son killed and caused him to come to me in anbsp;mishandled condition, and for that you shall get to see your sonnbsp;all bloody at my bidding; I shall also pick out the one whom Inbsp;know you would least of all wish to lose.� ^

(26) nbsp;nbsp;nbsp;Only a few of the ox-names which are transmitted in thenbsp;fulur and in various passages in the sagas interest us.nbsp;The following are connected with names for Othin: Jormuni,nbsp;related to the Othin-name Jormunr and probably understoodnbsp;here as �big ox,� Jormunrekr, similarly related to Jormunr,nbsp;and probably understood as �leader of the oxen,� Olgr, �snorter,� an Othin-name. Sveidur and Sveidudur are in ablautnbsp;relationship to the Othin-names Svidur and Svidudur. Thesenbsp;two ox-names are related to svida, �spear,� and probably meannbsp;�horned.� There is a similar relationship between the ox-namenbsp;Svigdir, �with curved horns,� and the Othin-name Sveigdir.nbsp;The ox-name Vingnir is a name of Thor. The ox-name Freyrnbsp;is the name of the god Freyr. I say �ox-name,� but in thenbsp;old sources uxi also refers to bulls, and since only an entirenbsp;male would be a suitable sacrifice in a fertility cult, these namesnbsp;may be meant for bulls rather oxen.

(27) nbsp;nbsp;nbsp;The following words referring to cattle were used as namesnbsp;of men: Kdlfr, �calf,� was frequent in Iceland, sporadic butnbsp;ancient in Norway, Kvigr, �bull-calf,� was used occasionally.

The details of the owner�s unsuccessful fight are like those of Hord�s successful fight {21) in FAS III, 498�501. Reykjavik edition III (1889), 383�85.

Finnur J�nsson, Den norsk-islandske Skjaldedigtningen (K0b. og Krist. 1912�15), A I, 650, 675. B I, 656, 669.

The kenningar for bull in Meissner, �Die Kenningar der Skalden,� Rhei-nische Beitrage und Hilfsbiicher z. germ. Philol. u. Volksk. I (Bonn u. Leipzig, 1921), III, sec. 30, contain nothing of religious import.

On all names connected with Othin see Falk, Odensheite, s. v. and on the others see his page 44 ff. For references to the occurrences of the same see Finnurnbsp;]�n%son, Lexicon poeticum antiquae linguae sepientrionalis (K0b. 1913�16), s. t).

The foregoing represents what we find on the subject. Equally important is what we do not find. (28) The protothemenbsp;blot- is recorded in the lexicons with only three deuterothemesnbsp;that are the names of animals: -kdlfr, -naut, -neyti, all referringnbsp;to cattle. That is; there is no such word listed in any dictionarynbsp;as *hl�thestr, *hl�tgoUr, *hl�tsvin. While such words may havenbsp;been used, they played such an unimportant role that they havenbsp;not been recorded, nor have they lived down to modern times.

In former times much of the material presented here was considered of little value by reason of its lateness, but the studynbsp;of current tales and customs has changed our point of view innbsp;this respect. In evaluating such sources we do not so much asknbsp;if a given event occurred at a stated time or place, as we ask ifnbsp;the story has preserved a non-christian tradition or usage.

It is of course difficult to tell how much of the Younger Edda is due to Snorri�s fondness for telling a good story, but since therenbsp;is not much of a story connected with Au�humla it is likely thatnbsp;the account of her is a bit of old tradition (i). The story of Gefjonnbsp;and her sons sounds like a local tradition, some skipper�s jestingnbsp;account of the origin of the island which his craft is approaching (2).nbsp;The agreement of the Cimbrian bull-cult with the Scandinaviannbsp;as to object and with the sonargoltr cult as to content confirmsnbsp;our belief in Plutarch�s account (3). Similarly the linguisticnbsp;equation Nerthus-Njordr satisfies us as to Tacitus (5). Nos. (6),nbsp;(7), (8), (9), similar in kind but in independent texts, confirmnbsp;each other as to contents, as do also Nos. (12), (13), (14), (15).nbsp;The story of the deification of a milch-cow (16) differs from thenbsp;others, Au�humla being too far removed in kind to be compared.

^ E. H. Lind, Norsk-islandska dopnamn och fingerade namn frdn medeltiden (Upsala och Leipzig 1905�15), and the same author�s Norsk-islandska personbinarnnbsp;framp;n medeltiden (Upsala 1920�21), s. u. Names used only as nicknames are notnbsp;included in (24).

The material in folk-tales and folk-customs is not included here. Many things in de Gubernatis� section, ��Der Stier und die Kuh in der germano-skandinavischennbsp;und frankisch-keltischen Sage� of his Die Thieve in dev indogermanischen Mythologie, 172�203, do not belong within the limits of this study.

Bldtvavgr means quot;a person given to profanity,� vargr having lost its original meaning, quot;wolf� Cp. Swedish slitvarg, quot;a person who tears his clothes.�

The passage which follows this in the saga says that people believed that the peculiar old man was the devil who had takennbsp;on the form of Othin. There is no adequate motivation for the insertion of the Ogvald incident, but contemporaries may havenbsp;known a reason for its inclusion. This story is hardly a deificationnbsp;of a source of food; it is more likely a sarcastic hit at some actnbsp;or habit of King Olaf�s which it was both unnecessary and unwisenbsp;to point out. I should not attribute to it any value as a source ofnbsp;information as to ancient beliefs. Nos. (17), (18), (20) containnbsp;tales of cattle that became enchanted by the worship offerednbsp;them and in this point show a mutually confirmatory agreement.nbsp;The behavior of Sibilia (17) is very like that of the two heifers (23)nbsp;in the same saga and unlike that of any other cattle in old Scandinavian tradition. Hermann suspects a foreign source, but onenbsp;wishes a closer investigation, for Hermann gives no reasonnbsp;for the faith that is in him when he attributes them to the Irish.nbsp;Both (18) and (20) are laid in Bjarmaland, which was inhabitednbsp;by Finnish tribes. The passage in the B�sa saga which justnbsp;precedes (20) says that the temple is that of J�mali. He was anbsp;Finnish deity. Thus (18) and (20) are intended to be descriptionsnbsp;of Finnish usages; but voyages to Bjarmaland were always rarenbsp;and had stopped entirely, at least from Icelandic ports, when thisnbsp;saga was written, and little was known of the country. Thesenbsp;accounts tell less of conditions in Bjarmaland than they do ofnbsp;what an Icelander could imagine. The main feature, the bull-cult,nbsp;was native to his own country. No. (19) is in the saga of Kingnbsp;Olafr Tryggvason, a work which contains much historical truth.nbsp;One suspects that the accusation of bull-worship was only anbsp;scheme to enable the king to seize Harek�s property. Had therenbsp;not been a general belief that people did worship bulls, the kingnbsp;would have invented some other charge. No. (21) is of course anbsp;wholly imaginary tale, though suggested by facts, and so ofnbsp;interest. The bull of which it tells reminds us of those in (7) andnbsp;(18). No. (23) has been discussed; (24) sounds like a local traditionnbsp;that had grown up around a story of a violent bull that had killednbsp;its owner. No. (25) may well be fact and Olafr pai may have had

quot; H. F. Feilberg, Bidrag til en Ordbog over den jyshe Almuesmal {K0b. 1886�1911) III, 908 a, 21. Paul Hermann, �Islandische Heldenromane�, Thulenbsp;XXI (1923), 155 note 2, 157 note i. Ewald Liden in Festskrift til Finnurnbsp;J�nsson {K0b., 1928), 361 If,

such a dream, supernatural origin having been suggested by the color of his bull and the stern woman by some such tale as thatnbsp;of the woman who cursed Hadingus for slaying a supernaturalnbsp;being after he had killed an inauditi generis heluam.

While the Icelander of the early Christian centuries officially rejected the heathen gods, he often lived in a lively fear of them,nbsp;and we are justified in thinking that cattle-names reminiscent ofnbsp;the old gods may well have been a bid for their protection overnbsp;the cattle (26). Such names for human beings as Kdlfr, Kvigr,nbsp;Uxi ma}^ have arisen from the nicknames given to the precedingnbsp;generation of men, a frequent development; but it is possiblenbsp;that they arose as cult-names, a possibility to which insufficientnbsp;attention has been given in Germanics.

It is probable that the annual journey of the goddess Nerthus w'as to insure the multiplication of crops and herds, and therenbsp;must have been a definite reason for yoking cows to her car,nbsp;whereas oxen were the normal draught-animals for Northernnbsp;Germany. At any rate sterile oxen which could signify onlynbsp;barrenness gave way to fertile females (5).

Bull-worship is a far flung fertility cult. We know it especially from the accounts in the Old Testament of bull-worship innbsp;Samaria, and from the Mithras worship in the lands held bynbsp;the Roman armies. The Swedes sacrificed �oxen,� more likelynbsp;bulls, to avert famine and secure fertility. The meaning ofnbsp;No. (8) becomes clear when we notice that the story is annbsp;attempt to explain the origin of the Fr�blod, the famous annualnbsp;sacrifice in Upsala, here named after Freyr, the god of fertility, the god whose name is given to a bull (26). Cattle then,nbsp;and primarily bulls, play a leading part in the Scandinaviannbsp;cult of fertility.

One of the stories included above (19) is of special interest in determining the character of the Scandinavian cattle cult. Itnbsp;is not in a legendary or romantic source such as the YngUnga

HansNaumann, �AltnoTdischelUcLmenstudien,��AciaGermanica, neueReihe, Heft I (Berlin 1912), 172. Note Christian cult-names, as Spanish names for men;nbsp;Jesus, Jos� Maria, Juan Bautista, for women; Asuncion, Circoncision, Concepcion, Maria de Jesus, for either sex; Encarnaci�n.

saga, which provides us with version 2 of (2), with (6) and (7), or the Younger Edda, which contains (i) and version i of (2),nbsp;or the Fornaldars�gur, from which we have (17), (18), (20), (23);nbsp;it is in a king�s saga, which carries the air of vei'isimilitude andnbsp;the assumption that its contents are historical facts. Here a mannbsp;was accused of worshipping a certain bull and the accusationnbsp;apparently seemed plausible to his contemporaries. The bull wasnbsp;the head of this man�s herd. The two bulls of supernatural originnbsp;(24), (25), also in non-romantic sagas, were likewise regardednbsp;as choice breeding animals and also apparently the heads of theirnbsp;respective owners� herds. Does not this point the way?

The bull is important; the modern scientific breeder says, �The bull is half the herd.�� His qualities affect all the young stock;nbsp;his virility is of special concern, for if he is infertile there arenbsp;neither calves nor milk and a year is wasted. Was not the worshipnbsp;of cattle primarily a domestic fertility cult, carried on at eachnbsp;farm, centering on the bull, and intended to insure an annualnbsp;crop of calves ? In heathen times it may well have been a publicnbsp;affair of greater proportions (7), (8), but in Christian times itnbsp;could survive only on a modest scale. Possibly each sort of domesticnbsp;animal that was of any economic importance enjoyed such anbsp;cult. In many countries the church took over this function andnbsp;annually blessed the domestic animals.

When a fertility cult was once established it could develop off-shoots, as in the cow Au�humla (i), a �cow of plenty,�� anothernbsp;side of fertility, and so important a figure that the systematizednbsp;mythology of Snorri gave her the prime place in the stories ofnbsp;origins as the first living creature, the nourisher of the infantnbsp;world, the bringer-to-light of the ancestor of the gods.

The Cimbrians regarded an oath as binding when it was taken on the image of a bull (3), a symbol of fertility and in Swedennbsp;associated with the god of fertility, Freyr. Similarly oaths werenbsp;taken in Sweden on the boar, also a symbol of fertility andnbsp;associated with Freyr, and in Iceland on a ring reddened innbsp;the blood of a sacrificed neat (4).

It is difficult to interpret the slaying of the bull in the duelling ceremony (12), (13), (14), (15). This was part of the holmganga,

For the horse, see quot;VolsiJ)4ttr� in Flateyjarb�k II, 331�36, for the boari Falk, Odensheite s. v.pr�r, for various animals. Nils Lid, quot; Joleband og Vegetasjons-guddom,� Vid. Ahad. Shr. II. H,-F. Kl. (Oslo), 1928, No. 4.

the ritualistic trial by battle, not of the informal einv'igi. Karl Miillenhoff considered it a thank-offering, which it may wellnbsp;have been, but our information is so slight that we can onlynbsp;surmise. At any rate the persistence of the name bl�tnaut in thesenbsp;cases is a witness to the religious character of the act of slayingnbsp;the bull.

We do not know what Porvar�r wished to gain by giving the carcass of the slain bull to an elf (9), but probably he desired thenbsp;strength to overcome Kormakr in the impending second duel,nbsp;an idea not far removed from the fundamental concept of thenbsp;bull-cult.

The pampered and therefore illnatured bulls that were in early times kept for sacrifice and the venerated herd-bulls thatnbsp;grew ugly with age gave rise to tales of dangerous bl�tnaut (7), (21);nbsp;the Icelandic romancers made such a bull part of the templenbsp;equipment of a foreign heathen religion (20) and by way of goodnbsp;measure added a heifer as ritual food for the priestess. For anothernbsp;heathen temple they provided an urarhorn (18) and added thenbsp;story of an urr whose behavior v/as like that of the bl�tnaut ofnbsp;(21). It is interesting that the horn of a wild animal, a thing ofnbsp;a culturally more primitive order than the horn of a domesticnbsp;bull, was considered appropriate ritual furniture. The inclusionnbsp;of a drinking horn in the ritual equipment was not chance, fornbsp;drinking was a fundamental part of the rite. Oxen were usednbsp;in the story of Gefjon and her sons (2) because they are thenbsp;strongest domestic animals known in the North. I see nonbsp;religious import in this tale, nor should I care to claim such fornbsp;Ogvald�s cow (16). quot;Our god� Sibilia (17) and the two heifersnbsp;in which men believed (23) seem to be a romancer�s insertion.

Among the ox-names (26) there is much that is interesting but nothing very convincing except the name Freyr. We havenbsp;already noted the bull sacrifices to this god (8).

The frequent use of bl�tkdlfr and bl�tnctut in the thirteenth and fourteenth centuries and the occasional use of bl�tneyti (28),nbsp;together with the accurate use of bl�tkdlfr in the latter half ofnbsp;the fourteenth century for the golden calf used in the Samaritan

** Maurice Cahen, ��tudes sur Ie vocabulaire religieux du vieux-scandinave. La libation.� Collection linguistique publi�e par la soci�t� de lingnistique de Paris. �nbsp;IX (1921).

bull-cult (22) is evidence that after several centuries of Christianity the Icelanders had some knowledge of cattle-worship. The persistence of hl�tneyti in Modern Icelandic with the meaning �uglynbsp;bull� shows that the mental picture of the vicious pamperednbsp;sacrificial bull of earlier times and the illnatured cult-herd-bullnbsp;of later times was so widespread and tenacious that it gave thenbsp;word a permanent foothold in the language. The complete absencenbsp;of other compounds of animal-names with the prototheme hl�t~nbsp;shows that cattle-worship, and in the light of the accounts wenbsp;have collected here, bull-worship mainly, was the chief animal-cult known in Icelandic tradition and practice (28). It wasnbsp;probably a simple domestic fertility rite intended to insure thenbsp;productivity of the farmer�s herd.

Diese Strophe ist in der Literatur haufig besprochen worden. Wenn wir uns an die im 20. Jahrhundert gefiihrte Diskussionnbsp;halten, sind �� abgesehen von den W�rterb�chern � vor allemnbsp;folgende Literaturstellen zu beriicksichtigen:

F. Better und R. H e i n z e 1, Scemundar Eddall., Leipzig, 1903, S. 136, Strophe 132.

Ernst A. Kock, Notationes Norrcence. Lunds Universitets Arsskrift, Andra delen 1924, � 207, S. 6 f.

B. S i j m o n s und H. Gering, Die Lieder derEdda. Dritter Band: Kommentar. Erste Halfte: G�tterlieder. Halle (Saaie), 1927,nbsp;S. 141 f.. Strophe 135.

Abgesehen von dem Ausdruck gefa bang und dem Pronomen pat glaube ich, daB unsere Strophe kein Wort enthalt, dasnbsp;nicht von mehreren der genannten Autoren richtig aufgefaBtnbsp;wurde.

156 ein Schlagbaum (F. J � n s s o n) gedacht werden soil �, ist einenbsp;Frage, von deren Beantwortung unsere Auffassung des Grund-gedankens der Strophe keineswegs abhangt. Beide k�nnen sichnbsp;drehen (Hda) und rn�ssen stark (vgl. ramt) sein, urn die betreffende Bewegung oft ausf�hren zu k�nnen. Abwegig scheintnbsp;mir ^Bj. M. �lsens Gedanke, daB unter dem tr� ein Kn�ppelnbsp;zu versteken sei. Ebensowenig kann ich Ernst A. Koeknbsp;beistimrnen, wenn er glaubt, daB ramt hier �zauberkraftig� be-deutet. Allerdings sagt ein altes danisches Sprichwort: �Ofte giel-der E)0r Stakkarls Vrede� was so zu versteken ist, daB die mitnbsp;den El�chen eines abgewiesenen Bettlers beladene T�rnbsp;groBen Gefahren ausgesetzt ist. In unserer Strophe wird abernbsp;gesagt, daB ein T�rzapfen, welcher sich oft dreht, um Bettlernbsp;hereinzulassen, stark sein muB. Es hat keinen Zweck,nbsp;das Wort ramt in diesem Falie als �zauberkraftig� auf-zufassen. Der freundlich empfangene Bettler flucht eben nicht.

Ernst A. Kocks Bedenken gegen die �bliche Uebersetzung von (ramt) er, d. h. �muB sein� � oder �ma vaere� scheinen mirnbsp;unbegr�ndet. Ein Vers kann gut sein, ohne den h�chsten An-forderungen der Logik zu gen�gen, und es ist erlaubt, ihn bei dernbsp;Uebersetzung zu verdeutlichen.

Mit Finnur J�nsson� und S i j m o n s � halte ich es f�r ausgemacht, daB sich pat auf den abgewiesenen Bettler bezieht.nbsp;Die neutrale Form ist ganz korrekt, weil das Geschlecht desnbsp;Bettlers unbestimmt ist Der Singular erklart sich durch dennbsp;Wunsch des Dichters, einen konkreten Fall zu schildern, undnbsp;steht also in keinem Widerspruch mit dem Plural ollom. F. J � n s-s o n (�folkene�) und S i j m o n s (�das Bettelpack�) scheinen demnbsp;Pronomen pat eine kollektive Bedeutung beilegen zu wollen, wasnbsp;ich f�r unrichtig halte. G e r i n g gibt die strenge richtigenbsp;Uebersetzung �er�.

� Auch andere Uebersetzungen sind vielleicht denkbar. Vgl. F r i t z n e r III, S. loi a (�Traeklinke�) und HugoPipping Studier nordisk filologi XVIII. 4,nbsp;S. 6. Gering (Sijmons-Gering S. 141) schreibt: quot;riegel�.

� Vgl. Detter-Heinzel, wo M o 1 b e c h als Quelle angegeben wird. 'Hugo Gering Die Edda {1892), S. 104.

Ich kann unm�glich Gering beistimmen, wenn dieser For-scher meint, da6 der baugr ein fiotorldss sei, mit dejn man die T�r zu versehen habe. Wie B j. M. Olsen S. 86 bemerkt, istnbsp;gefa bang ein, mindestens gesagt, sehr gesuchter Ausdi'uck fiirnbsp;�einen Ring anbringen�. Und vor allem versteht man nicht,nbsp;warum die Anbringung einer neuen SchlieBvorrichtung an dienbsp;T�r die �blen Folgen der Fl�che abwehren k�nnte. Allerdingsnbsp;lieBe es sich denken, daB die Form des neuen Schlosses an einennbsp;Gegenstand erinnerte, der eine abwehrende Kraft besaB. Abernbsp;wenn der baugr als ein Abwehrmittel betrachtet werden soli,nbsp;laBt sich eine gute Erklarung der Strophe herausfinden, ohnenbsp;daB man die Hypothese vom fiotorldss herbeizieht. B j. M. Olsen (S. 86) bestreitet �brigens, daB es jemals eine SchlieBvorrichtung gegeben habe, welche den Namen baugr f�hrte.

1. nbsp;nbsp;nbsp;Einerseits enthalt diese Halbstrophe eine Aufforderung, dennbsp;Bettler so zu behandeln, daB seine Fl�che kein Unheil anstiftennbsp;k�nnen.

2. nbsp;nbsp;nbsp;Anderseits kann die zweite Halbstrophe keine Aufforderungnbsp;zu Freigebigkeit enthalten, denn in der ersten Halbstrophe wirdnbsp;vor den Gefahren der �bertriebenen Gastfreundschaft gewarnt.nbsp;Auf keinen Fall d�rfen wir aus der zweiten Halbstrophe einenbsp;Aufforderung herauslesen, jedem Bettler einen Ring aus wert-vollem Metall zu geben, denn eine solche Praxis w�rde nochnbsp;teurer kommen als die Gastfreundschaft, vor welcher die erstenbsp;Halbstrophe wamt

Unser Dilemma ist indessen nicht so unbequem, wie man bisher gedacht hat. Es liegt hier ein Spiel mit dem doppelsinnigennbsp;Worte baugr vor. Baugar nannte man die Metallringe, welche dienbsp;F�rsten ihrem Gefolge verschenkten, aber baugr war auch dernbsp;Name des Afterringes, den man zu zeigen pflegte, wo es daraufnbsp;ankam, den �blen Wirkungen des b�sen Blickes vorzubeugen.

DaB baugr in zwei anderen Eddagedichten in der Bedeutung �Afteming� ben�tzt wurde, glaube ich in einer fr�heren Schriftnbsp;dargelegt zu haben Meine Lesung und Deutung von Alv. 5: 6

hat nachtraglich eine gute Bestatigung gefunden. Dr. E. H. L i n d teilt mir mit, daB man in einer Gegend von Varmland die tiefstenbsp;Verachtung in der Weise kundzugeben pflegte, daB man vonnbsp;einem andern sagte, er sei �auf die Welt geschissen.�

Der Volksglaube, daB man den b�sen Bliek abwehren kann, wenn man der gefahrlichen Person den Hintem zeigt, ist vonnbsp;Feilberg erortert worden

Der Sinn unserer Strophe ist also folgender; Jeden Bettler kann man nicht hereinlassen, ohne zu verarmen. Aber wenn mannbsp;gezwungen ist, einen Bettler abzuweisen, zeige man ihm dennbsp;Hintern, um den �blen Wirkungen des b�sen Blickes zu entgehen.

Man wird mir die Frage stellen, warum es gefa haug und nicht syna baug heiBt, wenn von dem eben beschriebenen Gestus dienbsp;Rede ist. Diese Frage ist nicht schwierig zu beantworten. Dernbsp;Dichter hat das Zeitwort gefa gewahlt, weil gefa batig den Ge-danken des Zuh�rers auf die Gabenspenden der Fiirsten lenkt,nbsp;wobei eine Verh�hnung des Bettlers beabsichtigt wird. DerFiirstnbsp;gibt seinem Gefolge Ringe � gib du dem Bettler einen Ring,nbsp;aber nur den Afterring. Vor seinen Fl�chen brauchst du keinenbsp;Angst zu haben, denn der Gestus sch�tzt dich.

Fin ahnliches Spiel mit dem Doppelsinn des Wortes baugr findet man in Hrbl. 42, wie ich a. a. O. S. 28 f. gezeigt habe.

Die richtige Uebersetzung der Strophe 136 40 Havamal d�rfte also folgende sein:

Stark mufi der T�rzapfen sein, wenn er sich drehen soil, um alle hereinzulassen. Zeige (dem Fremden) den Hintern, sonst wird ernbsp;alles Unheil auf deine Glieder herabw�nschen.

H. F. F e i 1 b e r g ,,Der b�se Bliek in der Volks�berlieferung.quot; ZeUschrift des Vereins f�r Volkskunde, XI.

THE USE OF OLD ENGLISH SWA IN NEGATIVE

CLAUSES!

In a recent article Professor Kemp Malone raises the question as to whether or not an adverbial (i. e. non-conjunctive) swanbsp;may be used to introduce a negative clause. The passage undernbsp;discussion is Beowulf 1138�1145:

Swa he ne forwyrnde woroldrasdenne, ponne him Hunlafing hildeleomannbsp;billa selest on bearm dyde

The swa in this passage Professor Klaeber lists under modal adverbs in his vocabulary, and in his notes he further refinesnbsp;it into quot;under these circumstancesquot;, quot;in this frame of mind�.nbsp;To this rendering Professor Malone objects, on the ground thatnbsp;quot;the use of an adverbial swa to introduce a negative clause isnbsp;excessively rarequot;. His judgment is that this swa is a causalnbsp;conjunction, to be compared with the swa in line 2184 of thenbsp;same poem.

The purpose of this study is to throw further light on the use of swa in negative clauses

This article constitutes a section of a longer study, soon to be published. On �The Use of Swa in Old English.�

^ So, Klaeber. Malone emends woroldrcsdenne to woroldrtBdende, and by the use of a comma after gemunde, makes the sie;a-clause subordinate. He followsnbsp;Olrik in considering Htmlafing the name of a sword. See Malone�s article onnbsp;Hunlafing�, MLN. XLHI, 300 ff. For Klaeber's reaction to Malone�s readingnbsp;Sue his second edn. of Beowulf: Sup. P. 430.

^ The following clarification of my least familiar abbrevations may be of assistance to thereader: AEx = TElfric�s Exodus-, AGen = yElfric�s Genesis:

Swa peak is used to introduce negative clauses, in three different ways; (i) as a true conjunction: �although�, (2) as a pseudoconjunction: �yet�, �nevertheless�; (3) as an adversative adverb:nbsp;�yet�, �however�.

. . he wundra worn wordum cydde, swa peak ne gelyfdon larum sinumnbsp;modblinde menn.

(He wrought many a wonder by His word, though men, blind of heart, believed not in his teachings. � Kennedy)

he weard ]pa |)urh Codes celmihtiges fultum mid pam maegene gestrangod to ]jam swide, paet se pe aer waesnbsp;ealdor waepnedmanna .... swa peah ne geswac he tonbsp;manienne his gingran, etc.

Filling in the blur in the manuscript with the part following wera ealdorman in Ms. C, I translate:

(He was by the help of the Almighty God so strengthened in that power, that he who ere now had been governor of men, after that undertooknbsp;also the instruction and overlordship of women, yet he did not desist fromnbsp;warning his disciples, etc.) �

ASHom. = Anglo-Saxon Homilies (Grein: Prosa, Vol. 3); BH = Blichling Homilies', BR = The Benedictine Rule (Grein: Prosa, Vol. 2); Chm. = Charmsnbsp;(Grein-Wiilker: Poesie, Vol. 2); DG = Dialogues of Gregory (first 75 pages ofnbsp;Hecht�s edition in Grein: Prosa, Vol. 4); LAS = Laws of the Anglo-Saxonsnbsp;(Liebermann); PB: = The Poems of Boethius; PPs. = Paris Psalter; PN�nbsp;= Hilfric�s Preface to the New Testament; POT = Preface to the Old Testament;nbsp;VPs = Vespasian Psalter.

� Another case is DG. 33, 12. There the rebuke administered to Peter (wite . . weorce) serves as the antecedent clause, to which the swa Peah clause is thenbsp;subsequent clause.

i6i

Other cases: Interior; Beow. 2g68. DG. 23, 13. 60, 33. 66, 23 (H). EH. 3, igig. Or. 56, I. 134, 16. 2g. 146, 55. ASHom. 4, 105. g, 230. Josh. 9, 18.nbsp;Final: Beow. ig2g. Guth. 464'�. ASHom. 4, 22g.

]Daet we him on |)am lande lad gefremedon, he haefcf us ]Deah Jjses leohtes bescyrede, etc.

2. nbsp;nbsp;nbsp;Ex. 80 ff.:

rroajja maestrapas men ne cu�'on . . (�though earth-dwellers knew not, etc.�)

Some difference of opinion exists among grammarians as to the status of the word yet. Certain authorities classify it as anbsp;coordinating conjunction, but even these would not contend thatnbsp;it is of the same rank as and, hut, and or. My opinion is, thatnbsp;yet is a concessive adverb, and that while it is often found in anbsp;conjunctive capacity, it is never a real conjunctive adverb likenbsp;where or when, both of which may be replaced by prepositionalnbsp;phrases; for example:

1. nbsp;nbsp;nbsp;No crops grow where (i. e. �in the place in which�) it never rains.

2. nbsp;nbsp;nbsp;He spoke when (i. e. �at the time at which�) his turn came.

� Thorpe (Codex Exoniensis) translates this one as �so that�, but the meaning is plainly concessive.

The ordinary concessive conjunction in English is although (though), but �upside-down� subordination often causes althoughnbsp;to be replaced by the concessive adverb yet. In such cases, whatnbsp;originally stood as the concessive clause is strengthened by beingnbsp;stated as an independent clause, the original independent clausenbsp;remaining, however, as before, and being attached loosely to thenbsp;new independent clause by the word yet with the meaning ofnbsp;�in spite of that� (i. e. �in spite of� the idea expressed in thenbsp;first clause). For a case of such reversed subordination, comparenbsp;the first Case 4, Page 163, with its Latin original:

Cumque moratus esset cum ea per annos viginti, filios aut filias ex ea non accipit.

The term �pseudo-conjunction� may be fittingly applied to the yet in such instances as the above. In Old English, swa isnbsp;sometimes used as such a pseudo-conjunction, in the sense of

Ic waes on gife�^e iu ond nu (Jpa) syxtyne si^um on saebate,nbsp;mere hrerendum mundum freorig,nbsp;eagorstreamas; (is dys ane ma),nbsp;swa ic sefre ne geseah aenigne mann,

com micel fyrbyrne on Romeburg, Jiaet Jiaerbinnan forburnon XV tunas, swa nan man nyste hwanannbsp;Jiaet f5ur com. (. .which burned fifteen wards;nbsp;no man knew, etc.)

J)aer waeron xxx M ofslagen, and at Jiam geate of-tredd, swa nan men nyste hwonon sio ivro(h)t com.

� So Root, Kennedy, and Krapp. Grein has quot;wie,� but in what sense it is hard to tell.

4. ASHom. 10, 73:

Hi do, WBeron samod drohtniende Eetga�dere twentig wintra, swa hi nan beam ne begeaton. . livingnbsp;together twenty years; yet they had begotten, etc.�)

III. Swa in Result Clauses.

1. nbsp;nbsp;nbsp;Finn. 41:nbsp;nbsp;nbsp;nbsp;�

2. nbsp;nbsp;nbsp;Beow. 1506:

swa he ne mihte, no he J)am modig waes, waepna gewealdan'h (�so that he could not . . .�)

3. nbsp;nbsp;nbsp;Az. 60:

tosweop and toswengde J)urh swi^^es meaht liges leoman, swa hyra lice ne scod. (�so thatnbsp;it injured not, etc.�)

4. nbsp;nbsp;nbsp;Or. 198, 8:

and J)a burg on niht abraec, swa J)a nyston ]De J)3erinne w�8eron. (quot;captured the city . . .in such a way that, etc.�)

3. AGen. 8, 12 b:

and abad swa �^eah seofan dagas and asende ut culfran; swa heo ne gecyrde ongean to him. (quot;sent out a dovenbsp;so that^'^ [with the result that] she did not return. .�)

Other cases: Beow. 2006. Gen. 381. 431. 611. 733. Dan. 659. Az. 187. El. 340. And. g86. Or. 296, 34. ASHom. 10, 2g8. AEx. 34, 38.

TV. Swa pest in Result Clauses.

1. nbsp;nbsp;nbsp;And. 260:

2. nbsp;nbsp;nbsp;EH. 3, 156:

swi^e geswenced waes, swa pmt he for J)y sare ne mihte furJ)on his hand to muj)e gedon.

The Laud and the Cambridge Mss. do not have this swa, and Grein-Wiilker omits it. The Latin is quot;quae non est reuersa ad eum.� The result notion fits asnbsp;well as any, although a bromidic transitional swa (= �and�) may be intended.nbsp;The translator apparently means to equate quae with heo.

. . to Cristes ]Deov/dome gecyrdon . . swa hie . . wendan noldan. (quot;. . turned to Christ's service . . so that they would not return.�)

and fcr wunode fit mon^a fsece, swa piBt olaer naenig aerendraca betweonan ne ferde him and his ge-maeccan, (quot;so that no messenger passed, etc.�)

For ^an 5e God gewitnode ealle hys wimmen, swa Jgt;cst heora nan ne mihte habhan aenig cild . . (�so that nonenbsp;of them could have a child, etc.�)

DG. 67, IT. 72, 9. EH. 2, JJ57. 4, iti6. 374^. 31431. 5, 396. VPs. 103, 35. BR. 70, 14. ASHom. 10, 402. AGen. 9, 23. AEx. 9, 10�ii. Deut. 9,9.

Jpaet he Eotena beam inne gemunde, swa he ne forwyrnde woroldraedendenbsp;Jponne him Hunlafing, hildeleoman,nbsp;billa selest, on bearm dyde;

Klaeber�s translation of the last three lines of this passage (Beowulf: note, P. 170) is as follows:

�Under these circumstances (or, in this frame of mind) he did not refuse (him, i. e. Hunlafing) the condition, when Hunlafing placed thenbsp;battle-flame (or: Battle-Flame), the best ofnbsp;swords, on his lap.�

Klaeber, then, considers the stfiai-clause an independent one, and the swa a modal non-conjunctive adverb.

Malone uses a comma after gemunde and takes the swa-clause to be a dependent causal clause modifying gemunde. He alsonbsp;emends woroldradenne to read woroldrcsdende. His translation is:

In this case, it seems to me that Malone is on the right track. Waiving the matter of woroldrcsdende, we may form a judgmentnbsp;on the basis of the position of adverbial and conjunctive swa innbsp;negative clauses. As I hope to show in the course of this article,nbsp;adverbial (non-conjunctive) swa occurs almost without exception

i65

in the interior or final position (See Pp. 167�174) Conjunctive swa, on the other hand, is invariably found in the initial positionnbsp;(See Pp. 166�167). The swa in Beow. 1142 must therefore be eithernbsp;a conjunctive adverb or a conjunction. The context calls for thenbsp;latter, a causal conjunction: �since�. Malone cites the swa innbsp;Beow. 2184 as evidence that causal swa is to be found in negativenbsp;clauses. Other cases he might have added are as follows:

and -waeron on heora modgepance swidlice afyrhte and gedrefde, swa hit naenig fyren waes. (quot;seeing thatnbsp;there was no sin.�)

Wiste he drihten ana, hwa hi laewend and myidra waes, swa swa him nan bemiden beon maeg. (quot;since nothingnbsp;can be concealed, etc.�)

for pon ic ne maeg eal pa monigfealdan yfel emdenes areccan, swa ic eac ealles pises mid-dangeardes na maran daeles ne angite. (quot;as I amnbsp;not acquainted, etc.�)

Nim nu, lareow Apolloni, swa hit pe ne mislicie . . (quot;f/ it is not displeasing to thee�)

For the use of �so that� {= �therefore�) at the head of an independent clause with back-reference to a preceding sentence of which it is semanticallynbsp;a dependent clause, see Pp 167�16S,

Tupper (Riddles of the Exeter Book [glossary]) uses �where� here, but I follow Grein, who makes an ohne daf} clause out of it. Similar cases are to benbsp;found in Old Saxon: Heliand 798. 813. 5777, et. al. See Sehrt�s Vollstandigesnbsp;IV�rterbuch zum Heliand, P.484, Col. i, line 23 ff.

i66

VII. Swa in Modal Clauses.

Here the swa is a conjunctive adverb of manner, and is always found in the initial position.

A. The simple Ste^a-clause.

1. nbsp;nbsp;nbsp;Beow. 2331:

2. nbsp;nbsp;nbsp;Gen. 1565:

3. nbsp;nbsp;nbsp;Rid. 3, i;

Other cases: Beow. 257^?. 2585. Gen. 90J. Dan. 20. CS. 412. El, 838. HP. 3, 43. PB. 26, 92. MM. 63. PPs. los, 22. iig, 5. Or. 254, 9. BR, 34, ig.nbsp;DG. 46, 6. 61, 6. ASHom. 8, 59. With swa swa: VPs. 38, 6. DG. 13, 13.nbsp;BR. 18, 14. PNT. X269. In an ellipsis: EH. 3, 4g.

B. Swa introducing the second member of anbsp;comparison of inequality.

1. nbsp;nbsp;nbsp;Or. 180, 18:

Jja wear�' Tiber seo ea swa fledu swa heo naefre aer waes (��so swollen as it never had been.�)

2. nbsp;nbsp;nbsp;POT. 60�61:

Other cases: Beow. 1046�1048 Jud. 67�68. EH. 3, i8g�go. 5, S59�� 62. POT. 4g3�5. DG. 69, ii. AEx. 9, 24. jo, 6. Swa . . . efne swa DG.nbsp;J5, 31�33. 25, 29. Swa cac.....swa: EH. i, 388�790

C. Swa in the sense of quot;as if�.

1. nbsp;nbsp;nbsp;WComp. 24:

2. nbsp;nbsp;nbsp;DG. 26, 29:

and of J)3ere tide he wear^ swa fremde Jpaere costunge, efne swa he naefre J)one had on his lichaman naefde. (f'as henbsp;had never had, etc.�)

Brooke (EELit. P. 142) has ��how�; Warren (Treas. Eng. Lit. I: 45), ��whenquot; These might also be considered swa . . . swa clauses of result.

D. Swa introducing the second member of as quot;as� . . . �so� correlation.

Not counting 16 cases of adverbial swah peak, there are 36 cases of non-conjunctive swa in the monuments examined. Ofnbsp;these latter, 20 occur in the interior position of the clause and 16nbsp;at the end. There is not one indisputable case of adverbial swanbsp;at the beginning of a clause. Six cases have been put forwardnbsp;as such, the twists and turns of which will be discussed in thisnbsp;section.

Caldea cyning, od 1)�^ cwelm gesceod, swa him oler eordan andsaca ne wassnbsp;gumena senig.

The swa in the above Grein translates as �wie�, and Kennedy obscures the matter with a transitional �and�. Bouterwek andnbsp;Thorpe both begin a new sentence with �so�. But no completenbsp;stop is necessary, if we follow Grein�s punctuation and considernbsp;the sic^a-clause one of �cause� or �evidence�. My translation is:

(He [Nebuchadnezzar] guarded the heroes� treasure and the lofty city, this wise and mighty leader ofnbsp;the folk, Chaldea�s king, until death overtook him,nbsp;for over all the earth there was no man to opposenbsp;him.)

Forpon hie leng ne magon healdan heofonrice, ac hie to helle sculonnbsp;on pone sweartan sid, swa pu his sorge ne pearftnbsp;beran on pinum breostum . .

Here Bouterwek and Grein both begin a new sentence with swa, and translate �so.� Even though this is done, the initial so

I except, in this statement, the adverb of manner used in correlation with the conjunctive (modal) adverb in the �as� . . . �so� combination. But in suchnbsp;cases the swa is protected by the conjunctive swa in the preceding clause.

of the sentence leans on the preceding clause. In modern English, a �so that� clause is sometimes used to introduce a pseudoindependent clause. Isay fseudo because the clause is semanticallynbsp;dependent, in spite of its standing alone. The �so that� has givennbsp;up its more exact result notion for the related one of inferencenbsp;or conclusion, with back reference to the preceding clause. As annbsp;example consider:

John Macy: World�s Literature, 165:

�The Pope at Rome was often the unifying personality at the head of the universal Church, who presided notnbsp;only over the spiritual life but over the temporal affairsnbsp;of the many countries and principalities of Europe; andnbsp;at least one great and strong man, Hildebrand, Popenbsp;Gregory VII, was a supreme monarch with kings undernbsp;his thumb. So that Europe did not quite cease to be Romannbsp;through these long centuries, etc.�

Similar cases in Old English are: EH. 4,3749 (See Latin original; also Sellars� translation) and AEx. 9, 10�ii (Cp. Latin).

What is one to call the swa (or swa pcet) in such cases? It is almost the toss of a coin. But since meanings are older thannbsp;grammatical classifications, I prefer to put such cases on a semantic basis, and to class them as subordinating conjunctions (�sonbsp;that�), even when they are at the head of an apparently independent clause.

(Hence they [Adam and Eve] may no longer hold their heavenly kingdom, but theynbsp;must travel the dark road to hell, sonbsp;that thou [Satan] needst feel no sorrow, etc.)

3. Gen. 288 (Grein-Wiilker punctuation): 20 ic maeg hyra hearra wesannbsp;raedan on pis rice. Swa me paet riht ne pinced,nbsp;paet ic oleccan awiht purfe

2� Professor Malone, in the article already cited, speaks of the possibility of such non-conjunctives at the head of a clause with �verbs of thinking.� Mislednbsp;by the punctuation of Grein-Wiilker, I once took Case 3 to be of that type. Butnbsp;Case 3 is a causal clause, and no others are at hand. I remain, therefore, skepticalnbsp;on this point.

(. . ich kann ihr Herr sein, herrschen in diesem Reiche; drum diinkt es mich nicht recht, daB ich irgend wie,nbsp;um eines Gutes willen, Gott schmeicheln sollte: nichtnbsp;langer will ich ihm unterthanig sein.)

(I may be their lord, and rule this realm. It seemeth no wise right to me that I should cringe a whit to God for any good.)

With these I disagree. My judgment is that the colon used by Grein-Wiilker after cBnegum should be replaced by a comma, andnbsp;that the �swa ... cenegum� portion should be read as the antecedent clause, with the �ne . . . wurpan� as the subsequent, in anbsp;causal sequence. My translation follows:

(Since it does not seem right to me that I should cringe a whit to God for any good, no longer will I remain his underling.)

ne past to nahte nyt ne bip peet man godne mete ete oppe pset betste win on gebeorscipe drince,nbsp;gif pa�t he hit eft spiwende anfortetep past henbsp;ier to blisse nam and to lichoman nyttnesse; swanbsp;we ponne pa gastlican lare unwaerlice ne sceolannbsp;anforlastan, pe ure saul big leofap and feded bid;nbsp;swa se lichoma buton mete amp; drence leofian nenbsp;maeg, swa ponne seo saul, gif heo ne bid mid Godesnbsp;worde feded gastlice hungre amp; purste heo bid cwelmed.

(It is all to no purpose for a man to eat good meat or at a feast to drink the best of wine, if it happeneth that he afterwards spews up and losesnbsp;that which he previously received for enjoyment, and for the advantagenbsp;of the body. So, then, we must not unwarily relinquish the spiritualnbsp;teaching by which our soul liveth and is fed. As the body cannot livenbsp;without meat and drink, so then the soul, if she be not spiritually fednbsp;with God�s word, will perish through hunger and thirst.)

It is the first swa that that we are concerned with, but I have italicized those in the sentence following because I feel that thisnbsp;second clause is a restatement, in concentrated form, of the onenbsp;immediately preceding it. I assume that the first swa is thenbsp;adverb of manner in a defective correlative modal pair (quot;as� . . .nbsp;�so�). A case of such omission of the conjunctive adverb is atnbsp;hand^B viz.

Cp. the following from OHG in E. Steinmeycr: Die kleineren Ahd. Sprach-denkmdler, P. 378, LXIX: Christ uuarth giuund | tho uuarth he hel gi ok gisund. that bluod forstuond: so duo thu bluod!

ForJ)on J)a oferliydigan blissia�^ for ]pam arum Jre hi begyta�', swa eac full oft Jja ea�'modannbsp;for heora forsewennysse.

(For (as) the proud are made glad because of their honor, so those that are humble very often rejoice over contempt and disgrace.)

(As it is all to no purpose for a man to eat good meat, or at a feast to drink the best wine, if it happeneth that he afterwards spews up andnbsp;loses that which he previously received for enjoyment and for the advantagenbsp;of his body, so, then, we must not unwarily relinquish the spiritual teaching by which our soul liveth and is fed.)

Judeas Crist ahengon, dydon das da Jia wyrrestan; haelon paet hy forhelan ne mihtan: swa peos daednbsp;naenige l)inga forholen ne wurjae J)urh pa haligannbsp;Cristes rode. Amen.

(The Jews hung up Christ, they did of deeds the worst, they hid that they could not hide; so may this deed be in no wise hidden throughnbsp;the holy rood of Christ. Amen.)

Judeas Crist ahengon, gcdidon him daeda pa wyrstan; haelon, p$t hi forhelan ne mihton; swa nsefre deos daednbsp;forholen ne wyrpe per crucem Christi

(The Jews hanged Christ, they did to him the worst of deeds; they concealed what they were not able to conceal. So never may this deed become concealed. Per crucem Christi.)

Grendon ^ is in practical agreement with Cockayne in the translation of these two cases.

Now these two examples seem at first glance to prove that an adverbial swa may introduce a negative clause, but closer scrutinynbsp;will show that the swa in each sentence is (as in Case 4, thisnbsp;section) the adverb of manner of a defective correlative modalnbsp;pair, an independent clause standing in each case for what wouldnbsp;normally be the ��as�-clause of an �as� . . . �so� correlation.

O. Cockayne; Lecchdoms, Wortcunning and Starcraft of Early England. Ill: 60�61; I: 390�393.

It is characteristic of most Teutonic charms that they are built on a simile frame, usually by way of reference to something innbsp;church or biblical tradition. Jacob Grimm long ago remarkednbsp;this :

�Viele formeln beruhen auf bloBer sympathie zwischen gleichnis und wirkung. Das blut, das feuer soil so still stehn, wie Christus still am kreuz hiengnbsp;(XLI sanguis, mane in venis, sicut Christus pro te in poenis; sanguisnbsp;mane fixus sicut Christus crucifixus); wie der Jordan still stand bei dernbsp;taufe (VIII); wie die menschen am j iingsten tag still stehn werden (XXXII).nbsp;das feuer soil seine funken behalten, wie Maria ihre jungfrauschaft behieltnbsp;(XXVI); dem wurm im fleisch soil so weh werden, als es Petro weh ward,nbsp;da er seines herrn marter sah (XXXVI), etc.�

In the High German charms given by Grimm and others, the simile frame is regular, that is, a correlative clause of comparison,nbsp;either complete or defective. Examples:

Soso Krist gibuozta themo sancte Stephanes hrosse thaz entphangana, so gibuozi ihc it mid Kristes fullesti thessemo hrosse.

behalt dein funk und flammen, wie Marie ir jungfrauschaft und er behalten hat vor alien mannen

ich gebut dir huf und horn, das du als lutzel zerbrechist als got der herr die wort zerbrach, do er himel und erd beschuof.

Professor Grendon in his study of Old English charms, likewise calls attention to this �sympathie� ... as he puts it,nbsp;quot;the mention of something which often bears only a remotenbsp;relation to the subject of the charm�. He cites in particular thenbsp;two Old English charms here under discussion' (Cases 5 and 6,nbsp;this section), giving as the formula, �The cross of Christ^wasnbsp;hidden and has been found�, and as the associated idea, �so maynbsp;these cattle be found�.

Now in expressing this �gleichnis� and �wirkung�, apparently what is known to English as �upside-down subordination� wasnbsp;often resorted to, just as in the case of concessive clauses (Seenbsp;above P. 164. By that process, the subordinate clause assumes suchnbsp;importance in the mind of the speaker or writer that it becomes

172 an independent clause. What results is an asyndeton, in whichnbsp;each clause may stand alone. A case in kind is a Pomeranian charmnbsp;cited by McBryde :

Das Kreuz ward verborgen ward wiedergefuiiden durch die Sankt Hell-man � Also wahr mufi der Dieb widerkehren und sich wiederfinden mit dem gestohlen Gut.

Except that the second clause here is affirmative, this charm is very much like the two we are considering, and one senses here,nbsp;too, what would be normally an �as� . . . �so� arrangement.

Another case is on record in which, although the same similitic idea is apparent, the construction does not show it in the slightest.nbsp;LWS. 1,392:

Adiuro te infans si es masculus an femina per patrem et filium et spiritum sanctum ut exeas et recedas et ultra ei non noceas neque insipientamnbsp;illi faceas. Amen. Videns dominus flentes sorores Lazari ad monumentumnbsp;lacrimatus est coram ludeis, et clamabat, �Lazare, veni foras!� et produit,nbsp;ligatus manibus et pedibus, qui fuerat quatriduanus mortuus.

(I charge you, infant, be you man-child or woman-child, in the name of the Father, the Son, and the Holy Ghost, that you come out and withdraw,nbsp;and further that you do not harm her, nor do any injury to her. Amen.nbsp;The Lord, seeing the sisters of Lazarus weeping at the tomb, wept in thenbsp;presence of the Jews and exclaimed, �Lazarus, come forth!� and he camenbsp;forth, wrapped hand and foot, he who had been four days dead.)

In a third charm, one for blood-stopping, the �gleichnis� alone is given, although the subsequent clause can be readily inferrednbsp;from it. The case is from Cockayne (LWS):

(Make a cross of Christ and sing this thrice over the place, together with a pater noster. The soldier Longinus pierced the Lord with his lance andnbsp;the blood ceased to flow and his pain ended. � Cockayne)

Cockayne gives us as a source for the above charm �a transcript forwarded by a friend�. His ending it with a semi-colon indicatesnbsp;that he probably considered it incomplete. If that surmise is

J. M. McBryde, Jr. MLN 21; 180�183. His source for this charm, as acknowledged by the author, is Ulrich Jahn: Hexenwesen und Zauherei in Pom-mern.

(As the Saviour�s blood ceased flowing and His pain ended when the soldier Longinus pierced His side with a lance, so may this blood ceasenbsp;flowing and the pain end.)

Since, then, the simile was a frequent device in the charm, and since (as has been shown in Case 4, P. 169�170) the conjunctivenbsp;adverb of the antecedent clause was sometimes omitted, thesenbsp;two cases (Chm. V�A, 13 and Chm. V�B, 10) may be considerednbsp;defective correlative (modal) pairs, to be translated as follows^�:

(As the Jews hanged Christ, did to him the worst of deeds, tried to hide what they could not hide, so may this deed never be concealed.)

(�Yet might these vain and miscreant men in no wise work his death/quot; � Kennedy)

Other cases: EH. 4. i8g. 5, 2346. VPs. 14J, 20. BH. 215, 12. BR. 46, 18. 48,3. 9. ASHom. 16, 316. AGen. 27, 27 (note). 18, 21. AEx. 8, 26. 10, 11.nbsp;21, 23. DG. 50, 20. LAS. 164, 21. 210, 8�I. 262, 4�I.

Other cases: Beow. 2ogi. And. 13^3. Gen. 718. PPs. 107, 10. VPs. 118, 85. SS. 3gi. ASHom. 10,614. BH. J75, 36. AGen. 4, 15. 18, 13. 23, 11. 42, 10.nbsp;48, 18.

2. nbsp;nbsp;nbsp;As a non-conjunctive adverb (modal) introducing the subsequent

Let me make it plain before drawing my conclusions that the cases discussed above cover every negative clause in the wholenbsp;body of Old English poetry, as well as those of a substantialnbsp;and representative portion of the prose. Of my findings withnbsp;regard to the poetry, therefore, there can be little doubt; andnbsp;while an examination of prose monuments other than those Inbsp;have studied may qualify my conclusions somewhat, that possibility is slight, seeing that the prose works I have examinednbsp;are tjqiical of the prose literature in general.

What conclusion, then, can be drawn from these 154 cases, as regards the position of swa in the clause ? As would be expectednbsp;in the 57 cases where the word is used as either a true or pseudoconjunction, as well as in the 42 conjunctive adverb cases, thenbsp;swa is found in the initial position. And of the 53 non-conjunctivenbsp;cases, only 3 (those of the �as� . . . �so� clauses of comparison)

are found at the beginning of the clause, and these are accompanied by an extenuating circumstance � the fact that the clauses in which they occur do not stand alone, but are accompanied by a preceding sze'fl-clause. Since, then, in this collectionnbsp;of representative cases, there is not a single clear case of adverbialnbsp;(non-conjunctive) ^wa in the initial position, one may with safetynbsp;conclude that the appearance of swa at the head of a negativenbsp;clause is -prima facie evidence of its conjunctive character.

CHANGE OF MEANING BY ANALOGY

In spite of the great progress which has been made in the field of lexicography within the last few decades there is probably nonbsp;phase of linguistic science which still leaves so much to be desirednbsp;as the matter of the investigation into the causes and the development of semantic changes as well as the proper arrangementnbsp;and recording of meanings.

Strangely enough the editors of some of the best High German dictionaries have been prone to regard these points less seriouslynbsp;than we should expect in view of the fact that the problems innbsp;the field of sematology are at the same time the most difficult andnbsp;most important of all the problems of lexicography. The editorsnbsp;of the Grimm Dictionary, following an antiquated medievalnbsp;tradition, carried out the plan of putting Latin synonyms innbsp;lieu of definitions in the German, a procedure justly criticisednbsp;by Hermann Paul ^ and fortunately corrected in later serialnbsp;issues. The defence of this plan by the editor, Jakob Grimm,nbsp;is quoted in part below since it throws an interesting light onnbsp;the views held on this matter at the time that great work wasnbsp;begun: �Auch wird man nicht die Verstandlichkeit aller imnbsp;Lateinischen gebrauchten Ausdriicke fiir alle Leser des W�rter-buches verlangen; die ihrer unkundig sind, h�pfen mit leichtemnbsp;FuBe daran vorbei und finden sich dennoch zurecht, wie sienbsp;vor�bergehen, wenn sie auf ein Wort gestoBen sind, dessen Gehaltnbsp;sie gar nicht anzieht . . . Jeder Leser bringt eine Menge Ver-standnisse mit sich, die ihm den Zutritt zu den W�rtern leichtnbsp;machen . . . Nicht zu verachten ist auch, daB der Gebrauch der

^ Ueber die Aufgaben der wissenschaftlichen Lexihographie. Sitzungsber. cl. philos.-philol. Klasse d. Bayer, Akad, d. Wiss. 1894, p. 63 ff.

fremden Sprache die Erklariing der unz�chtigen W�rter l�blich verdeckt und dem allgemeinen Verstandnis gewissermaCen ent-zogen wird� (DWB I XLI). If today these words make an impression almost of frivolousness in the face of so serious a tasknbsp;they may be excused, at least, on the ground that they werenbsp;written many years ago before the study of semantics had receivednbsp;any serious attention. But the fact that the user of the dictionarynbsp;is in one way or another to fill in the gaps in the semantic historynbsp;of a word seems scarcely justified in a work like Weigand�s Deutsches W�rterhuch, whose editor, Hirt, expresses himself as followsnbsp;in the introduction to this work: �Mancher wird vielleicht einenbsp;eingehendere Darstellung der Bedeutungsentwicklung vermissen.nbsp;Da aber die meisten heutigen Bedeutungen, die selbstandigennbsp;durch ; getrennt, au�erdem die mitt el- und althochdeutschennbsp;genau angef�hrt sind, so wird man sick hei einigem Nachdenkennbsp;die Bedeutungsentwicklung leicht klarmachen konnen.� ^ Certainlynbsp;as long as this phase of lexicography is taken so lightly we neednbsp;not expect a dictionary whose discussion of the development ofnbsp;meaning is on a par with other parts of the work. The users ofnbsp;the NED are much more fortunate in this respect in that thenbsp;editors of that great undertaking, early recognizing the difficulties and importance of this problem, have given us a muchnbsp;more careful and informative discussion of the semantic historynbsp;of a word than any that has appeared elsewhere.

It is undoubtedly much more difficult to trace the beginnings of semantic changes or the causes for these changes than it isnbsp;to trace phonetic forms. For this reason there will always benbsp;gaps in the semantic history of a word, especially in the oldernbsp;dialects. But by no means all possible phases of semantic historynbsp;have thus far been studied and it is hoped that many morenbsp;avenues of approach to the field may be found. The presentnbsp;paper is an attempt to discuss one of these neglected avenuesnbsp;of approach, namely change of meaning by analogy with anbsp;closely associated language, in this case Latin.

For more than one thousand years High German was most closely associated with Latin. Old High German prose literaturenbsp;consisted almost without exception of translations from Latinnbsp;originals. Our earliest records of the Old High German were the

glosses, the principal means by which the German novice who knew only his mother tongue, the lingua theodisca, was introducednbsp;to the lingua latina, the language of the Church. These glosses,nbsp;extending unbrokenly down almost to the end of the fifteenthnbsp;century, form the beginning of our science of lexicography. Thenbsp;invention of printing and the consequent appearance of our firstnbsp;dictionaries made no great change in the method of recordingnbsp;the meanings of words. Down to the eighteenth century, in factnbsp;to Adelung�s great work (1775), practically without exceptionnbsp;German dictionaries recorded meanings in Latin. The mostnbsp;important of these are the w'ell known works of Aler, Frisch,nbsp;Frisius, Kirsch, Maaier, Schottelius, Steinbach, Stieler, etc., sonbsp;often quoted by the DWB and others. The importance of thisnbsp;fact from a linguistic standpoint cannot be overestimated becausenbsp;it makes clear the condition of bi-lingualism which was universalnbsp;among the writers of the Middle Ages, that is, among those whonbsp;have left us the records of the older dialects. That this constantnbsp;association of two languages with each other in the minds ofnbsp;thousands of individuals through centuries should have leftnbsp;few or no traces in the semantic content of the High Germannbsp;seems inconceivable, and yet there is little evidence of the recognition of any such influence in our High German dictionaries.nbsp;A few most striking cases of analogical influence may occasionallynbsp;be mentioned, but on the-whole, meanings are universally regardednbsp;by the lexicographers as having developed spontaneously eithernbsp;from the fundamental or the secondary meanings within the wordnbsp;itself. This, for example, is consistently the attitude of Paul innbsp;his Deutsches W�rterhuch, in most other phases of its discussionnbsp;of reieanings a model of excellence, and for its size in this regardnbsp;probably the best German dictionary published. Possible analogical influences receive no consideration at all. If OHG zunganbsp;shows the development of meaning �speech; a peninsula��, thesenbsp;are regarded as natural semantic developments in the German,nbsp;Lat. lingua and Gk. ylSgt;aaa being mentioned only to supplynbsp;semantic parallels. The possibility of the borrowing of thesenbsp;meanings by the Latin from the Greek and in turn by the Germannbsp;from the Latin is not considered worth mentioning. That thesenbsp;three linguistic groups should have developed independently thenbsp;same figurative expression for �language� is not an impossibility, but if two of them have been closely associated and if it

can not be shown that these meanings had existed independently before this association, it is certainly not justifiable to disclaimnbsp;possible analogical influence; for the transference of the semanticnbsp;content of one language to another is no more remarkable thannbsp;the commonly accepted transference of the phonetic peculiaritiesnbsp;of one language to another.

The association of the highly developed Latin, the language expressing the advanced culture of the Roman civilization, withnbsp;the crude and undeveloped Old and Middle High German wasnbsp;of the utmost importance for the latter. Naturally the introductionnbsp;of the Rom.an-Christian culture and learning found the Oldnbsp;High German inadequate to express in the native form innumerable ideas familiar to the user of Latin. The result wasnbsp;a rapid development of the Old High German in two principalnbsp;directions, the one, the enlargement of the vocabulary throughnbsp;the introduction of numerous loan-words from the Latin, thenbsp;other, the enrichment of the semiantic content of the nativenbsp;vocabulary. We are concerned here only with the latter development. The semantically highly developed Latin increased usesnbsp;for native High Gernran words principally in three ways. Thenbsp;first thru the formation of new German compounds, usuallynbsp;literal translations of Latin compounds, as, for example circum-dare: umbigehan, umhisellan (cf. OE ynihsellan), which havenbsp;sometimes been called translation loans. In the second typenbsp;either a new or an old Germanic comipound was used to expressnbsp;less exactly the m.eaning of a Latin word or compound, as,nbsp;for example, �einfach�: simplex; �einzig�; unicus. In thesenbsp;two types the native High Germian vocabulary was increasednbsp;thru the formation, for the most part, of new native compoundsnbsp;needed to express ideas of the associated Latin. In the thirdnbsp;type the High German word, unchanged in form, is given newnbsp;meanings by its association with the Latin word and it is thisnbsp;type especially with which we are concerned here.

It is the nature of the human mind to associate linguistic phenomena in groups whether consciously or unconsciously. Notnbsp;only single words may be thus associated but phrases, wholenbsp;sentences, syntax, grammar, meanings, etc., all are grouped innbsp;the mind of the individual in an infinite variety of complications.nbsp;In grammar, for example, there is the grouping of the differentnbsp;cases of the single noun, the different tenses, modes, persons,

i8o number of the verb, or the ramifications of the word stem intonbsp;a great variety of forms related both phonetically and semantically. � For the bilinguist the strongest association is the synonymnbsp;group, t he word in one language recalling this or that synonymnbsp;in the other. Even tho these are not exact equivalents, and thisnbsp;is usually the case, the one readily recalls the other and the usesnbsp;of the one will be readily transferred to the other. This transference of usage often entirely unconscious in the bilinguist,nbsp;results in the creation of new meanings for words and thesenbsp;usually in the less highly developed language. ^

To illustrate, let us take the example used before. OHG. zunga and Lat. lingua both mean primarily that most useful part ofnbsp;our anatomy, the tongue. Let us suppose the OHG word wasnbsp;used only in the primary sense while the Latin word because ofnbsp;association with Gr. ykmaaa had developed also such meaningsnbsp;as �speech, language; a peninsula, etc.� The German bilinguistnbsp;would be likely to use zunga also in these special uses becausenbsp;of the perseverative tendencies created by his use of lingua.nbsp;Thus the semantically more highly developed language wouldnbsp;enrich the less highly developed, because the former would havenbsp;more uses for its words and these uses would be transferred tonbsp;the other language in the mind of the user of both. This associationnbsp;would tend to the new use of the word zunga in the OHG evennbsp;tho sprdhha might have been just as adequate to convey thenbsp;thought. The writer has noted repeatedly, especially in Notker,nbsp;that where the two languages were associated, as in translation,nbsp;zunga always translated lingua in the sense �speech�.

8 Compare Paul, Prinzipien der Sprachgeschichte p. 25 ff.: Borowitz, Die Uebersetzungstechnik Heinrich Steinh�wels, Hermaea XIII, p. 44 says:nbsp;�Unsere ganze Sprechtatigkeit beruht ja auf assoziativen Vorgangen. Gleich-artige Elemente verbinden sich im BewuBtsein zu Vorstellungsgruppen, undnbsp;wird nun eins dieser Elemente in der Sprache reproduziert, so stellen sich dienbsp;anderen auf assoziativem Wege, d. i. automatisch, ganz von selbst ein.�

* The majority of translations of the OHG and MHG period represented little more than literal, word for word transferences from one language intonbsp;the other. Translation had not been developed as an art. Even as late as thenbsp;fifteenth century this practice was common, as we may see from the wordsnbsp;of Heinrich Steinh�wel [Aesop, Oesterley, p. 276), who, agitating for a reformnbsp;in this matter, explains how he believes in translating �in ringem verstentlichemnbsp;tiisch on beheltne ordnung der wort gegen wort, ouch nit gelyche sinn gegennbsp;sinnen, sender offt mit zuogeletten worten.��

In order to show in an especially striking manner the influence of the Latin on the development of meaning in the High German,nbsp;let us take a case in which the Latin word represents a widenbsp;expansion of ideas new to the German, i. e. ideas introduced bynbsp;the Roman-Christian culture. Such a word is Latin ars. Thisnbsp;word meant primarily �skill in joining or making something��nbsp;(cf. artus, Gk. aQtvco dget)]) and developed the followingnbsp;meanings, practically all of them used in the words to be discussednbsp;below; i. any physical or mental activity, so far as it is practicallynbsp;exhibited; a profession, art. 2. Science, knowledge; (a) the theorynbsp;of any art, science, etc., grammar, rhetoric, etc.; (b) the knowledge,nbsp;art, skill, workmanship employed in effecting or working uponnbsp;an object; (c) the object artistically formed; a work of art;nbsp;(d) cunning, artifice, fraud; (e) a machine (Med. Lat. only). Beingnbsp;primarily a learned word, ars was acquired very early by thenbsp;Old High German cleric learning Latin because it was the namenbsp;for at least a part of his course of studies. The trivium, consistingnbsp;of grammar, logic and rhetoric and the quadrivium, made up ofnbsp;arithmetic, geometry, music and astronomy were^called thenbsp;septem artes liberates. � Since the Old High German had no wordsnbsp;to represent these ideas, either the Latin word itself (cf. Pipernbsp;Notker I, 10, ii, 65, etc.), or a German word representing an ideanbsp;synonymous with one of the meanings of ars might be used.nbsp;In the OHG period this word was list �Kenntnis�, i. e. �knowledge, science��, which began to be used to translate ars at first,nbsp;no doubt, only in this sense. But once associated with the Latinnbsp;Word in this meaning, it was easy to use it for the special branchesnbsp;of knowledge, the artes liberates, especially since there was nonbsp;other word to express these meanings. In the Regula Benedict.nbsp;we find artis spiritalis translated by listi atumlihhun (IV); ipsasnbsp;artes: dea selbun listi (LVII); artes diversos: listi missilihho (LXVI);nbsp;in arte aliqtio: in listi edeslihheru (XLVI). Other correspondencesnbsp;are: liberalium artium: buohliste (Notker, Boeth. V); theiwarinbsp;gougalares list (Otfrid, Evangel. IV 16, 33) cf. magicis artibusnbsp;facta (Hrab. Maurus in Matth. p. 147a); magicae artes: zoubirlistenbsp;(Notker, Schilter, Thesaur. I 10 a, 6); ludicra arte: spilelisti (Can.

� Cf. Specht, Geschichle des Untcrrichtswesens in Deutschland, Stuttgart 1885; Hoffmann, Dev mittelalterliche Mensch, Gotha 1922 and particularlynbsp;Burdach, Der Ackermann aus B�hmen. Vom Mittelalter zur Reformation III,nbsp;P- 334

i82 cod. tegernsee); exclamat artem: sciccant list (Isidor, Lib. de officinisnbsp;eccles.).

In order to express Latin derivaties of ars such as artifex, artificimn, artificiosus, etc., new High German words were formednbsp;on the stem of list, e. g. listarra: artifices {Reg. Benedict. IV);nbsp;listwirker: artifex (Dief. nov. gZoss. 36 a); listig: arteficiosus (Greg.,nbsp;Cura past. IV); listigemo; artificioso (Gl. monsee., Fez I 384);nbsp;listlihemo: artificiose (Greg., Ctira. past Yl); list: artificium (St.nbsp;u. Siev., Ahd. Gl. 474, 21). Once Notker (Piper I, 276) translatesnbsp;artifex by zimberman. For the MHG period we may add list-machaere, listwerkaere, listwerke, listwiirhte: artifex (Lexer s. v.);nbsp;tusendlistelaere: milleartifex (Berth, v. Reg., Pfeiffer 408). Othernbsp;words found in the poetic literature of the period and thereforenbsp;without their Latin counterpart, but clearty analogical in originnbsp;are: listsache quot;Zauberkunst� (arsmagica), h'siiiw�durchgeheimenbsp;K�nste bereitetes Feuer� (Lexer s. v.), der siben liste brunnennbsp;{Trof. Kg. 1958), from Tristan: arzetlist (7780), boslist (2903),nbsp;houbetlist (4780) jagelist (3420), schuollist (7971); der list vonnbsp;nigromanze {Parz. 617, 12), etc. The uses of list in the usualnbsp;senses of ars extend thru the thirteenth and fourteenth centuriesnbsp;but with ever diminishing frequency: e. g. astronomie den listnbsp;von dem gestirne (R. v. Ems, Weltchr. 1178), zouberlist (ars magica)nbsp;Die Apok. H. v. Hesler 1291) wen er can alle die liste (�Zauber-kiinste�) die je wurdent (Lucid. D. Texte d. Mas. 28, 66, i);nbsp;Wan Jubal der fand schmides list {D. T. d. Mas. 31, 1897). Gradually during the fourteenth and fifteenth centuries the variednbsp;meanings of list become restricted almost entirely to one sensenbsp;of ars, viz. 2 (d) above, the prevalent meaning in the NHG today.nbsp;The cause of this specialization of meaning is to be found partlynbsp;in the increased use of the word with such MHG adjectives asnbsp;bose, arc, valsch, ubel, etc. with the resultant compounds arclistic,nbsp;boeslistic, hinderlistic, �berlist, tr�gelist, etc., and partly to thenbsp;gradual assumption of the other meanings by a synonym of list,nbsp;namely kunst. By the sixteenth century the transference of thenbsp;whole group of meanings of list with the exception of 2 (d) tonbsp;kunst was accomplished and still other meanings of ars addednbsp;which list seems never to have had.

OHG kunst, a verbal abstract of kunnan, NHG konnen was an exact synonym of list (cf. Goth, lais �ich weiB�) meaningnbsp;primarily �Kenntnis, Wissen�. While list was associated with

ars in the OHG period there is no indication that kunst was thought of in connection with our Latin word. The OHG translator, for the most part, associated scientia with kunst as seennbsp;by the following parallels from Graff, Ahd. Sprachschatz s. v.nbsp;chimsti: scientia; chunst kclirnis: scientiam doctrinae; kmistigonbsp;kelerto: scientes. Once chunste renders sapientiae and kunstiger:nbsp;gnarus, potens in scripturis. At that kunst does not seem to havenbsp;been the word most usually associated with scientia, the translations of wisluom and wisseniheit being even more numerous,nbsp;at least with Notker. Now scientia was occasionally used innbsp;connection with the artes liherales, for instance by Notker (I, 65)nbsp;tiu scientia diu rhetorica heizet. But kunst was not associated withnbsp;ars in the OHG period as far as the writer was able to discover.

We come then to the MHG period. It is unfortunate that for the later OHG and the earlier MHG periods the lack of recordsnbsp;of translations makes impossible the tracing of any semanticnbsp;associations for our word. In the early MHG poetic works kunstnbsp;seems to have been rarely used. A hasty glance through severalnbsp;thousand lines of the Rolandslied, Kaiserchronik, K�nig Rothernbsp;and the Vorauer Alexander fails to reveal a single instance of itsnbsp;use, whereas list is found numerous times. In the heroic epicsnbsp;it is used very seldom, in the Nibelungenlied only once (2222, 2),nbsp;in the sense quot;Geschicklichkeit�, against eight uses for list. Thatnbsp;kunst had become popular by this time, however, is evidencednbsp;by the fact that derivatives such as kunstec, kunstriche, kunstlichnbsp;are used by the author of the Nibelungenlied. An examinationnbsp;iirto the uses of kunst by the poets of the court epic shows annbsp;increasing use as a synonym of list, especially in the meaningnbsp;�skill� (Geschicklichkeit, Fertigkeit, Kunstbegabung) which, asnbsp;we have seen, is the fundamental meaning of ars. The mostnbsp;definite evidence of this is seen in those cases where list and kunstnbsp;are used synonymously in the same passage, as in the following;

In vernam von jagerie solher liste nie niht nienbsp;dar zuo liez er im aelien enbsp;wie man den hirz enbesten sol

In Wernher�s Marienleben whose source was the Vita beate virginis Marie et Salvatoris rhythmica, the author, in referringnbsp;to Mary and the art of dressmaking (ars suendi), says: . . . kundenbsp;alle fr�liche ktmst 5509). In 778 list is used for the arts of sewing,nbsp;weaving, spinning, etc. When Wolfram von Eschenbach says of hisnbsp;hero Parzival (123, 13) do lac diu gotes kunst an ini he is thinking ofnbsp;God as the artifex, the artist. As a variant of the arzate list abovenbsp;he has dar zuo al der arzate kunst (643, 21) (cf. hov/ever, der listnbsp;von nigromanzi 617, 12). The same use of kunst as the ars medicinaenbsp;is found in a passage of an old work of this period on medicinenbsp;daz den week mit die rechten chunst geleret hat . . . die da geschribennbsp;habent die chriechischen arzet Yfocris, etc. Zwei deutsche Arznei-biicher aus dem 12. und 13. Jahrh. Pfeiffer, Wiener Sb. 42, 127.nbsp;Toward the middle of the thirteenth century list and kunst arenbsp;still often used as synonyms. Compare the following from R.nbsp;V. Ems Weltchronik:

This author rather favors list in referring to the artes liberates, judging from the foregoing uses and from the following; (1178)nbsp;astronomie den list von den gestirne; (15766) der siben listenbsp;hohe kunst. For the manual arts, however, kunst seems now wellnbsp;established. Compare the following from the same author:

(535) gelt, silbir, stahil, isin von dem selbin wisinnbsp;wart dirre hohe list irdahtnbsp;und zeiner kunst der welte braht.

(686) ir kunst mit vlize worhte zwo siule der einiu ziegelinnbsp;was und diu ander steininnbsp;.....zwaz kunst von in do fundin was.

i85

{539) des swester diu hiez Neoma .... diu was von erst die mit begunstnbsp;irdahte wipliche kunst,nbsp;mit nadiln unt mit drihinnbsp;nejen, brettin, rihinnbsp;diz vant ir kunst do bi den tagennbsp;nu kunde birsen, schiezen, jagennbsp;Lamech durch chunst und durcli bejacnbsp;dirre selbin chunst er pflac.

Thomasin von Zirclaria uses kunst and list interchangeably in reference to the septem artes liberates. Compare the following fromnbsp;his Der welsche Gast:

{8899) wir haben kiinste vil geschribcn der sint uz erwelt sibennbsp;liste heize wir die k�nst.

von den ich iu ban geseit sint ander zwo k�nste groznbsp;die enen sint �bergenoz:nbsp;die heizent davon liste nihtnbsp;wan in ze herschen gescliicht

Besides the two additional artes mentioned here Thomasin adds the Decrete and the Leges (9151). These with the original sevennbsp;he calls the einlef k�nstn (9667). After the expansion of thenbsp;educational curriculum beyond the original seven arts kunstnbsp;seems to have been used regularly, list evidently remainingnbsp;restricted to the original seven. From the fourteenth century on,nbsp;therefore, kunst has taken on almost all the senses which listnbsp;formerly had acquired from ars with the exception of the usenbsp;prevalent today. This, for instance, is true of the uses of thesenbsp;words in Hugo von Trimberg�s Renner. �

� An examination of Trimberg�s Renner indicated that by this time kunst had already taken over the meanings of list. The latter was used only in thenbsp;sense which it had in the NHG, whereas kunst while often used in rathernbsp;indefinite senses, is a favorite word with the author. A rapid survey of thenbsp;work showed kunst used eighty times to five times for list. A century earliernbsp;the proportion of usage for these words would have been the reverse.

\Ne have thus far attempted to trace the expansion of the m.eanings of MHG kunst thru the literature of the period butnbsp;without direct association with the Latin. This expansion maynbsp;have been due to its close association with list rather than withnbsp;ars, especially in the early period of its semantic development.nbsp;From the fourteenth century on, however, we have again thenbsp;Latin for comparison in the manuscript glosses and for the laternbsp;centuries the Latin-German dictionaries, these sources showingnbsp;an ever expanding use of kunst to parallel similar uses in thenbsp;classical and medieval Latin.

In Diefenbach, Glossarium, list is mentioned once only in connection with ars. Kunst, on the other hand, appears as follows:nbsp;kunst, konst (three times); ars cterna: ewige kunst; artes liberalesnbsp;fry (vrighe) kunste, frige kunste, frihe konste, frie konst, freynbsp;chunst, kunst; artifex (artista): kunstiger, konstiger (four times),nbsp;ktmsfner, kunstmeister, konstcner (twice), kuoster, kuosta; arti-ficium; cin werck der kunst; artificiose: kilnstlich; artificiosus:nbsp;kunstig, kinstig, konstig, kunstick, kunstich, kunstlich; Med. Lat.nbsp;artista: ein kunstiger, kinstiger, fri ktmstiger, frey chunstiger. Thenbsp;Novum Glossariumnbsp;nbsp;nbsp;nbsp;ars eter na: cunstekeit; artificimn: kunst-

heit; artista: fry kinstig, siihen k�nster; astrologia: kunst der sterne. Once ars is glossed by behendekeit and once by hantwerck, probablynbsp;with the connotation of the artes manuales cf. Dief. 352 c (mechanica).

The early German-Latin dictionaries are of importance in showing many special uses of High German compounds withnbsp;kunst to correspond to or even translate certain long establishednbsp;Latin expressions. From Stieler, Der deutschen Sprache Stamm-baum im Fortwachs, N�rnberg 1691, we take the following wordsnbsp;with Latin equivalents, many doubtless formed by analogy ornbsp;association with the latter; Kunst �ber alle Kunste: ars artium;nbsp;kunsibeflissen: artis studiosus; kunsthegleiter: comes et socius artis;nbsp;kunsthild: Opus artificis; kunstfeind: osor artium liberalium; kunst-hasser: id.; kunstfeuer: ignis artificialis (cf. Fr. feu d�artifice);nbsp;kunstgerat: instrumenta artium; kunstgeschmack: artis gustatus;nbsp;kunstgesellschaft: collegium artificium; kunsthand.,: manus artificiosa,nbsp;artifex; kunstlauf: stadium artium; kunstlehre: artijicium; kunst-probe: specimen artis; kunstquell: seminarium artis; kunstrede: orationbsp;artificiosa; kunstregel: regula artis, pl. principia artium; kunst-schade: artium labes; kunstschl�ssel: clavis artium; kunstschweifi:

studium artis; kunstspiegel: speculum politissima arte perfectim; kunstspi�l: Indus artificialis; kunstst�rer: abusus artis, kunstst�m-pelung: id.; kunstst�ck, kunstgriff: artificium; kunsttanz: saltationbsp;artificiosa; kunst�bung: artis agitatie; kunstverwandter: ejusdemnbsp;artis studiosus; kunstwasser: aquae artificialis; kunstwerker: artifex;nbsp;kunstzweck: finis artis. Other compounds of possible analogicalnbsp;origin found in the dictionaries of Aler, Kirsch, Schottel, Stein-bach and in the DWB are: kunstfrei: immunis ob artem; kimst-freund, kunstliebend: artium amans; kunstg�nner: fautor artium;nbsp;kunstmaler: pictor artifex; kunstpfeifer: tibicen artifex; kunstsprache:nbsp;sermo artificiosus; kunstverstdndig: artis peritus; kunstwerk: artisnbsp;opus. With kunst as the last element of the compound we havenbsp;the following analogies, which could easily be increased: apo-thekerkunst: ars pharmaceutica; baukunst: (ars) arckitectura; bild-hauerkunst: (ars) statuaria; dichtkunst: (ars) po�tica; kriegskunst:nbsp;(ars) duellica; kriegskiinstler: bellandi artifex (Livy); redekunst:nbsp;(ars) rhetorica; sprachkunst: (ars) grammatica, etc. (the othernbsp;artes liberates); sittenkunst: (ars) ethica; sehkunst: (ars) optica;nbsp;zergliederkunst: (ars) anatomia; scheidkunst: (ars) chymia; denk-kimst: (ars) logica (Leibnitz.); wapenkunst: (ars) heraldica; schau-spielk�nste: artes theatrales. An interesting analogical compoundnbsp;of this type is notenkunst, a direct translation of ars notoria,nbsp;found in Der Ackermann aus Bohmen 26, 30. (Cf. the note innbsp;Burdach�s edition p. 360 f.)

In later times these analogies extend even to French derivatives of ars, in kunstgegenstand, kunstsache: objet d�art; kunstgenie: genicnbsp;de Vart, etc. In the eighteenth century the analogical expressionsnbsp;die guten Kiinste: bonae artes and die schonen K�nste: les beauxnbsp;arts were in common usage. Semantic analogies, in fact, becomingnbsp;popular may attain a momentum which sometimes carries themnbsp;to extremes. This was the case with the word kunst in its use bynbsp;students of the seventeenth century who, facetiously identifyingnbsp;Latin ars with German ars, gave kunst the analogical meaningnbsp;quot;podex�. From this developed the compounds kunsthummelnbsp;�Kotfliege� (Luther), kunstkammer id. and kiinstjucken. (Cf. DWB,nbsp;Vol. 5 s. V.)

For the development of kunst �machine� by association with ars �machina� (Du Cange) we refer the reader to the author�snbsp;article on the semantic development of the OE creeft. This particular sense, found as early as the fourteenth century in Konrad

von Megenberg�s Buck der Natur, e. g. kiinstendingel: mechanicum aliquid 434, 20, was restricted mainly to mining machinery andnbsp;produced such compounds as feuerkunst, damffkunst, wasser-ktmst, etc. (Cf. DWB V Sp. 2683). Another analogical developmentnbsp;of meaning found in K. v. Megenberg is kunstig in the sensenbsp;�artificialis� contrasting with �naturalis�. In his Deutschenbsp;Sphaera we find him using, or shall we say translating, kunstignbsp;tag in contrast to nat�rleichen tag (31, 14; 32, 8; 36, 26�28 Dnbsp;Texte d. Mas.), kunstiger: artificialis (7, ii, id.), and in his Bucknbsp;der Natur (Pfeiffer, 478, 7) k�nstleich: artificialis, this, the usualnbsp;sense of NHG kunstlich.

The conclusions to be drawn from the foregoing discussion may be briefly expressed as follows. Semantic developmentnbsp;contrary to the prevalent conception does not always representnbsp;a direct line of associative changes from the fundamental meaningnbsp;of the word in the particular dialect. There are many crosscurrents of semantic influence, due to the association of synonymsnbsp;in the same or foreign dialects. These latter influences may be sonbsp;strong as to blot out entirely the fundamental meaning of thenbsp;word. This is illustrated in OE craft (cf. op. cit.) where thru thenbsp;influence of Lat. virtus and ars the fundamental meaning �power�nbsp;was entirely supplanted by the meanings of these Latin words.nbsp;It is true again of list and kunst whose fundamental sense �knowledge�, becoming associated with the same sense in the Lat.nbsp;ars, adopted most of the other meanings of the latter by analogynbsp;and discarded the original fundamental meaning altogether.

It is, therefore, erroneous to think of the meanings of a word as always derived the one from the other in a close concatenationnbsp;from a fundamental meaning as source. True, that is the usualnbsp;conception which the lexicographer has of his task today. Thisnbsp;is not enough, however. Even tho such a chain of meanings isnbsp;psychologically unimpeachable it may still be incorrect unlessnbsp;possible analogical influences are weighed in considering thenbsp;semantic history of a word.

If, then, the development depicted here is correct, a very commonly accepted test for semantic relationship must of necessity lose much of its value. We refer to the use of the semanticnbsp;parallel as a sign or proof of the correctness of a semantic association. For languages as closely associated as Latin and Greeknbsp;or Latin and the Germanic dialects it would have almost no

validity because such parallels would more naturally be explained as semantic analogies.

What has occurred with the words discussed here, has taken place in hundreds of similar cases, not always with the greatnbsp;abundance of deviations in meanings shown in the foregoing,nbsp;but nevertheless, with associative transferences just as convincingnbsp;as those seen in High German kunst.

DER WORTSCHATZ DES AHD. TATIAN UND DIE �BERSETZERFRAGE

Es ist langst anerkannt, daB die ahd. Tatian�bersetzung nicht die Arbeit �ines Uebersetzers sein kann. Die Unbeholfenheitnbsp;einiger Partien, die stellenweise, z. B. 77�82. 11 a und 99�103,nbsp;fast interlinear sind, gegen�ber anderen gut, wenn auch nichtnbsp;vorz�glich �bertragenen Kapitein, muB jedern unbefangenennbsp;Leser auffallen. ^ In seiner ersten Ausgabe des T ^ hat Sieversnbsp;es unternommen, die Abschnitte zu bestimmen, welche verschie-denen Uebersetzern zuzuweisen sind. Er hielt sich an die wech-selnde Uebertragung verschiedener haufig auftretender W�rter,nbsp;WO Zufall oder absichtliche Variation als bestimmende Faktorennbsp;leichter ausgeschieden werden k�nnen als bei gelegentlichen Aus-dr�cken. Er wahlte also solche W�rter wie quia = bithiu imanta,nbsp;uuanta, bithm, thaz, un�bers.; cufii = mitthiu, thanne, th�, s�s�, s�;nbsp;U.SW. Er kam zum Entscheid, daB die Abschnitte 18, 45, 67, 104,nbsp;119, 135,146, 175 als Anfangspunkte neuer Arbeiter zu betrachtennbsp;seien. In seiner Rezension von Sievers Ausgabe � hat Steinmeyer

1 Die Versuche von Arens, ZfdPh 29, 510 ff. und Walther, Die deutsche Bibel-�berseteung des Mittelalters, Braunschweig, 1892, Sp. 446, wieder einheitliche Verfasserschaft f�r den ahd. T zu beweisen, sind vcrfehlt. Das von Arens heran-gezogene Material beweist noch nicht, da� gar kein Uebersetzerwechsel statt-gefunden hat, nur daB die Grenzen der Uebersetzerabschnitte nicht alle richtignbsp;gezogen sind. Karl F�rster, Der Gebrauch der Modi im ahd. Tatian, Kiel (Diss.),nbsp;1895, kam zum SchluB (S. i); �Der Modusgebrauch erscheint im ganzen Werknbsp;einheitUch genug, so daB man sich auf diese Iiidizien hin f�r die Annahme dernbsp;Uebersetzung von �iner Hand erklaren m�6te.lt;?

� Paderborn, 1872, S. 48�53. Nachtrage in der zweiten Auflage, 1892, S. LXX�LXXV.

diese statistischen Aufstellungen weiter ausgebaut und korrigiert. Er legt �berzeugend dar, daB ein haufigerer Uebersetzerwechselnbsp;angenommen werden m�Bte und lieB die Abschnitte anhebennbsp;mit 4. 12; 13. 5; 17. 6; 45; 67; 77; y8. i; 79. 4; 82. na; 86;nbsp;88. 13; 91. 5; 96. 5; 104. i; 104. 6; 119; 125. jj; 132; 135; 146;nbsp;175; ig8. Die kursiv gedruckten sah er als nur schwach bewiesennbsp;an. Sievers lieB nur 45; 67; 104; 119; 135; 146; 175 als sichernbsp;geiten und wies die �brigen als unbewiesen zur�ck.

Die ganze Frage blieb nun lange ^ ruhen, bis sie wieder von Friedrich Kohier in einer Leipziger Dissertation � aufgenommennbsp;wurde, der auf Grund syntaktischer Indizien und durch Anwen-dung des Rutzschen Verfahrens zur Unterscheidung der Stimm-typen, eine groBe Menge kleiner Abschnitte machte, gegen drei-hundert an der Zahl, wobei er aber an nur wenige Arbeiter dachte,nbsp;die sich im Laufe der Uebersetzung gegenseitig abl�sten. SchlieB-lich hat Leo Kramp die von Steinmeyer gew�nschte stilistischenbsp;Untersuchung angestellt �, die es ihm erm�glichte, die folgendennbsp;Abschnitte festzustellen; 17. 2; 45; 67; 77. i; 82. na; 89; 96;nbsp;104; 119; 141 (Vers unsicher); 146'; 175; 188; 197.

Bemerkenswert ist, daB zwei oder m.ehr Arbeiter an den folgenden Stellen �bereinstimmen;

4. 12 Steinmeyer, Kohier; 17. 6 Sievers, Steinmeyer, Kramp, Kohier; 45. i Sievers, Steinmeyer, Kramp, Kohier; 67. i Sievers,nbsp;Steinmeyer, Kramp, Kohier; 82. n a Steinmeyer, Kramp, Kohier; 104. I Sievers, Steinmeyer, Kramp, Kohier; 104. 6 Steinmeyer, Kohier; 119. I Sievers, Steinmeyer, Kramp, Kohier;nbsp;135. I Sievers, Steinmeyer, Kohier; 146, i Sievers, Steinmeyer,nbsp;Kramp, Kohier; 175. i Sievers, Steinmeyer, Kramp, Kohier;nbsp;198. r Steinmeyer, Kohier, Kramp (ungefahr). Bei der groBennbsp;Zahl der von Kohier gemachten Einschnitte ist es beinahe un-vermeidlich, daB er immer mit einem der �brigen zusammentrifft.nbsp;Doch darf jetzt als einigermaBen sicher geiten, daB an den obigennbsp;zehn Stellen oder in deren unmittelbarer Nahe wirklich ein Uebersetzerwechsel stattgefunden hat.

* A. Hilscher, Die Verfasserfrage im ahd. T. I. Programm, Posen, 1901, war mir nicht zuganglich.

� quot;Der Wortschatz des ahd. T. in seinem Verhaltnis zum alts., angels., und altfries.��, PBB 39, iff., 229 ff.

hat Gutmacher reiches Material beigebracht, welches sich zur L�sung der Uebersetzerfrage verwerten laBt. Gutmacher hatnbsp;gezeigt, daB ein groBer Prozentsatz des Wortschatzes des Tnbsp;(461 W�rter, etwa 15 v. H. des gesamten Wortschatzes) Eigen-heiten aufweist, die dem T gegeniiber anderen ahd. Denkmalernnbsp;ein recht eigent�mliches Geprage geben. Darunter sind 41 Wort-stamme, die im ahd. nur im T belegt sind. Von diesen W�rtern,nbsp;welche durch ihr Fehlen oder ihrenbeschrankten Gebrauch im iibri-gen ahd. Schrifttum ausgezeichnet sind, finden sich eine groBe Mengenbsp;im alteng., alts., altfries, und altnfr. Nach Gutmachers Zahlungnbsp;sind von ca. 2030 W�rtern im T 280 den �brigen ahd. Quellennbsp;fremd und von diesen kehren 120 im alteng. bzw. alts., mnd.,nbsp;mnl. wieder. Die Tabelle am Schlusse seines Artikels macht diesenbsp;Beziehungen deutlich. �

Es fragt sich nun, ob diese Eigentiimlichkeiten des T gleich-maBig auf die ganze Uebersetzung verteilt sind, oder ob einige Par-tien im Wortgebrauch mehr mit dem �brigen Ahd. zusammen-gehen als andere. Haben die Uebersetzer im Kloster Fulda gemaB ihrer landschaftlichen Herkunft diese oder jene Eigent�mlichkeitnbsp;bewahrt ? Sollten sich die von Gutmacher behandelten W�rternbsp;sehr ungleichmaBig verteilt finden, so liegt die Verm.utung nahe,nbsp;daB hier wirklich der Sprachgebrauch verschiedener Individuennbsp;wiedergespiegelt ist.

Eine statistische Pr�fung des Wortschatzes in bezug auf die Verteilung der von Gutmacher besprochenen W�rter zeigt nunnbsp;eine �berraschende Gleichf�rmigkeit. Es fallen auf die Seite vonnbsp;Sievers Text durchschnittlich 2.5, auf die Seite der Hs. 2.3 W�rter. Mustert man die oben aufgezahlten Abschnitte, wo mehrerenbsp;Arbeiter �bereinstimmen, so zeigen die folgenden eine gr�Berenbsp;Variation von der Norm: Prol. 4. 12 (3 W�rter auf die Seite);

Wenn aber diese Grenzen nicht ganz genau gezogen sind, k�nnte die Verschiebung von nur einem Vers schon einen nicht unbe-deutenden Unterschied machen, da nicht selten drei bis viernbsp;dieser W�rter in ein paar Zeilen vorkommen, oder sie auf mehreren

* Dio Schl�sse, die aus diesen Verhaltnissen gezogen werden k�nnen, gab Braune in der umfassenden und methodisch wichtigen Abhandlung quot;Althoch-deutsch und Angelsachsisch�, PBB 43,361�445. Es handelt sich nach ihm beimnbsp;T lediglich um einen altert�mlichen Wortschatz und nur in einigen Fallen umnbsp;angelsachsischen EinfluB.

Seiten ganz fehlen k�nnen. Es scheint daher geboten, Seite f�r Seite der Hs. den Wortschatz zu pr�fen, und um so eher, da,nbsp;wie schon Sievers bemerkte, der Uebersetzerwechsel fast immernbsp;mit dem Anfang einer neuen Seite der Hs. zusanimentrifft. Sonbsp;gezahlt finden sich gr�Bere Abweichungen von der Norm, alsnbsp;wenn man langere Abschnitte als eine Einheit betrachtet undnbsp;der Text zerfallt in Abschnitte, die durch gr�Bere oder geringerenbsp;Haufigkeit der tatianischen W�rter ausgezeichnet sind. Bezeich-nen wir erstere als Typus A, letztere als Typus B. Es folgen hiernbsp;diese Abschnitte, gezahlt nach Seiten der Hs., und daneben innbsp;Klammern die entsprechenden Kapitel und Verse des Textesnbsp;sowie auch die auf die Seite der Hs. fallende Durchschnittszahlnbsp;der von Gutmacher besprochenen W�rter.

25 (Prol. 1�2. i; o); 26�27 (2. i�ii; 4); 28 (2. ii�3. 8; 2); 29�32 (3- 8�4-17; 4); 33�36 (4. 18�6. 6; 1.25); 37�44 (6. 6bisnbsp;13- 6; 3-37); 45 (i3- 6�12; o); 46 (13. 12�18; 4); 47 (13. 18�23;nbsp;i); 48�50 (13. 23�16. i; 3.3); 51�54 (16. I�19. 4; 0.75);nbsp;55�57 (19- 4-21- 8; 3); 58 (21. 5�22.2; i); 59�61 (22. 2�25.4;nbsp;5); 62�64 (25- 4-31- 2; 1.3); 65�66 (31. 2�33.2; 3); 67�68nbsp;{33- 2�35- 2; 1.5); 69�70 (35. 2�38. 5; 4); 71�81 {38. 5�45. 5;nbsp;1.36); 82�85 (45. 6�51. 3; 3.25); 86�88 (51. 3�53. 14; 2); 89nbsp;(54. 1�8; 4); 90�92 (54. 8�57. 3; 2); 93 (57- 3�8; 6); 94�96nbsp;(58. 1�60. 16; 2); 97�103 (60. 17�67. 7; 3.71); 104�106nbsp;(67. 7�69. 7; i); 107�112 (69. 7�76. 5; 4.5); 113�116 (76. 5 bisnbsp;79.10; 1.5); 117 (79- 10�80. 3; 4); 118 (80. 3�7; 2); 119 (80. 7 bisnbsp;8r. 3; 6); 120�121 (81. 3�82. 5; 2); 122�123 (82. 5�ii; 3.5);nbsp;124�127 (82.11�84. 5; i); 128�129 (84. 5-85- 4; 5); 130�132nbsp;(86. 1�87.7; 2); 133�134 (87- 7�88. i; 4); 135�146 (88. ibisnbsp;91. 5; 1.58); 147�148 (91. 5-92. 5; 5-5); 149 (92. 5�8; 2);nbsp;150�152 (92. 8�95. 5; 3.66); 153�154 (96. 1-97- i; 2); 155 bisnbsp;157 (97. 1�98. 3; 3.6); 158�162 (98. 3�102. i; 1.2); 163�164nbsp;(102.1�103. 5; 3-5); 165 (103. 5�104. 2; 2); 166�167 (104. 2�6;

4) ; 168 (104. 6�8; i); 169�170 (104. 8�106. i; 4); 171�172nbsp;(106.1�6; i); 173�174 (106. 6�107. 3; 2.5); 175�176 (107. 3 bisnbsp;108. 5; 0.5); 177 (108. 5�109. i; 4); 178�179 (109. I�iio. i;nbsp;i); 180�182 (iio. I�112. 2; 3); 183�185 (112. 2�113. i; 1.5);nbsp;186�188 (113. I�116. i; 5); 189 (116. I�3; o); 190�191nbsp;(116. 3�6; 4); 192�194 (116. 6�118. 2; 1.66); 195 (118. 2�4;

5) ; 196�201 (119. I�123. 2; 1.16); 202 (123. 2�7; 3); 203�204nbsp;(123. 7�125. 2; 1.5); 205�207 (125. 2�127. i; 3); 208�209

(127- 1�4; 1-5); 210�212 (i28. 4�129. 8; 4); 213�219 (129. 8�131. 24; 1.33); 220 (131. 24�132. 4; 3): 221�226nbsp;(132. 5�134- 3; o. 66); 227 (134. 3�8; 4); 228 (134. 8�135. i;nbsp;o); 229�231 (135. I�19; 3-3); 232�237 (135. 19�138. 6; 0.66);nbsp;238 (138.6�11; 4); 239 (138.11�139. i; o); 240�243 (139. ibisnbsp;141. II; 3.75); 244 (141. II�-16; 2); 245 (141. 16�20; 2); 246 bisnbsp;248 (141. 20�143. 3; 1.33); 249 (143. 4�8; 4); 250 (144. i�145. 3;nbsp;o); 251�253 (1.45.4�16; 4.3); 254 (145.16�19; i); 255 (145.19 bisnbsp;146. 2; 3); 256�266 (146. 3�152- 3; 1-09); 267 (152. 3�7; 3);nbsp;268 (152.7�153.4; i); 269 (153-4�155- 2; 4); 270�274 (155.3 bisnbsp;159. 6; i); 275�276 (159.6�^161.2; 2.5); 277�280 (161. 2 bisnbsp;165. 4; 1.25); 281�282 (165. 4�167. i; 3-5); 283�285 (167. 2 bisnbsp;170. 6; 1.66); 286�288 (170. 6�^174. 6; 3.66); 289 (175. i�176. 3;nbsp;2); 290 (176. 3�177. 4; 5); 291 (177.4�178.5; 2); 292 (178. 5 bisnbsp;179. 2; 5); 293�294 (179. 2�181. 4; i); 295�296 (181. 4�184. i;

4.5) ; 306 (194. 3�196-1; i); 307 (196- 1�7; 6); 308 (196. 7 bisnbsp;197. 6; i); 309�310 (197. 6�199. 3; 3); 311 (199- 4�ii; 2);nbsp;312�313 (199. 11�202. 2; 3.5); 314�318 (202. 3�209. 4; 1.6);nbsp;319 (210. 1�5; 4); 320 (211. 1�4; I); 321�323 (212. I�217. i;nbsp;2.66); 324�327 (217. I�221. 7; i); 328 (222. 1�223. 3; 3);nbsp;329�331 (223. 3�228. i; 1.66); 332 (228.1�229. 3; 3); 333�335nbsp;(229.3�233- 6; 0.66); 336 (233. 6�235. 3; 3); 337�339 (235- 3 bisnbsp;239- 3; i); 340�342 (239- 3�244- 4; 3)-

Wie erklart sich nun diese Verteilung, wonach in kleineren Abschnitten ein bedeutender Unterschied hervortritt, der in dennbsp;langeren wieder verwischt wird? DaB mehrere Uebersetzer be-teiligt waren, darf als Tatsache geiten. Wenn diese sich so in dienbsp;Arbeit teilten, daB jeder nur einen kleinen Abschnitt �bersetzte,nbsp;ehe er abgel�st wurde, dann darf man in den kleineren Abschnittennbsp;Unterschiede erwarten, die in den langeren wieder verschwinden.nbsp;Das durchschnittliche Auftreten auffallender W�rter in langerennbsp;Abschnitten muB sich notwendigerweise dem Durchschnitt innbsp;der ganzen Uebersetzung nahem. Je gr�Ber der Abschnitt, destonbsp;geringer der Abstand Vom Durchschnitt des ganzen Werkes. 1stnbsp;nun aber ein bedeutender Unterschied bei den kleineren Abschnitten zu merken, muB gefolgert werden, daB der Sprachgebrauchnbsp;eines oder mehrerer der Uebersetzer sich weiter vom ahd. Sprachgebrauch entfernte als der von den �brigen.

die leicht ein Verzerrtes Bild geben. Die Uebersetzung ist selbst-verstandlich durch die Wortwahl der lateinischen Vorlage bedingt. DaB ftmstsizzento nur an einer Stelle (45. 6, 7) vorkommt, erklart sich durch das einmalige Auftreten von architricUnus.nbsp;Ob ein anderer Uebersetzer einen andern Ausdruck (Gutmachernbsp;S. ii) gebraucht haben w�rde, entzieht sich unserer Kenntnis.nbsp;Ein und derselbe Uebersetzer wird auch wohl freiwillig den Ausdruck variiert haben und ausnahmsweise ein veraltetes odernbsp;selten gebrauchtes Wort herangezogen haben. Ueber solchenbsp;Variation in den ersten 16 Kapitein handelt Kramp ausf�hrlich.nbsp;Ferner v/erden gewisse W�rter in der Klostersprache von Fuldanbsp;Gemeingut geworden sein, so daB der heimische Ausdruck desnbsp;Uebersetzers in Vergessenheit geriet. Das ist sicher der Fall beinbsp;jluobar und Sippe.� Immerhin ist die Haufigkeit der fremdennbsp;Ausdr�cke in einer Partie und die sparsarne Anwendung der-selben in einer anderen von Bedeutung, denn die Wahl geradenbsp;dieser Ausdr�cke zur Uebertragung von selten oder nur einmalnbsp;auftretenden lateinischen W�rtern k�nnte nicht durchweg aufnbsp;reinem Zufall beruhen. Auch ist die Verteilung der W�rter, welche,nbsp;was den Wortstainm anbetrifft, dem Ahd. sonst fremd sind, vonnbsp;groBejn Interesse. Sie linden sich beinahe ausnahmslos in Partiennbsp;vom Typus A. Wo das nicht der Fall ist, laBt sich aus stilistischennbsp;Gr�nden leicht eine Erklarung geben oder der Anfangspunkt desnbsp;neuen Uebersetzerabschnittes ist um ein weniges zu verschieben.

Die iin Ahd. sonst unbelegten Wortstamme (Gutmacher 68 ff.) finden sich im T wie folgt;

asni 133. II (Seite der Hs. 225). Typus B von S. 221�'226 (132.5bis 134. 3). Kohier macht Einschnitte bei 132. 6 und 134. 3.nbsp;asneri 97. 3 (2) (S. 155). Partie Typus A reicht von S. 155�157nbsp;(97. 1�98. 3).

berd: 141. 28 (S. 247). S. 246�248 geh�ren Typus B an. Gerade davor sind die fremden Ausdr�cke haufig und wieder danachnbsp;auf S. 250. Kohier macht einen Abschnitt bei 141. 20 (S. 246),nbsp;aber nicht 143. 3 (S. 248).nbsp;biril: 80. 6 (S. 118) Abschnitt Typus B.

� Es ist aber bemerkenswert, daB die Sippe von fhtobar mit einer Austiahme nur in Partien vorkommt, wo die fremden Ausdr�cke sonst haufig sind (Typus A).nbsp;Die Ausnahme ist 135. 19. Wenn man aber die Grenze vom Seitenanfang bis annbsp;den SchluB des Satzes verscliiebt, dann fallt fluohritun in eine Partie lypus A.nbsp;Ueber fluobar und Sippe siehe Anm. 9.

gifehan: Die einzige Uebertragung von gaudere, und ist daher wenig beweiskraftig. Wie fluobar ist das Wort in die Sprachenbsp;aller Insassen des Klosters �bergegangen. Dasselbe gilt vonnbsp;gifeho.

girdinon: 74. 8 (S. iio); 97. 2 (S. 155). Beide Stellen in Partien vom Typus A, S. 107�-112 und S. 155�157 (69. 7-�76. 5nbsp;und 97. I�^98. 3). Kohier macht Einschnitte bei 70. i; 77. i;nbsp;97- 4; 99- 2.

hansa: 200. i (S. 312). Typus A von 312�313 (199. ii�202. 2). gikeuuan: 141. 7 (S. 243). Typus A von S. 240�243 (139. i bisnbsp;141. ii). K�hlers Einschnitte: 139. 2 und 141. 12.nbsp;landen: 199. 8 (S. 311). Typus B (199. 4�ii).nbsp;liodar: 145. 15 (S. 253). Typus A von S. 251�253.nbsp;manduuan: 22.9 (S. 60); 67.9 (S. 104); 116.3 (S. 190). Dienbsp;zweite Stelle in einer Partie Typus B, doch ist der Grenz-punkt des Uebersetzerwechsels vielleicht zu verschieben.nbsp;Steinmeyer, Sievers und Kramp machen Abschnitt bei 67. inbsp;(S. 103); Kohier bei 67.3 und 67.8. Die anderen beidennbsp;Stellen in Partien vom Typus A.nbsp;manzon: 58. i (S. 94). Typus B von 94�96 (58. i�60. 16). Dochnbsp;steht das Wort gegen Anfang der Seite und der vorgehendenbsp;A-Abschnitt reicht vielleicht bis 59. i.nbsp;melmi: 44. 9 (S. 77). Typus B. Beinahe sicher ist aber in der Mittenbsp;des langen Abschnittes S. 71�81 ein Uebersetzerwechsel.nbsp;Kohier faBt 44. �-14 zusammen, also ungefahr gerade diesenbsp;Seite der Hs.

gimunt: 138. 6 (S. 238); 160. 3 (S. 276). Die erste Stelle in einer Partie Typus A, die diese Seite allein umfaBt. Bei Kohiernbsp;wird 138. 7�'10, ungefahr die S. 238, als ein Abschnittnbsp;zusammengefaBt. Die zweite Stelle auch in einer Partienbsp;Typus A.

tuomo: Von den 6 Fallen kommen vier in Partien vom Typus A vor. Die Stelle 122. 3 (S. 201) fallt in eine Partie Typus B,nbsp;doch fangt mit 123. 2 (S. 202) ein neuer Abschnitt Typus A

an. Dieser beginnt aber wohl schon vor tmmo. Kohler macht einen Abschnitt von 122. 2�123. 6. Da der andere Fall vonnbsp;tuomo in 122. i zu linden ist, spricht dies daf�r, daB dernbsp;neue Uebersetzer mit dem Kapitel 122 anhob. Zu beachtennbsp;ist aber, daB tuomo die einzige Uebertragung ist f�r �iudex�.nbsp;furumrgen: 92. 2 (S. 148). Partie Typus A reicht von S. 147�148.nbsp;Der andere Fall, 129. 9 (S. 213) in einer Partie T5rpus B,nbsp;aber gegen Anfang der Seite und SchluB des Kapitels.nbsp;Kohler macht einen Abschnitt 129. 8�ii. Der dritte Fall,nbsp;152. 6 (S. 267), in einer Partie Typus A.nbsp;aruuizan: Sechs Falie, davon 22. 4 (S. 59), 92. 2 (S. 148), 113. inbsp;(S. 186), 244. 2 (S. 342) in Partien von Typus A. Bei 42. 3nbsp;(S. 74) steht das Wort am Schlusse eines Kapitels und amnbsp;Anfang einer neuen Seite und es ist wohl mit einem neuennbsp;Einschnitt zu rechnen. Bei 152. 6 (S. 267) haben wir esnbsp;wohl mit einer kurzen Partie Typus A zu tun, die nur diesenbsp;Seite der Hs. umfaBt, 152. 3�-152. 6. Kohler machte Ein-schnitte am Schlusse von 152. 3 und 152. 6.nbsp;uozarnen: Die f�nf Falie nur in Partien vom Typus A. Die Stellenbsp;118. 2 ist sicher zum folgenden A-Abschnitt zu ziehen, dernbsp;dann mit dem Anfang des Verses anheben wiirde.nbsp;uuaharsiuni: 210. 3 (S. 319) Partie vom Typus A.

Von den 120 Einschnitten, die hier gemacht werden, stimmen 98 genau oder ziemlich genau zu den von Kohler anerkannten.nbsp;Und da die obige Einteilung vorerst mechanisch vorgenommennbsp;wurde, ohne Riicksicht auf Vers oder Satzschliisse, ist diesenbsp;Uebereinstimmung sehr �berraschend. Eine scharfere Abgrenzungnbsp;kann nur erzielt werden durch Heranziehung anderer Indizien,nbsp;die haufiger auftreten als die speziell tatianischen W�rter. Einenbsp;solche Untersuchung aber wiirde weit fiber den Umfang einesnbsp;Aufsatzes hinausfiihren. Ich gebe also im folgenden nur einnbsp;paar Proben von der naheren Bestimmung der Grenzen. Letztennbsp;Endes wird doch wohl die Schallanalyse angerufen werden m�ssen.

Grenze zwischen 3. 2 und 3. 4. Ob vor SchluB von S. 32 der Hs. (4. 17) eine Grenze zu ziehen ist ? Im Prolog undnbsp;dem ersten Kapitel kommen keine von Gutmachers Wortennbsp;vor, in dem langeren zweiten aber acht, namlich rehtjesti, uuirouh-hrunst, ungiloubfol, menigi, hlidida, tougilta sih, lazzen, gifeho.nbsp;Vier von diesen und die lateinischen Worte, welche sie iibersetzen,nbsp;kommen nicht wieder vor, lassen sich also f�r unseren Zweck

igS nicht venverten. Umrouhbrunsf, 2.4, �bersetzt incensum, dasnbsp;aber 2. 3 durch uuirouh und rouhennes wiedergegeben ist. Diesnbsp;scheint also eine freiwillige Variation eines einzigen Uebersetzersnbsp;zu sein. Das sonst im Ahd. unbelegte Adjektiv imgiloubjol f�rnbsp;incredibilis 2. 7, das nur an dieser Stelle der Vulgata vorkommt,nbsp;findet sich auch 21. 8 und 233. 6 f�r incredulus. Die ebenfalls aufnbsp;den T beschrankte Ableitung ungiloubfidli {incredulitas) stehtnbsp;92. 5 und 241. 2; dagegen das gelaufige ungiloubo 78. 6 und 92. 8.nbsp;Kohier macht zwischen 92. 5 und 92. 8 einen Einschnitt, wasnbsp;auch zu dem verschiedenen Sprachgebrauch ungiloubfulli �� ungiloubo paBt. Das Adjektiv ungiloubfol gehort also einem Ueber-setzer vom Typus A an. Gifeho, 2. 6, die einzige Entsprechungnbsp;von gaudium, scheinbar allen Uebersetzern gemeinsam be-kannt, und menigi, 2. 3, das nicht auf den T beschrankt istnbsp;(Gutmacher 27 f.) sind f�r unseren Zweck nicht zu verwerten.nbsp;Huldi 3. 4 = gratia ist zwar auch nicht speziell tatianisch, dochnbsp;w'eicht der Gebrauch hier von anderen Stellen ab. Gratia heiBtnbsp;sonst geba 3. 2; 12. i, 9; 13. 7, 9 oder thanc 32. 5, 6; 82. 3; 89. 2;nbsp;III. 2; I18. 2; 135. 25; 160. 2. Ob diesem Sprachgebrauch nachnbsp;zwischen 3. 2 und 3. 4 ein Einschnitt anzunehmen ist ? Ich glaubenbsp;in der Tat, daB nach 3. 3 ein neuer Uebei'setzer anfing. Weiterenbsp;St�tzen f�r diese Annahme f inden sich in uuis sin 3. 6 f�r co-gnoscere gegen�ber dem �blichen furstantan Prol. 4, obwohl dienbsp;Bedeutung 3. 6 wie auch 5. 10 eine spezielle ist; tixor = quenanbsp;2. I, 5, 8, II und spater haufig, gimahha nur 5. 12 (Gutmachernbsp;S. 24); benedicere = segenon 3. 2 gegen�ber uuihen 4. 12, 14; salu-tatio = uuolaqiieti 3. 3, = heilezunga 4. 2. Bemerkenswert istnbsp;auch der Vergleich zwischen ginemnis 3. 4 und nemnis 2. 5; undnbsp;der Orthographic intjieng 2. ii und infhahis 3.4.^^

Grenze um 4. 12 bis 4. 15. Eine Grenze um 4. 12 scheint mir schon aus den von Steinmeyer und Kohier vorgebrachten Gr�ndennbsp;sicher. Kramps Beweise f�r die Einheit von i�17. i sind un-gen�gend. Auch die gr�Bere Kompetenz zugegeben, die in diesennbsp;ersten Kapitein zutage tritt, bleibt doch nicht ausgeschlossen,nbsp;daB wir es hier mit mehr als einem Arbeiter zu tun haben. Krampnbsp;betont selbst die Unbeholfenheit des Prologs und die Fehlernbsp;4. 7, 10. I und 10. 3, die schlecht zu seiner Annahme passen.

Zu einer Grenze um 4. 12 ist aber noch einiges hinzuzuf�gen. In 4. 8 wird memorari durch gimuntigon, aber 4. 15 durch gihugen

�bersetzt. Letzterer ist der gelaufige Ausdruck. Aferire = giof-janon 4. 12, = intuon 7. 2, spater, 8. 7, wieder gioffanon. Bene-dicere = nuihen 4. 12, aber = lobon 7. 5; gleich darauf, 7. 7, wieder iiuihen.

Grenze um 5. i bis 5. 6. Da die ersten f�nf Verse dieses Ab-schnittes aus einer Namenliste besteken, ist kaum festzustellen, ob sie dem Uebersetzer angeh�ren, der bis SchluB von Kapitel 4nbsp;ging, dem von 5. 6 ff., oder noch einem driften. Die tatianischennbsp;Worte sind S. 34�-36 (5.6�^6.6) selten: nur girado, gimahha,nbsp;gijeho und menigi. Das erste �bersetzt 5. 8 und 9. i ecce, wasnbsp;sonst durch senu, nu, tho, thana wiedergegeben wird. Gimahhanbsp;f�r coniunx ist auf diesen Abschnitt (5. 8, 10, 12) beschrankt;nbsp;das synonym uxor heiBt 2. 5, 8. ii und spater haufig quena.nbsp;Bemerkenswert ist auch das t in t�trida 6. 3, sonst, schon 6. 7,nbsp;haufiger d. Dieser Wechsel braucht nicht vom Schreiber herzu-r�hren. (Vgl. Sievers � 29.)

Grenze um 6. 6. Nun werden die tatianischen Worte haufiger; gommanbarn 7. 2 und gotjorht 7. 4 kommen nur hier vor. Zunbsp;gommanbarn vgl. Gutmacher S. 4. Der Kontrast zu thegankindnbsp;9. 2 ist bemerkenswert. Unkombiniertes barn kommt offers vornbsp;(2. 6; 3. 7; 13. 14; 22. 14; 62. 10; 127, 3; 131.15; 139. 10), thegannbsp;unkombiniert gar nicht. Progenies viper arum = cimni natrononbsp;13. 13 gegen barn natrono 62. 10 kommt gerade vor barn 13. 14,nbsp;ist also vielleicht stilistischer Wechsel. Von den drei Uebertra-gungen von filius wird sun durchweg gebraucht, um das Vater-Sohn-Verhaltnis auszudr�cken, besonders haufig von Christus.nbsp;Kind ist ohne R�cksicht auf das Geschlecht einfach �Nach-komme� und dr�ckt entweder die geistliche Verwandtschaftnbsp;(54. 4; 107. 3) oder die leibliche (alle �brigen Stellen) aus. Dienbsp;einzige Ausnahme ist 67. 10, wo es im Kontrast zu jilia-dohternbsp;gebraucht wird. Ba.rn wird an keiner Stelle gebraucht, um dasnbsp;Vater-Sohn-Verhaltnis auszudr�cken, also nirgends = sun, son-dern = Kind. Nur 3. 7 von Christus; 22. 14 und 127. 3 vom geist-lichen Verhaltnis zu Gott; 2. 6 von den Kindern Israel. Also amnbsp;Wechsel von barn und kind ist wahrscheinlich auch ein Ueber-setzerwechsel zu erkennen. Solcher Wechsel findet sich wie folgt:nbsp;barn 2. 6 �� kind 2. 7; barn 3. 7 �� kind 4. 2 und 4. ii; barnnbsp;22. 14 � kind 22. 6; barn 127. 3 �� kind 127. 2, 3; barn 131. 15 �nbsp;kind 131. 16.

mit acc. rei 6. 5, aber mit iihar 7. 7. Auch sonst ist der syntaktische Gebrauch dieses Verbums innerhalb ein und desselben Ueber-setzer-Abschnittes konstant.

Grenze um 13. 23 (S. 48). Steinmeyer macht einen Einschnitt bei 13. 5 wegen des Auftretens der Praposition fur- statt for-\nbsp;Kohier aus syntaktischen Gr�nden. Letzterer nimmt aber auchnbsp;13. 23 einen Einschnitt an, den ich ebenfalls anerkenne, weil hiernbsp;die tatianischen Worte wieder haufig werden.

Grenze um 16. i (S. 51). Hier ist sicher ein Einschnitt, da die tatianischen Worte bis S. 55 (19. 4) beinahe ganz verschwinden,nbsp;nur mittilgart, st., sihuuaz, gotspellon. Dazwischen liegt wenigstensnbsp;noch ein Abschnitt bei 17. 6, wie Steinmeyer auf Grund desnbsp;Auftretens von antuuurtan annahm.

Grenze um 19. 4 (S. 55). Eine Partie A reicht von 19.4 bis 20. 9. Vielleicht kommt der Einschnitt zwischen 19. 2 und 19. 3. Vgl.nbsp;die Orthographie forlazanen nezzin 19.2 mit forlazzanen nezin 19. 3.

Grenze um 27. i (S. 63). Diese Partie reicht von 27. i bis 31. i, doch sind wohl dazwischen weitere Abschnitte anzunehmen.nbsp;Z. B. hebt mit 28. i wieder das Prafix fur- an. Das eine for- amnbsp;Schlusse von 28. i ist wohl dem Schreiber aus Gewohnheit in dienbsp;Feder geflossen.

Grenze um 31. 2 (S. 65). Der SchluB dieses Abschnittes fallt wohl um 33. 3, denn in diesem Verse wird reddet falsch mit geltenbsp;�bersetzt, aber 34. 2 richtig mit giltit.

Grenze um 34. i (S. 67). Diese ist ohnehin auf Grund des Wortschatzes anzunehmen. Auf einen Wechsel deutet auch dienbsp;Uebertragung von merces durch mieta 34. i und 35. i gegen Ionnbsp;33- 2.

Aus Mangel an Raum m�ssen wir hier auf eine eingehende Analyse der hypothetischen Grenzen verzichten. Diese Grenzen,nbsp;wie sie oben angedeutet wurden, sind aber nicht als endg�ltigenbsp;Resultate anzusehen. Erstensist die Statistik nach SeitenderHs.nbsp;aufgestellt worden und die Seitenanfange der St. Galler Hs.nbsp;stimmen nicht notwendigerweise genau mit denen der Urhand-schrift �berein noch ist anzunehmen, daB mechanisch bei jeder

Ziemlich genau wird die Uebereinstimmung aber doch sein, wie aus dem Schlusse von S. 164 erhellt. Hier ist effenbar vom Kopisten ein Wort ausgelassen.nbsp;Die lat. Vorlage lautet: gaudebat in universis (165) quae . . ., die Uebersetzung:nbsp;gifah in then (165) thiu. Der neue Schreiber [6) beginnt mit quae � thiu amnbsp;Seitenanfang, denn y war mit then mit seinem Pergament zu Ende gekommen.

neuen Seite ein anderer Uebersetzer anhob. Wie schon gesagt, w�rde die Statistik deshalb hier und da der Korrektur bed�rfen.nbsp;Solche Verschiebungen w�rden aber, wie die gegebenen Probennbsp;zeigen, eher die Hypothese starken als untergraben. Sogar nachnbsp;der provisorischen Einteilung ist die Uebereinstimmung mitnbsp;K�hlers Resultaten �berraschend genau. Er gibt in seiner Dissertation � 105 ein Beispiel f�r eine Gruppe von Abschnitten, dienbsp;�inem Uebersetzer angeh�ren k�nnten. Eine Pr�fung derselbennbsp;erweist, daB von den 53 darin vorkommenden tatianischennbsp;W�rtern 49 nur in A-Partien vorkommen oder A- und B-Partiennbsp;gemeinsam sind. Von dem Rest, gimahha, diacan, ahafurhouuan,nbsp;widarscouuan, kommen diacan und ahafurhouuan in unmittelbarernbsp;Nahe von Uebersetzereinschnitten vor, so daB bei einer geringennbsp;Verschiebung der Grenze sie einer A-Partie zufallen w�rden.nbsp;K�hlers hypothetischer Uebersetzer ware also einer vom Typus A.

Eine scharfere und definitive Abgrenzung k�nnte erst ge-schehen nach Aufstellung umfangreicher Tabellen, ahnlich der von K�hler gebotenen, aber mit Hinzuf�gung aller Eigent�mlich-keiten des Wortschatzes, Stiles und der Orthographie, sofernnbsp;letztere nicht vom Schreiber allein herr�hrt. Auf diese Weisenbsp;k�nnte auch dann der Wortschatz der einzelnen Uebersetzernbsp;bestimmt werden.

1. nbsp;nbsp;nbsp;Die an der Uebersetzung des T beteiligten M�nche lassennbsp;sich in zwei Hauptgruppen teilen. Die eine (A) bediente sich einesnbsp;Wortschatzes, der erheblich vom Sprachgebrauch im Ahd. desnbsp;fr�hen neunten Jahrhunderts abwich. Ob diese Abweichungennbsp;einfach als Altert�mlichkeiten aufzufassen sind (Braune) odernbsp;auf Dialektverwandtschaft mit dem Niederdeutschen und Alt-englischen hinweisen, ist f�r unseren Zweck einer lei. Die anderenbsp;Gruppe (B) stand sprachlich dem Oberdeutschen naher. Einenbsp;Anzahl Ausdr�cke aber, wie milten, gifehon, fluobar u. a. m.,nbsp;die dem Hochdeutschen sonst fremd sind, scheinen beiden Grup-pen gemeinsam gewesen zu sein und sind als Bestandteile der innbsp;Fulda �blichen Klostersprache anzusehen.

2. nbsp;nbsp;nbsp;Die Zahl dieser Uebersetzer genau zu bestimmen ist bei demnbsp;gegenwartigen Stand der Untersuchung nicht m�glich, doch wird

Das Versehen ist leicht zu erklaren, wenn universis und seine ahd. Entsprechung (allen?) auch in der Vorlage am Seitenschlusse stand.

3. Die Uebersetzerabschnitte sind von ungleicher Gr�Be, aber iiberschreiten selten oder nie die Lange von einer Seite der Hs.nbsp;Die von Sievers, Steinmeyer und Kramp angenommenen langennbsp;Abschnitte sind alle weiter einzuteilen. Die Mehrzahl der Ab-schnitte heben annahernd oder genau mit dein Seitenanfang an,nbsp;so daB die von Sievers (Einleitung LXXIV, � 125) und Kohlernbsp;(S. 89�90) ausgesprochene Vermutung, die Uebersetzer battennbsp;sich in der Arbeit beim Beginn einer neuen Seite abgel�st, wohlnbsp;den Tatsachen entspricht

Oben S. 197, Z. 8 v. u. hatte auf Sievers Aufsatz in PBB 50, 416 ff verwiesen werden sollen. Er macht da den Versuch, einc Anzahl in Sagversennbsp;abgefaBte Abschnitte des T herauszuschalen, von denen einer (124. 7 � 125,11)nbsp;aus klanglichen Gr�nden Hrabanus Maurus zuzuschreiben sei. Merkw�rdig ist,nbsp;daB in diesen Abschnitten nur zwei der speziell tatianisclien W�rter {tuomonbsp;und nozarnen) vorkommen,

Erst nach der Drucklegung des Aufsatzes fand ich Ernst Schr�ter, Walah-frids deutsche Glossienmg und der aUhochdeutsche Tatian, Halle, 1926 und Baeseckes Artikel in ZfdA 58, 241 ff. Ich glaube, daB sich meine Schliisse mitnbsp;denen von Schr�ter und Baesecke gut vertragen werden.

RANDGLOSSEN ZUM MORIZ VON CRAON

Das vortreffliche kleine Gedicht eines unbekannten Verfassers aus der Bl�tezeit des 13. Jahrhunderts, woriiber man vergleichenbsp;neuerdings Ehrismann, Gesch. der deutschen Lit. 2, 2, i, S. 127 f.,nbsp;bietet gerade durch den Umstand, daB es allein in dem bekanntennbsp;Ambraser Codex aus dem Anfang des 16. Jahrhunderts �ber-liefert ist, noch manches zu erwagen in bezug auf seine textlichenbsp;Gestaltung. Durch seine ausgezeichnete Ausgabe in den �Zweinbsp;altdeutsche Rittermaren� hat Edward Schroder 1894 das Gedichtnbsp;dem Dunkei und der Vergessenheit entzogen und somit allgemeinnbsp;zuganglich gemacht. Von diesem Werkchen erschien 1913 einenbsp;zweite, 1920 bereits eine dritte Auflage. Wohl im Hinblick aufnbsp;seine erste Bestimmung als Uebungsmaterial im Seminar hatnbsp;Schr�der einen m�glichst konservativen Text geliefert und ausnbsp;eben diesem Grunde ist es auch zu er klaren, wenn die spaterennbsp;Auflagen keine wesentlichen Unterschiede gegen�ber der erstennbsp;aufzuweisen haben. Nur gelegentlich sind die Besserungsvorschlagenbsp;andrer in den Text auf genommen worden. Es sind also verschie-dene Stellen stehen geblieben, die, wenn nicht gerade unverstand-lich, doch zum mindes ten undeutlich oder dem Sprachgebrauchnbsp;des Dichters zuwider sind. Im folgenden teile ich eine Reihe Vor-schlage mit, die mir bei wiederholter Lekt�re der Dichtung ein-gefallen sind, mache aber durchaus keinen Anspruch darauf, allenbsp;Schwierigkeiten gehoben, noch wo ich etwas gewagt, immer dasnbsp;Richtige getroffen zu haben. Es bleibt immer noch die dankbarenbsp;Aufgabe, die Stellung der Dichtung innerhalb ihres Kreises undnbsp;ihrer Periode genauer zu bestimmen. DaB wir es hier mit demnbsp;Werke eines begabten Dichters zu tun haben, leuchtet jedem ein,nbsp;der die letzten hundert Verse durchliest.

98. 99. hat Schr�der die Verderbnis so zu heilen versucht: man zinste in, n� gernt sie hulde (daz ist doch ein ungelichez lehen).

Mit teilweise naherem AnschluB an die Hs. m�chte ich lesen; man zinste in, n� gebent sie hulde, (daz ist dechein ungeltcheznbsp;leben): wodurch die Stelle erst verstandlich wird, besonders dienbsp;Aussage in Vers loo.

120. ein lop erk�s sin hant ist mindestens eine auffallige Wendung und soweit ich sehe, sonst nicht belegt; vielleicht ist er ze hantnbsp;statt sin hant zu lesen.

233. Itbes unde guotes ist der formelhafte Ausdruck, wozu vgl. Wilmanns zu Walther 88, 2.

272. nbsp;nbsp;nbsp;h�s in der Bedeutung �Familie, Geschlecht� scheint erstnbsp;im 14. Jahrhundert vorzukommen. Um 1300 schreibt Heinrichnbsp;V. HeBler, Apok. gegen den SchluB: Heinrich ist min rechter nam,nbsp;Hesler ist min h�s g�nant. Unsere Stelle ware somit der fr�hestenbsp;Beleg f�r diesen Gebrauch.

273. nbsp;nbsp;nbsp;Vorzuziehen ware der was ir dienstes hereit mit Bezug aufnbsp;die grcBvinne (268). Wegen des gew�hnlichen Dativs vgl. Wigaloisnbsp;3149, 6226 usw.

277. dne l�n ist auffallig und dr�ckt gerade das Gegenteil aus von dem, was man erwartet, denn der Held hat doch immer dennbsp;Preis davongetragen, wie wir aus Vers 282 erfahren. umhe l�nnbsp;w�rde dem Sinne nach besser passen.

301. Man setze einen Punkt nach rat gegen das Komma der drei Auflagen Schr�ders; ebenso Punkt statt Kolon nach maze 354.

357. 358. Der Sinn dieser beiden Verse ist nicht klar. Der Ausdruck als in daz mer ein slac bedeutet ebensoviel wie �nichts, etwas, was nichts zu bedeuten hat� und paBt gar nicht zum Ge-danken. Vielleicht ist der Stelle geholfen, wenn man leider stattnbsp;lieber 358 liest. Danach ware der Gedanke etwa folgender: �Es istnbsp;ihm einharter Schlag, denn nichtsk�nnteihm unangenehmer sein�.

363. Ich m�chte mit Haupt an �re lesen. Diese Wendung ist formelhaft; vgl. i. B�chl. 1583!. der frume wirt niht m�re wannbsp;der schade an �re.

373. Es ist wohl den statt die zu lesen. In der zw^eiten und dritten Auflage hat Schroder, einer Konjektur Wilmanns gemaB, einenbsp;L�cke von zwei oder vier Versen angesetzt. Es scheint uns abernbsp;dem Sinne nach nichts zu fehlen, besonders wenn man baz gemeinenbsp;374 in komparativer Bedeutung nimmt, eine Konstruktion, dienbsp;dem 13. Jhrh. durchaus nicht fremd war; vgl. u. a. Gottfr. Tristannbsp;3630 baz lobebcere, 7744 baz gemuot, Iwein 3969 baz ein scelic man =nbsp;sceliger m. si 374 bezieht sich nat�rlich aul �ppic und Irre (367).

429. Relatives so ist der Sprache Hans Rieds gemaB, nicht aber der des Dichters; es ist also mit daz zu ersetzen.

444. Die Form mac ist wohl durch Attraktion an gemach ent-standen. Man lese dafiir nmoz, nach zahlreichen Stellen in den W�bchern.

477. In den beiden ersten Auflagen hat der Herausgeber das handschriftliche aber beibehalten und ich sehe nicht ein, inwiefernnbsp;das alter der dritten den Vorzug verdient. Ungefahr denselbennbsp;Gedanken hat Freidank 117, 10 swem dicke hit geschikt dennbsp;enwirret truren niht.

490 f. Schr�ders Text scheint der Hs. genau zu folgen, da er nichts in den Laa. bemerkt. Den Sinn kann man h�chstens erraten.nbsp;Ich m�chte die Stelle etwa so verbessern: miner sorgen, der ichnbsp;tusent hdn, diu machent uz ein iewederem dine . . ., d. h. �durch dienbsp;vielen Sorgen, die ich bereits trage, wird jedes [weitere] Dingnbsp;(Ungluck) verdreiBigfacht�. Ich muB aber gestehen, daB ich mitnbsp;diesem Versuche nicht ganz zufrieden bin.

496. Der Genetiv darf eigentlich nicht fehlen; vgl. Erec 6041 got si der mirs ein ende gebe.

607. Den �berladenen Vers kann man durch Streichung von ,der hand beseitigen. Vgl. Wig. 6331 er zoch ir ahe ein vingerlin,nbsp;was sich vielleicht auf Parz. 131, 16 f. bezieht.

661. In Anlehnung an die Hs. m�chte ich mit Haupt zu alle samt zur�ckkehren; vgl. noch 699.

718. Lexer, Wbch. s. v. strecken zitiert: ,Ein segel hiez er ze iegelichem maste strecken . daz 717 darf aber nur auf baniere 714nbsp;bezogen werden. DanachheiBt es vielmehr: )gt;EineFahnelieB er annbsp;jeden Speer binden; die [Fahnen] waren [wie der Segel] alle weiB.�

727. Der Artikel ist entbehrlich, auch wird der Vers glatter, wenn man ihn streicht und daf�r unde liest.

875. Die hs.liche Lesart gewieret, ebenso wie Vers 953, verdient den Vorzug. Vgl. Lexer s. v. wieren, Haupt zu Erec 4636.

1062. Es ist nicht einzusehen, weshalb man die hs.liche Lesart gegangen mit MaBmann und Haupt nicht beibehalten sollte. Einnbsp;,man gevangen� hatte doch nicht wohl zum Helden gehen k�nnen.nbsp;Vgl. noch Junker u. der treue Heinr. 1330 da kam ein ritter gegangen.

ii6o. Die schwache Flexion des Namens Veldeke ist die vor-herrschende, wie bei Gottfr. v. StraBburg 4725 von Veldeken Heinrich, wonach meister an unsrer Stelle zu streichen ware.nbsp;Audi Wolfram Parz. 292, 18 sagt einfach her Heinrich v. V.,nbsp;wogegen aber Herbert im Troj. Krieg 17381 ihn meister nennt.

1311. Bech will lieber niht entbehren und zwar mit Recht. Vgl. die Laa. zum Iwein 588 sowie Erec 261, wo es ebenfalls �ber-fl�ssig scheint.

1313. Wie der Konjunktiv lebe 1311 durch den Reim bewahrt blieb, so ist gediene hier zu lesen.

1513. belangen statt verlangen ist vorzuziehen; letzteres wird nur zweimal belegt, eins davon mit ndch Engelh. 15.

1551. In den Laa. stellt Schr�der die Frage ,erklancte er?� Das Transitivum hatte aber ohne weiteres in den Text gesetztnbsp;werden sollen, doch vgl. Nibel. 2285, 4 Nibelunges swert das guotenbsp;vil l�te �f Dietrich erklanc.

1645. Seit der zweiten Auflage hat Schr�der wert f�r das �ber-lieferte wort gesetzt. Dieses ist aber wohl beizubehalten, denn es hat hier die Bedeutung �Ruf, Leumund�. Vgl. Wbeh. 3, 808,nbsp;Erec 830, 2695. Lanz. 3414. Hartm. Lieder 211, 24 usw.

1691. Lies lo�p, bl�emen unde gras, wof�r Belege wohl �ber-fl�ssig sind.

1780. 1st nicht rede aus 1777 geholt und rime daf�r zu lesen ? rime rihten Engelh. 213. Otte 751. Heinzlein iii, 9 usw. rimenbsp;iingerihtet Reinh. Fuchs 2249.

KLEINE BEITRAGE ZUR TEXTKRITIK UND erklArung von �des MINNESANGS FR�HLING�

Die nachstehenden Bernerkungen zum Text und zur Erklarung von �Des Minnesangs Friihling� waren urspr�nglich als Teil einernbsp;Besprechung von Friedrich Vogts vorz�glicher Ausgabe (3. Auf-lage, 1920) bestimmt, kamen aber dann, als die feindliche Kriegs-stimmungdeutsche Studi�n in Amerika vernichtet zu haben scliien,nbsp;leider nicht zur Niederschrift. Ob sie es heute noch verdienen, be-kannt gemacht zu werden, mag der Leser selbst entscheiden.

In seinem ausgezeichneten Aufsatz: �Die K�renberg-Literatur und die Anfange des deutschen Minnesangsd {Germanisch-Roni.nbsp;Monatsschrift Sept./Okt. 1927) hat Gustav Ehrismann k�rzlichnbsp;so ausf�hrlich �ber den K�renberger gehandelt, daB dem dortnbsp;Gesagten nur wenig hinzuzuf�gen ist. Zu den beiden Strophennbsp;des ersten Tones m�chte ich jedoch bemerken, daB die Weise innbsp;7, 5, reimend mit 7, 14, beim Sammeln im 13. Jahrhundert wohlnbsp;als K�rner angesehen wurde, wodurch die beiden liet in diesemnbsp;Tone f�r uns gerettet wurden.

7, I. Bartschs Lesung ))fremden� ist nach meiner Meinung wegen des Gegensatzes zu �behaltenlt;lt;. 7, 3, Vogts ncheidem vorzuziehen.

7, 2. Zu den Reimen �schedelich: lobelich� vgl. �gremelich: hei-melich�. Rol. 23, 5.

7, 17, 18. Die Stelle im Rother 2931: daz ist mir daz minnist � ist mir einerlei, scheint mir die Ueberlieferung in C zu bestatigennbsp;und Vogts Aenderung unn�tig zu machen. Der Sinn der beidennbsp;Zeilen ware daher: meine Frende wie alle anderen Manner sindnbsp;'niir dann einerlei. Es ist das trostlose Elend nach verlorener Liebe,

das der Sanger ausdr�cken will. Die Zeilen w�rden dann lauten: daz mm jroide ist daz minnist und alle andere man.

'], 19, 20. Lachmanns geschraubte Erklarung dieser Zeilen, wie Vogt sie umschreibt, ist nach diesem �freilich nicht ganz so plannbsp;im Ausdruck, wie man es beim K�renberger zunachst erwartet�.nbsp;Viel einfacher und einleuchtender ist es, in den Zeilen mit Ehris-mann eine bekannte sprichw�rtliche Wendung, den Gegensatznbsp;von leit und lief, Liebesleid und Liebesfreud zu sehen.

8, I. Zu dem mir ware Anno 614: d� stuont imi �f der vilinbsp;guote man, sowie Heiland 202 b (Heyne): ging imu heranzu-ziehen.

8, 9 ff. Nach Ehrismanns Ausf�hrungen darf auch diese Strophe unserem Dichter zugesprochen werden.

8, 16. �so sprach daz wip� scheint so nachgebracht, als k�nnte es auch fehlen, ahnlich wie: 39, 7; 5, 6; 32, 3. 7.

8, nbsp;nbsp;nbsp;33. In betreff des viel behandelten Falkenliedes und seinernbsp;vermeintlichen italienischen Quelle sei noch bemerkt: das An-deuten, die Bl�te des deutschen Liedes, fehlt im italienischennbsp;Sonett, so daB Entlehnung von kelner Seite passen will, zumalnbsp;das deutsche Lied, knapp und rasch, ganz anderen Sachgangnbsp;hat als das italienische. Bei K�renberg ist der Falke nur in dienbsp;Freiheit entflohen, wieder wild geworden, im italienischen Sonettnbsp;dient er nun bald einer Anderen. Da k�nnte also nur dieses jenemnbsp;nachgemacht sein, mit Weiterf�hrung des Gedankens. Auch ver-gleiche man den reinen SchluB des deutschen Liedes mit dernbsp;verh�llten Zweideutigkeit des italienischen.

9, nbsp;nbsp;nbsp;19. 20. Nach versuonde sollte Komma stehen und nach wolnbsp;Ausrufungszeichen. IFo/ = euge!

10, nbsp;nbsp;nbsp;I. K�nnte der tunkel sterne wohl auch Kornet bedeuten ?nbsp;Daf�r scheinen mir die von Pfeiffer u. a. beigebrachten Stellennbsp;weit eher zu sprechen als Vogts Deutung, �der schwach leuchtende,nbsp;der sich dem Blicke zeitweilig ganz entzieht�. Besonders dienbsp;Zitate aus Bertsch mitteld. Gedichte S. 16, 531 und S. 15, 511 ff.nbsp;(bei Pfeiffer, Germania 12, 224 ff.) w�rden schlecht zu diesernbsp;Deutung stimmen, da hier der tunkelsterne gerade seines Fichtesnbsp;wegen ger�hmt wird, dem an Leuchtkraft und Sch�nheit dasnbsp;Licht aller sieben Planeten zusammengenommen nicht gleich-kommt. Wozu auch die Matmung K�renbergs an die Geliebte,nbsp;sich zu verbergen wie ein ohnehin �schwach leuchtender� Stern ?nbsp;FaBt man dagegen unser Liedchen als eine Art Morgenstandchen

mit dem Hinweis auf den heil leuchtenden Kometen, der sich bei anbrechendem Tage zu verbergen scheint, dann wird die Mahnungnbsp;des Dichters weit eher verstandlich. Auch will mir der Vergleichnbsp;der Dame mit einem glanzenden Kometen viel passender er-scheinen als der mit einem �schwach leuchtenden� Sternchen.

i6, 15, 22. Vogt �bersetzt die Strophe in seiner Mittelhoch-deutschen Literaturgeschichte mit Beiseitelassung der Lesung in AC also: �Ich lag den Winter allein. / Schonen Trost gab mirnbsp;ein Weib, / seit mir Erende k�ndeten / die Blumen und dienbsp;Sommerzeit. / Das neiden mir Aufpasser. / Davon ist mein Herznbsp;wund. / Hilft mir nicht eine Frau mit ihrer Minne, / so wird esnbsp;nimmermehr gesund.lt;(

Ich glaube nicht, daB dies den Sinn des Dichters trifft, �ber den �brigens weder Lachrnann-Haupt, noch Bartsch, der �v�rnbsp;sie mir vr�ide kuntenlt;( liest, ein Wort bemerken. Was sollte auchnbsp;der Wunsch des Dichters, daB ihm eine Frau sein wundes Herznbsp;mit ihrer Minne heile, wenn eine andere ihn doch erfolgreichnbsp;getr�stet hat! Und w�rden ihm die Aufpasser nicht auch dasnbsp;zweite Liebesverhaltnis verleiden ?

Ganz anderen und, wie ich glaube, den richtigen Sinn erhalt die Strophe aber, wenn wir mit A C lesen:

v�ere sie mir mit, vr�iden kunten die bluomen und die sommerzit,nbsp;und mden, ebenfalls mit A C, als Conjunctiv praet. fassen.nbsp;Was ihm das �herze wunt� macht, ist die Erkenntnis, daB dienbsp;Aufpasser ihm die Vertr�stung auf die Sommerzeit mit ihremnbsp;Neid vereitelt haben w�rden. Darum hofft der Dichter am Endenbsp;der Strophe auf Heilung seines wunden Herzens durch diewirk-1 i c h e Liebe der Frau.

F�r seine Annahme, daB 20, 17�21, 4 einen Wechsel bildeten, hat Vogt keinerlei Beweis vorgebracht.

20, 25�-21, 4. Die schone, wohl im Dienste eines edlen Ge-schlechts, dem er nach einem Ungl�ck Trost zuspricht, gedichtete Strophe zeigt uns den Sanger als Berater, ganz wie auch 24, 25.nbsp;In ahnlicher Weise beziehen sich ferner die Strophen 23, 5 und

24, 9 wahrscheinlich auf sein Verhaltnis zu derselben Gruppe von Freunden oder Genossen.

23, 32. stro, gew�hnlich = Halm, auf dem Felde stehend mit der Aehre, vgl. Walther 66, 5,7. Zugleich aber bezeugt dienbsp;Waltherstelle 17, 35: von grase wird ein halm ze stro, dab stronbsp;auch die gedroschenen Halme, d. h. Stroh in unserem heutigennbsp;Sinne bedeuten kann, wobei an Stroh als Lager zu denkennbsp;ist, wie auch Wilmanns in seiner von V. Michel besorgtennbsp;4. Ausgabe von Walthers Gedichten S. 104 betont. Die Stellen,nbsp;die er hier f�r diese Bedeutung von stro heranzieht, undnbsp;zu denen noch Winsbeckin 14, 2 zu f�gen ware, weisen sichernbsp;auf die Bedeutung Stroh, Strohlager. Ebenso hatte die bitterenbsp;Bemerkung am Ende der Strophe, daB das stro wieder ze mistenbsp;werde, wenig Sinn, wenn halm nicht auch ein Synonym vonnbsp;stro ware.

�6, ^.7. Trifft meine Vermutung, daB der Sanger in dieser Sti'ophe die Streu (den Haufen Stroh) besingt, auf dem er innbsp;seiner Armut schlafen muB, dann batten wir in dem kleinennbsp;Kunstwerk ein Beispiel seines genialen Humors, der sich in sieg-hafter Geistesfreiheit �ber seine und seiner Genossen oft beklagtenbsp;d�rftige Lage hinwegsetzt.

26,10. Hinter erhen sollte Doppelpunkt stehen, um anzudeuten: d. h. einen solchen (erben).

26, nbsp;nbsp;nbsp;21. Herg�r. DaB wir in Herger den Namen des Dichters zunbsp;sehen haben, k�nnte von Vogt starker betont werden als er esnbsp;tut. Unter Spervogel ware dann die Bezeichnung einer Geschmacks-richtung zu versteken, der Fahrende verschiedenen Namens huldigt en.

29, 20. Dem Dichter schwebt augenscheinlich ein Herrenhof vor, in dem zweierlei Obst wachst, s�Bes und saures.

29, nbsp;nbsp;nbsp;24. wir suln das ohez teilen, gemeint ist das scheiden dernbsp;Guten von den B�sen, das auch Walther oft verlangt, z. B.nbsp;58, 35; 48, 29; 45, 28.

30, 21. stuont sich, vgl. 8, i ich stuont m i r nehtint spate; �f erstuont sich Gorijo dar, M. S. D.� 37, 29 (Vom heiligen Georg);nbsp;do stuont imi �f der vili guote man. Anno 614.

33, 27. bezzer hier nicht, wie Scherer D. St. 2, 43, 66 will, imnbsp;eigentlichen moralischen Sinn gemeint, sondern = Uwer, w er der,nbsp;wie bei Johansdorf 94, 14 und Meinloh ii, 7.

33, nbsp;nbsp;nbsp;34. In der Handschrift C sind diese beiden Verse ganznbsp;anders gemeint, ohne B waren sie, wie sie nun stehen, kaumnbsp;herzustellen gewesen.

34, nbsp;nbsp;nbsp;19. Gedanke . . . fri, vgl. dazu Hildebrand, DWb. IV. i. I.nbsp;1961 unter Gedanke. Erst von diesem Gedichte an finden sichnbsp;�berschlagende Reime und Waisen.

35, nbsp;nbsp;nbsp;32�36, 4. Die Strophe zeigt einen merkw�rdigen Kampfnbsp;des Frauenherzens um die minne und maze.

36, nbsp;nbsp;nbsp;2. also auch das gew�nschte senen 35, 35 nur mit wohl-t u e n d e r maze.

37, nbsp;nbsp;nbsp;22. miniu wol st�nden ougen � meine v o r h e r schonennbsp;Augen.

39, 2. s� h�h �w%, wohl soviel wie: hohe klagen, h�hiu klage, wof�r der Ausruf in der 3. Strophe spricht. Mit Sch�nbach u. a.nbsp;in dem Refrain einen Schifferruf zu erblicken, kann ich mich nichtnbsp;entschlieBen.

Auch �ber den Text und die Erklarung der �brigen Dichter des �Fr�hlings� ware mancherlei nachzutragen, doch glaubte ich michnbsp;zunachst auf die am meisten umstrittenen Sanger der deutschennbsp;Fr�hzeit beschranken zu sollen.

THE PHONETIC AND MORPHOLOGICAL SETTINGS OF THE MIDDLE HIGH GERMAN CLIPT PRETERITS

Professor Collitz in two clear and admirable articles ^ has discust the history and probable origin of the MHG clipt preteritsnbsp;gie, fie, lie, beside the full forms gieng, fieng, liez. The presentnbsp;inquiry is based primarily on Collitz�s findings, but attempts tonbsp;trace more fully (i) their basis in OHG preexistent clipt forms;nbsp;(2) their phonetic background in MHG, and (3) to show how thenbsp;shifting of this background conditioned both the original development and later disappearance of these forms. It thus constitutesnbsp;a concrete study of the trend of linguistic development as shownnbsp;in these forms. ^

A brief summary of Collitz� articles may serve to refresh the memory of those who have already read them or to whet thenbsp;appetite of those who have not yet done so. He points out thenbsp;following facts:

The forms gie, fie, lie are the i and 3 sg. prets. which arose toward the end of the eleventh century in the transition period between OHG and MHG asnbsp;doublets of the older historically correct preterit forms gieng, fieng, liez ^ fromnbsp;the reduplicating-preterit verbs gdn (g�n), fdhen (fan), lazen (Ian).

� This continuation of Professor Collitz� work along these special lines was begun several years ago with his kind permission and approval. The subject wasnbsp;first workt out under a Harrison Research Fellowship in Linguistic Psychologynbsp;at the University of Pennsylvania in 1922�3, forming part of a larger study ofnbsp;the normal direction of structural development in language. Its only publicnbsp;presentation hitherto was in a lecture before the Germanic Society of the University of Pennsylvania. The present article is a revision and extension of certainnbsp;portions of this lecture.

� For convenience the MHG forms are cited to represent both the OHG and MHG forms.

These dipt forms first appear clearly in the Merigario and the Vienna Genesis. Collitz� investigation of the latter shows the dipt forms to be fully developt andnbsp;used in alternation with the full forms according to more or less consistent rules.

As verse or sentence finals the dipt forms greatly predominate �for gie ry; 2, fie 9:1, and lie 8:5 � most of the exceptions being explainable by the exigenciesnbsp;of rime or poetic style.

As verse or sentence medials, the dipt forms are used before a following consonant, the full forms before a following vowel, as gieng er; gie der, inpfiengnbsp;in; impfie non, liez uns; lie si.

Collitz� figures show 42 regular to 8 irregular occurrences. If he should class iu in giench ludas as a diphthong instead of a consonant -|- vowel, his figuresnbsp;would become 43: 7. Of these seven irregular uses, only one (giench uon) is a fullnbsp;form used before a consonant, the other 6 being encroachments of the dipt formsnbsp;before vowels. There are in all 23 instances where Collitz� rule calls for full forms;nbsp;they occur in 17 of these, while in 6 instances the dipt forms occur in their stead.nbsp;This clearly proves Collitz� contention that these clippings were fully developt formsnbsp;already intrencht in the speech consciousness of the folk. They also prove that thenbsp;dipt forms were still extending their use at the expense of the original full forms.

Next, as to the possible origin of these forms: prior to their extant literary appearance, the verb lazen developt a dipt 2 sg. impv. Id from the present stem Idz. This appears in Notker, being used by him as the tiefton or reduced gradenbsp;of impv. Idz Paul � and Braune � explain this form by analogy of the -mi verbsnbsp;gdn and stdn. Collitz rejects this explanation as a starting point, believing thenbsp;form Id was due to the contaminative influence of the closely associated andnbsp;partly synonymous 2 sg. impv. tuo. The development of Id, however, necessarilynbsp;brought Idzen into touch with gdn and stdn, so that from this point on Collitz�nbsp;explanation agrees with and amplifies that of Paul and Braune.

Following the impv. Id came the dipt pret. sg. lie, which occurs once in Notker and twice in the OHG Glosses; then the 3 sg. pres. ind. Idt in the Memento mori,nbsp;the Prudentiiis Glosses and the OHG Glosses. ^

The dipt preterit gie makes its first extant appearance in Merigarto, * while fie (entfie) is first extant in the Vienna Genesis. ^ By the opening of the MHGnbsp;period the whole present and preterit systems of the three verbs gdn, fan and Idnnbsp;had been leveled, and in the present this parallelism extended to the verb stdn. �

The rise of these dipt preterits was favored by the fact that the resultant distinction between singular and plural preterit stems brought them into linenbsp;with the ablaut verbs generally, with which they had long been groupt in speechnbsp;feeling. � Thus gie: giengen, lie: liezen are, in a general way, comparable to sang:nbsp;sungen, was: waren, etc.

With the NHG leveling of singular and plural pret. stems came also the disappearance of these dipt preterits gie, fie, lie and the revival and retriumphnbsp;of the original forms gieng, fieng, liez (ging, fing, liefi).

So much for a brief and therefore inadequate summary of Collitz� investigation, which the reader should by all meansnbsp;know at first hand.

' The pret. stiiont was not of the reduplicating type and so did not admit of a similar levelling.

� This grouping was in fact primitive, as the reduplicating verbs were ablaut verbs also in Gc. and IE,

Owing to the tendency of verb forms to take a strong sentence stress, they are apt to be followed, especially in verse, by wordsnbsp;of weaker stress. This is particularly true of monosyllabic verbnbsp;forms such as the i and 3 sing, preterit forms. Furthermore thenbsp;requirements of sentence connection most often favor the bringingnbsp;of a subject or object pronoun or an article belonging to a nounnbsp;subject or object, or an adverb or preposition next to the verb.nbsp;For both these reasons the verb is most apt to occur beforenbsp;unstrest sequence, especially before enclitics or proclitics.

As a result, even in authors (such as Walther von der Vogelweide and Gottfried von StraBburg) who follow the rule of simple consonant and vowel sequence without regard to stress or sentencenbsp;position, most of the actual combinations bring an unstrest wordnbsp;(most often a clitic) after the monosyllabic preterit form. Thus innbsp;Tristan und Isolde, the sentence medial occurrences of thesenbsp;dipt and full preterits and of Idjldz (including verse finals usednbsp;as sentence medials) show 90% of unstrest consonant or vowel sequence �. Such unstrest sequence was therefore the main sandhinbsp;environment in which these forms developt.

Linking, or the tendency of words to run together in connected speech, is a familiar phenomenon in many languages, as in Frenchnbsp;liaison and Sanskrit rules of euphonic combination Thisnbsp;tendency is strongest between adjacent strest and unstrestnbsp;words, as Hitt, ammu-ga �mihi�: Gr. e^Moiye.'Go. mi-k; Lat. hun-c,nbsp;pot-est; F. donner-ai; Eng. for-give, home-ward; Ger. zu-m, andnbsp;especially such MHG combinations as gruoztes {= gruozte si)nbsp;Nib. 1410 (Holzm. Ausg.), zehant (ibid.), sine 1409, zir (= ze ir)nbsp;161, zen (= ze den) 1417, giengez 432, gienger 653, ranger 641,nbsp;hieng in (pronounced as hiengin) 642, etc. That is to say, a fullnbsp;w'ord followed without any break by an enclitic or proclitic ornbsp;other unaccented word tends to combine with it after the phoneticnbsp;pattern of a dissyllabic word, this fusion taking place most easilynbsp;with a consonantal final followed by a vocalic initial, or a vocalicnbsp;final followed by a consonantal initial. This fusion may or maynbsp;not be represented in writing, but in either event the phonetic

pattern is the same. Thus gienger, gieng er, gie der, are all of the same general rhythmic and phonetic pattern as the dissyllabic 2nbsp;sing. pret. ind. gienge, the plural pret. ind. forms giengen, gienget,nbsp;and the pret. subjunctive forms gienge, giengest, giengen, gienget.

In other words, the full preterit forms were regularly preserved only in sandhi combinations which were for speech feeling identicalnbsp;in rhythmic and general phonetic pattern with the dissyllabicnbsp;preterit forms and whose sandhi combinations were practicallynbsp;internal. Where the full forms could not preserve this pattern,nbsp;it was preserved by the development of the dipt forms. In casenbsp;of encroachment in either direction, this rhythmic pattern isnbsp;destroyed. The rise of the dipt forms was favored by thesenbsp;phonetic analogies, which could not however function independently to create them, but only as accessories to the inflectionalnbsp;groupings and levelings which C. notes in his article.

A large number of dipt present and preterit forms were preexistent in the full OHG period. Such were: (i) the reduplicating dipt presents (the ga�-group); (2) the sixth class dipt presentsnbsp;(the siaw-group); (3) the quot;pure vowel� verbs (the �/aw-group);nbsp;(4) verbs of the first weak conjugation with long stem-vowelnbsp;followed by stem-final -w-; (5) the first ablaut class dipt presentsnbsp;and preterits (the sft(w)anlsp�o group); (6) the second ablaut classnbsp;dipt preterits (the blou-group); (7) the reduplicating dipt preteritnbsp;�hieb�; (8) the dipt present and preterit forms of hdbenjhdn.nbsp;Groupt with these in speech feeling, tho not involving syncopation,nbsp;were (9) verbs showing consonant variation between sing, andnbsp;plur. stems (the xo/i/^MgMW-group). � With the exception ofnbsp;hdn from hab�n, these forms were the result of normal phoneticnbsp;shifting, so that syncopation was merely apparent � which was,nbsp;however, the same thing for speech feeling as genuine syncopation.

I. The reduplicating verbs, because of their peculiar vocalization, formed in all periods of OHG (and earlier) a compact associativenbsp;group. They were in IE straight ablauting verbs, and a considerablenbsp;number were still so in primitive Germanic. All were inflectionallynbsp;groupt with the ablaut verbs; and, after the leveling out of theirnbsp;initial reduplication, they agreed with the other strong verbs innbsp;their general consonantal pattern, and also came again to show

2i6 a difference of vowel between the present and preterit stems,nbsp;which was certainly felt as ablaut. They were, therefore, for speechnbsp;feeling simply a special class of strong verbs.

Within the reduplicating group, a smaller sub-group of three verbs, gangan (gdn), fdhan and hdhan, showed still more intimatenbsp;associations with each other. Their full stems and their inflectionsnbsp;rimed thruout, and all showed in the present system phoneticnbsp;variation based on the apparent weakening or syncopationnbsp;of the final stem consonants f-ng). This parallelism was furthernbsp;increast (tho it never became absolutely perfect) by the fact thatnbsp;medial -k- was sporadically syncopated in all periods of OHGnbsp;so that completely syncopated forms of fdhan and hdhan occurrednbsp;occasionally in writing and probably far more often in speaking. Thenbsp;resultant analogical proportion for speech feeling was therefore:nbsp;gangan j(nbsp;nbsp;nbsp;nbsp;) jgdn jgiang jgiangun jgigangan

2. The closest external association of the reduplicating verbs was with the verbs of the sixth ablaut class, owing to the partialnbsp;similarity of vowel gradation. In this class also there was a subgroup, consisting of stantanjstdn and slahan/(sldn), also showingnbsp;apparent syncopation in their present systems, and hencenbsp;closely associated with the gaw-group both by this syncopationnbsp;and by the resemblance of the phonetic type of stdn to gdn andnbsp;of slahan to fdhan, giving the approximate phonetic proportionsnbsp;gangan: gdn:: stantan: stdn and fdhan: fdn:: slahan: sldn. Their analogical proportion for the full and dipt forms was stantan: ( ):nbsp;stdn: stuont: stuontun: gistantan:: {nbsp;nbsp;nbsp;nbsp;): slahan: (sldn) : sluog

Gdn of course was not a syncopation of gangan but an independent word going back to Prim. Gc. *gai- and perhaps to IE *co-i-. See Kluge Et. Wtb.nbsp;s. V. gehen. (However, see Streitberg, Urgerm. Gramm., p. 319.) � Similarly,nbsp;fdhan and hdhan were differentiated from the participial stems (gi)fang(an)nbsp;and (gi)hang(an) not by weakening but by grammatical change (Verner�s Law).nbsp;�� See Braune, Ahd. Gramm. �� 33 and 128 A 1 with further references therenbsp;given, Wright Gram, of the Go. Lang. �� 59 and 126 f.

Braune, Ahd. Gramm. � 154 with A i and 7. � For the full development of these dipt present forms before the close of the OHG period, cf. the free usenbsp;of such present syncopations as fdn, gesldt, hdn, hat, etc. in Memento rnori. �nbsp;See Collitz art. cit., p. 456 with n. ii.

Stdn lt; IE *std- was wholly independent of stantan, so that here also syncopation was apparent only, but none the less real for speech feeling; slahan lt;nbsp;Gc. *slahan- has h by grammatical change {Verner�s Law) � see Braune � 346 A 2.

3- Closely associated with the gdn-stdn groups were the pure -d- verbs, which showed apparent syncopation in their presentnbsp;formations, as bldhen (bldjen) jbldn �blahen�, drdjenjdrdenjdrdnnbsp;�drehen�, kndhenjkndenjkndan quot;kennen�, krdhen (krdjen) jkrdennbsp;quot;krahen� (cf. noun krdja (krdwa) jkrd quot;Krahe�), ndhen (nawen,nbsp;ndjan) jnden �nahen�, sdjan (sdhen, sdwen) jsdenjsdn �saen�, etc.nbsp;Originally they were reduplicating verbs wdth strong past participles showing apparent syncopation, as gibldan (gibldn), gidrdannbsp;(gidrdn). Cf. also apparent syncopation in such weak formationsnbsp;as bikndt �bekannt�. These syncopated forms continued intonbsp;MHG, as bldnjgebldn, drdnjgedrdn (gedrdt), beside such full formsnbsp;as gedrcejet and new syncopations as gedrcet.

The new analogical proportions offered (bldhen:bldn::sdhen:sdn, etc.) w�ere such as to associate themselves closely with fdhan:nbsp;fdn::slahan:sldn, etc., and added the important patterns ofnbsp;syncopated past participles and new consonantal syncopationsnbsp;and variations.

The possibility of speech feeling holding such forms as bldn/ gibldn for syncopated forms depends on the stem increments,nbsp;-h-, -j- and -w-, being not merely graphic, but actually pronounced.nbsp;For the proof of the actual pronunciation of the -h-, see Braunenbsp;� 152 b). For the reality of the compare (i) the early shiftnbsp;of these stems to the -jan conjugation, (2) their uniform umlautnbsp;in MHG sajen, wcejen, etc., (3) 0. Sax. and Dutch forms in -ian,nbsp;(4) such Germanic -f-stem formations, as Go. saian, waian (: Lith.nbsp;v�jas quot;wind�: Skt. vayiis quot;wind�, etc.), Go. daddjan quot;saugen�nbsp;(: OHG tden lt; Gc. *daian- lt; IE *dh�i-jdhn-gt; Skt. dhdyati:nbsp;Armen. diem:0^\i\g. doja:01r. dinim, etc.). �� The genuinenessnbsp;of the stem final -w- appears (i) in its dialect survival in MHG,nbsp;as in Upper Sax. and Thur. m�wen, s�wen, w�wen for MHGnbsp;mcejen, scejen, weejen 1�, (2) in its OSax. survival in the old reduplicated forms seu (OEng. seow) and threu (OEng. pr�ow) beside thenbsp;newer weak forms saida, thraida, (3) in its uniform preservationnbsp;in OEng. blawan, prawan, sawan, etc.

Furthermore, many of these verbs show ablaut and stem-final variations in IE relatively identical with those appearing in Germanic and OHG and MHG. Variation in phonetic form by ablautnbsp;and suffixal extension was, as is well known, characteristic of IE

2i8 verbal roots generally, as in the Sanskrit and Greek present stemnbsp;formations. Their occurrence in the pure vowel verbs in Germanicnbsp;is therefore not an anomaly from the IE standpoint, but annbsp;interesting survival of a normal IE type of verb formation. See,nbsp;for instance, Walde-Pokorny Vergl. Wth. d. idg. Spr. under thenbsp;roots 2bhel-jbhl�-jbhl�- ��ppig sprieBen�, the obviously identicalnbsp;4bhel-jbhl�-jbhl�-s-jbhl�-jbhl�-s- �aufblasen, schwellen, strotzen�,nbsp;the related (extended) roots 2bhlei-jbhlei-s- �aufblasen, schwellen,nbsp;strotzen, �berflieBen�, whose dehnstufe bhl�i- is not listed bynbsp;VWIS but occurs in Lat. flare, OHG bidjan, etc., bJileu-jbhl�u-(dehnstufe in Greek citations) quot;aufblasen, schnauben, briillen,nbsp;schwellen, strotzen, iiberwallen�, 8bhel-(fbhel-n- or bhel-s-) �briillen, bellen�. The -k- and -q- extensions � which would yieldnbsp;Germanic forms in -h- (or -w- by Sievers� Law) � are cited bynbsp;VWIS only under hhleig- and bhleu-(k)-, extensions respectivelynbsp;of ibhlei- and bhleu-, which are in turn extensions of ibhel-�glanzen�, and as such obviously related to the other bhel- rootsnbsp;cited above. (For connection of meanings cf. Eng. burst, Fr.nbsp;�clater, Eng. blaze beside Ger. blasen, etc.).

These IE variants yielded Gc. forms as follows: IE bhel- gt; MDu. beien; IE bhel-s- gt; OHG hellan; IE bhl�- gt; OHG bldan (distinguish from bidjan gt; later blden); IE bhl�i- gt; Gc. bl�i- gt; OHGnbsp;bidjan (later blden); IE bhl�-s- gt; Gc. bl�s- gt; OHG bldsan; IEnbsp;bhl�-u- gt; Gc. bl�u- gt; OHG *bldwan (accidentally not extant,nbsp;but cf. OSax. and OEng. forms, and the -w- grade in other verbsnbsp;of this group in OHG); IE bhleiq- (bhleik-) gt; Gc. bl�hfw)- gt;nbsp;OHG bldhen; IE bhl�- gt; Gc. bl�- gt; OHG bluoan: IE bhl�-i- gt;nbsp;OHG bluojen (pluogen); IE bhl�-u- gt; OHG blouwen; IE bhl�-k- gt; OHG bluohen; IE bhl�-s- gt; MHG bluest; OE blostma; etc.

In other words, the stem variations of bldanjbldjanl*bldwanj bldhen, bldsan, bluoanjbluohenjbluowen, and other phoneticallynbsp;and semantically resembling verbs, agreed with and presumablynbsp;were derived from similar variant formations in Indo-European.nbsp;Theoretically bldan, bidjan, bldhen and *hldwan might as wellnbsp;have given separate words in OHG and other Germanic languagesnbsp;as did bluoan and bldsan � or speech feeling might have includednbsp;the IE -�- and -es- grades bluoan and bldsan in the group as variantnbsp;forms of bldan jbldjan � cf. the Gc. -�i-, -�- and -�- grades in Go.nbsp;saianjsais�j-s�ps, and the inceptive -s- suffix as a tense sign in thenbsp;sigmatic aorist and future in Sanskrit, Greek, Latin, Tocharian, etc.

Similar stem variations occurred in other IE roots � see, for instance VWIS sub rad. mai- �hauen, schnitzen� with itsnbsp;extended form maid- beside (s)me(i)-j(s)mi- �hauen, abhauen,nbsp;etc.�, with which compare OHG mdenlmdjen, meizan, etc. Seenbsp;also (s)n�(i)- �spinnen, nahen� gt; Skt. snayati �umwindet�:nbsp;Gr. vfj {*av^iei) �spinnt�: Lat. n�re �spinnen�: Mir. sniid (fr.nbsp;*snei-) �spinnt�: OHG ndjanjndenindwen; etc.

The pure vowel verbs were, therefore, contaminative OHG groupings of resem^bling stem formations reduced in speechnbsp;feeling to the level of phonetic variants. Naturally these variationsnbsp;did not occur equally in all verbs of the group, some originallynbsp;having several variations, others one or none (as tden: Go. daddjan),nbsp;but the group once formed, variations tended to spread and assimilate in all verbs of the group. The name �pure a� is a misnomer.

Only one verb beside hldhen could have conceivably furnisht a starting point for the -h- formations, namely dr den gt; NHGnbsp;drehen. Intrinsically, the etymological equation drehenjdrechsel lt;nbsp;OilG drd(h)enjdrdchsil i^�Vgt;rech.s\eT:�) lt; Gc. *prehv- lt; IE* tereq-gt; Gr. xq�nm: Lat. torqueo, is both phonetically and semanticallynbsp;flawless. But, because of the character of drden as a �pure� -d-verb, this connection is usually given up in favor of Lat. tero:nbsp;Gr. rslgojl*r[TQrjf.u �rub, abrade, wear out, wear holes thru�/nbsp;roQfws, rQfjf.ta �aperture, hole� (: Ger. Darm �gut�) and thenbsp;supposedly connected x�qvoq �carpenter�s compass, lathe-chiselnbsp;(for wearing or rounding off)�. Both Kluge and Boisacq furthernbsp;connect the root with Skt. tarati �goes thru, crosses�.

But the -d- j-dh- verbs offer no bar to the connection of drehen with drechsel, but only a confirmation. IE g' gt; OHG hwjhjw,nbsp;as IE *quis gt; OHG hwerjwer, IE Hiq- gt; OHG lih�mjliwum,nbsp;IE *seq- gt; OHG gisehanjgisewan; cf. also ahajouwa �water,nbsp;stream�, and other variations of hjw in orthography or derivation. The sound which could thus tempt double spelling mustnbsp;have been a velar h or ch [hv or with a -u- offglide. With thenbsp;above noted occasional syncopation of h we should thereforenbsp;expect a Gc. *^r�hvian to appear in OHG with the spellingsnbsp;drdhen (drden, drdn), drdwen, drdjan, several of which actuallynbsp;occur. Such a verb would fall automatically into the -d-j-dw-j-dj-group, or rather help to create it, and would tend to spread bothnbsp;h and w spellings and pronunciations to other verbs of the group.

3 b. The pure -�- (OHG -uo-) verbs blnoen, druoen, gluoen, gruoen, tuoen, muoen, s-puoen, showed the same extensions asnbsp;the pure -d- verbs, frequently inserting -h- (as muohen, bluohen) ornbsp;-w- (as bluowen), and showing apparently syncopated forms (asnbsp;muon, gluota, spuon, spuot) in the present and preterit.

3 c. The verb tuon (Braune � 380 f.) inflects in the present system as a dipt -uo- verb, in the preterit and perf. ptc. approximately as a dipt pure -d- verb.

3 d. The pure vowel verbs therefore made several important contributions to the development of syncopating verb types;nbsp;(i) They establisht types of (apparent) syncopation in pret. andnbsp;perf. participle forms as well as in the present, thus strengthening the tendency to dipt formations as such, and also makingnbsp;a beginning of a connection in speech feeling between presentnbsp;and preterit syncopations. But their syncopated preterits (drdta,nbsp;sdta, spuota, t�ta, etc.) could not associate closely with strongnbsp;preterits (sluoc, giang, etc.) because of the difference of phoneticnbsp;and inflectional pattern. (2) But their present syncopations closelynbsp;paralleled and therefore strengthened those of the gdn and sldnnbsp;groups. Compare, for instance, the phonetic and functionalnbsp;proportions jdhan:fdn::slahan: sldn::bldhen:bldn (::muohen:muon),nbsp;etc. (3) They establisht in advance a type of past participle syncopation suitable both for the reduplicating dipt presents and fornbsp;hab�n in such forms as gibldn, gitdn, etc.

4. Verbs of the first weak conjugation with long stem vowel followed by stem-final -w- showed the same connection betweennbsp;present and preterit syncopations as did the pure vowel verbs.nbsp;In these verbs also apparent syncopation came as the result ofnbsp;normal phonetic shifting, the stem-final -w- disappearing necessarily before a consonant, and optionally before a vowel. As anbsp;result the preterit regularly showed syncopation, while the presentnbsp;and the perf. participle varied between syncopated and full forms.nbsp;The two following verbs occur:

a) The verb hiwen (Men) IhitalgiMwit (giMt) was analogous thruout to the -w- forms of the pure vowel verbs, but because ofnbsp;the difference of vowel and the absence of an -h- grade, it couldnbsp;Braune � 359 A 4.

� Braune �� io8, no A i. These rules do not affect verbs of this class -with short stem vowel, which therefore remained unsyncopated, as frewenjfrewita,nbsp;gifrewit. Cf. also the unsyncopated pret. plur. and past ptc. in verbs of the spian-group.

not show the same close parallelism to the gdn and sldn groups. It necessarily contributed, however, to the feeling for syncopationnbsp;in all tenses and to the connection between present and preteritnbsp;syncopations.

b) nbsp;nbsp;nbsp;The verb Idwen (Iden) jldtajgildwit (gildt) �verraten�. occursnbsp;in the pret. subjunc. forms firldti and gildti in Otfrid. The remaining forms are positively inferable from these and from Go. l�wjan:nbsp;OEng. lawen quot;verraten� i�. Observe the close parallelism to thenbsp;-w- and dipt grades of the hldn-group, to which Idwen couldnbsp;be admitted unchanged.

c) nbsp;nbsp;nbsp;The verb huwan (buan) jbutajgibuwan (gibuan) ^ originallynbsp;belonged to the reduplicating conjugation, from which it retainednbsp;the strong perf. ptc. into the MHG period. The old reduplicatingnbsp;pret. plur. biruun (from earlier *bibu(w)un) occurs in Otfrid. Butnbsp;the pres, and pret. form.s buwan (b�an) jbuta wholly parallelnbsp;Mwen (Men) Ihita, bldwen (blden) jbldta, etc. � The -w- in buwannbsp;may be excrescent � see Feist op. cit. s. v. bauan.

5. Verbs of the first ablaut class with stem-final -w- suffered optional syncopation in the present system and regular syncopation in the preterit singular, according to the rules for the dropping of -w- noted in the preceding section They retained thenbsp;full unsyncopated forms in the pret. plur. and perf. ptc. becausenbsp;of the short stem vowel of these forms. The verbs originally innbsp;this group were spiwan (spmn) jsp�ojspiwum jgispiwan andnbsp;sniwan (snian) jsneolsniwumjgisniwan. By contamination of thesenbsp;verbs scnan created forms in -w- giving the gradation scnwannbsp;(scnan) jscreijscriwum (scrirun) jgiscnran.

Besides further strengthening the feeling for both present and preterit syncopation and for the connection between them, thenbsp;sptan-gronp added the further contribution of connecting thesenbsp;syncopations with the vowel gradation of ablauting preteritnbsp;singular and plural forms. The later preterit singular forms of thesenbsp;verbs were spe, sne, sere, giving a combined vowel and consonantnbsp;variation of -� (zero)-: -iw-. Such consonantal variation mightnbsp;at any time come to be felt as functional, just as the vowel variation was, for distinguishing the pret. sing, and pret. plur. stems.

Brauno, Ahd. Lesebuch, vocab. s. v. gildwenjfirldwen, and Ahd. Gramm. � 363 A 4 d; Feist, Etym. Wtb. d. goth. Spy. s. v. l�wjan.

� Branne �� 353 A 3, 354 A 3 d.�The perf, ptc. forms are inferred from the MHG forms.

6. nbsp;nbsp;nbsp;Verbs of the second ablaut class with stem-final -w- showednbsp;apparent syncopation in the preterit singular under the rule ofnbsp;dropping final -w. Long -u- appears as the stem vowel of thenbsp;preterit plural and perfect participle by the rule that -uww- gt;

The verbs involved axe hliuwanlhloujU�wunjgihl�wan, hriuwanjbroujbr�wunlgibr�wan, kiuwan / kou j k�wun / gik�wan,nbsp;riuwan / rou / ruwun / giruwan, niuwan / nou / nuwun / ginuwan.nbsp;By rule there should be optional syncopation after the long stemnbsp;vowel in the present, the pret. plur. and the perf. participle, andnbsp;this actually occurs in some instances, as bliuan, bluun, gibluan,nbsp;but this is rare because of the opposing tendency to develop annbsp;excrescent -w- between -u- and a following vowel. �� Their chiefnbsp;contribution was, therefore, that they reinforced the verbs of thenbsp;spian-gvowp in offering a combined vowel and consonant variationnbsp;between the pret. sing, and plural, namely, ~ou 4- zero: -ti -f w-.

7. nbsp;nbsp;nbsp;The reduplicating dipt preterit hio (later hie) from houwannbsp;(with stem-final -w-) was the sole instance of preterit syncopationnbsp;remaining among the reduplicating verbs after the above notednbsp;shifts of pure-vowel and stem-final-w verbs to the first weaknbsp;conjugation. Tense stems houwanjhiojhiowunjgihouwan showednbsp;as yet no such variation between the dipt and full preterit singularnbsp;forms as later developt in MHG hiejhieb, but only the distinctionnbsp;between sing, and plur. stems, which was consonantal only. Thonbsp;not as common in literary use as gdn, fdhan and Idzan, the verbnbsp;houwan must have been common in the spoken language; so thatnbsp;the preexistence of this dipt preterit form is important both innbsp;itself and as the only preexistent form of the later reduplicatingnbsp;dipt preterit group: giejfiejlielhie (�hing�)/Are (�hieb�).

8. nbsp;nbsp;nbsp;In the later OHG period the third class weak verb hab�nnbsp;developt the syncopated present forms hdn hast, hdt, hdnt,nbsp;and the syncopated preterit Aafe. The development of these shorternbsp;forms was favored (i) by the general speech feeling revealed bynbsp;the various dipt present, preterit and participle forms noted innbsp;the preceding sections; (2) by weakened sentence stress, as thesenbsp;dipt forms of hab�n jhdn occurred especially in the auxiliary usenbsp;of the verb; (3) by the specific analogy of other syncopated forms,nbsp;as hab�n, hdn, hdt, hdte, etc. beside bldwen, bldn, bldt. Mate, etc.,nbsp;or Idwen, ld(e)n, Idta, etc. This analogical influence must have

come into play still more easily owing to the close resemblance in phonetic type, the distinction between the voiced bilabial spirantnbsp;-h- and the voiced bilabial semi-vowel -w- being very narrow.

The dipt present forms of hah�n jhdn became closely associated in speech feeling with the corresponding forms of gan, stan, etc.,nbsp;and as in the case of hldnjhldta, hienjhtta, sfianjspe, etc., helpt tonbsp;establish still more firmly the connection between present andnbsp;preterit syncopations in speech feeling.

That the dipt forms of hah�njhdn may have been much older in actual speech than in their common literary usage is indicatednbsp;by the emergence of the auxiliary hdt in the Petruslied (Unsarnbsp;trohtin hdt farsalt. . .) in the ninth century. This in no wise affectsnbsp;the close relation of the hdn forms to those of the gdn, stdn andnbsp;hldn groups, since these latter (apparent) syncopations werenbsp;primitive.

9. nbsp;nbsp;nbsp;The dipt preterits came under the larger category of consonantnbsp;variation between the sing, and plur. preterit stems, under whichnbsp;they constituted the special case in which the first variant is zero.nbsp;They were therefore strongly reinforced by the strong verbsnbsp;showing grammatical change in the pret. sing, and plural. Thenbsp;consonantal equations involved were djt, hjg, sjr The followingnbsp;verbs occurred: (a) in the first ablaut class, sneidjsnitum, leidjli-tum, meidjmitum, z�hjzigum, d�hjdigum, r�hirigum, w�hlwigum,nbsp;l�hjligum, reisjrirum; (b) in the second ablaut class, s�djsutum,nbsp;z�hjzugum, k�sjkurum, jr�sjfrurum, firl�sjfirlurum; (c) in thenbsp;fifth ablaut class, las jldrum, gas jgdrum, ginas jgindrum, was Iwdruni.

These variations, going back to Prim. Gc., show an establisht feeling for consonantal variation as a normal means of distinguishing sing, and plur. pret. stems.

10. nbsp;nbsp;nbsp;To sum up, both present and preterit syncopations occurrednbsp;in a considerable number of common verbs and were well in-trencht in OH G speech feeling. Present syncopations occurred in thenbsp;gdn and stdn groups, and preterit syncopations occurred alone innbsp;the -w- verbs of the second weak conjugation and in houwanjhio.nbsp;Both present and preterit syncopations occurred together in allnbsp;the other groups. In the weak verbs (5/a�-group, weak -w- verbs,nbsp;hah�njhdn) syncopation made no difference betw'een singularnbsp;and plural preterit forms; in the strong verbs [spian-group. blou-group. Mo) it brought a new means of distinguishing pret. sing.

and pret. plur. stems by consonant variation, which was groupt in speech feeling with consonantal variation by grammaticalnbsp;change.

Certain other components of the later reduplicating dipt present and preterit group were preexistent, but were still isolated in thenbsp;main OHG period. Especially to be noted are (i) the isolation ofnbsp;houivanjhio from the gaw-group; (2) the isolation of Idzan fromnbsp;all syncopating groups, and especially from the ga�-group; (3) thenbsp;isolation of the gaw-group from all preterit syncopations; (4) thenbsp;phonetic isolation between the gaw-presents and preterits; (5) thenbsp;isolation of the ga�-preterits from all other preterit groups;nbsp;(6) the resultant inhibition of a comprehensive dipt present andnbsp;pi'eterit group.

1. nbsp;nbsp;nbsp;Houwanjhio was for several reasons not closely associatednbsp;with the gdnifdhanjhdhan group. They had (apparent) presentnbsp;stem syncopation or weakening, Jiouwan had only the full stem;nbsp;they had only full preterit sing, forms, houwan had only the diptnbsp;form; the present stems were not of the same phonetic pattern;nbsp;their vowel gradations differed thruout (a jia jia ja and ott jio jio jou)nbsp;in the main OHG period. So that, tho the type of a reduplicatingnbsp;dipt preterit was preexistent in hotmanjhio, it did not yet touchnbsp;the verbs gdn, fdhan and hdhan.

2. nbsp;nbsp;nbsp;The verb Idzan did not in the full OHG period belong to anynbsp;of these syncopating groups, and had no special associations withnbsp;the verbs gdn, fdhan, hdhan or houwan. It had, to be sure, severalnbsp;general points of contact with these verbs in that it was (i) anbsp;reduplicating verb (2) of the same vowel gradation (a jia jia ja)nbsp;as gdn, fdhan and hdhan, and (3) in its main meaning of �los-lassen, freilassen� the exact antonym of fdhan �fassen, fangen�,nbsp;and therefore already semantically associated with this word andnbsp;with the phonetically resembling fazzon quot;ergreifen�. But it lacktnbsp;the prime essential of membership in the gdn-fdhan group innbsp;that it had no dipt present forms and it equally lackt the diptnbsp;preterit forms which might have associated it with hiojhiowunnbsp;and the other groups of preexistent dipt preterits noted above.

with the gaw-group, none exerted any analogical pressure on the gdn preterit formations. Stan and slahan did not form diptnbsp;preterits; the hldn-grou^ and hdn formed the weak preteritsnbsp;hldta, hdte, etc., whose phonetic pattern was both in stem andnbsp;inflection exactly the reverse of a strong preterit, and hencenbsp;furnisht no model for dipt strong formations. But, if a specificnbsp;pattern and actual startingpoint should be found elsewhere, thesenbsp;forms would exert a strong general pressure toward its acceptancenbsp;and spread.

4. The phonetic discrepancy between gdn, fdhan, hdhan and their preterits giang, fiang, hiang prevented their forming anbsp;proportional group. Fdhan and hdhan had no present stems innbsp;-ng to group with the preterits fiang and hiang. Gdn was sonbsp;thoroly dominant over gangan in ordinary use that the normalnbsp;functional association was certainly gdnjgiang and not gangan jnbsp;giang, just as English is gojwent and not the historically correctnbsp;wend (went (cf. send fsent). Had the apparently syncopated presentsnbsp;really arisen out of S3mcopation of stems ending in -ng- theynbsp;might have carried the preterit stems with them. As it was withnbsp;the present and preterit stems phonetically isolated from eachnbsp;other, there was no urge in any form to connect up the twonbsp;systems. The situation was exactly parallel to that in NHGnbsp;gehenfging, stehenj stand or English gojwent, stand j stood, Er. jenbsp;vaisjj�allai, Gr. eQxo/.mj'^Xamp;ov, and many others. In these verynbsp;common words, each heterothematic form is strong enough tonbsp;maintain its separate and self-sufficient existence. No Englishnbsp;speaker, for instance, feels any urge to say I *goed, nor Germannbsp;to say ich *gehte nor French speaker fe *vayai.

This isolation was strengthened by their completely riming stems and inflections. Their phonetic pattern in -iang was sharednbsp;by no other OHG words. As a result their association with each

This explains why thru hundreds of years gdn refrained from forming the preterit gie except temporarily under the very special circumstances we arenbsp;considering. It is so obvious a priori, however, that gdn must have formed itsnbsp;natural preterit gie as the starting point of the whole development! Thus, one of thenbsp;best and keenest of American philologists wrote me: � quot;The outward mechanismnbsp;of the genesis of those forms is simple enough: starting with the forms gdn-gie,nbsp;and the analogical infinitives fan, Idn � the shorter forms jie, lie were addednbsp;as parallels to gdn-gie-gienc. Gie, of course, is originally the preterit of gdn, whilenbsp;gienc is the preterit of gangan.quot; � At the time I fully shared this seeminglynbsp;self-evident view. Actual investigation shows it to be so wholly erroneous thatnbsp;I now refrain from giving the name of the honored friend who wrote it.

other was very strong, which drew them still more completely away from phonetic influence of their present forms. Furthermorenbsp;these words were all of such common everyday use that each ofnbsp;them (especially giang) would have been able to stand alone asnbsp;an irregular form if necessary. As a result these three verbs formednbsp;a compact, self-contained and inter-reinforcing group, which stillnbsp;holds a thousand years later in NHG gingjfingjhing.

5. nbsp;nbsp;nbsp;For these same reasons OHG giangjfiangjhiang stood aloofnbsp;from all other strong preterit groups. They offered no hold fornbsp;specific analogical pressure from the strong dipt preterits spe,nbsp;blou, etc., such a proportion (?) as gdn: giang:: spian: spe beingnbsp;absurd. � Their similar immunity against specific pressure fromnbsp;the weak dipt preterits hldta, hdte, etc. was noted under � 3 above.

6. nbsp;nbsp;nbsp;In spite therefore of the preexistence of these types of diptnbsp;presents and preterits and of several forms of connection betweennbsp;them, these isolations had prevented the development of anbsp;consistent group of syncopating verbs as such. A comprehensivenbsp;group leveling all syncopating verbs was manifestly impossiblenbsp;because of the heterogeneity of the conjugations involved. Asnbsp;to the preterits of the ga?j-group, their above-noted phoneticnbsp;segregation both from their own present systems and from allnbsp;syncopating preterits, apparently precluded the possibility ofnbsp;leveling them to the dipt type or of creating variant dipt formsnbsp;beside them.

Obviously these gaps between the gdn presents and preterits, and between the preterits and those of other verb groups, couldnbsp;be bridged, if at all, only from within the group by a verb withnbsp;a phonetically regular present and preterit stem.

1. The bridge from the present forms ganganjgdn, fdhanjfdn, hdhanjhdn to their preterit forms giang, fiang, hiang, and alsonbsp;the bridge connecting these clipt preterit forms with those ofnbsp;other syncopating groups, were furnisht by the accidental entrancenbsp;of Idzen into the group in the late OHG period (about the turn

As this movement overlaps the transition to the MHG period, it will be better at this point to change to MHG spellings of cited forms.

of the tenth and eleventh centuries) Collitz shows conclusively that the dipt impv. 2 sg. Id originated independently of thenbsp;influence of gdn and stdn, at least as far as appearance in writingnbsp;is concerned. That this holds also for spoken usage would seemnbsp;to be shown by the original disconnection between lazen and gdnnbsp;noted above (IV 2) Both these conclusions are further bornenbsp;out by facts found by Starck and Sehrt in the index which theynbsp;are making of Notker�s works. Starck notes (in a personal letter)nbsp;that in his part of the work (all except Boethius and the last fiftynbsp;Psalms) 21 instances of Id occur, but no other contract forms ofnbsp;Idzen, and none at all of gdn and fdhen. Sehrt informs me that henbsp;finds in the Boethius 7 instances of Id, none of Idz and the singlenbsp;instance of lie (126, 2) referred to by Collitz; in his share of thenbsp;Psalms he finds 5 instances of Id, 6 of Idz and none of lie.

It was noted above that Id first appeared as the unaccented grade of the 2 sg. impv. Idz. It is a significant coincidence thatnbsp;the dipt forms habenjhdn, whose increast use (noted in III 8nbsp;above) came in this same century, generally occurred under nonstress conditions in the auxiliary use of the verb 22. Compare alsonbsp;such syncopations as the dropping of -z- of unstrest daz in daznbsp;ist gt; deist, etc.

2. nbsp;nbsp;nbsp;As soon as the present tense forms were establisht in speechnbsp;feeling, the verb Idzen jldn thereby became an equal membernbsp;of the reduplicating dipt present group, with its formationsnbsp;thruout exactly parallel to those of gdn, less perfectly so to thosenbsp;of fdhen and hdhen. The proportions may be shown as (gangennbsp;gdn:gang:gd:gieng::ldzen:ldn:ldz:ld:liez::hdhen:hdn:hdh:hd:hieng, etc.

3. nbsp;nbsp;nbsp;The connection which Idzen formed between the present andnbsp;preterit tenses rested on the parallelism in the phonetic patternsnbsp;of Idz and liez, a parallelism wholly lacking mfd(h) jfieng and hd(h) jnbsp;hieng and seriously disturbed in gd(ng) jgieng by the larger usenbsp;of the short form. The analogical proportion of Idzen was ldz:ld::

We must assume that in all cases spoken forms first developt and formed the basis of their literary appearance. Also we should expect OHG literary usagenbsp;to agree generally with spoken usage, because it was based directly on the folknbsp;dialects and not on any accumulated or artificial literary tradition.

This form was dying out, but the stem continued in undiminisht use in the past ptc., and in the impv. gang, derivative noun ganc, etc.; so that the proportionnbsp;was not destroyed, altho weakened.

liez:{x), in which the speech feeling value of x was necessarily lie. It is not strange therefore that lie was the next form to emerge

4. nbsp;nbsp;nbsp;With the creation of the form lie beside Id, giving the proportion ldz:ld::liez:lie, connection was completed between thenbsp;present and preterit system of the gaw-group. Therewith the oldnbsp;reduplicating dipt present group became in nascent form thenbsp;new Reduplicating Clipt Present and Preterit Group. The preteritnbsp;lie at once establisht connection with the old clipt pret. Mo, nownbsp;shifted to Me, from the verb houwen, which thereby became anbsp;member of the group, tho lacking present syncopation Thenbsp;group thus came to contain the verbs gdn, fdhen, hdhen, Idzennbsp;and houwen.

5. nbsp;nbsp;nbsp;With the linking together of present and preterit syncopationsnbsp;of this group, each of these types of syncopation was greatlynbsp;strengthened by the other, since they now fell for the first timenbsp;into proportional groups of present and preterit forms. And fromnbsp;this came the impulse to the analogical creation of new forms,nbsp;since most of the new proportions were defective in either theirnbsp;present or preterit formations as judged by the new patterns,nbsp;and stem analogies also exerted pressure for the extension of thenbsp;new system of syncopations to the perfect participle. For illustrations, see the schema of the reduplicating clipt present andnbsp;preterit group given below.

6. nbsp;nbsp;nbsp;The changing of the gaw-group to the clipt present andnbsp;preterit status brought it into line with the general plan of thenbsp;preexistent syncopating verbs, most of which had both presentnbsp;and preterit syncopations coordinated with each other. Thisnbsp;again workt to the mutual strengthening of both the old and newnbsp;groups. This general connection was made concrete and precisenbsp;by the pret. MejMewen which was already associated in formationnbsp;with the other -w- syncopating preterits, as blou jbluwen, spejs-pi-wen, etc. This also resulted in a tendency to analogical extensionsnbsp;and the creation of new forms which will be noted below.

7. nbsp;nbsp;nbsp;The individual verbs of the reduplicating clipt group werenbsp;further reinforced by other words of similar phonetic form and

That it was actually felt as in this group is shown by the later leveling of houwen to hawen on the pattern of lazen, etc. � The present syncopation ofnbsp;houwen was resisted by the diphthong -on-, as explained for the verbs of thenbsp;6/oM-group in III 6.

meaning which showed the same or similar syncopations and gradations. These were partly true cognates or derivatives, asnbsp;ganc (gen. ganges), fane, hanc, showing the missing full presentnbsp;stems of these words; houwe (houwe)jhou (h�u), showing thenbsp;missing syncopation of the present system of the verb. Note alsonbsp;lazjldzjlazzen beside lazen, hdhel (hdl) �kesselhaken� besidenbsp;hdhenfhdn, etc. Others were contaminative or congeneric associates showing accidental resemblances, as gdhefgd �eile� (schnellesnbsp;Gehen), etc. In some cases such cognate and contaminativenbsp;associates offered a new bond of resemblance between differentnbsp;verbs, as gazze (cognate of gdn) and faz quot;laQ�jfazzen quot;fangen�nbsp;(contaminates of fdhen) beside lazjlazzejlazzen (cognates ofnbsp;Idzen)

8. The interaction of all these factors will be seen more clearly and quickly in the following.

Preexistent forms (that is forms which had originated independently in any way up to the nascent formation of the new group) are printed in capitals. The missing forms whose newnbsp;creation was demanded and subsequently effected by the analogi-

NHG lassen has inherited the phonetic form of lazzen, the inflection of lazen, and combines in its derivatives the meanings of both �- cf. lassen, ahlassen,nbsp;Idssig, unabldssig, etc.

230 cal proportions of the new group are put in pointed brackets. Cognates of gdn, fdhen, hdhen, lazen and houwen are put in parentheses.nbsp;Contaminatively associated words are put in square brackets.nbsp;Spaces in parentheses indicate that the verb was phonetically unfitted to create the corresponding form. A dotted line denotes annbsp;actual gap in the series of formations. A solid line indicates thatnbsp;there is no real gap and that the space is to be ignored.

9. nbsp;nbsp;nbsp;The schema is in general self-explanatory, but certain observations may help:

(a) nbsp;nbsp;nbsp;The forms gdhenjgdhejgd, gazze/fazzenjlazzen, etc., had ofnbsp;course nothing to do with developing the dipt forms of gdn,nbsp;fdhen and lazen, but after the dipt forms were developt,nbsp;such side-associations helpt to bind these words into still firmernbsp;association.

(b) nbsp;nbsp;nbsp;The forms fangen and hangen were in no sense preexistent,nbsp;but were later analogical recreations, since the present stems hadnbsp;from Prim. Gc. on terminated in h. The dipt inf. hdn �hangen�nbsp;was largely inhibited by the preexistent and far commoner hdnnbsp;�haben�. Also hie �hing� and hie �hieb� must have hamperednbsp;each other, tho both were used.

(c) nbsp;nbsp;nbsp;The new formations hawenfgehawen for the original formsnbsp;houwen I gehouwen were analogical extensions on the pattern ofnbsp;the other four verbs of the group, and not a premature NHGnbsp;shifting of -ouw- to -au-. The leveling proportions which producednbsp;the new forms were gangen: gie: giengen: gegangen::fdhen: fie: fien-gen: gefangen: :ldzen: lie: liezen: geldzen, etc., which analogicallynbsp;demanded hawen: hie: hiewen: gehawen. Similarly, the full sing,nbsp;pret. form hi(e)w is a new formation on the pattern of lie: liez.nbsp;Both hawen and houwen yielded NHG hauen by regular shifting. ��nbsp;The -h- grade of the present stem was impossible for hawen,nbsp;because the -w- was original and not for a velar. � The diptnbsp;inf. *hdn for hawen was inhibited by hdn �haben� and hdnnbsp;�hangen�. Both *gehdn �gehangen� and ^gehdn �gehauen� werenbsp;completely inhibited by gehdn �gehabt�.

(d) nbsp;nbsp;nbsp;The phonetic analogical proportions of the other newnbsp;formations gie, fie, hie �hing� and (later) gegdn, gefdn and gehdnnbsp;are self-evident.

10. nbsp;nbsp;nbsp;The distinctive feature of the new group was its developm.entnbsp;of this special variation in the preterit singular with dipt andnbsp;full forms side by side, as against the full form only in the pret.

plural. Variant syncopation in the present was shared with the stdn, sldn, hlcen (earlier hldn), Men and spien groups and the verbnbsp;hahenjhdn.

1. nbsp;nbsp;nbsp;The fact that the other classes of preexistent syncopatingnbsp;verbs continued and strengthened their association in speechnbsp;consciousness with the verbs of the reduplicating clipt groupnbsp;is shown by their tendency to retain or create forms which agreednbsp;with those of the new group. Old forms were preserved whichnbsp;agreed with the pattern of the gdnjldn group, tho no longernbsp;regular in phonology from their own standpoint, as drdtejgedrdtnbsp;(gedrdn) beside the newer regular drcBtejgedrcet. New forms werenbsp;created on the particular or general pattern of the ga�-group, asnbsp;gesldn beside geslagen, U beside Uch from Uhen. Especially to benbsp;noted is the retention or creation of strong past participles of thenbsp;gaw-type for weak verbs, as gehldn beside gebleet, gehdn besidenbsp;gehdt, etc., thus showing the continuance of the OHG associationnbsp;of the gdn, stdn, sldn, bldn and hdn (hab�n) groups.

2. nbsp;nbsp;nbsp;The aggressive vigor of the now reinforced syncopating typenbsp;is further shown by the tendency to level other verbs to itsnbsp;pattern of total conjugation. Thus liken (lien) jU (Uch) jlihenjnbsp;gelihen (geligenigelien) was completely leveled to the new type,nbsp;which is the more significant in view of its belonging to anothernbsp;ablaut series. Partial levelings also occurred in which consonantnbsp;variation replaced syncopation (cf. Ill 9 above), as ligen (liken,nbsp;lian, lin) jlaclldgen I gelegen {-c due to devoicing of final), sehennbsp;(sen) jsachjsdgenlgesehen (ges�n)^^, geschehen (geschen) jgeschachjnbsp;geschdgen jgeschehen (geschen), slahen (sldn) jsluoc jsluogen jgeslagennbsp;(gesldn), twahen (twdn)jtwuocjtwuogenlgetwagen (getwdn), tragennbsp;(Iran) jtruoc (tmogen jgetragen (getrdn) (-g- due to grammatical

By Verner�s Law the pret. plur. of se/ien should contain a voiced velar spirant as stemfinal (OHG sajuum, MHG sajuen). This sound would be essentiallynbsp;a velarized or quot;throaty� -w-, which (lacking exact representation in the Latinnbsp;alphabet) would most naturally be spelled in the various Germanic languagesnbsp;with the closely resembling -w-. (Incidentally this suggests a possible explanationnbsp;of Sievers� Law; see PBB 5, 149 and Kluge, Urgerm. � 47.) But a throaty -w-is intermediate in sound between -w- and spirant -g-. � Could the variant spellings OHG sdwun, MHG sagen be therefore simply approximations of the samenbsp;sound ? � Compare Gc. initial w- from IE initial velar voiced aspirate, and thenbsp;Romance borrowing of this w- as gu- in Fr. guerre, guichet, etc.

change). � Part of these levelings were due to ordinary phonetic changes, so that speech feeling had merely to accept accomplishtnbsp;results. �� It is to be noted that in most of these verbs the pret.nbsp;sing, showed only the weaker stem, the plural the full form.nbsp;This did not however separate them from the clipt preterits innbsp;speech feeling, because, as Collitz notes, already in the transitionnbsp;period, the clipt singular stem had become the regular popular form.

1. nbsp;nbsp;nbsp;The larger phonetic background both of the reduplicatingnbsp;clipt present and preterit verbs and of the other syncopating verbnbsp;groups was formed by the general system of phonetic variationnbsp;which characterized OHG and MHG phonology. The vowelnbsp;and consonant syncopations and variations in these variousnbsp;groups were themselves an essential part of the larger scheme,nbsp;and were dependent on and sustained by it. The system wentnbsp;back in its beginnings to IE, and received successive accretionsnbsp;in both the Germanic and OHG periods. Its extreme developmentnbsp;fell in the later OHG and the MHG periods and formed thenbsp;necessary basis for the rise and spread of those groups.

2. nbsp;nbsp;nbsp;For the scientific and historical discrimination of these variantnbsp;formations see Braune, Ahd. Gramm., division on �Lautlehre�nbsp;and Paul, Mhd. Gramm., chapter on �Lautwechsel�. Our presentnbsp;purpose is rather to consider how these variations presentednbsp;themselves to speech feeling, which is neither scientific nornbsp;historical. For naive speech feeling all these variations stoodnbsp;on a par with each other and were not discriminated accordingnbsp;to historic origins.

3. nbsp;nbsp;nbsp;These variations were due to diverse causes, partly corresponding to the different periods of their origins. Among thenbsp;causes of consonant variation we may note: (i) Phonetic shiftingnbsp;(all periods), as deckenjdahte, hiejhiewen, etc.; (2) grammaticalnbsp;change or Verner�s Law, fdhenjgefangen, z�chjzugen etc.; (3) de-voicing of finals, as tacjtage, lampjlemher, etc.; (4) syncopationnbsp;(all periods), as fdhenjfdn, gibistjgist, etc. �� By the action andnbsp;interaction of these factors many words had come to shownbsp;phonetic variation either in the same inflected form (as fdhen jfdn)nbsp;or between different forms of the same word (as z�chjzugen).

4- Among consonantal equations or interchanges which had thus developt we may note: (i) the optional or necessary omissionnbsp;of various intervocalic and final consonants, such as h, g, j, w;nbsp;(2) inflectional, derivational and sandhi interchange of consonants,nbsp;especially of h, ch, k, g, nk, and ng; (3) combinations of these intonbsp;more extended series. It will be well to quote sufficient illustrationsnbsp;to show how widely and firmly these variation series were in-trencht in MHG phonetic feeling.

5. nbsp;nbsp;nbsp;Omission of intervocalic or final -h- appears, for instance,nbsp;in such verbs as sehen/sen, geschehenigesch�n, hdhenjhdn, fdhenjfdn,nbsp;slahenjsldn, twahenjtwdn, ziehenjzien, lihenjltn, jehenjj�n, etc.; innbsp;such nouns as maheljmdl, UhenjUn, blahelbld, staheljstdl, etc.; innbsp;such adjectives and adverbs as h�chjhohejh�, ndchindheind, etc.

For omission of -g- compare magetlmeit, jagetjjeit, slagejsld, gegenfgein, ligenjlien, etc.

For -w- compare schatewejschate, /�/gen. Uwes, krdwelkrd(e), farwejidr, Icewejldwlld quot;lau�, hiwejhie- �gatte�, S'pmmjspun,nbsp;iwerjiur, etc.

Sporadically occurred other real or apparent syncopations and clippings, such as -t- in sitzenjsaz, netzenjnaz, etc., -z- in fazzenjfdnnbsp;(viewed as folk-cognates), gazze (deriv. o{)lgdn; -nd- in standenjstdnnbsp;-ng- in gegangenjgdn: -r- in ddjddr, wdjwdr, m�rjm�, zer-jze-; -d-in geschadetjgaschdt; -b- in habenjhdn, gibitjgU; -t- in leitenf'piet.nbsp;lei-te; -p- in scharfjscherpfe, sleifejsleipje, etc.

6. nbsp;nbsp;nbsp;It is clear that the average MHG speech consciousness mustnbsp;have had a deep-rooted feeling that all non-initial consonants,nbsp;especially between vowels and in final consonants matching thesenbsp;in inflection or derivation, were regularly or sporadically subjectnbsp;to syncopation or elision. Hence MHG phonetic feeling wasnbsp;predisposed on the basis of a host of apparently slurred or diptnbsp;forms to accept and preserve new formations of the same kindnbsp;wherever other causes tended to their formation, or to regroupnbsp;originally independent but resembling forms (such as standen jstdn,nbsp;gangenjgdn) as sjmcopation variants of the �same� word.

7. nbsp;nbsp;nbsp;Completely interwoven with the phenomenon of consonantalnbsp;omission was that of consonantal variation and equivalence. Thenbsp;two occur conjointly in many v/ords, so that the syncopated form

MHG in this developmental sense includes the same tendencies in OHG �-especially in later OHG � leading into MHG.

comes to count as the zero grade in the consonantal variation series, and consonantal weakening (such as -ng-j-hj etc.) is feltnbsp;as a partial or incipient syncopation. Such consonantal variationnbsp;series thus combined in varying proportions the results of syncopation, grammatical change, devoicing of finals, normal soundnbsp;shifting, and contaminative blending of originally separate words.nbsp;The distinction between these various causes was, as alreadynbsp;noted, largely blurred or obliterated for speech feeling, and theirnbsp;results were felt as belonging together, the more so as the phoneticnbsp;equations resulting from them frequently overlapt or duplicatednbsp;each other, and hence necessarily blended into one series whennbsp;they occurred in different forms of the same word.

The simple phonetic equations due to devoicing of finals {bjp, djt, gjc, as in wt'pjwihes, fiencjfiengen, etc.) are too common tonbsp;need extended illustration. They tended automatically to blendnbsp;with similar phonetic equations resulting from grammatical changenbsp;and in many cases with those due to phonetic shifting. Thus,nbsp;devoicing blends with grammatical change to produce such equations as djt in lidenjleitjliten, siedenjsotlsuten, etc. It blends withnbsp;phonetic shifting to produce such series as kjgjh in mac jmaht jmagennbsp;and toucjtugenjtohte. It blends with phonetic shifting and grammatical change in the forms of hdhen, falien, denken, hringen, etc.,nbsp;to produce the more extended series Ojhlnkjng. Simple shiftingnbsp;produces h/chjk in such groups as dachjdeckenldahte (dacte).nbsp;Compare such further equations (due to simple or mixt causes)nbsp;as h/k in wurken jworhte; g/hjOjk in ligen j liken /lien (lin) /lac; hlOjckjnbsp;gj in the forms of sehen, gesche(c)hen, dihen, zihen, etc. It is clearnbsp;that these and very many other similar equations must havenbsp;developt the fixt feeling that 0Ihjchlklglnkjng were freely equa-table and interchangeable in any order, selection or combination.nbsp;Other equations due to mixt causes were s/r in forms of wesen,nbsp;risen, etc.; zjO in such contractions as daz istjdeist, ez istjeist;nbsp;bjpjO in habenjhdnlgehabetjgehapt j gekat, gibejgUlgip, etc.; njO innbsp;einjeime, dinjdime, etc. djtjO in forms of queden, reden, baden,nbsp;schaden, etc. and their derivatives; and many others.

8. Because of these mainfold blendings and overlappings, all of these series of consonantal variations cohered in a general systemnbsp;of consonantal variation in MHG speech feeling.

g. Like consonantal variation, MHG vowel variation went back to several different origins. The following historical forms of

vowel variation may be noted; (a) ablaut, going back to primitive Indo-European, a.sbUenlbeitlbitenlgebiten: Go. beidanlbaipjbidumjnbsp;bidans: Gr. nei�o}, nenoid'a, �'m�ov, Ttfamp;av�c', (b) interchange of ejinbsp;and of oju before various vowels in the following syllable, withnbsp;or without an intervening nasal consonant, as helfenjhilfe,nbsp;herdejhirte, zugen I gezogen, hof jhiibesch, liehtjliuhten (OHG liohtjnbsp;liuhtan), etc.; (c) umlaut, beginning in late OHG and developingnbsp;further in the MHG period, as gastjgeste, h�rtefhoeren, h�sjhiuser,nbsp;houwejh�uwe, etc.; (d) contraction (all periods, often fused withnbsp;and practically a part of syncopation), as lienjlin, krdejkrd,nbsp;fdhenjfdn, gibistjgist, etc.; (e) results of leveling out primitivenbsp;reduplication, with the following contraction and phoneticnbsp;shifting, as halten j hielt, ruofenjrief, etc.; (f) phonetic shiftingnbsp;leading to vowel variation, as gefangenjfdhen, m�rejmeist, etc.

10. These several types of vowel variation blended together in varying combinations. Thus, original ablaut and the primitivenbsp;Germanic variation eji, oju, iejiu blend into larger series ofnbsp;vowel gradation in the ablaut verbs generally, as nemenjnimenbsp;(Germanic Ysoriation) jnamj namen j genomen (ablaut); ziehenjziuhenbsp;(Gc. var.) / zoch (abl.) / zugen j gezogen (abl. -f Gc. var.), and manynbsp;others. In other words, Germanic vowel variation was normallynbsp;superposed on preexistent ablaut, and subsisted only as annbsp;inseparable aspect of the same, the two together forming fornbsp;speech feeling an enlarged ablaut series.

Similarly, umlaut is inseparably superposed on both ablaut and Germanic vowel variation, blending with them into still larger composite series. Thus it combines with ablaut in grabenjgrebetjgruopjnbsp;griiebe, in tuon jtete jtdten jtcete jgetdn, and in many o thers. It combinesnbsp;with Germanic vowel variation and ablaut in soljs�lnjsoltejs�lte,nbsp;wellen jwil jwolte jw�lte, mac jmegen jmugen jmahte jmohtejmahtejm�hte,nbsp;nam jnceme jndmen j genomen, ziehen jziuhe jzoch j zugen jzuge j gezogen,nbsp;etc. It occurs alone simulating ablaut in setzenjsatzte, which wasnbsp;surely associated also with sitzenjsazjscezejgesezzen.

The results of phonetic shifting combine with ablaut in numberless instances, as in the preterits of the reduplicating verbs, such as haltenjhielt {-ie- lt; OHG -ia- lt; earlier -ea-, etc. ^�) besidenbsp;ruofenjrief {-ie- lt; OHG -io- lt; earlier -eo-).

I venture to base this name on Braune�s �urgermanische wandlungen�� Ahd, Gramm. � 30 c.

11. nbsp;nbsp;nbsp;Both consonant and vowel variation frequently alternatenbsp;or concur in the same word or in the same functional form andnbsp;systems.

Thus in nouns the stem-distinction between nom.-acc. singular and plural forms may, in addition to variation of inflectionalnbsp;endings, show; (i) consonantal variation only, as in tacjtage, etc.;nbsp;(2) vowel variation only, as bachjbeche; (3) both consonant andnbsp;vowel variation, as lampjlember, burcjbilrge, etc. Similar illustrations could be given for adjective stems, as blintjblinde, volljv�lliu,nbsp;etc.

In the comparison of adjectives arise (i) such consonantal equations as krefticjkreftiger, etc.; (2) such vowel equations asnbsp;altjelter, etc.; (3) such combined vowel and consonantal equationsnbsp;as juncjj�nger, lane jlenger, etc. In the formation of adverbs fromnbsp;adjectives, we may note (i) lancjlange, and (2) schoenejschone,nbsp;vestejvaste, etc. In verbs occur (i) such consonantal equationsnbsp;as gruopjgruoben, truoc/truogen, slakenjgeslagen, etc.; (2) suchnbsp;vowel equations as haltenjhielt, giengenjgegangen, etc.; (3) suchnbsp;combined vowel and consonant equations as siudejgesoten, z�chjnbsp;zugen, fdhenjgefangen, etc.

12. nbsp;nbsp;nbsp;In the case of noun and adjective and adverb forms, vowelnbsp;equations (chiefly umlaut) constituted the dominant form ofnbsp;phonetic variation, with consonant variation loosely associatednbsp;with it as a possible minor variant. Essentially the same situationnbsp;holds for the present indicative and imperative of strong ornbsp;ablaut verbs, as wirdejwirtjwerden, gibe jgipjgebet, grabejgrebetjnbsp;grapjgrahet, etc. (The form of vowel variation here involved is ofnbsp;course only partly umlaut, including also the primitive Gc. vowelnbsp;variation e: i, 0: u.)

13. nbsp;nbsp;nbsp;But in the preterit system of the ablaut verbs, vowel andnbsp;consonant variation were more strongly intrencht and morenbsp;closely associated together. Vowel variation in the strong preteritnbsp;included, not only umlaut as in noun, adjective and adverbnbsp;formations, but original ablaut, original reduplication, Germanicnbsp;vowel variation and umlaut. Consonant variation included notnbsp;only devoicing, but also the historically older and functionallynbsp;more vital phenomenon of grammatical change (Verner�s Law) andnbsp;sometimes the results of phonetic shifting. Phonetic variation ofnbsp;either or both kinds was felt as fundamental as a means of distinguishing singular and plural stems. Vowel variation alone

distinguished the two stems in such verbs as beitjbiten, halfjhulfen, namjndmen; consonant variation alone in truocjtniogen, stuont,nbsp;sUionden, gruo-p Igruoben, fiencjfiengen; neither variation in vuorjnbsp;vuoren, hieltjhieUen, liefjliefen, etc.; combined vowel and consonantnbsp;variation in d�chjdigen, z�chjzugen, wartjwurden, etc.

14. To sum up, extreme phonetic variation, partly functionless, partly serving to express functional distinctions, was the mostnbsp;characteristic feature of MHG phonology. It found its fullestnbsp;manifestation in the preterit of strong verbs, whose phoneticnbsp;equations may however be abundantly duplicated from othernbsp;parts of speech. � This sweeping system of phonetic variationnbsp;was the necessary background and milieu of the developmentnbsp;and continuance of the various dipt present and preterit systems.

1. nbsp;nbsp;nbsp;All of these phonetic and functional variations, includingnbsp;those of the dipt preterits, subsisted in a still larger milieu ofnbsp;�regular� forms which showed no variation of stem, but usednbsp;suffixes to indicate grammatical distinctions, as nageljnagele, zaljnbsp;zaln, botejboten, machenjmachte, kerenjkerte, wonenIwonete. As anbsp;result the dipt preterit verbs were exposed to two opposingnbsp;tendencies, a narrower and at first intenser tendency to preservenbsp;the phonetic variation of singular and plural stems and to createnbsp;new analogical forms on the same or resembling patterns, and anbsp;broader tendency to level out all useless stem variations. Thenbsp;basic strength of this levelling tendency obviously lay in itsnbsp;conformity to the fundamental analogical principles of language,nbsp;and in its conducing to clearness and mental economy in individualnbsp;instances. Its initial weakness lay in the fact that in all thenbsp;declensions and conjugations in which the �regular� forms occurred, other forms with manifold phonetic variations occurrednbsp;beside them, as gastjgeste, l�jl�wes, geschadetjgeschat, etc., etc.

2. nbsp;nbsp;nbsp;The strength of this narrower tendency as applied to thenbsp;reduplicating dipt present and preterit group rested on bothnbsp;internal and external factors: (i) Its forms were of frequent andnbsp;familiar occurrence and able to hold their own individually;nbsp;(2) they formed a compact group strengthened by (3) associationnbsp;with other syncopating present and preterit verbs and (4) association with verbs making phonetic distinction between the

preterit singluar and plural stems, and (5) fortified by the still larger background of phonetic variation noted above. As longnbsp;as these interlinking factors continued unchanged, they maintained the dipt verbs as an impregnable group.

3. nbsp;nbsp;nbsp;The intrinsic weakness of the group lay in (i) the impossibility of conforming some stems (as those ending in a vowel) tonbsp;the formation, (2) the limited numbers of distinct dipt formsnbsp;(thus hdn must theoretically do duty for haben, hawen (houwen),nbsp;hdhen), and (3) the consequent frequent ambiguity in dipt formsnbsp;leading to the partial or complete inhibition of some forms,nbsp;(4) the impossibility of securing phonetic consistency and homogeneity in a group so constructed. By preexistent accident gdn,nbsp;fdhen and hdhen formed the compact group above noted. In spitenbsp;of their strengthening effect in other ways, the addition of lazennbsp;and houwen to the group (not to mention such later additions asnbsp;l�jl�ch) greatly decreast this compact homogeneity. � Thenbsp;possible grovdh of the group was therefore very limited, and thenbsp;above noted tendency to exceed these limits, tended also tonbsp;break down the group.

4. nbsp;nbsp;nbsp;A more serious blow to the group was the general tendencynbsp;to wipe out the distinction between preterit singular and pluralnbsp;stems which occurred in later MHG in the ordinary course ofnbsp;phonetic development and as a result of the leveling of vowelnbsp;quantity according to position��.

As a preexistent basis for the levelling of preterit singular and plural stems, these already had the same stem in the sixth andnbsp;seventh class strong verbs (as fuorjfuoren, Meltjhielten) and innbsp;weak verbs generally (as machte/machten, satztejsatzten, wolte-n,)nbsp;and in all preterit subjunctives (as nceme/ncemen, bilgejbilgen,nbsp;Ueze/liezen, etc.). Likewise the 2 sing pret. ind. in all classesnbsp;had the stem of the plural indicative.

5. nbsp;nbsp;nbsp;In the first class, as a result of the phonetic coalescence of i,nbsp;and ei such verbs as sttgen: steig, etc. produced the confusingnbsp;equation steigen/steig/stiegen/gestiegen. The necessity of beingnbsp;understood in speaking forced the elimination of the preteritnbsp;singular stem steig and the consequent levelling of the preteritnbsp;singular and plural to stieg: stiegen.

See Behaghel, Geschichte d. deutschen Sp.^, pp. 427�438; Cvirme, Gram of the Ger. Lang. 197�204 and p. 14.

In the 4th and 5th classes, as a result of the positional levelling of vowel quantity the feeling that there was any distinctivenbsp;difference between singular and plural stems was lost � thenbsp;variation in namindmen, gabjgdben became the necessary variationnbsp;of the same stem in different syllabic positions. The later analogicalnbsp;leveling of vowel quantity in all forms of the same inflected stemnbsp;completed the levelling to NHG ndhm: ndhmen, gab: gaben, etc.

Thus five out of the 7 classes of strong verbs came independently of each other to have identical singular and plural preterit stems. These five classes (I, IV, V, VI, VII) contained about 60%nbsp;of all strong verbs.

6. nbsp;nbsp;nbsp;The 3d, 4th and 5th class preterits were strongly associatednbsp;in speech feeling by the vowel of the present singular in gradationnbsp;with that of the preterit singular in all three classes and also thenbsp;perfect participle in the 3d and 4th classes. Compare geiten: gilte:nbsp;gait: gegolten, nemen: nime: nam: genomen, scaAgeben: gibe: gap: (ge-geben), etc. As soon as the singular and plural of the 4th and 5thnbsp;classes were levelled, such 3d class plurals as guiten, sungen, etc.,nbsp;necessarily began to feel incorrect and were corrected (probablynbsp;both unconsciously and consciously) to galten, sangen, etc., thusnbsp;leveling the 3d class preterits.

In the 2d class, b�t: buten: ppc. geboten became partly leveled when mere phonetic shifting gave b�t: buten: geb�ten, etc., whosenbsp;seemingly irregular plural yielded to the combined pressure ofnbsp;these forms and of all other classes and levelled to b�t: b�ten.

7. nbsp;nbsp;nbsp;Thus, owing to phonetic shifting and consequent analogicalnbsp;readjustment, the distinction of singular and plural stems disappeared from the strong preterits, where it had originally beennbsp;intrencht most strongly of all. With it most of the analogicalnbsp;support for the dipt preterits gie, fie, lie, hie �hing� and hienbsp;�hieb� had disappeared from speech feeling.

8. nbsp;nbsp;nbsp;These levelings of the singular and plural preterit stems werenbsp;an integral part of a general movement in later MHG whichnbsp;leveled out most of the syncopations, vowel and consonantnbsp;variations and phonetic doublet stems, and thus led to the NHGnbsp;period of stabilized sounds and leveled stems. The leveling ofnbsp;quantity according to position, the falling together of t and ei,

Save for such vanishing remnants as waydjwttrden, sangjsungen, and such pret. subjunctives as giilte, sHirbe, etc.

and the analogical levelings of variant preterit stems in the second and third ablaut classes were noted above. By the regularnbsp;shifting of both ou and -au-, the doublet pair houwenfhawennbsp;coalesced. By the normal mutescence of -h- between vowels, andnbsp;the general leveling out of final -chj-h-, such variations as sehenjnbsp;sachjsdhenjgesehen disappeared. By contaminative and analogicalnbsp;new formations, non-functional and antifunctional phoneticnbsp;variations in noun, adjective and verb stems generally werenbsp;leveled out, as kraftjkrefte, schatejschatewes, r�chjr�hes, ndchjndhes,nbsp;bldjhldwes, fUegenjfHuge etc. As a part of this last tendencynbsp;syncopated forms generally tended to yield to the correspondingnbsp;full forms, as gesldn to geschlagen, treit to traget, deist to daz ist, etc.

9. nbsp;nbsp;nbsp;By the successive accumulation of such individual changesnbsp;the phonetic character of High German speech became more fixtnbsp;and gradually ceast to be prominently characterized by excessivenbsp;phonetic variation. This change was partly precipitated, as notednbsp;above, by normal phonetic changes which either happened tonbsp;level variant forms or made them so dissimilar that subsequentnbsp;analogical leveling was necessary. The need for clearness alsonbsp;helpt in such ambiguous forms as hdnjhie, etc. Other factorsnbsp;calculated to help were the impulse to purism and formalismnbsp;connected with some forms of literary composition, and alsonbsp;with the development of an interdialectal and formal Kanzlei-sprache, and the tendency to precision and correctness whichnbsp;necessarily accompanied the more or less conscious and artificialnbsp;spread of the Schriftsprache.

10. nbsp;nbsp;nbsp;These cumulative changes constituted not only a normalizingnbsp;and stabilizing of particular words and forms, but also a generalnbsp;shifting of MHG phonetic speech feeling and of the phonetic basisnbsp;and background on which the syncopated verb formations hadnbsp;previously rested. The result was the collapse of syncopationnbsp;as a system of stem formation and conjugation. � But here anbsp;caution is necessary. J ust as the development of the reduplicatingnbsp;dipt preterits at the close of the OHG period involved regroupingsnbsp;but not the wholesale wiping out of older forms, so the downfallnbsp;of this system at the opening of the NHG period did not meannbsp;the sweeping destruction of all dipt forms. Very many of thesenbsp;still survive in present-day German, but always at the price of

But in separate stems the variation remained as Uehtjliuhtefi gt; NHG licht leuchien.

various regroupings. Furthermore it does hot mean that slurred or slovenly or abbreviated pronunciations of verbs and othernbsp;words cannot take place in German as they do in other languages,nbsp;but only that such slurring is no longer recognized as basic ornbsp;standard by speech consciousness.

11. nbsp;nbsp;nbsp;Among present-day survivals of dipt forms the followingnbsp;may be noted. The dipt presents gehn and stehn, with their fullnbsp;preterits ging and stand, have reverted to the OHG status, butnbsp;they are no longer felt as dipt stems but as full stems with diptnbsp;endings and irregular preterits. Schlan, emffahn and han �haben�nbsp;linger on as vanishing poetic echoes; hast, hat, hatte survive asnbsp;accepted irregularities, but they are short-voweled and no longernbsp;felt as syncopations. The dipt preterits spie and schrie survivenbsp;but they are no longer felt as dipt since all other forms of thesenbsp;verbs now have stems equally shortened. The same is true of thenbsp;weak dipt preterits bldhte, drehte, etc. � It is worth noting thatnbsp;all the survivals are long-time preexistent forms; the brand newnbsp;dipt reduplicating forms Id, Idn, lie, gie, fie, hie �hing� are cleannbsp;gone, and with them the preexistent hie �hieb�, which was feltnbsp;as one of them.

12. nbsp;nbsp;nbsp;The revived vigor of the full or leveled stem type of conjugation is shown, as was formerly that of the syncopated forms,nbsp;by analogical new formations, back-formations and restorations,nbsp;and by undoing not only newer dipt formations but even somenbsp;primitive OHG syncopations. We may note (i) the wiping outnbsp;of the dipt forms gie, fie, hie �hing�, lie and hie �hieb�; (2) thenbsp;preservation of the full preterit hiew gt; NHG hieh; (3) the fullnbsp;restoration of the erstwhile languishing forms gi(e)ng, fi(e)ng,nbsp;hi(e)ng and liefi; (4) the displacement of the dipt forms of haben,nbsp;except hast, hat, hatte, hatte; (5) the displacement of Idn by lassen;

(6) the displacement of slahenjsldn, fdhenffdn, hdhenjhdn and all their dipt present forms by full forms of the type of schlagen,nbsp;fangen, etc.; (7) the displacement of the dipt past participlesnbsp;gegdn, gefdn, geldn, gehdn (�gehabt�), gesldn, gehldn, etc. by thenbsp;full forms gegangen, geschlagen, etc.; (8) the creation of thenbsp;dissyllables gehen, stehen, sehen, gehe, stehe, sehe, etc. beside thenbsp;old dipt forms gehn, stehn, sehn, etc., which are now felt merely asnbsp;colloquial forms of the normal dissyllabic words, with a clippingnbsp;of endings which can be applied colloquially to any verb ifnbsp;euphony allows. (9) By the contaminative inclusion of tun in this

dipt ending group, the restored full formations have wiped out the last remnant of the -mi inflection as a primary formationnbsp;in German speech feeling

VIII. The Bearing of the MHG Clipt Verb Formations on the Problem of the Normal Direction of Linguistic Development.

1. nbsp;nbsp;nbsp;This double reversal of the direction of linguistic developmentnbsp;� gdn and its kindred clipt forms first refraining for centuries tonbsp;encroach on preterit formations, then thru a space of three or fournbsp;centuries forming an aggressive and conquering group, then againnbsp;rereversing the direction of their development and building againnbsp;the forms they had begun to destroy �� is the most classicnbsp;illustration of the fact that not only the elements and specialnbsp;groups of a language are subject to shifting and displacement,nbsp;but that along with them speech feeling itself is subject eithernbsp;to a slow drift (often cumulative), or (with the breaking down ofnbsp;old forms) to a more rapid displacement and re-formation. Thenbsp;system of phonetic variation which reacht its overdevelopmentnbsp;and culmination in the late OHG and early MHG period, hadnbsp;develop! slowly and cumulatively since Indo-European times.nbsp;At last its strength became great enough to inaugurate the newnbsp;formations and groupings of the clipt preterit type, at variancenbsp;with the older types of stem formation, but not disagreeing withnbsp;the accumulatingtotalof phonetic variations. Then, at the close ofnbsp;the MHG period, with the development of a new trend in phoneticnbsp;feeling, the new groupings based on former speech feeling collapsed, and the direction of development was again reverst. Butnbsp;every one of these diverse developments lay in the direction ofnbsp;the then prevailing speech feeling.

2. nbsp;nbsp;nbsp;This throws light on an important moot question, namely,nbsp;that of the normal order and direction of structural and morphological development. Do inflections and affixes normally formnbsp;by excretion and analysis, as -er in kalber, -en in frauen and -z- innbsp;liellie-z(en) ? Or do they normally form by accretion and composition, as in Fr. donner-ai, triste-ment and Ger. ge-macht, faul-heit ? Fortunately or unfortunately the adherents of both viewsnbsp;can bring an inexhaustible array of irrefragable illustrations andnbsp;proofs of their respective views.

Bin is not an exception to this statement, as it is no longer felt as an analyz-able inflected form.

3- The fact that all these divergent OHG, MHG and NHG developments agreed with the then drift of the language, andnbsp;were in accord with the speech feeling and establisht norms whichnbsp;formed their basis and background, suggests a different principlenbsp;which easily reconciles and includes both these opposing views:nbsp;Speech feeling determines the form and direction of specificnbsp;linguistic developments, and these are always in the directionnbsp;of the establisht norms and patterns of the language in whichnbsp;they occur. The survival of new formations depends on theirnbsp;conformity to this rule.

4. It is for this reason that phonetic, structural, inflectional and syntactical types and shiftings usually occur, not detachtnbsp;and in disorder, but in analogous and orderly groups, as in thenbsp;case of the different groups of syncopating verbs considered above.

OLD PRETERITES OF THE FIRST ABLAUT CLASS IN THE 1671 WITTENBERG REVISION OF THEnbsp;LUTHERAN BIBLES

One of the most interesting archaisms of the Lutheran Bible is the retention of the old preterite singular in ei of the first classnbsp;of ablaut Verbs. Rudolf von Raumer in his Vorschlage zur Revisionnbsp;von der M. Luther-Bibeliibersetzung, Jena 1884, states that thenbsp;Liineburg Bible of 1677, the N�rnberg edition of 1692 andnbsp;the Wittenberg edition of 1703 still preserve preterites likenbsp;bleib, treib, etc. Virgil Moser in his interesting article: Friih-neuhochdeutsche Studi�n ^ compared the 49 instances which henbsp;gives with the forms of all the editions at his disposal and pointsnbsp;out that an Ulm edition of 1671 introduced the new forms withnbsp;ie in two sevenths of the cases, e. g. always erschien, biesz, schriebnbsp;and predominantly blieb. He suggests that the disuse of thenbsp;preterite in the Upper German dialects may have been the reasonnbsp;for the early disappearance of the form ei. It gave way, however,nbsp;very slowly. Nearly thirty years later the N�rnberg edition of 1698nbsp;has only one new form to every four of the old. The Wittenbergnbsp;edition of 1702 has the same proportion, whereas in the N�rnbergnbsp;edition of the same year the new forms predominate in thenbsp;proportion of three fifths to two fifths.

^ This article is a rearrangement of one section of a longer paper: The language of the i6yi Wittenberg Revision of the Lutheran Bible which the authornbsp;hopes will shortly appear in the Germanic Review. This revision was undertaken by the theological faculty of the University of Wittenberg and wasnbsp;published by Balthasar Christoph Wust. In the preface they state: �Zu demnbsp;Ende, dasz Lutheri Arbeit rein und wol erhalten werde, ist die H. Bibel nochnbsp;unlangst widerum auffs neue von uns mit FleiB durchgegangen und revidiertnbsp;Worden.� This Vorrede is dated Wittenberg March loth, 1671 and is signed bynbsp;the �Dechant, Senior, Doctores und Professores der theologischen Facultatnbsp;daselbst*.

Just why Moser confined himself to examining only the 49 cases he gives I fail to see. They were certainly not the onlynbsp;occurrences of the sing. pret. of verbs of this class, for I noted nonbsp;less than 160 instances, counting those given by Moser which Inbsp;carefully compared. If he desired to make such an examinationnbsp;at all, it should have been based on all the instances occurringnbsp;and not on just a few.

The edition of 1671 has made considerable progress in introducing the new forms. Moser�s statement for 1622 is that ei is retained in all but two cases (Ex. 16, 10; Judith 4, 5) whichnbsp;already appear in the edition of 1545. In the Bible of 1671 66 newnbsp;forms appear to 532 old forms, or about in the proportion of inbsp;to 8. StiU this is much less than that in 1698 which has i to 4nbsp;and that of the N�rnberg edition of the same year which hasnbsp;3 to 5, as we have seen. In the last quarter of the 17th centurynbsp;the old forms disappeared quite rapidly, still a goodly numbernbsp;continued over into the i8th century.

Taking the individual verbs in alphabetical order the statistics for the Bible of 1671 are as follows:

bleiben: hleih 114 times: Num. 21, 9; 2 Sam. 13,20, etc.; hliehi^ times: Gen. 28,11; Tob. 2,13, etc.

(er)greiffen: (er-jgm//38 times: Gen. 44, 6; Matth. 8, 15, etc.; one of gyi]]: 2 Mac. 14, 17.

reiten: reit 9 times: Deut. 22, 22; i Sam. 25, 20, etc.; once du rittest: Hab. 4, 8.

(er)scheinen: (er)sc^ew25times: Gen. 12,7; 18, i, etc.; {er)schien 22 times: Ex. 16, 21; Matth. i, 20, etc.; one of erschienenbsp;2 Mac. II, 8. One sees that the new forms are almostnbsp;as numerous as the old.

(be)schneid: schneid ii times: 2 Sam. 10, 4 etc.;; beschneid 5 times: i Mac. 2,46, efc.

schreiben: scAm'�42times: Ex,24, 4, etc.; schrieb 8times: Dan. 5, 5, etc.; unterschrieb once: Dan. 6,10; schriebe once: Judithnbsp;4. 5-

schweigen: schweig lo times: Gen. 34, 5; Matth.26, 63; Acts 15, 12, etc.; schwieg or\ce\ Gen. 24, 21. This instance is alsonbsp;found in the Bible of 1622, but not in that of 1545.nbsp;steigen: steig 31 times: Josh. 4,19; 2 Sam. 7, 18, etc.; none of stieg.nbsp;streichen: bestrich once: Judith 16, 10; no other form,nbsp;streiten: streit 34 times: Ex. 17, 8; Josh. 10, 14, etc.; twonbsp;false preterites: streitet: Jdg. 9,45; stritte: Judith 5,14.nbsp;treiben: treib 37 times: Gen. 3,24; Jdg. 1,20, etc.; trieb 15nbsp;times: Mk. 5, 40; Acts 7, 19; 18, 16, 18, etc.nbsp;weichen: weich 31 times: Gen. 12,9; John 6,15, etc.; no other form.nbsp;The preterite plural of verbs of this class conforms in everynbsp;respect to the modern norm. Those with a long vowel are:nbsp;blieben: Num. 11,5; 14,38, etc.nbsp;schienen: 2 Mac. 10, 29, etc.nbsp;schryen: Ps. 22, 6, etc.nbsp;schwiegen: Acts ii, 18, etc.nbsp;stiegen: Josh. 4, 18, etc.nbsp;trieben: Ps. 107, 23, etc.

With short vowel: bissen: Num. 21, 6, etc.nbsp;litten: Wis. 18, i or lidten'. i Sam. 2, 5, etc.nbsp;wichen: Acts 1,4, etc.

Weak verbs now belonging to the first ablaut class are: gleichen: ich gleichct: Wis. 7, 9; vergleichet, ppl.: Prov. 27, 15.nbsp;Strong forms of this originally weak verb do not appearnbsp;before the 17th cent. Weak forms are found occasionallynbsp;as late as Schiller.

Jer. 15, 10, both instances being in Luther, meiden: meidete: Tob. i, 10; meidet (pret.): Tob. i, 5; Job l, i.

Weak forms are not found before the i6th century, preisen; freiseten: Gen. 12, 15; gepreiset: Luke 4, 15. This verbnbsp;from French priser was originally weak. Weak formsnbsp;continue in part to the 19. century, tho strong formsnbsp;appear as early as the 15th cent.

scheinen has the weak pret. scheinete: i Mac. 6, 39, as in L., otherwise the verb is strong, as above. The weak forms,nbsp;which do not occur before the i6th cent., may be undernbsp;the influence of the MLG.

schmeissen: in the meaning of �befoul�, is conjugated weak: schmeyste: Tob. 2, ii. This verb which appears alreadynbsp;in OHG. continues down to the i8th century.

speien: speyete: Jonah 2,11; Matth. 15,9; Wis. 11,19; aus-gespeyet: Lev. 18, 28; Num. 12, 13; 2 Peter 2, 22. Weak forms appear as early as the 12. cent, and are very commonnbsp;in the i6th and 17th centuries.

weisen: anweisete: Gen. 46, 28; hewciseten: 2 Mac. 14, 22; ppl. heweiset: Ps. 31, 22; �berweiset: 2 Mac. 4, 45; unterweiseten:nbsp;Acts 14, 21; geweiset: Matth. 3, 7. Strong forms of thisnbsp;originally weak verb begin in the 15th cent, and predominate since the i6th century.

A few instances of the old strong forms of verstegen were noted, which in older NHG. was verseigen and verseihen, e. g. inf.nbsp;verseihen'. Jer. 51, 36. The imperative is verseige: Is. 44, 27; thenbsp;past ppl. versiegen: Ps. 107, 33; Hos. 9, 14; this latter form isnbsp;used as late as Herder. The present tense of the new weak verbnbsp;is also found inf. versiegen: Is. 19, 5; 3rd plu. versiegen: Sir. 40, 13.

�DER RIHTER UND DER TEUFEL�

The tale which Chaucer�s Friar tells on the road to Canterbury has recently been used to illustrate the proper editorial handling of traditional materials, �� a subject involving thenbsp;general reliability of Asbj�rnsen and Moe�s collection, �� andnbsp;the differences between literary and popular narrative stylenbsp;Since the tale has thus acquired an impoitance perhaps beyondnbsp;its deserts, it is worth while to list and discuss briefly the dozennbsp;or so versions which have lately come to my attention Thesenbsp;new texts show the longevity of the story versified by the Strieker

1 R. Th. Christiansen Festskrift til Hjalmar Pettersen (1926); Jan de Vries Het sprookje Antwerp [1929] pp. 91 ff.

� About 30 versions are listed and 20 of them are discussed iii my article quot;The Devil and the Advocate�� PMLA XXXVI (1921) 35 ff. In that articlenbsp;certain matters call for comment either by way of correction or addition: {Notenbsp;3): The MHG text was first printed in Lassberg Liedersaal II (1846) 341 ff.nbsp;There exists a modernization by K. Simrock which I have not seen; see F. W.nbsp;Wander Deutsches Spvichw�rterlexikon, s.v. �Anderes 7.�� (Note 11): Weber�snbsp;text may be found in Reclam�s Universalbibliothek. (Note 13); The story is reprinted in Tewaag Erzahlungen aus Hessen {1888) p. 87. (Note 15): Usteri�snbsp;text contains an incident involving goats which I failed to mention. (Note 22):nbsp;The story is found in P. Kennedy Legendary Fictions of the Irish Celts (Londonnbsp;1866), pp. 147�8 = W. B. Yeats Fairy and Folk Tales of the Irish Peasantrynbsp;(London 1888). (Note 25); The story is found in Zaunert Deutsche Marchen seitnbsp;Grimm I 368. (Note 27): Correct the reference to read: p. 104 � 48 ^ 149. (Note 33):nbsp;Insert at the end of the first paragraph; Chauvin III 69 no. 39.

To the list in note 2 of my article I add several which I have not seen: (i) E. von Wolzogen Schwankbuch (1922) p. 75; (2) Th. van Rijswijck Po�tischenbsp;Luimen 44: {3) Niedersachsen I no; {4) Bienenkorb 6, c. 69; (5) J. D. Ernstnbsp;Bilderhaus (1675) I 769.

References are conveniently assembled in Feilberg Bidrag til en ordbog over jyske almuesmdl (1886�1912) s. v. ��ladefoged��, �ridefoged� and Pauli�snbsp;Schimpf und Ernst (ed. Bolte, 1924) n. 81. The texts in the Dansk Folke-mindesamling (cited under no. 81 in Lunding FF Commun. 2) are printed innbsp;Kristensen Danske Skjcemtesagn (notes 17 and 28 in my article). Feilberg�snbsp;reference to �Vang Regio� deals with a different story, as does Kristensennbsp;Fra Bindestue og Kolle I no. 17.

and by Chaucer. It will be remembered that the story tells how the Devil, who is accompanying an avaricious advocate, refusesnbsp;everything that is offered him until someone curses the advocate.nbsp;The Devil seizes and bears off the advocate because the cursenbsp;is sincere. In the Strieker�s story we learn why the advocatenbsp;is cursed, viz., because he has unjustly possessed himself of annbsp;old woman�s cow. This explanation is not found in Chaucernbsp;and the exemplum tradition. Accordingly we may say withnbsp;confidence that the three Finnish tales taken down in the lastnbsp;few years � are descendants of the oral tradition on which thenbsp;Strieker drew *. This form of the story appears, therefore, tonbsp;be very old and to have maintained itself only on the extremenbsp;border of the area in which the story was current.

Within that area another form became widely known, a form first reported by Caesarius of Heisterbach, later employed bynbsp;Chaucer, and firmly established in exemplum tradition. Thisnbsp;form, which is recognizable by the absence of the cow, has persistednbsp;in tradition on the shores of the Baltic. In addition to thosenbsp;previously known, I note three Danish versions � and one fromnbsp;Pomerania �. It is^ furthermore, reported from what was formerlynbsp;Austrian Silesia in a form which appears to have undergonenbsp;literary embellishment Altered in still a different way, thisnbsp;traditional version is found in the homiletic writings of Peternbsp;Bornemisza (Abstemius), which appeared between 1573 and 1579.nbsp;As Kirjaly points out, this version is derived from the Hungariannbsp;Esopic tradition, for only there does the Devil hire himself outnbsp;as a laborer and demand as pay whatever is seriously given tonbsp;him�. Jan de Vries makes it clear that the Limburg version isnbsp;a derivative of the exemplum tradition, as is also the Norwegian

� See the references in Aarne FF. Communications 33 no. 821 C*. Copies of these three versions, which were taken down between igo8 and 1918, are innbsp;the University of Chicago library.

^ E. g., quot;Ni vaimol ei ollu raha maksa ja ku henel ol yks lehm, ni see tultti ja myytti ja henee lehmas men.� (The old woman had no money to pay andnbsp;since she had a cow, so they came and sold it and her cow was lost). The passagenbsp;is quoted from Kallio no. 77. Similar passages occur in the other texts.

� Skattegraveren II (1884) 105 no. 470; E. T. Kristensen Jyshe Folkeminder IV {1880) 341; E. T. Kristensen Danske Sagn VI (1900) 755 no. 761.

� ��Deutsche Sagen und Schwanke in einem ungarischen Teufelsbuche� Ungar. Jahrb. I (1921) 245.

250 story taken down by Hans Ross �. Beyond pointing out that thenbsp;oral Norwegian tradition, which forms the basis of the storynbsp;printed by Asbjornsen and Moe, follows the exemplum tradition,nbsp;we need not enter into the discussion of the evidence so clearlynbsp;and carefully presented by Christiansen. In this classificationnbsp;we cannot distinguish sharply between tales derived from thenbsp;exemplum and tales derived from the oral tradition in whichnbsp;the exemplum finds its origin. I have little doubt that thosenbsp;versions which have bookish antecedents, e. g., Bornemisza, gonbsp;back ultimately to manuscripts of the exemplum. On the othernbsp;hand, it is altogether likely that many of these north German,nbsp;Baltic, and Scandinavian versions derive directly from the oralnbsp;tradition which gave rise to the exemplum.

An abbreviated form, which is first met with in the writings of Hans Sachs, is found again in recent Danish tradition Innbsp;this form we have merely the weeping child which is offered tonbsp;the Devil and the carrying off of the lawyer. This variation isnbsp;rather frequent in the region where the story is most frequentlynbsp;told: it occurs three times in Danish and once in German tradition.nbsp;A connection with Hans Sachs is unlikely, inasmuch as thenbsp;omission of a single incident, and particularly of an incidentnbsp;forming part of a series, is an ordinary mishap in a tale�s life.

A very corrupt version is found in a Danish manuscript play of the first quarter of the eighteenth century, but it offers nonbsp;points of interest and appears to have suffered mutilation for thenbsp;purpose of adapting the story to use as a sub-plot

The distribution of these texts agrees well enough with the origin of the story in northwestern Germany. There it was earlynbsp;recorded and there it has remained in popular tradition, innbsp;Limburg, Westphalia, Hesse, Oldenburg, Bremen, Pomeranianbsp;and Denmark. Elsewhere the story seems not to have become

� J. de Vries Het sprookje pp. 91 ff. (A text reformed from P. Kemp Limburgs sagenboek, which I have not seen) and pp. 95�96 .{for the Norwegian tale). Seenbsp;also the references in R. Th. Christiansen Norsks folkeminne II (Kristiania 1921)nbsp;no. 1183. The story has received the number 1186 in Aarne-Thompson FF Com-mun. 74, where the reference to Christiansen�s article should be corrected tonbsp;read Festskrift til Hjalmar Pettersen and the number 1183 should be insertednbsp;after �Norw�.

See PMLA XXXVI 5tff.; Jens Kamp Danske Folkeminder, Aeventyr, Folkesagn, Gaader, Rim og Folketro (Odense 1877) pp. 286�7 no. 898.

See S. B. Smith Studier pd del gamle danske Skuespils Omrdde (Copenhagen 1883) pp. 117�23.

thoroughly acclimated; the oldest version, the Strieker�s, leaves no trace in the popular literature of southern and eastern Germany Versions reported from elsewhere than northwesternnbsp;Germany are either definitely literary in origin and associationnbsp;or extremely corrupt.

Apart from the connection of the story with Chaucer, �- a point on which these new versions throw no light, �� the mostnbsp;interesting detail in the story�s life is the preservation by Finnishnbsp;narrators of the form found in the thirteenth century versenbsp;of the Strieker. Since his narrative was entirely inaccessiblenbsp;until almost the middle of the nineteenth century and thennbsp;became known only to a small group of scholars, �� Lassberg�snbsp;Liedersaal is a rare book and von der Hagen�s Gesamtabenieuernbsp;cannot have enjoyed a wide circulation, � we have one morenbsp;example of the tenacity of folk-memory. The nice questionnbsp;remains whether the folk-versions of northwestern Germany andnbsp;Denmark likewise represent the independent preservation of anbsp;story disseminated throughout European lettered circles bynbsp;Caesarius� exemplum.

The text printed by Schmeller is, as I have pointed out {PMLA XXXVI 49 n. 20), a translation of Pauli�s jest into the dialect of Aschaffenburg. Under nonbsp;circumstances is it evidence for the persistence of the tradition in Bavaria. Thisnbsp;disposes of the notion mentioned by de Vries (Het sprookje p. 90) that the storynbsp;is Bavarian.

A VIEW OF LESSING

On Lessing�s birthday two centuries ago George Berkeley, then Dean, later Bishop, was aboard ship off the coast of Northnbsp;America. Three years earlier he had written of this region:

For the better part of the ensuing three years Berkeley dwelt on the shores of Narragansett Bay, reconciling himself as best henbsp;might to the necessary abandonment of his charitable plan fornbsp;a college at Bermuda, but finding life in New England congenial,nbsp;and leaving behind him a fragrant memory.

So far as I am aware, Lessing displayed no interest in our New England. The course of empire never engaged his attention, nor wasnbsp;he by nature endowed with missionary zeal. In spite of his passionnbsp;for doing things and getting things done in his time and place,nbsp;he was primarily a literary, philosophical, and theological scholarnbsp;who rather kept within range of the Mediterranean beaconsnbsp;of civilization than turned his gaze towards these plantationsnbsp;as yet not very productive in contributions to learning. Nevertheless, it is pleasant to recall that in Lessing�s time, even asnbsp;Berkeley had found, there were institutions here respectable fornbsp;age, as for other attributes, and a population of which an acutenbsp;observer 1 could say that it had �retained to an incalculable degreenbsp;qualities which had faded from ancestral England with the daysnbsp;of Queen Elizabeth��; namely, �spontaneity, enthusiasm, and

^ Barrett Wendell; A Literary History of America^ New York, 1900; cf. pp. 19, 77-

versatility.� To a people so enduringly Elizabethan the venerator of Shakespeare would not have been indifferent. But he mightnbsp;have discovered additional characteristics: first, a sturdy individualism which, as to the spiritual realm, made every man,nbsp;disdaining a mediator, bent upon proving himself one of God�snbsp;elect, and as to the material, made him watchful over his interestsnbsp;in trade and at the town meeting; and secondly, so persistentnbsp;a particularism that the smallest of the states became the lastnbsp;to ratify the Federal Constitution. Could any German fail tonbsp;perceive herein a certain affinity between this people and hisnbsp;own ? The Germans are a nation of individualists, and their proneness to particularism, keeping them for generations internationallynbsp;impotent and to this day so inconveniently multiplying theirnbsp;political parties in the Reich, has more than any other singlenbsp;cause brought to pass, on the one hand, an infinite variety ofnbsp;separate existences � so picturesque and romantic when reflectednbsp;in art � and has contributed, on the other hand, to create fornbsp;them that fecundity of the inner life which qualifies them to benbsp;a people of poets and thinkers.

It is no accident that the Protestant Reformation originated in Germany. The Protestant Reformation was an assertion ofnbsp;confident, self-sufficient individualism; and Martin Luther, thenbsp;Reformer, was a German to the marrow. So was Gotthold Ephraimnbsp;Lessing. On many accounts the two names belong together. Suchnbsp;parallel as we draw may begin with the purely historical observation that without Luther there could have been no Lessing.nbsp;It has been prettily said that of the many consequences of thenbsp;Reformation �� not all, certainly, salutary � one was the creationnbsp;of a new institution of inestimable advantage to the community,nbsp;the Protestant parsonage Nothing, surely, prevented the parishnbsp;house of the Roman Catholic Church from being a source ofnbsp;social as well as spiritual light and warmth; it often was and stillnbsp;is; but the parish priest being celibate, the house was sterile. We ofnbsp;New England, at least, do not need a demonstration of what thenbsp;difference means. We have only to remember how many of thenbsp;makers of our history have been scions of clerical stock. Lessingnbsp;too was the son of a Protestant parson.

To be sure, Johann Gottfried Lessing early became alarmed at his boy�s free-thinking and no doubt remained, as Lowell

^ Cf. W. Brecht: �Einfiihrung in das i6. Jahrhundertlt;s, GRM, III, pp. 340 ff.

shrewdly surmises, �like many other fathers, permanently astonished at the fruit of his loins�. Other ministers among thenbsp;orthodox were more than astonished as time went on; they waxednbsp;indignant and grew not a little apprehensive, lest the citadelnbsp;in which they felt secure against open assault might come tumblingnbsp;about their ears because a sapper and miner was dislodging itsnbsp;foundations. I should exceed the sphere of my competence if Inbsp;undertook to define Lessing�s theology, perhaps so if I evennbsp;expressed an opinion of his religion, though the evidence fornbsp;this is spread before us in sundry writings, whereas his theologynbsp;is largely a matter of inference from imperfect testimony. In anynbsp;case, I assume, we are more concerned with the spirit than thenbsp;letter of his doctrine and should be most solicitous to discern hisnbsp;attitude towards the faith and the speculation of his time. It w'asnbsp;a time when Protestant orthodoxy was far more rigid than Romannbsp;Catholic dogma, for all its positiveness, had ever been �� a time,nbsp;accordingly, when a generation of rationalists either abandonednbsp;Revelation altogether, as incapable of accommodation to reason,nbsp;or firmly upheld it as susceptible of confirmation by logic andnbsp;of proof on the basis of evidence furnished by the senses. Withnbsp;the former class Lessing was not impatient. At the age of twentynbsp;he composed his comedy Der Freigeist, describing the titular heronbsp;as �without religion, but of an abundantly virtuous disposition�.nbsp;In the course of the action this hero is made to acknowledge thatnbsp;a clergyman is not necessarily a hypocrite, and he is none toonbsp;gently converted from his hasty intolerance, though not exactlynbsp;converted to religion. With the latter class, on the contrary,nbsp;Lessing contends as with those who abuse the greatest of ournbsp;natural gifts, who subject to slavery that faculty through whichnbsp;alone man is capable of being free. He is of a mind to save Revelation from its friends. In other words. Revelation, he holds, isnbsp;not to be preserved by any attempt to demonstrate that in reasonnbsp;it is a final formulation of truth; rather the truths of Revelationnbsp;invite translation into truths of reason; and they can be sonbsp;translated, if we perceive that Revelation is not a product butnbsp;a process, is as yet only a partial, progressive communicationnbsp;of such truths as mankind has been successively ready to grasp,nbsp;conformably with a Providential plan for the education of thenbsp;human race. Shall we ever arrive at the entire comprehensionnbsp;of truth ? We have all eternity in which to try. And what, though

we do not ? It is the endeavor and not the attainment that shall make us also perfect.

I would not mar by a single superfluous word the noble passage in which Lessing declares his choice But I may be permittednbsp;to adduce a private communication recently made to me by annbsp;eminent theological colleague who thus confesses to an obligationnbsp;both professional and personal;

Some forty years ago, when I began to lecture on the Philosophy of Religion, I found in The Education of the Human Race and in Nathan the Wise the rootsnbsp;of the modern liberal theology, and had the happiness of impressing this viewnbsp;on successive classes. Lessing took an epoch-making step in the interpretation ofnbsp;religion, and this it was which so long ago arrested my attention and has remainednbsp;my reassurance.

Lessing, then, who paid homage to Luther not the less sincere because occasioned by polemics with the Lutherans of his day,nbsp;is a continuator of Luther in Luther�s most important line ofnbsp;activity, that is, in the cause of religious emancipation. But ifnbsp;we take a large view of the two hundred years separating them,nbsp;and properly appraise what each effectuated in his day andnbsp;generation, we shall see Lessing as a second Luther in respect tonbsp;a still more comprehensive achievement. It will not have escapednbsp;notice that I used just now an expression of Matthew Arnold�snbsp;defining the relation in which he conceived Heinrich Heine stoodnbsp;to Goethe, that of a continuator. I should be the last to proclaimnbsp;Heine a safe guide in either criticism or history. But he nevertheless often reveals no common penetration in his aper^us, and,nbsp;journalistic though his language is apt to be, he has a knack atnbsp;neat formulation. Heine � brings Luther and Lessing togethernbsp;as I have done, albeit for reasons that are not altogether mine.nbsp;He calls Lessing the continuator of Luther; he says, �Germanynbsp;has produced since Luther no greater and better man that Gottholdnbsp;Ephraim Lessing. These two are our pride and our delight��.nbsp;He avers �that Luther is not only the greatest, but also the mostnbsp;German man of our history; that in his character are magnificently united all the virtues and the faults of the Germans, sonbsp;that even in his personality he is a representative of wonderfulnbsp;Germany��. Heine particularly refers to those mundane constitu-

^ �Vater, gib! die reine Wahrheit ist ja doch nur f�r dich allein!� Fine DupUk (1778).

ents which made Luther a rough and ready controversialist as well as a tender husband and father, a boon companion, and anbsp;poet. He speaks of these elements as ��that earthly admixture,nbsp;without which Luther could not have been a man of action. Purenbsp;spirits cannot act.� We should say much the same of Lessing.nbsp;He also was a man of action, the tenderest husband, a goodnbsp;companion, by no means a pure spirit, but a being of quicknbsp;sympathies and sanguine temperament. Condemned by circumstances to a life prevailingly solitary, he greatly preferred thenbsp;market-place; and the deeds to which he aspired, which in factnbsp;he performed, were no more than prepared in the closet. Lessing�snbsp;historical mission was to be a Siegfried and to revive the Sleepingnbsp;Beauty of German nationality.

The first half of the sixteenth century in Germany fell, admittedly, far short of the spacious times of great Elizabeth at the end of the century in England. But it was none the less a buoyant timenbsp;of national vigor, of youthful promise, and of intrepid leadershipnbsp;towards better things for all mankind. Two forward movementsnbsp;got under way and for a while were fortunately united: one, thenbsp;religious reformation, originating within the country and earlynbsp;assuming the form of a national crusade; the other, humanism,nbsp;coming from without, but almost from the first appearing ��nbsp;by contrast to many of its aspects in Italy � as serious, moral,nbsp;pedagogical, altruistic, democratic, national, intent upon raisingnbsp;German life to the plane of newly discovered humanity, not atnbsp;all conspiring to set up an antique idol for indiscriminate worship.nbsp;Erasmus and Reuchlin are the oculi Germaniae, Hutten is Ger-maniae Libertatis Propugnator, Melanchthon, the Praeceptornbsp;Germaniae. Nobody could regard Luther as a humanist otherwisenbsp;than in principle. But in principle he is one: criticism, foundednbsp;upon the independent interpretation of an authoritative text,nbsp;is his instrument in theology, as it is that of Erasmus or Reuchlinnbsp;in philology. Nobody, on the other hand, could be blind to thenbsp;disintegrating tendency of classical scholarship in and of itself.nbsp;It encourages the learned to smug satisfaction over against thenbsp;unlearned; and if it lives in the world of Greek or Latin literature,nbsp;it is eo ipso alienated from home. The gap between the learnednbsp;and the unlearned which began to yawn wide almost immediatelynbsp;after the death of Luther was not invisible during his life-time,nbsp;but it was not yet portentous. On the contrary, if Germany did

not then produce a Shakespeare, she did give birth to a Hans Sachs �� like the Englishman armed with small Latin and lessnbsp;Greek, but like him also in command of multifarious literaturenbsp;through translations. There is in the six thousand poems of thenbsp;N�rnberg shoemaker an incontestable epic amplitude � morenbsp;than that, there is a picture of contemporary Germany, such anbsp;conspectus of a civilization as Grillparzer declared to be thenbsp;distinguishing mark of an epic poem Nor did Hans Sachs standnbsp;alone. Herder � was mayhap over-enthusiastic, but he found thenbsp;Reinke de Vos of 1498 the greatest epic since Homer. And Huttennbsp;discovered an epic hero in Arminius.

It will be perceived that I now have in mind several of the suggestive pronouncements with which Goethe enriched the seventh book of Dichtung und Wahrheit. He speaks of the lapse of Germanynbsp;into barbarism during the interval between Luther and Lessing,nbsp;declaring that there had never been any lack of literary talent,nbsp;but that there had been want of a national content for the formsnbsp;of poetry, and adding that every nation which lays claim tonbsp;greatness must have its epopoeia, �for which indeed the formnbsp;of the epic poem is nowise necessary��. We should identify thenbsp;Elizabethan epopoeia, in this sense, with what is unfolded innbsp;Shakespeare�s plays. Similarly, there is an epic energy stirringnbsp;in the days of Luther; we see it in Hans Sachs; it is the samenbsp;spirit when it prompts the leaders to utterance in the morenbsp;immediate, more characteristically German, and less objectivenbsp;form of the lyric. The national content might be phrased as thenbsp;ideal of a free church in a free state Goethe divined an epicnbsp;hero in G�tz von Berlichingen; and that Faust �� also of thenbsp;sixteenth century � in whom he symbolized striving, erring,nbsp;and finally triumphant humanity is, for indomitable individualism,nbsp;spiritually akin to the author of the war-song of faith, )gt;Ein festenbsp;Burg ist unser Gott�, and likewise akin to that other patrioticnbsp;warrior, �watch-dog� � of the commonweal, as he modestly describednbsp;himself, who undismayed reiterated to the end his slogan, �Ichnbsp;hab�s gewagt!lt;(

* �Lyrik, Epos, Drama, Aussicht, Umsicht, Ansicht�; cf. Werke, ed. Sauer, Wien, 1909 ff., Tgb. IV, Nr. 3186.

� Cf. Hutten, Opera, passim, and the �Wolfaria� in the F�njzehn Bunds-genossen of Joh. Eberlin von G�nzburg (1521).

I submit: the spirit animating Luther and Hutten is national, the tasks with which they grapple are present, and the vigor withnbsp;which they express themselves gives earnest of illimitable achievement. Humanism has meant for them a broadening of viewnbsp;indeed, but chiefly a vivifying of what lies closest at hand.

I would not be iinderstood to allege that when the German spirit, as above defined, fell into a torpor, it was in all respectsnbsp;a Sleeping Beauty which then awaited the coming of the appointednbsp;Prince. Confining myself to literature, I am aware of the crudenessnbsp;of most of the prose and verse of the period which Wilhelmnbsp;Schlegel characterized as �biirgerlich� and of which Wilhelmnbsp;Scherer said that while it lasted aesthetic cultivation was at anbsp;low ebb. Certainly. But the crudeness was not a sign of degeneration; it was due partly to inexperience, partly to youthfulnbsp;indifference; the gravity of the issues at stake called for force andnbsp;did not suffer luxurious elegance. Although beauty had not fullynbsp;developed, character and capacity were there, when Germanianbsp;was driven to hibernation through many winters of discontent.

It would be fantastic also to entertain the notion that Lessing was in all respects a representative German. He was not. As a very individualist he exercised his right to variation from the type. Henbsp;was no lover of nature, of music, or of the formative arts; he wasnbsp;no sentimentalist or visionary of any sort; figuratively and alsonbsp;literally he was not a dreamer; he had no fondness for lyricalnbsp;poetry �� poetry was to him a means for the treatment of action:nbsp;epic, therefore, or still better dramatic. In likening him to Siegfried I have regard to one feat only. Lessing read the Nibelungen-lied; there is no indication that he was especially edified by it;nbsp;certainly he had no share in Romantic enthusiasm for the Middlenbsp;Ages, or in Rousseau�s and Herder�s enthusiasm for the primitive.

When Lessing came to the fore there were baroque extravagances to curtail and pseudo-classical errors to redeem �� a thicket of thorns and brambles impeding progress to the Sleepingnbsp;Beauty. But there was also a still more formidable obstacle. Afternbsp;a year�s experience had shown the futility of trying to maintainnbsp;a national theatre at Hamburg Lessing exclaimed �, not withoutnbsp;bitterness:

What a na�ve idea, to provide a national theatre for the Germans, when we Germans are not yet a nation! I am not talking of a political constitution, but

259

solely of moral character. One would almost have to say, the moral character of the Germans is the determination not to have any of their own. We are stillnbsp;the sworn imitators of everything foreign, especially the humble admirers of thenbsp;never sufficiently to be admired French.

And yet a twelvemonth before this lament, in the very spring when the national theatre was started (1767), Lessing had himselfnbsp;given his fellow-countrymen, whether a nation or not, an examplenbsp;of national literature in Minna von Barnhelm.

Remarkable on many accounts, Minna von Barnhelm is in nothing more remarkable than for not being what we shouldnbsp;expect a play to be which bears the sub-title �das Soldatengl�ck�nbsp;and is advertised on its title page as �verfertigt im Jahre I763�.nbsp;Should we not at the very least expect the hero, Tellheim, tonbsp;be every inch a soldier ? Since none but the brave deserves thenbsp;fair, we surely anticipate that the hand of Minna shall be thenbsp;reward � perhaps delayed by some mischance, but still ultimatelynbsp;the reward � of conspicuous gallantry; and since we are at thenbsp;height of an epoch glorious in Prussian history � glorious fornbsp;all Germany, inasmuch as Prussia, by beating off a whole Continent of enemies, gave a new dignity to the German name, wenbsp;instinctively raise our hands to applaud military glory effulgentnbsp;in a worthy member of the victorious army. We are confronted,nbsp;in fact, by nothing of the kind. Tellheim is as far from being anbsp;typical Prussian officer as a wearer of the king�s coat could verynbsp;well be. He is not even a Prussian subject � he is a volunteernbsp;in this service, he hardly knows why; he campaigns like a grandnbsp;seigneur with a retinue of unreliable servants; the discipline in hisnbsp;battalion would scarcely pass muster in a provincial militia; henbsp;has been more than once under fire, he has been wounded, butnbsp;we do not hear of his presence at Rossbach or Leuthen, and thenbsp;great day for his command saw a brush with the enemy concerning which the chronicles are silent. This soldier, in short,nbsp;deserves his fortune by exhibitions of magnanimity for whichnbsp;the war and its sequel give him opportunity, but which arenbsp;altogether personal, in no sense military acts.

Nevertheless, the Seven Years� War is the epic stream which has caught up our magnanimous hero and which bears him, hisnbsp;comrades, their dependents and superiors along in its irresistiblenbsp;course. Somewhat after the manner of Diderot we have in Minnanbsp;von Barnhelm a particular class of society isolated and its characteristic virtues displayed: personal honor, hardihood, mutual

dependability, loyalty, bonne camaraderie �� likewise some of the defects of these virtues exhibited. But Lessing avoids allnbsp;commendation of the war in respect to either cause, conduct,nbsp;or result. Instead, he gives us a vivid picture of its effects uponnbsp;personal lives, such a picture as we have recently been ablenbsp;ourselves to contemplate in sordid reality. The principal figurenbsp;in the foreground is distressed and distressing. Tellheim, dishonored and resentful, appears to be as little heroic as Achilles in hisnbsp;wrath; but both are heroes just the same; from the obscure retreatnbsp;of the one as from the tent of the other we look out upon thenbsp;theatre of mighty actions; and as Homer peoples the backgroundnbsp;of his picture with gods, so Lessing fails not to indicate, in thenbsp;background to his, the potent personality of that Frederick whomnbsp;the world has agreed to call the Great.

Goethe describes Minna von Barnhelm as �the first theatrical production with a content, specifically of the time, taken fromnbsp;life of serious import�, �the veriest offspring of the Seven Years�nbsp;War, of perfect North-German national substance�, �the revelationnbsp;of a higher world than that of civic actuality or literary tradition�.nbsp;It is the earliest modern German drama and the prelude to thenbsp;epopoeia of German classical literature. Furthermore, it is thenbsp;masterpiece of a rationalist, a humanist whose acquaintancenbsp;with Aristotle and Plautus made him but the more capable ofnbsp;being a German and the more eager to do for Germans whatnbsp;Plautus, in the light of Greek example, had done for Romans.nbsp;But by �Germans� Lessing understood a community of individuals, not the constituent members of a state. Even love of countrynbsp;he confessed he could not comprehend; he felt no attachmentnbsp;to the soil � his nature was as little political as Goethe�s wasnbsp;warlike. True, he never said patriotism was �the last refuge ofnbsp;a scoundrel� � he was too liberal for that � but he did thinknbsp;it �at most an heroic weakness� What he wished, therefore,nbsp;to do with Minna von Barnhelm bore no semblance to adulation,nbsp;had no part in propaganda; it was something more intimate,nbsp;higher in aspiration and at the same time more vital; for he wasnbsp;a humanitarian � and now we recall that �earthly admixture�nbsp;which Heine attributed to Luther. Lessing was a warm lovernbsp;of his kind, who wished to move his fellows to quickly reactive

sympathy, and who knew � in the words of Gottfried Keller �� that quot;you can bring round the bone and sinew of the people onlynbsp;with the full momentum of truth�. In the ripeness of age Goethenbsp;admonished his disciples, �Let every man be in his oWn way anbsp;Greek, but see that he be one!� Lessing, before Goethe, was anbsp;Greek in his way, and this was the way of an enlightened spiritnbsp;fructifying with formative life the responsive reality of experience. Reality so informed arose as a new art for the Germansnbsp;of 1767. Criticism had penetrated the secrets of creation.

The distinctive excellence of Minna von Barnhelm resides in its authentic personalities, individuals, particularized, conceivednbsp;in their integrity and produced before us in unaffected selfevidence. quot;What an admirable piece!� exclaims Grillparzer ^2.nbsp;quot;Evidently the best German comedy. Comedy? Well, yes; whynbsp;not ? So genuinely German in all its characters, and precisely innbsp;that regard unique in German literature.� Such praise from thenbsp;censorious Austrian who, on the whole, looked askance at ournbsp;North-German author, is doubly noteworthy, for its greaternbsp;weight as coming from Grillparzer and for the technical questionnbsp;that it raises. What kind of comedy is Minna von Barnhelm ?

Although Lessing in his youth was ambitious to become a German Moli�re, it does not appear that he assiduously soughtnbsp;instruction in the plays of that master. Ever5d;hing in the German'snbsp;theory and practice points to a more democratic ideal than cannbsp;be found even in the splendid naturalness of him who studiednbsp;the court and knew the city of Louis XIV. But in Minna vonnbsp;Barnhelm Lessing had a happy thought, comparable for fruitfulness with the fundamental idea of VAvare or le Misanthrope,nbsp;though not unaffected by the exaggerated virtuosity of le Filsnbsp;naturel or le P�re de familie �� I mean that for the springs ofnbsp;action in his drama he envisaged not simply an interminablenbsp;series of those incongruities, maladjustments, mishaps, andnbsp;misunderstandings � including amazing credulity under thenbsp;stress of passion �� which prompt us to the reflection �Whatnbsp;fools these mortals be!� � but also a fundamental incongruitynbsp;upon which to erect a world quite out of joint. Something ofnbsp;the kind is suggested by the baffling sub-title to which I havenbsp;already alluded. The initial impulse, however, which turns thisnbsp;whole world topsy-turvy is this: during a time of public warfarenbsp;Tgb. II, Nr. 1142.

private virtue flourishes and lovers� joy reigns; when public warfare ceases righteousness and peace do not kiss each other,nbsp;but peace brings in her train public wrong, private sorrow, andnbsp;personal estrangement. Everything is now awry. Of the Prussiannbsp;major, by nature generous to the point of lavishness, we mightnbsp;say � paraphrasing a gruesome verdict of Hebbel�s passed uponnbsp;another man � that pertinacity has usurped the place ofnbsp;his soul; and of his Saxon betrothed, by nature bountiful as anbsp;May morning, that she becomes as capricious as April. Wishingnbsp;to convert her lover toher own conception of realities, she institutesnbsp;a stratagem in which she herself all too soon becomes inextricablynbsp;ensnared. The lovers� quarrel eventuates indeed in a reintegrationnbsp;of love, but more because the Gordion knot is cut for them thannbsp;because by themselves they could possibly have disentanglednbsp;the snarl.

Criticizing Kleist�s Prinz Friedrich von Homburg, Hebbel lays down a maxim that may be found generally useful in the interpretation of plays. The drama, he says, represents a thoughtnbsp;which seeks to become a deed either through action or throughnbsp;patience. The protagonist, that is, is impelled by an idea, andnbsp;antagonism to its realization gives rise to the action. If we shouldnbsp;ask what Tellheim�s idea is, the immediate answer would shapenbsp;itself into some such formula as this: honor is the supreme good;nbsp;without it, nothing is possible, and when it is impugned, no meansnbsp;are too costly, no sacrifice is too great, to restore it. Similarly,nbsp;Minna cherishes the idea that love is the supreme good; withoutnbsp;it, nothing is worth much; and everything inconsistent with itnbsp;is negligible. It is obvious that both positions are tenable; thatnbsp;they are mainly characteristic of the two sexes; that in this absolute form they are irreconcilable; that, therefore, they might, asnbsp;well as not, be factors in a tragedy; and that they are fit tonbsp;yield a comedy when taken with uncompromising, not to saynbsp;exaggerated, exclusiveness by persons who in reality are not onlynbsp;lovers, but also respectively a man and a woman of honor. Lovenbsp;is the fulfillment of the law; but Minna�s means are totallynbsp;inadequate to prove that love may defy the law. Doubtless, anbsp;good name is rather to be chosen than great riches; but there

His own father : �ein herzensguter, treuer, wohlmeinender Mann; aber die Armut hatte die Stelle seiner Seele eingenommeno; Werhe, ed. Werner,nbsp;Berlin, 1913 ff., Tgb. I, Nr. 1323.

is something more substantial than the bubble reputation, as Tellheim, in his right mind, well knows: self-respect is a higher-good than the respect of others. Thus the contestants are bothnbsp;in the right, and both in the wrong �� wherefore, we may parenthetically remark, we have a comedy instead of a tragedy.nbsp;Easier, however, though it be to sympathize with the untroublednbsp;gaiety of the good-hearted maiden than with the purblind stubbornness of the sorely troubled man, we are left with a sense thatnbsp;after all Tellheim is justified and has reinforced the truth of annbsp;assertion which, as we attribute it to him, has a far deepernbsp;meaning than on the lips of Lovelace:

Furthermore, it is plain that the foregoing generalities, that any such rationalistic formulations, not only disconcert aestheticnbsp;reaction to this drama, but also obscure what is most meritoriousnbsp;in the artistry employed. We have before us nothing so simplenbsp;as protagonist, antagonist, victory or defeat, nothing so mechanicalnbsp;as thesis, antithesis, and synthesis. Neither Tellheim nor Minnanbsp;is the mere embodiment of an idea, though Minna is unmistakablynbsp;much the less complex and significant personality � as such,nbsp;perhaps, the less entitled to prevail. And the significance ofnbsp;Tellheim�s more ample endowment of human qualities is notnbsp;made manifest by what he does in pursuit of an abstract ideal.nbsp;The play is, indeed, not devoid of action, nor �� especially in thenbsp;conduct and the successive situations of the sharply individualizednbsp;minor characters �� of those appeals to indulgent humor whichnbsp;are the prerogative of comedy. But the plot, with its jugglingnbsp;of rings, in which Minna involves Tellheim, is neither very clevernbsp;nor very useful, except as it furnishes occasion for exposition andnbsp;argument: Tellheim seems to be in singularly unstable equilibriumnbsp;and Minna to be singularly lacking in both perspicacity andnbsp;foresight. These are, however, but the symptoms and the signsnbsp;of passion. The real action is in the breast of the hero; the conflictnbsp;between love and honor he alone has fought out to a victory

By �honor� I here mean a subjective standard of right conduct. Tellheim�s self-respect will neither permit Minna to share his disgrace, nor permit him tonbsp;live on her bounty. For the most part, however, � and notably in Act IV, Scenenbsp;vi � Tellheim and Minna mean by �honor� something that can be taken awaynbsp;and restored, i. e. social recognition, objective approbation. Tellheim shares thenbsp;extreme sensitiveness of the aristocratic and military caste on this score: quot;Caesar�s

which he thinks secure, until the sudden apparition of Minna renews the struggle in an aggravated form.

That Tellheim, then, with all his imperfections on his head, is a man and a brother makes the deepest impression and, confirmed by the impressive verisimilitude of such another characternbsp;as Werner, constitutes the revolutionary excellence of thisnbsp;composition. Such was, moreover, precisely the effect aimed at.nbsp;Why, queries Lessing, all the toil and trouble of writing andnbsp;enacting a drama, unless we desire to attain in the fullest measurenbsp;what is fully attainable only on the stage ? Like any depictionnbsp;of life, the drama will convey, at least by implication, a moralnbsp;lesson � it may legitimately be produced for that explicit purpose.

Hebbel, and sometimes Schiller, will construct a plot that almost literally conforms to the following definition:

If we refer a general moral precept to a particular case, attribute reality to this particular case, and out of it compose a story in which the moral preceptnbsp;s intuitively perceptible, such a composition is called a fable.

These are the well-known words of Lessing 1�, the maker of definitions. If we remember that Fabel means either �fable� ornbsp;�plot�, and if �� as without unseemly violence we might do �nbsp;we render the phrase �daraus eine Geschichte dichten� by �inventnbsp;an action involving it�, we may at the end substitute �drama�nbsp;for �fable� and arrive at the definition, �such a composition isnbsp;called a drama�. But not in Lessing�s sense; and he is a makernbsp;of distinctions before he makes definitions. To the dramatist, henbsp;declares, it is immaterial whether or not a general truth can benbsp;inferred from the action of his persons Again :

The drama lays no claim to a single, definite doctrine emerging from its plot; it aims either at the passions which the course and the peripeteia of its

wife must be above suspicion�; so long as his mind is befogged by egotism or hardened by opposition he makes no distinction between rectitude and repute.nbsp;But when he clearly hears the call to altruistic duty, he casts to the winds thenbsp;talisman to which he has been clinging: �Nein, das ist dringender� (IV, viii),nbsp;sLieber hier alles im Stiche gelassenlt;lt; (V, i). This instinctive turning to the reallynbsp;preferable proves that his allegiance is after all neither to a phantom nor to anbsp;fetish. G. Kettner, in his admirable Lessings Dr amen (Berlin, 1904), is too littlenbsp;discriminating on this point (pp. 96 ff.).

plot are capable of inflaming and sustaining, or at the pleasure which a true and vivid depiction of morals and characters affords; and both aims require a certainnbsp;rounding out of the action, a certain satisfying ending � something that we cannbsp;dispense with in the moral narrative, because in it, all our attention is concentrated upon the general proposition of which the particular case gives so illuminating an example.

Since the illusion is far more potent in a drama than in a mere narrative, the persons in the former interest us far more than in the latter, and we arenbsp;not satisfied with seeing their fate settled just for the present moment, but wishnbsp;to rest forever content with the settlement.

And now We are prepared to answer the question, what kind of comedy Minna von Barnhelm is. It is a true comedy: it incitesnbsp;to laughter at manifold human frailty � in no wise to ridiculenbsp;thereof � and it also moves to pity for well accredited people innbsp;affectingly real tribulations. Secondly, it is a serious comedy ��nbsp;Lessing himself authorizes the apparent oxymoron ; he who,nbsp;for reasons that need no further exposition, welcomed the trag�dienbsp;bourgeoise and was not inhospitable even to the com�die larmoyante.nbsp;Minna von Barnhelm is but one degree removed from tragedy. Asnbsp;serious, it produces the effects peculiar to drama, and these are,nbsp;in brief, the poignant impression of significant human experience.nbsp;The more nearly this play approaches the gravity of tragedy,nbsp;the more insistently it challenges consideration of the questionnbsp;whether it might not properly be classified with that speciesnbsp;of drama extensively cultivated in the nineteenth century andnbsp;found particularly adapted to the debate of problems, the speciesnbsp;which the French call drames and the Germans Schauspiele.nbsp;Such a piece is Sudermann�s �Schauspiel�, Die Ehre � immeasurably inferior to Minna von Barnhelm, but not altogether dissimilarnbsp;in subject Another such is Kleist�s Prinz Friedrich von Homburg, which Hebbel denominates a tragedy notwithstanding itsnbsp;happy outcome, but with which I am in the habit of associatingnbsp;Minna von Barnhelm. As to frankly laughable scenes, these, innbsp;full accord with Lessing�s theories �� and with Shakespeare�snbsp;practice �� may have their place in any drama. But, after all,nbsp;classification is a question only of convenience. It is pedantry

Sudermann would have us believe not only that �honor� has at best no more than a social sanction, but also that each social class has its own conventionsnbsp;in the matter. Cf. note 14 above.

and routine, teste Grillparzer that judges by categories; true artistic sense knows only individuals. As a true and seriousnbsp;comedy, then, Minna von Barnhelm is one of not more than fournbsp;in all German literature, the first in time and in quality � suchnbsp;perhaps as could possibly be achieved even by an unusuallynbsp;competent author only when an unusually fortunate conjunctionnbsp;of circumstances made ready to his hand a matter so actual, sonbsp;national, so capable of transcendent purposes.

If one undertakes to speak of an excellent work of art, it is almost necessary to speak of the whole of art; for it comprizes the whole; and each of us can, tonbsp;the extent of his ability, develop the general from such particular example.

Thus Goethe^�; and the principle so enunciated must be my excuse for dwelling so long upon a single work of Lessing�s.nbsp;If the whole of art is comprized in it, so also may the wholenbsp;artist be therein discoverable. Other points of view than thosenbsp;taken would be obvious and would be requisite to an exhaustivenbsp;discussion; other works doubtless exhibit some of the author�snbsp;qualities more conspicuously; and qualities to which we have notnbsp;adverted might in all probability also be illustrated from thenbsp;drama so characteristically his and so unmistakably indicativenbsp;of a turn for the better in the course of German literature. Butnbsp;my purpose has been limited in respect both to the play and tonbsp;the pla5rwright. If our examination has correctly determinednbsp;the significant aspects of the one, and thereby put in evidence the salient features of the other, we may at last dismissnbsp;our witnesses and concurrently take leave of Lessing himselfnbsp;�� of that author who professed not to be a poet, who wrotenbsp;most abundantly as a critic, but who showed in Minna von Barn-helm for the first, though not for the only time, that superiornbsp;intelligence ennobled by benevolent intent can accomplishnbsp;something indistinguishable from creation. Grillparzer �� fornbsp;all the difference between himself and Lessing � also a humanist, and, by his own confession, both a fantastic poet and anbsp;rationalist of the toughest fibre, handily supplies us with thenbsp;following diagnosis ;

What constitutes the worth of Lessing is the union of artistic sense with logic. To be sure, neither the artistic sense is so pure, nor the logic always so

sound; but in this combination they are perhaps unprecedented; �� in fact, they are ordinarily even mutually exclusive.

For my part, I find no flaws in Lessing�s logic; on the contrary, I find him frequently conducted by logic to revelations usuallynbsp;made only to intuitive perception. That his artistic sense shouldnbsp;be inferior to Grillparzer�s would seem not unnatural in onenbsp;deficient in imagination. In him we discover no tumultuousnbsp;flights of fancy; his eye never rolled in a fine frenzy. But neithernbsp;was he prosily sober. Far from it; his very prose is alive in everynbsp;turn, and alert reason is what made it so. There is the flavor ofnbsp;an ancient rhetorician�s rule in this downright declaration ofnbsp;his

He who correctly reasons also invents, and he who wishes to invent must be able to reason. Only those believe that one can be separated from the other whonbsp;are not qualified for either.

We should, however, be doing Lessing an injustice in the very act of glorifying the clarity of his intellect. He could reason;nbsp;but he was also the happy man that getteth understanding, andnbsp;the still happier man that findeth wisdom; for wisdom is a moralnbsp;acquisition � it is knowledge tempered by good will.

Berkeley abounded in good will and was wise, but his grandiose scheme for �planting arts and learning in America�� does littlenbsp;credit to his practical understanding, or common sense. For onenbsp;thing, arts and learning had been �planted�� here by the firstnbsp;colonists � it would have been more expedient to foster thenbsp;gardens already under cultivation than to clear the ground for anbsp;new one six hundred miles off-shore. The experience of sixnbsp;generations demonstrates this for the white settlers; �the coursenbsp;of empire�� dealt in its own way with the red men. As to thenbsp;multitudinous academic establishments of the present day, theynbsp;have more to learn from the critical scholarship of Lessing thannbsp;from the idealistic philosophy of Berkeley. Both men, be it saidnbsp;at once, were shining examples of devotion to unselfish, reasonablenbsp;service. Berkeley, however, is an historical figure, whereas Lessingnbsp;speaks to us almost as a contemporary. This is largely because ofnbsp;the eminently personal tone of all his utterances � the mannbsp;everywhere stands forth greater than the sum of his achievements.nbsp;But Lessing�s humane learning is especially fit to provide for us,nbsp;as it provided for him, a corrective of the individualistic and

particularistic vagaries to which the idealist is exposed and from which, we may add, the materialist seldom escapes. Whichevernbsp;we are, we need an anchorage in information; in order to masternbsp;the present, we need to make the past live again, as he knew sonbsp;well how to make it; and if peradventure we also invent, wenbsp;still shall find, as he found, in the old the most reliable guidenbsp;to the new.

REPETITIONS

AS AN ELEMENT IN LESSING�S WORKS

The creative writer, endowed as he must be with an imagination that is at once plastic and agile, broadens his outlook, if he doesnbsp;not basically modify his point of view, with increasing experience.nbsp;He evolves in that he abandons the immature and championsnbsp;the mature. Despite the surface difference, for example, betweennbsp;the mood or matter of Kleist�s Hermannsschlacht and that ofnbsp;Prinz von Homburg, it becomes obvious, even to the undergraduate mind, that the same genius wrote both plays. In thisnbsp;there is nothing unique; for it is always possible to trace anbsp;direct connection between the various and successive works ofnbsp;a great author.

It is only rarely possible, however, to locate so many complete repetitions as occur in the works of Lessing. In Shakespeare therenbsp;is a similarity in tone and technic between any two of his comedies,nbsp;or tragedies, or historical plays; the search for actual verbal ornbsp;conceptual repetitions will lead nowhere. Goethe repeated ^nbsp;himself in the letter to Kestner regarding the liberal young lovernbsp;in Werther and in the Gretchen catechization scene in Faust]nbsp;search for repetitions elsewhere in Goethe will be idle. Such mennbsp;as Thomas Hardy ^ and Jakob Wassermann � repeat but withnbsp;them it is an affair of general atmosphere. The same applies tonbsp;Knut Hamsun. With negligible exceptions his major works arenbsp;very much alike. Lessing, on the other hand, overworked certainnbsp;cardinal words, repeated pivotal expressions, exploited favorite

settings in his dramas, and drew again and again on set pictures for illustrations in his critical writings. With the possible exceptionsnbsp;of Hebbel and Schnitzler, who are also given to repeating themselves. Lessing�s hardened and persistent habit in this regard isnbsp;unique, while the silence on the subject that has thus far obtainednbsp;on the part of those who have admired Lessing greatly andnbsp;studied him diligently is without easy or immediate explanation.

In his Lessing, Erich Schmidt points to the connection between Die Juden and Nathan, speaks of the Tempelherr as �ein j�ngerernbsp;Bruder Tellheims�, refers to Nathan as �der zw�lfte Anti-Goeze�,nbsp;and maintains that only Lessing could have written Ernst undnbsp;Falk, Beweis des Geistes und der Kraft, and Erziehung des Men-schengeschlechts. There is nothing in Erich Schmidt�s chapter onnbsp;style and language that would reveal the fact that he was innbsp;any way impressed by what remains a really striking feature ¹nbsp;of Lessing�s works. The case of Heinrich Bulthaupt � is wellnighnbsp;bewildering. There can be no doubt of his thorough knowledgenbsp;of Lessing�s dramas. He finds it �auffallend� that there is anbsp;similarity between Miss Sara Samfson (11, 8) and Emilianbsp;Galotti (IV, 5), and between the close of Acts I and III of Minnanbsp;von Barnhelm, I and II of Miss Sara, and III and IV of Emilianbsp;Galotti. The similarities to which Bulthaupt refers exist; but theynbsp;are not as noticeable as many others which have hitherto beennbsp;ignored. Indeed some German writers seem to have dreaded thenbsp;thought of emphasizing Lessing�s repetitions, whether of words,nbsp;dramatic settings, ideas, or critical illustrations. Hans Altmann �nbsp;writes; �Man kame also in Versuchung, was unser Problem betrifft,nbsp;eine Parallele zwischen der Sara Sampson und Emilia Galottinbsp;zu ziehen�. The quot;temptation� in this case is an idle one; Lessingnbsp;repeated himself much more frequently than any other great, ornbsp;classical, German writer.

Lessing himself was aware, to a degree at least, that he indulged in repetitions. In his Beweis des Geistes und der Kraft he wrote �:

Under this caption cannot come such a writer as Paul Albrecht who, in his Leszing�s Plagiate, 5 Bde., Hamburg u. Leipzig, 1888�1901, was not evennbsp;generous enough to allow Lessing the established spelling of his own name.nbsp;It is in a way fortunate that the author never turned to Lessing�s �plagiarisms�nbsp;from himself, otherwise his voluminous study would have been greatly increased.

� Vol. XVIII, p. 29, of the Cotta edition, the edition which, for convenience sake, is referred to consistently. No attempt is made tliroughout the paper to

�Und so wiederhole ich, was ich oben gesagt, mit den namlichen Worten�. In one of bis Gegensatze he wrote (XVII, 276): �Ichnbsp;muB bekennen, daB ich von einigen Gedanken dieses Aufsatzesnbsp;bereits w�rtlich Gebrauch gemacht habe . . . Die Indiskretion . . .nbsp;weiB ich zu verantwortenlt;(. In his Gedanken �her die Herrenhuter ¹ ²nbsp;he wrote (XVII, 16): �Man schreibt, wie man denkt; was mannbsp;an dem geh�rigen Ort ausgelassen hat, holet man bei Gelegenheitnbsp;nach; was man aus Versehen zweimal sagt, das bittet man dennbsp;Leser das andre Mal zu �bergehen�. He had too the idea ofnbsp;repetition in mind when, in his Theatralische Bihliothek, henbsp;wrote (VII, 12): �Die Regel, daB man das, was bereits getannbsp;ist, nicht noch einmal tun sollte, wenn man nicht gewiB wiiBte,nbsp;daB man es besser tun werde, scheint mir so billig als be-quem.�

Before attempting to enumerate and explain a representative list of Lessing�s repetitions, a few words of warning are indispensable : the greatest of German scholars contradict each other as tonbsp;the way in which Lessing�s mind grew; and he himself made anbsp;number 01 statements about his own method of working and hisnbsp;own style, that ill accord with the established facts. Moreover,nbsp;Lessing was given to obvious banter, so that it is not alwaysnbsp;easy to know precisely where he stood.

At the anniversary of his death in 1881, Wilhelm Scherer� came out with his studied survey and initiated the custom ofnbsp;dividing Lessing�s active or creative life into sharply staked-offnbsp;periods; but if we study his creative works along with his criticalnbsp;ones it becomes obvious at once that there were no such quot;periods�,nbsp;for they connote, and make imperative, breaks in his life. Lessingnbsp;evolved with smoothness and a continuity as but few writers.nbsp;Heinrich von Treitschke � contended that Lessing was quitenbsp;indifferent to his works once he had written them. It is a misleading assertion. Lessing would write a work and then seeminglynbsp;abandon it but only because he somehow knew that he wouldnbsp;eventually return to the same theme, do it over, and makenbsp;something immortal out of it. His Juden (1749) and Nathan

list the places in which Lessing quoted himself literally and insisted that his publisher reprint his quotation.

* nbsp;nbsp;nbsp;Lessing said the same thing in his first Anti-Goeze (XVIIl, 136).

� Cf. Historische und politische Aufsdtze, 6. Aufl., Leipzig, 1885, vol. i, PP- 56�74-

(1779) constitute a case in point. Gustav Kettner contended that he hurt himself by beginning to publish when he had verynbsp;little to say, and that he did not have much to say before he hadnbsp;passed the age at which Goethe and Schiller first acquired lastingnbsp;fame. The truth is, Lessing developed very young; the ideasnbsp;which were to make him famous were virtually all in his mind,nbsp;in immature form, when he was thirty-one years of age, in 1760.

Then there was his lifelong habit of saying things apparently in order to throw people off their guard. In his personal lettersnbsp;he derided without ceasing the bookish person; yet his privatenbsp;library contained 6,000 volumes in 1760. He himself lived withnbsp;books; yet he lampooned in his dramas the individual who soughtnbsp;the company of books rather than of men. In his tv/elfth Literatur-brief he wrote: �Man prahlt oft mit dem, was man gar nicht hat,nbsp;damit er es wenigstens zu haben scheine�. Precisely the samenbsp;idea is expressed in Minna.

An amusing illustration of his unwillingness, or his unconscious inability, to see himself as others saw him, is in his Rettung desnbsp;Cardanus. He has become excited; he writes: �Wenn ich ein Mannnbsp;von Ausrufungen ware, so w�rde ich mich ganz und gar darinnenbsp;ersch�pfen. Ich wiirde mit manchem O! und Ach! zu verstehennbsp;geben, daB . . .� But this is exactly what he did throughout hisnbsp;life. His use of exclamations, interrogations, and dashes makesnbsp;one of his pages look like that of a. contemporary Expressionist.

Although an explanation of his indulgence in repetitions falls quite beyond the scope and purpose of this paper, it may be thatnbsp;it was his very vigor, his belief in himself, his wish to drive anbsp;few thoughts home rather than break a lance for a great numbernbsp;of thoughts, that led him into the habit of repeating himself.nbsp;In his Philosophischer Nachlafi he stated (XIX, 224) that it wasnbsp;much better to know one thing than several, provided one werenbsp;certain that the thing in question was really known. Despite hisnbsp;seemingly wide range of interests, Lessing was not a man ofnbsp;marked versatility. In ethics, his one best and greatest thoughtnbsp;was that character existed long before the laws that are supposednbsp;to control character were formulated. In art, he spent his lifenbsp;trying to show that a man cannot use the tactics in writing that

Cf. Lessings Pddagogik, dargestelU auf Grund seiner Philosophie, by Georg Mann, Langensalza, 1894.

are essential to success in the others arts. He worked these two ideas over and over, and in so doing repeated himself with anbsp;frequency that militates against pronounced claims to originality

Lessing�s theories of art are not germane to the present subject. As to his idea that the thing comes before the classification � it was a part of his noble plea for tolerance � one illustrationnbsp;out of many must suffice. He wrote (XV, 107): �Es hat unzahligenbsp;Dichter vor dem Martial bei den Griechen sowohl als bei dennbsp;R�mem gegeben, welche Epigramme gemacht, aber einen Epi-grammatisten hat es vor ihm nicht gegeben. Ich will sagen, daBnbsp;er der erste ist, welcher das Epigramm als eine eigene Gattungnbsp;bearbeitet . . .�

The works on Lessing�s language with the exception of Lehmann to be noted later, confine themselves to themes thatnbsp;concern in no way the present investigation, which has to do,nbsp;particularly at this point, with Lessing�s vocabulary This wasnbsp;not large. He himself wrote in Eine Duplik (XVIII, 58): �Mirnbsp;graulet, eine Menge unn�tiger Worte machen zu m�ssen�. Bynbsp;�Worte� he really means �W�rter�; the distinction was rarelynbsp;observed by him. In his Rettung des Horaz he wrote; )gt;Parcusnbsp;selten ? Und infreqnens auch selten ? So verschwenderisch mit dennbsp;Worten ist Horaz schwerlich gewesenlt;(. Both statements arcnbsp;apposite although, as has been stated above, caution is alwaysnbsp;in place when Lessing comments on what someone has or does.

But taking him at his word, his vocabulary is limited, and the limitation is revealed first of all in his verbal repetitions. Anynbsp;vigorous and voluminous writer has his favorite expressions; wenbsp;think of Goethe�s fondness for �gelassenlt;(. But here is the point:

Cf. studi�n zti Lessings Stil in der Hamburgischen Dramaturgie, by Max R von V/aldberg, Berlin, 1882. This study deals only with repetitions for emphasis,nbsp;as when Lessing wrote: �Der Bediente hat ihn gerettet; dem Bedienten gehortnbsp;das erste Wort�. Martin Schiitze published a somewhat similar treatise in 1907.

His theory of art in Laohoon might be discussed to this extent: The one theory he laid down in Laohoon could be expressed in a single paragraph; whatnbsp;he did in this work of approximately 50,000 words was to repeat illustrationsnbsp;rvhich, in his judgment, confirmed his theory.

Cf. Sprachliche Kleinigkeiten zu Lessings Jugendwerhen, by W. Creizenach in Zeitschrift f. d. Wortf., StraBburg, 1901, pp. 31�32. The article deals largelynbsp;with Lessing�s attitude toward the use of Latin as opposed to German words.

A vocabulary for Nathan would with negligible exceptions serve also as a vocabulary for Minna-, but a vocabulary for Herm. u. Dor. would not be ofnbsp;great service to the student of Clavigo.

Goethe uses �gelassen� in Hermann und Dorothea and also in Clavigo, two works that bear not a shred of similarity to eachnbsp;other. Lessing on the other hand never wrote works that seem tonbsp;contradict each other in tone and theme.

The observing reader cannot go far in Lessing without being struck by his exploitation of certain pivotal words, as opposednbsp;to mere particles that have no particular signification. We thinknbsp;of )gt;erraten�, �schlechterdings�, �rechtschaffen�, and �retten� i�.nbsp;The word �schlechterdings� � occurs thirteen times in Wie dienbsp;Alten den Tod gebildet (1769), a pamphlet of forty (Cotta) pagesnbsp;at least seven of which are taken up with illustrations and directnbsp;quotations from the Latin and Greek. There is no question butnbsp;that Lessing used the word effectively, as when he wrote (thenbsp;passage is relevant): �Zwar muB ich gestehen, daB ich damals, alsnbsp;ich den Ort im Laokoon schrieb, schlechterdings keine Auslegungnbsp;kannte . . . �But there are least a dozen German words thatnbsp;would have served as adequate synonyms.

His attachment to the word �retten� is even more striking. He was under no obligation to use the word �Rettung� in connection with all those vindications which he wrote of men who, in hisnbsp;judgment, had been wronged, from Simon Lemnius on. He couldnbsp;have varied his captions and produced the same, if not evennbsp;greater effect. But he loved the idea embodied in �retten�. Hisnbsp;greatest �Rettung� is Nathan. In it he endeavored to write anbsp;�Rettung� of a great idea. And impelled to frame an adequatenbsp;reply to the embarrassing question Saladin had raised regardingnbsp;the comparative worth of the three great religions, Nathan,nbsp;recalling the story of the rings, exclaims: �Das wars! Das kannnbsp;mich retten!� The drama might have been called �eine dramatische Rettung�. But even Lessing seems to have gro^vn tired

Lessing never used, seems indeed never to have known, the words static and dynamic, as Fichte, and other Romanticists, made them popular. He believed,nbsp;however, that everything is dynamic, that is, capable of being changed, exceptnbsp;death. In Die Matrone von Ephesus (V, no) Antiphila says: �Aus den eisernennbsp;Armen des Todes ist keine Rettung�.

Cf. Forschungen Uher Lessings Sprache, by August Lehmann, Braunschweig, 1875. Lehman writes of �schlechterdings�: �Gehort zu den auserkorensten allernbsp;Lieblinge Lessings�. He points to �gegen 150� examples. There are more. Henbsp;says nothing of �rechtschaffen�.

In his Untersuchungen iiber Ramlers und Lessings Bearheitung von den Sinngedichten Logaus, Leipzig, 1903, W. Heuschkel takes the stand that Lessingnbsp;wrote a �Rettung� of Logau just as truly as of anyone else.

of certain of his own repetitions. In the 77th Literaturbrief he wrote: �Und wie oft werde ich dieses abermals, aberxnals brauchennbsp;miissen!�

Lessing revealed however the trait under discussion niost strikingly in his use of �rechtschaffen� There is no case onnbsp;record among writers of the first class like it. It occurs in all ofnbsp;his works, creative, critical, and personal or private. He usednbsp;it five times in Der Schatz (1749). Sometimes he used it merelynbsp;in the sense of �ehrlich�, more frequently in the sense of �gediegen�nbsp;or �tiichtig�, �griindlichlt;( or �groBmiitig�, and occasionaly innbsp;the sense of �geh�rig� or �recht�. On December 3, 1772, Henbsp;wrote to Eva K�nig: �. . . so m�ssen Sie ein Ende damit zunbsp;machen suchen. Auf die rechtschaffenste Art; das versteht sich:nbsp;aber nicht auf die skrupul�seste.� In the 49th Literaturbrief henbsp;discussed the question: �Kann man ohne Religion rechtschaffennbsp;sein ?� The word occurs there nineteen times in three pages.

Lessing�s most informative use of �rechtschaffen� however is in his translation of Plautus� Captivi He regarded Plautus asnbsp;the model writer of comedies and Captivi as his best work. Henbsp;even rose to such heights of admiration as the following (VI, 143):nbsp;�Ich bleibe also dabei, daB ,Die Gefangenen' das sch�nste Stiicknbsp;sind, das jemals auf die B�hne gekommen ist, . . . weil es dernbsp;Absicht der Lustspiele am nachsten kommt und mit den �brigennbsp;zufalligen Sch�nheiten reichlich versehen ist�. He Was to be surenbsp;very young when he wrote this. In the translator�s note he tellsnbsp;how carefully he followed the original, reproducing, wherenbsp;possible, even the stubborn puns of Plautus. Then we read hisnbsp;translation, in prose. The word �rechtschaffen� occurs ten times.nbsp;What Latin terms did he render by it ?

In his first use of it, there is no direct Latin equivalent: talem adulescentem is rendered by �so einen rechtschaffenen J�ngling�.nbsp;There can be no objection to this: talis is a colorless word. We maynbsp;translate talis homo by �such a good man� or �such a bad

The Deutsches W�rteybuch, which devotes three columns to �rechtschaffen� says it was common (�haufig�) in the i6th Century. Since words of this sort,nbsp;as contrasted with slang or technical expressions, rarely come into use suddenly,nbsp;it may be inferred that it was known in the fifteenth Century. Luther used it,nbsp;though Grimm does not cite cases in which it was used by Lessing's more important contemporaries.

The text of the Captivi used is that of E. P. Morris, Ginn amp; Co., Boston, 1898.

Cf. Lessings Ansichten von der deutschen Sprache, by K, Behschnitt, Breslau, 1915. Behschnitt discusses Lessing as a translator.

man�, depending on the context; and Lessing is accurate: the youth referred to in this case was �good�.

In the second case �rechtschaffen� is used adverbially: sed utrum strictimne adtonsurum becomes �so wird er ihn rechtschaffennbsp;zerkratzen�. In the third instance, hominem unmodified is translated �rechtschaffener Mensch�. The fourth case is literal enough:nbsp;optnmusque hominum es homo becomes �du bist der recht-schaffenste Mann�. The fifth and sixth eases are both appropriatenbsp;contextual forms of bonus. The seventh likewise: htiic homininbsp;optumo is done into �diesem rechtschaffenen Manne�. In thenbsp;eighth case it is again used adverbially: irrogabo multam becomesnbsp;�ich will sie rechtschaffen strafenlt;'. Likewise in the ninth: satin,nbsp;containing the root of �genug�, becomes �rechtschaffen�. Thenbsp;tenth and last case shows Lessing�s accurate freedom: quando egonbsp;te exemplis pessumis cruciauero is translated �wenn ich dich rechtschaffen habe martern lassen�. There is no thought of accusingnbsp;Lessing here, in any of these cases, of inaccuracy; but there arenbsp;numerous s5mon5nns for �rechtschaffen�, and a careful writernbsp;avoids so far as possible the betrayal of either verbal prejudicenbsp;or mental inertia.

In the matter of rhetorical figures and dramatic settings, the present writer is not minded to accuse Lessing of either of thesenbsp;weaknesses; but he did take it easy. Having worked out a scene ornbsp;situation in one drama he exploited his antecedent labors in thenbsp;following drama. Of this there can be no doubt, as a selected list ofnbsp;illustrations will prove.

The least important of Lessing�s finished dramas, Damon (1747) not excepted, is Die alte Jungfer (1749). In this comedy �das leidigenbsp;Geld� plays a large role. We think of Minna and Nathan. The oldnbsp;maid�s friend is referred to as a �Verschwender�, and the noun isnbsp;used here in much the same sense as later in connection with Tell-heim. The word ))rechtschaffen� also occurs three times. And J. C.nbsp;Edelmann, the theologian who had the intrepidity, at the very beginning of the eighteenth century, to doubt whether the entirenbsp;Bible was inspired and whether every word of it should be acceptednbsp;literally, is given a place of honor within the play. These mightnbsp;be mere coincidences; but the following cannot be.

The question has arisen as to who will make the best husband for the old maid. A certain Kapitan von Schlag is in the offing.nbsp;After years of military experience he has been given an honorable

discharge because of a wound; he can no longer engage in foreign service. He stands in high repute with the King, however. He isnbsp;Tellheim.

But the old maid has her doubts. Herr Oront thereupon replies: �Verlangen Sie denn einen Mann, der stets zu Felde liegt ? . . .nbsp;Die abgedankten Offiziere sind die besten Ehemanner; wenn sienbsp;ihren Mut nicht mehr an den Feinden beweisen k�nnen, so sind sienbsp;desto mannhafter gegen ihre . . .� The word to be supplied isnbsp;�Frauen�. Compare this with Minna. Tellheim has made out quitenbsp;a case for his unfitness as a husband; he is a cripple. Minna says:nbsp;�Ein SchuB hat Ihnen den rechten Arm ein wenig gelahmt. Dochnbsp;alles wohl iiberlegt, so ist auch das so schlimm nicht. Um so vielnbsp;sichrer bin ich vor Ihren Schlagen.� Two speeches precedingnbsp;this Minna rejoices in the fact that Tellheim has been discharged;nbsp;she can have him entirely to herself. Herr Oront advises the oldnbsp;maid, in case she is determined to marry an officer, to many anbsp;discharged one, otherwise she will see but little of her husband.

Take the case of Philotas (1759). It is not like the Lessing we know; Heinrich von Kleist might have written it: it is patriotic.nbsp;And it is a �Rettung�. Can Philotas, a prisoner of the enemy,nbsp;accept the liberty that has been offered him without jeopardizingnbsp;the cause of his father�s forces ? He cannot; so he takes his ownnbsp;life. There are three themes in the play that Lessing expanded,nbsp;with striking literalness, in Minna and Nathan] the naturalnbsp;inability of those in high power to reward all men according tonbsp;their merits, the significance of family resemblances, and thenbsp;approach of Parmenio, the soldier, to Philotas in his cell. Philotasnbsp;takes the same attitude, at first, toward his visitor that thenbsp;Tempelherr and Saladin take toward Nathan in their initialnbsp;interviews.

Let us look rather closely at Minna and Nathan. Fifteen years lay between the writing of these two plays. In both a war is justnbsp;over. In Minna the two outstanding kingdoms of Germany arenbsp;reconciled; in Nathan the three outstanding faiths are reconciled.nbsp;In Minna there are two rings, and it makes but little differencenbsp;which one either of the leading characters wears so long as it isnbsp;worn in the belief that it is genuine. In Nathan there are threenbsp;rings and all the world knows the story that is written about

The description of the stones in the ring is based on, or parallel to, Lessing�s discussion of this subject in Briefe antiquarischen Inhalts.

them. Money is now a much needed and now a despised thing in both plays. Paul Werner will go to the Orient and take part innbsp;whatever war Prince Heraklius is waging. In Nathan the charactersnbsp;are already in the Orient, and Heraklius was Patriarch therenbsp;at the time. The Uncle in Minna arrives before it is too late.nbsp;The Uncle in Nathan was the cause of the clearing up. In Minnanbsp;Riccaut gambles. In Nathan Saladin and Sittah play chess. Bothnbsp;recreations were indulged in, harmlessly, by Lessing. In Minnanbsp;the titular heroine says to Tellheim: �Der K�nig war eine un-gliickliche Karte fiir Sie: die Dame wird Ihnen desto giinstigernbsp;sein.� This figure from the card table awas taken over into Nathannbsp;and adapted to the chess board.

A quite typical similarity of a broader nature, and one that does not become apparent without a measure of thought, is thenbsp;rescue of the poodle by Just in Minna and the rescue of Rechanbsp;by the Tempelheir in Nathan. Just is walking along the canalnbsp;one winter day when he hears a voice (the Stimme motif is commonnbsp;to both plays). He goes down in the direction of the voice andnbsp;saves a dog from drowning: �Es ist ein haBlicher Pudel, aber einnbsp;gar zu guter Hund.� Mistreat that dog as he may, it refuses tonbsp;leave him. It is precisely so that the Tempelherr rescues Rechanbsp;from the fire. He is not proud of his deed, for she is only a �Juden-madchen�, but she is a �Mensch� and should therefore be saved.nbsp;He scorns her, but the more intense his hatred the less she isnbsp;minded to leave him. In the end he comes to love her. Contrastnbsp;this with the idea regarding the acceptance and recognition ofnbsp;gratitude expressed in the foreword to Lessing�s Sophokles'. �Ichnbsp;kann nicht bewundert werden, aber ich werde Dank verdienen.nbsp;Und die Vorstellung, Dank zu verdienen, muB eben so angenehmnbsp;sein als die Vorstellung, bewundert zu werden.�

Just in Minna proposes to Werner that they waylay the Wirt as he comes from the �Tabagie� some evening and assassinatenbsp;him, set fire to his house, or ruin his daughter. Werner will listennbsp;to no such schemes. The Klosterbruder in Nathan, followingnbsp;instructions from his hated chief (Lessing could never work withnbsp;ease for or under someone else), proposes to the Tempelherrnbsp;that they waylay Saladin some evening and dispose of him fornbsp;good and all. The Tempelherr is incensed at the thought of suchnbsp;a crime.

manner in which she gives them to him, with saucy impudence and more in detail than he had cared for, finds its parallel in thenbsp;Tempelherr�s detailed reply to the Klosterbruder in Nathan. Thenbsp;curt replies given Daja by the Tempelherr are a reproductionnbsp;of the replies that Just gives in Minna.

In each drama a friend�s friend is also a friend; the wish to pray is a prayer in itself; there is someone who cannot read ornbsp;who claims this inability; someone �hat Galle�; the significancenbsp;of friendship on short acquaintance is commented on; the principle of Interessen or Zinsen is set forth; the related roles playednbsp;by the head and the heart are discussed; in each play there isnbsp;an eavesdropper who is badly repaid for listening-in; the )gt;Reit-knecht� is present in each play and in each Persia is in the background; the Tempelherr stands outside and refuses to enternbsp;Nathan�s house precisely as in Minna Just disrelishes the thoughtnbsp;of re-entering the hotel; Minna is dressed )gt;v�llig und reich, abernbsp;mit Geschmack� while Nathan describes the things he has broughtnbsp;back with him from Babylon as �so reich, und mit Geschmack sonbsp;reich�; in Minna the �kostbarer Ring� had been received �vonnbsp;lieben Handen� while in Nathan the original ring was �aus liebernbsp;Hand� ; the way the heroine tries to put herself in the right moodnbsp;to receive someone who was loved long since and lost a while isnbsp;identical in both plays, as is the manner in which someone isnbsp;sent for by a certain character who may be in a position tonbsp;influence the speedy arrival of the desired person.

Then there is Emilia Galotti. It is admittedly a new type of play: there is in it an element of fatalistic chance Of it Erichnbsp;Schmidt says; �Als Komposition ist es die auBerste Leistungnbsp;eines peinlich rechnenden Kunstverstandes, mit eisernen Klam-mern festgebunden, Szene auf Szene festgemauert, . . . kein Wortnbsp;entbehrlich, -^eil nur das unentbehrliche gesagt wird.� If this benbsp;true, we should expect to find Emilia wholly different, in wordingnbsp;and setting, from Lessing�s other plays. This is not the case,nbsp;as some illustrations will show.

The widow Marloff in Minna says: �Sie wiirden sein letzter Gedanke, Ihr Name der letzte Ton seiner sterbenden Lippennbsp;gewesen sein.lt;( Claudia says in Emilia: �Marinelli war das letzte

^ Cf. Lessings Dramen, by Gustav Kettner, Berlin, 1904, pp. 230�237. Kestner�s contention cannot be refuted, but after all chance, mere quot;Zufall�,nbsp;plays a heavy role also in Minna and Nathan.

Wort des sterbenden Grafen.� At the opening of the play, the Prinz �will ansfahrend, accompanied by Marinelli; in Minnanbsp;Minna �wi]l ausfahren� accompanied by Tehheim. The Grafinnbsp;Orsina comes to town very much as Minna did and for a somewhatnbsp;similiar purpose. The �Schatzmeister� plays a role in Emilianbsp;not wholly dissimilar to the one he plays in Nathan. The Prinznbsp;is impressed by the smile on Emilia�s lips; the Tempelherrnbsp;discusses the same theme in Nathan. And in Emilia, as innbsp;virtually all of Lessing�s dramas, the characters quote eachnbsp;other, somewhat after the fashion of the cut-back in the motionnbsp;picture.

In Emilia, Claudia bemoans the �rauhe Tugend� of her husband much as Franziska and Minna bemoam the same trait in the mennbsp;in Minna When Emilia refers to the Prinz as �ihn selbst�,nbsp;she does precisely what Nathan and Daja do: �Bei ihm? Beinbsp;welchem Ihm?lt;( Emilia tells Appiani how to dress exactly asnbsp;Minna coaches Tellheim on this point. There are dreams in all ofnbsp;Lessing�s leading dramas, though Lessing himself is alleged tonbsp;have passed dreamless nights. Marinelli�s attempt to have Appianinbsp;sent away on an important mission reminds of the services thenbsp;Patriarch wished to have rendered by the Tempelherr. The takingnbsp;off of Appiani is the actual carrying out of the plan, proposednbsp;but rejected, inMwnawith the Wirt and in Nathan with Saladin.nbsp;Marinelli insists that the Prinz has not read Orsina�s letter;nbsp;Minna and Franziska stage a similar scene. Orsina�s referencenbsp;to a happy, laughing �Gesch�pf� finds its counterpart in Minna,nbsp;where Minna contends that a happy creature is the most pleasingnbsp;object that can be thought of. Orsina catches Marinelli in a lienbsp;just as Tellheim catches Werner in a lie and in both cases thenbsp;word ))ertappen� is used. This however was a common verb withnbsp;Lessing.

At the beginning of the last act, Marinelli tells the Prinz that Odoardo can be seen from the window, going up and down thenbsp;arcade, turning now to the right and now to the left. It is a scenenbsp;that is repeated literally in Nathan in which Daja and Rechanbsp;watch the Tempelherr. And throughout the entire Emilia the

The present writer regards it as quite remarkable that Otto Spiess could write such a book as his Die dramalische Handltmg in Lessings �Emilia Galotliquot;nbsp;und �Minna von Barnhelmquot;, Halle, 1911, and never say a word about repetitions,nbsp;for this phase of Lessing�s work has a direct bearing on Spiess�s theme.

same expressions occur: �mein rechtschaffener Galottk, �ein unbekannter Freund ist auch ein Freund�.

As to Miss Sara Sampson, it is obvious that Mellefont is the prototype of the Prinz and Marwood of Orsina. The Wirt of Missnbsp;Sara is the prototype of the Wirt in Minna] both use the samenbsp;expressions. Sir William comes to the hotel �aus den recht-schaffensten Absichten�. Sara, like the Prinz in Emilia, hasnbsp;nothing but �Klagen� to report to Mellefont who asks her, precisely as it is done in Emilia by Marinelli in his interview withnbsp;Appiani, why she should be so disturbed over the �Verschiebungnbsp;einer Zeremonie�, the ceremony being the wedding. And throughout the entire play it is the setting familiar to the student of thenbsp;older plays: �Vernunft� and �VerMand�, the �Bilder� created bynbsp;the �Phantasie�, �Tugend� and �Laster�, the bad role playednbsp;by �Schwarmerei�, the place in gallantry of �Schmeichelei�, andnbsp;the distress caused a father by the fact that nothing separatesnbsp;him from his endangered daughter but �eine Wand�, a situationnbsp;that re-occurs in Nathan.

Then there are Lessing�s critical works. An analysis of them, by way of showing how his mind evcdved with regard to thenbsp;working out of the few truths that interested him, is quite outnbsp;of the question: space makes such an analysis impossible whilenbsp;the aim of this essay makes it unnecessary. Let us again merelynbsp;enumerate a typical list of obvious repetitions.

As has already been indicated, Lessing was interested in the truth that the species precedes the class. In 1776 he exclaimednbsp;(XIX, 136): �Baum ist doch sicherlich alteren Ursprungs alsnbsp;Eiche,Tanne,Linde.� He proved in Nathan that Menschnbsp;came first and was then followed by such classifications as Jude,nbsp;Christ, Mohammedaner. When twenty years of age. May 31, 1749,nbsp;he wrote to his father: �Die christliche Religion ist kein Werk,nbsp;das man von seinen Eltern auf Treue und Glauben annehmennbsp;soll.� The Mohammedan says precisely the same to the Jew innbsp;Nathan.

Lessing�s most interesting use of �'rechtschaffen� in connection with a given individual is in Laohoon (X, 134): �Der weise und rechtschaffene Aesopnbsp;wird dadurch, dafi man ihm die HaBlichkeit des Thersites gegeben, nicht lacher-lich.� Moreover, there can be no doubt but that his general interest in the wordnbsp;moved him to write as follows in connection with Logau�s Sinngedichte (VIII,nbsp;204): �Weil Recht ein Knecht itzt ist, dem Frevel hat zu schaffen.�

Und will sie so � so bar, so blank � als ob Die Wahrheit M�nze ware! � Ja, wenn nochnbsp;Uralte M�nze, die gewogen ward! �

It is a world-famous passage, quoted time out of mind, and as though it were an original figure. The truth expressed is, ornbsp;was at the time, largely original. The figure used to express thenbsp;truth was based without the slightest modification on Lessing�snbsp;experience while secretary to Tauentzien in Breslau (November,nbsp;1760 �April, 1765). And in 1769 Lessing wrote (XIII, 192): �Esnbsp;ware doch sonderbar, wenn nur der reich heiBen sollte, der dasnbsp;meiste frisch gemiinzte Geld besitzet.�

It was also in 1749 that Lessing wrote the preface to his elaborately planned work, Beitrdge zur Histone und Aufnahme des Theaters, in which he worked out the very idea that constitutesnbsp;the core of his last major work. Die Erziehung des Menschen-geschlechts; the people as a whole can stand only so much truthnbsp;at a time; it is not safe to go from milk to wine with a child. Innbsp;the same year, 1749, Lessing wrote in Der Freygeist: �Ihm [d. h.,nbsp;dem P�bel] die Religion nehmen, heiBt ein wildes Pferd auf dernbsp;fetten Weide losbinden, das, sobald es sich frei f�hlt, lieber innbsp;unfruchtbaren Waldem herumschweifen und Mangel leiden, alsnbsp;durch einen gemachlichen Dienst alles, was es braucht, erwerbennbsp;will.� The two ideas ^ are so identical that comment is uncessary ;nbsp;and yet there lay no fewer than thirty-one years between thenbsp;writing of them.

There is room for but one more illustration of Lessing�s repetitions bearing on his search after moral truth. In Eine Duplik (1778) he wrote (XIII, 42)nbsp;nbsp;nbsp;nbsp;: �Wenn Gott in seinerRechten alle

� In a letter to his brother Karl, Lessing also wrote: �Das zahme Pferd wird im Stalle gef�ttert und muB dienen, das wilde in seiner W�ste ist frei, verkommt abernbsp;vor Hunger und Elend.�� Goethe uses the same figure in Faust, lines 1830�1833.

The figure the Tempelherr uses (lines 3243�3249) about the block of marble washed on to the shore and the form chiseled from it is a direct echonbsp;from the 29th Stiick of the Hamb. Dram. (XI, 218). A slight variant of the samenbsp;figure occurs in the 40th Brief antiquarischen Inhalts,

Wahrheit und in seiner Linken den einzigen immer regen Trieb nach Wahrheit, obschon mit dem Zusatze, mich immer und ewignbsp;zu irren, verschlossen hielte und sprache zu mir: .V/ahle!' ichnbsp;fiele ihm mit Demut in seine Linke und sagte; ,Vater, gib! dienbsp;reine Wahrheit ist ja doch nur fiir dich allein!�

Few statements of this sort have been more frequently quoted; but Lessing made the same point, and made it effectively, onnbsp;a number of other occassions and in other connections. In hisnbsp;Hypothesen �her die vier Evangelisten, likewise 1778, he wrote:nbsp;�Den wahren Weg einschlagen, ist oft bloBes Gl�ck; um dennbsp;rechten Weg bek�mmert zu sein, gibt allein Verdienst.�

If we attempt to explain this repetition by saying that the figure merely chanced to be on his mind at the time we encounternbsp;grave difficulty, for in his Rettung des Cardanus, written twenty-four years earlier, he said: �Man bediene sich des Gleichnissesnbsp;nicht, daB, wenn man einmal den rechten Weg wisse, man sichnbsp;nicht um die Irrwege zu bekiimmem brauche.� The case is typical:nbsp;with each repetition the expression generally became clearer andnbsp;more striking. Throughout his entire life, Lessing regarded himselfnbsp;as perpetually in quest of the Holy Grail of Truth; for the mannbsp;who contended that he had already found the Grail he had butnbsp;little use. Consequently he wrote in Nathan (lines 2150�2151):nbsp;Ich will mit Mannem lieber fallen, alsnbsp;Mit Kindern stehn.

Just as Lessing wrote a long line of plays in which he depicted a pert maid before this type of character found its culminationnbsp;in Franziska, so did he write four separate works on literature,nbsp;with especial reference to the drama, before he had the matternbsp;culminate in the Harnhurgische Dramaturgie; and before he wrotenbsp;Laokoon he became the author of various shorter studies thatnbsp;culminated in Laokoon. Had he written only the Hamb. Dram.nbsp;his stature, except for the motive of diligence, would hardly benbsp;diminished. His idea in all five of these works was: the Greeksnbsp;are to be studied; the English imitated; the French avoided. As

29 Beitrdge zur Historie und Aufnahme des Theaters (1749), Das Neueste aus dem Reiche des Witzes (1751), Theatralische Bibliothek (1755), and Briefe, dienbsp;neueste Literatur betreffend (1760). In the Brief of May 22, 1760, the word quot;recht-schaffen� occurs fourteen times in four pages and quot;unrechtschaffen� occursnbsp;once. Throughout all of these works �retten� and quot;schlechterdings� occur withnbsp;established frequency, while quot;frischerdings� and �platterdings� occur oncenbsp;each.

to Laokoon and its predecessors, it was a lifelong contention with Lessing that the genres must not be mixed. There is only one placenbsp;in his entire output where he seemed to favor the mixing of typesnbsp;(XI, 301): �Weil der Maulesel weder Pferd noch Esel ist, ist ernbsp;darum weniger eines von den nutzbarsten lasttragenden Tieren ?�nbsp;Lessing had a superb sense of humor; and in this one case he wasnbsp;defending Euripides: �Was geht mich es an, ob so ein Stiick desnbsp;Euripides weder ganz Erzahlung, noch ganz Drama ist.� In thenbsp;Greeks Lessing could see nothing but perfection. But the distinctlynbsp;unusual feature of all these five works is the fact that Lessingnbsp;drew on the same men (Greek, English, French, and German)nbsp;for his illustrations, or he discussed the works of the samenbsp;men. The case is paralleled only by his lifelong references tonbsp;and interest in James Thompson.

There is space for the discussion of only a few important repetitions in Laokoon and the Dramaturgie, such as Milton�s blindness and �Nachahmung der Natur�. As to the latter, it was rathernbsp;carefully treated in Laokoon] but that did not prevent Lessingnbsp;from covering the same ground in the Dramaturgie As tonbsp;Milton�s blindness, Lessing admitted that the pictures of lightnbsp;in Paradise Lost were different from what they would have beennbsp;had Milton been able to see; but he greatly admired Milton�snbsp;ability in this connection. Such cases always interested him. Innbsp;Emilia, Conti exclaims: �Oder meinen Sie, Prinz, dab Raffaelnbsp;nicht das gr�Bte malerische Genie gewesen ware, wenn er ungl�ck-licherweise ohne Hande geboren worden?� Under the captionnbsp;of Johannes Vermaasen, in the Kollektaneen (XX), he wrote ofnbsp;the blind man wLo reached the point where he could differentiatenbsp;colors by the sense of touch. He discussed the possibility ofnbsp;learning to speak if born without a tongue. He took a persistent interest in deaf-mutes. Such cases are so many �Ret-tungen�.

Charles Harris calls attention to this in his edition (Holt), 1906, p. 329. In a note on the 7otli Stiick of the H. D., Harris writes: �The first paragraphsnbsp;of this number, in fact, sound like an echo of that work [Laokoon]The notenbsp;is well founded, although it is strange that Harris was moved to comment onnbsp;this repetition and remain silent about the others. It is strange, for �echoesquot;nbsp;of Laokoon and the Hamb. Dram, can be found in the most out-of-way placesnbsp;in Lessing's works. In Die Matrone von Ephesus (V, 124) Pliilokrates says: �Allennbsp;ziemt nicht alles. Dem Mann, dem Krieger ist eine Thrane verg�imt, aber keinnbsp;Strom von Thranen.� That idea runs through all of Lessing's major critical works.

But there is a much more important phase of the matter than this. It will be recalled that the late William James of Harvardnbsp;created almost a sensation, at the beginning of the present century,nbsp;by his theory of emotions, according to which we do not crynbsp;because we are hurt, rather we are hurt because we cry. Lessing,nbsp;whom William James never mentioned, antedated James onnbsp;this point by nearly one hundred and fifty years. In the 3rdnbsp;Stuck of the Dramaturgic, he made it plain that an actor, if he actsnbsp;as though he were angry, will come to feel and actually be angry.nbsp;In his Auszug aiis dem Schauspieler (1757), he made preciselynbsp;the same point; and in his Rettung des Horaz (1754), he made thenbsp;identical point, apparently for the first time. It is improbablenbsp;that there is another case on record like this. Moreover, the ideanbsp;is exploited in the very heart of Nathan, when Recha says thatnbsp;�Ergebenheit in Gott von unserm Wahnen �ber Gott so ganz undnbsp;gar nicht abhangt�. That is to say, a really pious dispositionnbsp;comes from good acting and not at all from theological speculation.

In Eine Duplik, one of the most valuable theological works Lessing ever wrote, he said (XVIII, 91): �Wer gewisse Dingenbsp;nicht sogleich fiihlt, dem sind sie auf keine Weise fiihlbar zunbsp;machen.� Six years prior to the writing of this, he had Orsina innbsp;Emilia say: olch fiihle so was! �� Und glauben Sie, glauben Sienbsp;mir: wer �ber gewisse Dinge den Verstand nicht verliert, der hatnbsp;keinen zu verlieren�.

In a letter to Moses Mendelssohn (April 17, 1763) Lessing tells the story of the two barbarians who, seeing themselves for thenbsp;first time in a mirror, are impressed by the fact that as theynbsp;move the reflections in the mirror move. They conclude that thenbsp;movements in the mirror and those of their own bodies derivenbsp;from the same cause. Lessing carried this idea in his mind fornbsp;nearly thirty years, and made most excellent use of it in thenbsp;73rd paragraph of the Erziehung des Menschengeschlechts. Probablynbsp;he never fancied that his letters would become literature; itnbsp;may be that he never intended to create the situation that hasnbsp;resulted from the publication of his complete works. There is notnbsp;the slighest reason to believe however that had Lessing lived to

Lessing himself, however, seems minded to refute this argument in the foreword to Sogenannte Brie/e an verschiedene GoUesgelehrten (XVIII, 252).

be a really old man he would have brought out an edition of his works in which the writings of his younger years would have beennbsp;eliminated on the ground that the ideas contained in them arenbsp;found in the later volumes.

In the 49th paragraph of the Erziehung des Menschengeschlechts, Lessing gives his idea of an appropriate style for an �Elementar-buch� that is to be read by �Kinderlt;( and a �kindisches Volk�.nbsp;He says; �Bald plan und einfaltig, bald po�tisch, durchaus vollnbsp;Tautologien, aber solchen, die den Scharfsinn �ben, indem sienbsp;bald etwas anders zu sagen scheinen und doch das namlichenbsp;sagen, bald das namliche zu sagen scheinen und im Grunde etwasnbsp;anders bedeuten oder bedeuten k�nnen.�

With the reading public that Lessing here visualizes, he himself had, directly, very little to do; he wrote neither for �children�nbsp;nor for a �childish race�. He gives, nevertheless, a fair descriptionnbsp;of his own style. It may after all be that he had his own age, hisnbsp;own contemporaries in mind when he insisted that in order to benbsp;understood it was necessary to be plain, picturesque, and repetitive.

On December 16, 1771, he wrote to Eva K�nig; �Lassen Sie uns ruhig sein, und das Beste hoffen, und jeden Augenblicknbsp;nur immer das tun, was Rechtschaffenheit und Klugheit fiir dasnbsp;Gegenwartige von uns fordern. Rechtschaffenheit und Klugheit �nbsp;beide zugleich, meine Liebe!� Lessing was as honest a man asnbsp;Germany knew in his day; and he strove without ceasing notnbsp;merely for personal intelligence but for intelligence on the partnbsp;of his contemporaries.

Conclusion; In 1766, Lessing published Laokoon, the germinal ideas of which can all be found in the various works he hadnbsp;previously written on the general subject of the plastic arts. Innbsp;1769, he published the Hamburgische Dramaturgie, which is anbsp;rounding-up and a practical application of what he had alreadynbsp;written on dramatic criticism and related themes. In 1779, henbsp;brought out Nathan der Weise, which is a condensation of hisnbsp;dramatic practices to that date. In 1780, he finished Die Erziehungnbsp;des Menschengeschlechts, which is a r�sum� of his old ideas on thenbsp;broad subject of human and civic evolution. His contention thatnbsp;the Old Testament was good enough for a primitive people butnbsp;that the New Testament became necessary as mankind grew innbsp;mental and spiritual ability, and that a third Testament would

in time become indispensable is nothing more, than an expansion of the idea set forth in Nathan about the Judge who is to comenbsp;in thousands of years and hand down a decision in the light ofnbsp;such wisdom as will by then have accumulated. And his strangenbsp;excursion, in the Erziehung, into the domain of metempsychosis ?nbsp;That too is a matter of repetition.

WIELANDS BRIEFWECHSEL MIT JOHANNES GOTTFRIED GURLITT

Johannes Gottfried Gurlitt, geboren den ii. Marz 1754 in Halle, gestorben den 14. Juni 1827 als Direktor des Johanneums innbsp;Hamburg, war von 1779 bis 1801 Rektor der Schule des Klostersnbsp;Bergen bei Magdeburg, wo auch Wieland die Jahre 1747�1749nbsp;verlebt hatte. Gurlitt widmete sich besonders dem Studium dernbsp;klassischen und orientalischen Sprachen, und war wahrend seinernbsp;ersten Jahre in Kloster Bergen besonders mit der Uebersetzungnbsp;Pindarischer Oden beschaftigt. Neun derselben sind in Wielandsnbsp;Teutschem Merkur abgedruckt: die Hefte 5�8, 12 desnbsp;Jahres 1785 bringen je eine Ode, das 9. und 10. Heft je zwei;nbsp;abgesehen vom 9. Hefte (September), welches mit Wielandsnbsp;eigenem Aufsatze: Ueber die Rechte und Pflich-ten der Schriftsteller er�ffnet wird, steht jede Odenbsp;am Anfang des betreffenden Heftes.

Vor etwas mehr als Jahresfrist erwarb ich zwei Briefe Wielands an Gurlitt, die �ber die Pindarischen Oden handeln; Bernhardnbsp;Seuffert steilte mir dann seine Ausziige aus zwei Briefen Gurlittsnbsp;an Wieland zur Verfiigung, die im Germanischen Museum zunbsp;N�rnberg aufbewahrt sind. Spater verglich ich selbst die Originalenbsp;im Germanischen Museum, um Seufferts Exzerpte, die zu anderennbsp;Zwecken gemacht worden waren, zu vervollstandigen.

Die Wielandschen Briefe liegen nicht im Original, sondern in alter Abschrift vor. Der Schreiber scheint des Griechischen undnbsp;Franz�sischen unkundig gewesen zu sein: die Briefe sind dahernbsp;nachtraglich von anderer Hand korrigiert und wie es scheint,nbsp;zum Druck vorbereitet worden. Sie scheinen jedoch unver�ffent-

licht zu sein, wie auch Gmiitts Name bei Goedeke IV, i, 539�544 nicht vorkommt.

Das Format der Briefe ist Quart: das Papier des ersten tragt als Wasserzeichen den Namen J Kool amp; Comp, dar�ber ein�nnbsp;gekr�nten L�wen; der zweite Brief weist denselben Namen auf,nbsp;aber ohne L�wen.

Der Text folgt hier in der vom Korrektor hergestellten Fassung; kleinere Versehen des Sclireibers sind nur ausnahmsweise notiertnbsp;worden;

Die Einr�ckung ihrer Uebersetzung der zweyten Isthmischen Ode ^ Pindai's in No. 5 des Merkurs d. J. � wird Ihnen bereitsnbsp;bewiesen haben, dab mir Ihre Zuschrift mit den drey Pindarischennbsp;Oden richtig zugekommen ist. So nachtheilig es auch diesemnbsp;principi novem Lyricorum bey uns Teutschen, so wie bey dennbsp;�brigen heutigen Erdbewohnern ist, dab der In halt seinernbsp;Gesange auf die Siege in den �ffentlichen Kampf-Spielen, ausnbsp;bekannten und ganz nat�rlichen Ursachen, beynahe gar nichtsnbsp;interessantes f�r uns hat, und so viel er auch in einer Uebersetzungnbsp;(von dem Unterschiede der Sprache nichts zu gedenken) blosnbsp;durch den Umstand verliehrt, dab er in der Uebersetzung blobnbsp;gelesen, und nicht gesungen wird, da es doch zum Wesen dernbsp;lyrischen Poesie der Griechen, und ganz besonders der Pindarischen Oden, gehort, nicht nur gesungen und mit Instrumenten begleitet, sondern sogar getanzt zunbsp;werden: so bin ich doch, in Ansicht des Nutzens, den eine gutenbsp;Uebersetzung des kostbaren Restes, der von den mannichfaltigennbsp;Werken dieses Dichters auf uns gekommen ist, unserer Litteraturnbsp;verschaffen k�nnte, ganzlich der Meinung des Herrn v. Chabanonnbsp;in dessen, im 58ten Bande der Memoires de l�academie des B. L.nbsp;et I. einger�ckten. Discours sur Pindar e et la poesie lyrique ich dasnbsp;Gr�ndlichste und scharfsinnigste gefunden habe, was meinesnbsp;Wissens �ber den Werth dieses Dichters, �ber die Dichtart, worinnbsp;er gesungen und �ber den Gebrauch, der von seinen Werken

1 'Isthmischen' vom Korrektor in freigelassenem Raum nachgetragen, �Pindars� am Rande.

* Pindars zweyte Isthmische Hymne auf des Xenokrates, des Akragantiners, Wagensieg steht im Mai-Heft S. 97�103. Bis cinschlieBlich August werdennbsp;Gurlitts Beitrage als �Hymnelt;lt; betitelt, von September an steht lt;gt;Ode�.

(ungeachtet des wenigen Verhaltnisses, so sie zu uns zu haben scheinen) wenigstens von denjenigen gemacht werden kann,nbsp;welche sich zur musicalischen Dichtkunst gebohren fiihlen.

'So wenig Gewicht ich auch meinem Urtheil �ber die mir anvertrauten Proben Ihrer Uebersetzung beylege, so kann ichnbsp;doch nicht umhin, Sie, meines Ortes, zur Fortsetzung einer Arbeitnbsp;aufzumuntern, wobey Sie zwar, wie nahe Sie darin der Voll-kommenheit immer kommen mogen, weder auf den Dank dernbsp;Nation, noch auf Belohnung von unseren F�rsten und GroBen,nbsp;noch auf ein zahlreiches Publicum von Lesern rechnen k�nnen,nbsp;WO durch Sie Sich aber � gewiB um unsere Litteratur, um dienbsp;kleine Anzahl der Liebhaber der Alten, welche in ihrem Studionbsp;derselben fremden Beystand n�thig haben, und besonders umnbsp;die edlere Klasse unserer studierenden J�nglinge, die in dienbsp;Mysteri�n der griechischen Musen eingef�hrt zu werden w�n-schen, ein wahres und groBes Verdienst machen werden.

Der Merkur taugt allerdings wegen seiner Verbreitung durch den ganzen Umkreis unserer Sprache, zu einer B�hne, wo einnbsp;Gelehrter, der sich mit Untersuchungen dieser Art zu befassennbsp;Lust hat, Proben seiner Arbeit auf stellen kann, um das Urtheilnbsp;der Kenner dar�ber zu horen. Da Ihnen aber dieses letztere nichtnbsp;anders, als in einem hohen Grade g�nstig seyn kann, so kann esnbsp;wohl schwerlich Ihre ¹ Meynung seyn, sich blos auf die Isth-mischen � Oden einzuschranken: Sie werden uns den ganzennbsp;Pindar geben, und da dieses ein h�chst m�hsames Werk denbsp;longtte haleine ist, vielleicht f�r gut finden, von Zeit zu Zeit dasnbsp;Publicum durch Einr�ckung eines St�ckes in den Merkur, odernbsp;ein andres Journal daran erinnern, daB Sie Sich jnit dieser Arbeitnbsp;beschaftigen, ohne darum den ganzen Pindar nach und nach sonbsp;st�ckweise geben zu wollen, wodurch Sie einer k�nftigen Ausgabenbsp;mehr schaden, als n�tzen w�rden. Vielleicht w�rde auch einenbsp;vorlaufige Abhandlung �ber Pindar, den Geist und Characternbsp;seiner Oden (warum Hymnen?�) und in wiefern er, ungeachtet der Wahrheit des Horazischen Pindarum quisquisnbsp;studet imitari auch noch izt von Lyrischen Dichtern mit Nutzennbsp;studiert werden k�nnen (zu welchem allen der oben ber�hrte

Discours, des Mons. de Chahanon Ihnen vielleicht einige Dienste thun k�nnte) nicht �berfl�ssig seyn, das Publicum auf Ihrenbsp;Arbeiten aufmerksam zu machen und dazu vorzubereiten. Solltenbsp;Ihnen dieser Vorschlag der Ausf�hrung werth scheinen, so bietenbsp;ich Ihnen den Merkur dazu mit Vergn�gen an, wenn Sie andersnbsp;nicht durch das honorar von 5 rthl per i Bogen abgeschrecktnbsp;werden, welches bey dem dermaligen zerr�tteten Zustand desnbsp;teutschen Buchhandels, alles ist, was der Herrausgeber einesnbsp;schon 12 Jahre daurenden Journals thun kann.

Das g�tige Schreiben, womit mich Eu. Wolgeboren beehrt haben, hat mich sowol durch seinen lehrreichen Inhalt, als durchnbsp;die nachsichtige u.'^ f�r mich vorteilhafte Beurteilung meinernbsp;Arbeit u. durch die wolwollende Bereitwilligkeit meine Arbeitnbsp;zu f�rdern, ganz auBerordentlich erfreut: ich sage Ihnen dahernbsp;den verbindlichsten Dank daf�r. � Den ganzen Pindar zu geben �,nbsp;bin ich nicht Willens, weil ich es nicht besser tun zu k�nnen glaube,nbsp;als Gedike; und wenn ich auch manches wenigstens anders darstellen u. erklaren d�rfte, so mag ich �, da ich ihn u. er michnbsp;kennt, doch wenigstens nicht den Schein eines Wetteifers mitnbsp;ihm annehmen. Ich habe daher blos die Nemeischen u. Isthms.nbsp;Oden �bersetzt, welche ich schon seit Jahr u. Tag fertig liegennbsp;habe. Vor ungefar einem Jahre lieB ich die e r s t e und zehntenbsp;Nems. in die Dessauischen Berichte � etc. (ich glaube, im Aug.

Gemeint sind: Berichte der allgemeinen Buchhandlung der Gelehrten, Dessau und Leipzig, die vom April 1781 bis Dezember 1784nbsp;erschienen. Gurlitts Beitrage stehen im 8. und 9. St�ck, 1784.

u. Sept.) einr�ken, welche zweie der besten sind; u. vor etwa 3 Monat sendete ich 3 Isthms. an einen Freund, um sie insnbsp;Museum einr�ken zu lassen, babe sie aber noch nicht darinnenbsp;gefunden; welches jedoch an der Saumseeligkeit meines Freundesnbsp;liegen mag. Noch liegen mir also 10 Oden ungedrukt da, dannnbsp;habe ich alle noch nicht von Gedike und von niemand, als demnbsp;steifen ganz unpoets. Dani, �bersetzten Oden geliefert. Eu. Wolgeboren w�rden sich nun um meine geringe Arbeit sehr verdient machen u. mir die gr�Bte Gefalligkeit erzeigen, wennnbsp;Sie diesen 10 Oden Raum in ihrem Journale verg�nnten. Dennnbsp;ich weis, daB meine Arbeit alsdann allgemeiner beka��t wird,nbsp;als wenn ich sie besonders herausgebe: auch haBe ich das wider-holte Umherschreiben u. Pazisziren mit den Buchhandlern. Innbsp;dem Vertrauen auf Ihre G�te u. in R�cksicht auf Ihr Anerbieten,nbsp;wenigstens noch einige aufzunehmen, lege ich hier die Halftenbsp;derselben bei, u. ersuche Sie ergebenst, mir, wo m�glich, baldigst,nbsp;mir in zwoon Zeilen zu 1* melden, ob ich den Rest �bersendennbsp;solle. Denn ihn sogleich beizuschlieBen achte ich f�r unbeschei-denen MiBbrauch Ihres /Vnerbietens. Willigen Sie ein, so bin ichnbsp;gesonnen, es in der Jenaischen Litteraturzeitung bekaht zunbsp;machen, daB ich nun alle von Gedike noch nicht �bersezten Odennbsp;des Pindar �bersezt habe, u. wo sie befindlich sind. � Ziemlichnbsp;deutlich, d�nkt mich, habe ich sie abgeschrieben, damit demnbsp;Sezer u. Korrektor hinf�ro nicht wieder dergleichen Fehlernbsp;entwischen, als in No. 6, welche den Sinn verunstalten kontennbsp;z.B.gleichwonnigi� statt gleichnahmig;ergreiftnbsp;statt ergriffi'^; Erzbandigen den Wezstein statt

Gemeint ist Boies Deutsches Museum: Gurlitts Beitrage finde ich dort nicht. Dagegen enthalt der Jahrgang 1786 drei seiner Nemeischen Oden.

Friedrich Gedike, seit 1779 Direktor des Friedrichswerderischen Gymnasiums zu Berlin, hatte Pindars Olympische Siegshymnen (Berlin, 1777) und Pindars pythische Siegshymnen, mit erklarenden und kritischen Anmerkungennbsp;(Berlin, 1779) herausgegeben. Chr. Tob. Damms Versuch einer pro-saischen Uebersetzung der griechischen Lieder desnbsp;Pindar war 1770, 1771 in Berlin erschienen.

Fine Bekanntmachung Gurlitts habe ich in der Jenaischen Literaturzeitung 1785 nicht gefunden.

Die ger�gten Druckfehler finden sich im Mcrkur wie folgt: �gleichwonnig� S. 200, 8; �ergreifft� S. 198, i. 2; �Erzbandigen den� S. 202, 15; �Olympics�nbsp;S. 195, 6.

Erzbandigenden; Kampfeserretter Olympics statt Olympias. Jezt kann ich den von Eu. Wolgeborennbsp;mir angegebenen Stoff zu einer Abhandlung iiber die Brauchbar-keit Pindars f�r den heutigen Lyriker nicht sogleich benuzen, weilnbsp;es mir an Zeit gebricht. Denn teils kosten mich meine Schul-arbeiten den gr�sten Teil der Zeit, u. diese verrichte ich dochnbsp;gern mit Treue u. FleiB, weil schriftstellerische Arbeiten demnbsp;Beamten vielleicht mehr Glanz u. Ruhm, aber nur die gewissen-hafte Abwartung des Berufs an jedem Abend von neuem Ruhenbsp;u. Zufriedenheit giebt. Teils habe ich auch jezt schon meinen

AbriB der Geschichte der Philosophic meinen Vertragen unter die Presse gegeben, und habe also bisnbsp;Micha�lis nicht die mindeste Zeit etwas anders zu unternehmen.nbsp;Aber auf die Micha�lisferien gedenke ich nicht nur die benanntenbsp;Abhandl. gr�stenteils auszuarbeiten, sondern nebenher auch nochnbsp;eine andre, in welcher ich mich bem�hen will, die groBen unnach-amlichen Sch�nheiten der 4ten Pythischen Ode, wo die herrlichenbsp;Episode vom Argonautenzug ist, auseinander zu setzen, wen ichnbsp;anders nicht iiber der Arbeit finde, daB leztre Abhandlung in dienbsp;erste sehr bequem verwebt werden kann. Den Band der Memoiresnbsp;wo Chabanons Abhandlung steht, werde ich aus der Wolfenbiittl.nbsp;Bibl. erhalten.

Ich nehme mir die Freiheit, Eu. Wolgeb. hierbei auch mit ein Paar kleinen Schriften aufzuwarten. Ich wiinschte Ihr Urteilnbsp;von denselben in einer Rezension im Merkur, oder im Anzeigernbsp;zu vernehmen, wenn Sie es der Miihe wert achten. � Ziehennbsp;Sie aber aus dem alien ja nicht den SchluB, daB ich Lust zumnbsp;Vielschreiben u., welches sehr haufig damit verbunden ist. Hangnbsp;zum NachlaBigschreiben hatte. Ich mag vielleicht schlecht schrei-ben; aber, ist dieB, so ist es nicht die Folge der NachlaBigkeit,nbsp;sondern des Mangels an Talenten, die ungeachtet alles Bem�hens,nbsp;ungeachtet der behutsamsten u. iiber legtesten Fortschritte imnbsp;Arbeiten, doch nichts mittelmaBiges hervorzubringen vermogen..nbsp;Als ich den Pindar ausarbeitete habe ich ein Vierteljar des Abends

Das im Jahr 1786 bei M�ller in Leipzig erschienene Buch ist im A n z e i g e r des Teutschen Merkur, November 1786, S. CLXXXIII ff. angezeigt.

Eine Anzeige der Schriften habe ich im Merkur nicht gefunden: es handelt sich wohl um Gurlitts Einleitung und Anmerkungen zu Rosen-felds Uebersetzung der zwey Brautgesange des Catullsnbsp;und zweyer Oden des Horaz, Leipzig, 1785.

nicht gegeBen, um nur einmal wieder neben den gehauften Amts-arbeiten mit etwas hervortreten zu k�nnen. Deun ich hatte in 6 Jahren gar nichts schreiben k�nnen, und auch vorher warensnbsp;nur kleinere Piecen u. zwar in der morgenlandischen Litteratur,nbsp;die ich nachher gegen die griechische u. r�mische Litteratur u.nbsp;Philosophie ganz vertauscht habe, wie es mein Amt samt meinernbsp;Neigung von mir heischte. Denn wer wird nicht lieber die groBennbsp;Griechen u. Romer, als die Araber lesen? Eu. Wolgeboren ver-zeihen mir diesen Egoismus. Ich hatte dabei zur Absicht, michnbsp;Ihnen einigermaBen darzulegen, wie ich bin; u. einem Manne,nbsp;wie Sie sind, etwas naher bekaht zu werden, muB mir h�chstnbsp;schazbar seyn. �� Unter Versicherung meiner unbegranzten Hoch-achtung u. Ergebenheit verharre ich

N. S. Mit dem mir von Eu. etc. bestimten Honorarium bin ich v�llig zufrieden; nur ersuche ich Sie mir dasselbe, erst nach-dem alles einger�kt ist, im Ganzen zu �bermachen, weil ich es zunbsp;einer Reise nach Gottingen bestiiht habe. � Noch wage ich es,nbsp;Ew. etc. um eine Gefalligkeit im Namen meines Kollegen, desnbsp;Oberlehrer L o r e n z , zu ersuchen. Dieser w�nscht das M i n e -ralkabinetchennebstdendrei(besondersge-drukten) Briefen �ber die Gebirge von Hn.nbsp;Bergse kretair Voigt zu besizen. D�rfte ich bitten dasselbe durch Ihren Bedienten an mich einschlieBen zu lassen u. mirnbsp;die Auslage am Honor ar. abzuziehen ?

Ware es m�glich, zuweilen zwei Oden auf einmal einzur�ken, so ware mir dieB noch angenehmer; und dann w�rde auch fr�hernbsp;Plaz f�r die 2 Abhandlungen gewonnen. Besonders sahe ich esnbsp;gern, daB noch eine neben der zweiten Nems. Ode einger�ktnbsp;w�rde, weil diese sehr kurz. � Ich w�nschte �brigens, daB auchnbsp;diese Uebersezungen einigermaBen Ihren mir schazbaren Beifallnbsp;erhalten. M�he habe mir deBhalb gegeben, u. selbst zu denkennbsp;gesucht; welches Eu. Wolgeboren auch den Anmerkungen hinnbsp;u. wieder ansehen werden. (SchluB des ersten Briefes.)

Da ich aus fortdaurenden Einriiken meiner Pindarischen Oden ersehe, daB Eu. Wolgeb. nicht unzufrieden mit denselben sind;nbsp;so laBe ich mir das eine desto gr�Bere Aufmunterung zu den ver-sprochenen Abhandlungen seyn, an welchen keinen FleiB sparennbsp;will. Ungeachtet ich nun zwar in den ersten zwei Monaten nochnbsp;nicht m�chte ganz dar an gehen k�nnen, so habe ich doch schonnbsp;vorlaufig wegen der Abhandlung von Chabanon nach Wolfenbiittelnbsp;geschrieben, aber daher die Antwort erhalten, daB diese Abh.nbsp;weder in dem von Eu. Wolgb. angegebenen 58 Bande der Memoir,nbsp;del�Acad, des B.L.d; I. noch inirgend einemandern Bande stehe.nbsp;Nun muB ich freilich sagen, daB ich in meinem Apparat, wo ichnbsp;ziemlich alles gesamelt habe, was iiber Pindar geschrieben ist,nbsp;nichts von einer Abh. im 58 Bande der genanten Mem. fand,nbsp;wol aber eine Bemerkung antraf, daB Chabanon im 32 u. 35 Bandenbsp;etwas iiber Pindar, ich glaube aber bios Uebersezungen dernbsp;Pythisch. Gesange eingeriikt habe, welche er a. 72 zu Paris unternbsp;dem Titel; les Odes Pythiqucs de Pindar, traduites avec desnbsp;remarques nebst einer Abh. iiber das lyr. u. Pindarische Gedicht,nbsp;herausgegeben. Ob nun diese Abh. schon im 32 oder 35 Bandenbsp;der Mem. gestanden habe, Weis ich nicht. Besizen Eu. Wolgb.nbsp;die Mem. selbst, so ersuche ich Sie ergebenst mir gelegentlich nurnbsp;mit ein Paar Worten Auskunft zu geben, damit ich den Bandnbsp;einem Bibliothekar mit Gewisheit bestimmen kah. Wenn alles,nbsp;auch die Abhandlungen, eingeriikt sind, werde ich dem Hofr.nbsp;Heyne alles zusamen �berschicken, und zur Beurteilung vor-legen, wovon ich ihm bereits schon vorlaufig Nachricht gegebennbsp;habe, als ich neulich an ihn schrieb.

Mit diesem Schreiben nehme ich mir die Freiheit, dem Herrn Hofrath die Musikalien, deren ich neulich schon gedachte, zunbsp;iibersenden, mit dem ergebensten Ersuchen, sie, wenn es anderernbsp;Umstande halber wol angeht, dem T. Merkur stiikweise, wenn u.nbsp;wie viel Ihnen auf einmal gefallig ist anzufiigen. Da Ihrenbsp;Demoisells T�chter, wie ich aus dem im Okt. des T. Merk.nbsp;befindlichen Tonkiinst und auch aus Hn. v. Alxingers Gedichte

2� Tonk�nst ist wohl Schreibfehler f�r Tonst�k, welches wciter unten vor-

2g6

im T. Museum ersehe, selbst Kunstverstandige sind, so werden Sie dieselben bald k�nnen entscheiden lassen, ob diese St�ke dernbsp;Publikazion werth sind. Ich hoffe aber, so wol derselben, als desnbsp;Herrn Hofraths eignes Urteil wird f�r dieselben ausfallen. Auchnbsp;das langste Stuk, Gerstenbergs vortrefl. Mobrenlied, wird kompreBnbsp;gedrukt sehr gut k�nnen einger�kt werden. � Das Honorar f�rnbsp;diese Tonst�ke �berlaBe ich ganz Ihrer Bestimmung; denn ichnbsp;bin ganz �berzeugt, daB Eu. Wolgeb. daf�r so viel bestimmennbsp;werden, was Sie nach den ganzen Unstiinden bestimen k�nnen.nbsp;� Wenn ich �ber die Folge der St�ke in der Einr�kung meinnbsp;Urteil sagen darf, so glaube ich w�rde es gut seyn mit zweennbsp;der besten St�ke den Anfang zu machen; etwa mit No. 7 u. 8nbsp;mit Klaudius u. Goethens zwei herrlichen Liedern. Ich habe dernbsp;Nachricht, welche ich den erstern St�ken beizuf�gen bitte,nbsp;darnach eingerichtet, �berlasse es aber iiner noch des Hn.nbsp;Hofraths Gutbefinden, obs so seyn soil. Mit dem schlafen-denMadchenu. demMohrenlied No. 10 u. 12 w�rdenbsp;etwa k�i�en der BeschluB gemacht werden, so daB die Liebhabernbsp;von dem jungen Tonk�nstler mit ihrer Zufriedenheit an seinernbsp;musikal. Tafel empfangen u. demittirt v/�rden. So rathete un-gefar, we�i ich dem Ernste etwas Scherzhaftes beif�gen darf,nbsp;die Rhetorik des vorigen Jahrhunderts u. auch Vv^ol noch dienbsp;jezige hin u. wieder mit starken Argumenten zu beginnen u.nbsp;zu enden u. die schwachen Kinder derselben in der Mitte vonnbsp;vorne u. hinten zu deken.

Eu. Wolgebor. w�rden mir eine groBe Gefalligkeit erzeigen, wen Sie mir von jedem beifolgenden Tonst�ke ein Exempt a r ^0 besonders abziehen lieBen, weil ich sie nicht mehr besize,nbsp;weshalb ich auch ergebenst ersuche, sie so aufbewaren zu lassen,nbsp;daB nichts davon verloren geht.

In einem meiner vorigen Briefe war ich so frei, den Herrn Hofr. zu ersuchen, das Voigtische Werkgen nur durch Ihren Bedientennbsp;einsiegeln u. mit einer Aufschrift an mich begleiten zu lassen.

kommt. Das Oktober-Heft des M e r k u r s enthalt eine Musik-Beilage: Lied, mit Melodie von Caroline W. Der Text: Glanzender sinket die Sonne dort in dasnbsp;wallende Meer, ist Caroline R. unterzeichnet.

In meinem Exemplar des Deutschen Museums 1785 linde ich keine Komposition eines Alxingerschen Liedes.

mit, darnach �dre� gestrichen; zween k�nnte auch als zwoon gelesen werden, wahrend die Form zweien zu erwarten ware.

Ich hatte noch vorschlagen k�nnen, daB Sie blos die G�te haben mochten, Hn. Voigt den Zettel nebst der Aufschrift an michnbsp;zuzuschiken u. ihn bitten zu lassen es abzuschiken. Es ist hier innbsp;solchen Dingen mit unserm Buchhandler Scheidhauer nichts anzu-fangen, u. Hrr. K. Lorenz braucht es notwendig zu seinem Fache.

AuBer den Musikalien sende ich Eu. Wolgeb. noch einige kleine Uebersetzungen von einem hiesigen Lehrer, welcher michnbsp;ersucht hat sie beizuschlieBen, u. es dan Eu. etc. Urteil zunbsp;�berlassen, ob Sie die Bekantmachung in Ihrem T. M. verdienennbsp;oder nicht, so wie er auch das Honorar daf�r Ihrer g�tigen Be-stimung �berlaBt. Ich habe die Uebersetzungen nicht mit demnbsp;Original verglichen, aber aus seiner mir bekanten Starke in dernbsp;Engl. Sprache u. aus seiner Uebung im deutschen Styl vermutenbsp;ich gewiB, daB sie gut sind.

Ich empfele mich Eu. Wolgb. Gewogenheit u. Zuneigung u. verharre unter Bezeugung meiner unbegrenzten u. unverander-lichen Hochachtung und Ergebenheit

Wahrscheinlich kamen Wieland die verschiedenen Zumutungen Gurlitts etwas zudringlich vor: in dem folgenden Antwortschreibennbsp;werden weder die �Paar kleinen Schriften� Gurlitts, noch dienbsp;kleinen �Uebersetzungen von einem hiesigen Lehrer�, noch Voigtsnbsp;Mineralkabinetchen erwahnt, um dessen Zusendung Gurlitt zweimal gebeten hatte.

Was werden Sie von mir denken, daB ich Ihnen den Rest Ihrer Pindarischen Oden, nehmlich soviel davon, als in demnbsp;nun zu Ende gehenden Jahrgang 1785 des T. Merkurs nicht Platznbsp;gefunden, wieder zur�k schicke, ungeachtet ich in Ihrer Arbeit

Am Raiide ist hier nachgetragen�noch erlauben Sie mir vier hitte um akkuraten Druk der Musikalien. Ich oder mein verstorbener Freund mochtennbsp;sonst die Fehler auf unsrc Rechnung kriegen�.nbsp;urspr�nglich �Phindarischen�.

die Hand eines Meisters erkenne, und Ihre Uebersetzung f�r die einzige halte, die einen der griechischen Sprache wenig kundigennbsp;und �berhaupt einen jeden, der (aus welcher Ursache es sey) dennbsp;Pindar nicht im Original s t u d i e r t hat, mit dem Geiste undnbsp;der Manier dieses Dichters v�llig bekannt machen, und ihm Ver-standniB und GenuB seiner oft so schwer zu verstehenden Gesangenbsp;beybringen kann. Ich muBte aber den Gr�nden, die mir meinenbsp;associ�s vorlegten, diesen Artikel im Merkur, der nun bereitsnbsp;7 Monatsst�cke durch contimiirt worden, mit dem heurigennbsp;Jahrgange zu beschlieBen, um so mehr nach geben, da wir (wienbsp;alle Herrausgeber solcher Journale, die mehr von Ungelehrten,nbsp;als Gelehrten gekauft werden) nicht blos auf die innere G�tenbsp;eines Artikels, oder auf das, was ad palatum der kleinen Zahl dernbsp;Leser ist, sehen k�nnen, sondern noch dem Geschmack undnbsp;den W�nschen des groBen Haufens uns accomodiren mussen, f�rnbsp;welche (wie uns von mehreren Orten gemeldet wurde) diesenbsp;Pindarische deliciae w�rklich schon zu oft aufgetischt wordennbsp;sind. Wie sehr ein Gelehrter, der f�r Pindars lyrisches Genienbsp;und die Sch�nheiten seiner Oden Sinn und Empfanglichkeit hat,nbsp;auch immer Freude daran haben kann, ihm in seinen Adlers-fl�geln zu folgen, so k�nnen wir uns doch nicht verbergen, daBnbsp;diese Siegesgesange, im Ganzen genommen, f�r teutsche Lesernbsp;aus dem iS*'quot; Jahrhundert, beynahe gar kein Interesse mehrnbsp;haben, noch haben k�nnen; und daB also ein Journal, das f�rnbsp;eigenliche gelehrte und Philologische Artikel sich noch weitnbsp;weniger qualificirt, als das Deutsche Museum, eigentlichnbsp;kein schicklicher Platz ist, eine lange Suite von Pindarischennbsp;St�cken dem Publico aus zu stellen. Ich bin �berzeugt, daB Sie,nbsp;mein hochgeschatztester Herr, dieses selbst einsehen, und sichnbsp;also durch die Zur�cksendung der hiebey gelegten St�cke, aufnbsp;keine Weise beleidigt finden werden.

Meine unmaBgebliche Meynung ware, daB Sie nun wohl am besten thaten, Ihre Bem�hung auf eine eigne Ausgabe Ihrernbsp;Uebersetzung Pindars mit alien dazu n�thigen Commentarien

'associ�s' in freigelassenem Raum vom Korrektor nachgetragen. �heurigen� in freigelassenem Raum vom Korrektor nachgetragen.nbsp;�Leser� vom Korrektor iiber ein jetzt unleserliches Wort geschrieben.nbsp;�deliciae� aus 'delicie' korrigiert.

und Abhandlungen zu wenden; wo vielleicht wohl gethan ware, wenn Sie den Text neben Ihrer Uebersetzung mit abdrucken lieBen.nbsp;An einem Verleger dazu wird es Ihnen schwerlich fehlen k�nnen.

Die Ausgabe der Memoires de l�academie des Belles-Lettres, die ich besitze, ist in GroB 12 oder vielmehr median 12�, d Parisnbsp;chez Pancouke 1772, nach und nach continuirt bis zum Bostennbsp;volume, welches das letzte ist, so ich von meinem Buchhandlernbsp;in Strasburg bis dato erhalten habe. In dieser Ausgabe linden sich

1) nbsp;nbsp;nbsp;eineDissertationvomAbh�Fraguier �ber Pindar, imvol. 2.

2) nbsp;nbsp;nbsp;eine Uebersetzung der i. 2. 4. 5. 12. und 14 Olympischennbsp;und der i und 2 Isthmischen Ode, vom Abb� Massieunbsp;(im 61 8l und 14I volume zerstreut.

3) nbsp;nbsp;nbsp;Recherches sur la vie et les ouvrages de Pindare im 23*nbsp;volume.

4) nbsp;nbsp;nbsp;Ein discours sur Pindar et sa poesie lyrique im 58�* volume

5) nbsp;nbsp;nbsp;Uebersetzungen nebst Einleitung und Anmerkungen dernbsp;7I Olympischen und der i. 2. und 3I Pythischen Ode vonnbsp;Mons. de Chabanon in besagten 58sten volume, sehr interessant und sch�n gedacht und geschrieben

F�r die in den T. M. in diesem Jahre einger�ckten St�cke lege hier eine kleine Erkenntlichkeit bey, womit f�r Heb u. willennbsp;zu nehmen bitte. Arbeiten dieser Art m�ssen durch das Ver-gn�gen, das man selbst darin findet, belohnen, oder sie w�rdennbsp;die undankbarste unter allen seyn. Ich muB auf das Hesiodischenbsp;und Sokratische m��vvaniv �Q�siv provociren; ware ich einnbsp;K�nig oder Nabob von Bengalen, so sollte weder ein Pindar,nbsp;noch ein Uebersetzer desselben, wie Sie, �ber meine Munificenznbsp;gegen seltne Talente zu klagen haben.

Der Satz �sehr . . . geschrieben� wurde vom Schreiber am Rande nachgetragen; der Korrektor strich diesen nicht sehr deutlichen Nachtrag, um ihn auf dem unteren Rande zu wiederholen, wobei er zuerst die Worte �beides (nurnbsp;4 und)� schrieb und dann strich, wahrend �u. sch�n� vergessen und spater nachgetragen wurde.

�F�r� kein Absatz; der Korrektor bemerkt: �NB abgesezt und einger�ckt�.

Die mir neulich ebenfalls zum Gebrauch des Merkurs zuge-schickten Lieder-Compositionen muB icb aucli wieder zur�ck gehen lassen, weil beschlossen ist, kiinftig (zumal seit dem Todenbsp;des Herrn von Seckendorf) gar keine Musik mehr im Merkurnbsp;zu liefern. Icb versichre Sie aber bey meiner Ebre, daB diese mirnbsp;anvertrauten Compositionen bestandig im meinem Schreib-tische eingeschlossen gewesen sind, und daB von keinem derselbennbsp;Abschrift genommen worden ist.

Bin ich auf irgend eine Weise im Stande Ihnen einige Gefallig-keiten zu erweisen, so soli es wenigstens an meinem Willen nie fehlen. Denn ich schatze und ebre die leider so selten unter unsnbsp;werdenden Gelehrten Ihrer Art, und werde daher nur mit demnbsp;Leben aufh�ren, wiewohl ich des Vortheils Ihrer pers�nlichennbsp;Bekanntschaft entbehre, einer Ihrer aufrichtisten und warmstennbsp;Freunde zu seyn.

Die von Wieland zur�ckgeschickten Oden hat Gurlitt dann im Deutschen Museum und in Wiedeburgs humanistischem Magazinnbsp;untergebracht. Nach seiner Uebersiedelung nach Hamburg be-nutzte er sie dann noch einmal zu Schulprogrammen; im Jahrenbsp;1816 wurden diese Programme unter einem Sammeltitel vereinigt:nbsp;Pindars Pythische Siegsgesange ii b e r s e t z tnbsp;mit Anmerkungen von J. Gurlitt, Dr. dernbsp;Theol. , Professor und Director der Lehr-anstalten des Johan neums zu Hamburg, wienbsp;auch Professor am acad. Gymn. und Mitgliednbsp;der Schu1deputat i o n eines hochansehn-lichen Scholarchats daselbst. in ii Program-men. Angehangt ist Pindars erster und zehnter Siegsgesang.nbsp;Hamburg, 1816 (M�nchen, Staatsbibliothek A. gr. a. 783. 4).nbsp;Das auf diesen Titel folgende Vorwort ist vom 28. April 1816nbsp;datiert. Der folgende Text besteht aus elf besonders paginierten

Programmen aus den Jahren i8io�^1816, die jedoch nicht in numerischer Reihenfolge stehen. Eine Anmerkung zeigt die ur-spr�nglichen Erscheinungsjahre an; von den nicht im Merkurnbsp;ver�ffentlichten waren erschienen: Nem. i. in den Ber. cl. Buchh.nbsp;cl. Gel. iy84, St. 8; Nem. 8 im d. Museum iy86 St. 3; Nem. 9 das.nbsp;St. 4; Nem. 10 in den Ber. der Gelehrtenhuchh. iy84 St. 9; Nem. iinbsp;im d. Museum iy86, St. 9; Isthm. i in Wiedeh. hum. Mag. Neuiahrnbsp;iy88, S. 39; Isthm. 3 in Wiedeb. hum. Mag. Johannis iy8y S. 280;nbsp;Isthm. 4 das. Ostern 1788, S. 165; Isthm. 7 in Wiedeb. h. Mag.nbsp;O'Stern iy8y, S. 345; Isthm. 8 das. Johannisst�ck 1788 S. 199.

A LETTER OF GOETHE

iibergebe, auf Sereniss. Befehl, beyliegendes Pro Memoria mit der Bitte davon Gebrauch zu machen. Noch fiige ich bey daB Hr.nbsp;Lips w�nscht, wenn es thulich ware, vorerst ohne Charackternbsp;hier zu existiren. Es wiirde ihm nach seiner Denkungs art an-genehm und in seiner gesellschaftlichen Lage vortheilhaft seyn.nbsp;Im AdreBkalender k�nnte er bey der Zeichen-Academie auf-gef�hrt werden. Es ist ein gar wackerer Mann und ttichtigernbsp;fleiBiger K�nstler, der viel Nutzen stiften wird.

Our letter was acquired by purchase during the latter part of 1923 from the firm of Geering in Basel. As a supplement to thenbsp;letters of the Weimar Edition it would bear the number 2793a.nbsp;It is written, entirely in Goethe�s hand, on a single sheet measuringnbsp;18^/2 by 23^/4 cent. The water-mark shows a dancing bear withnbsp;trainer, both in profile. Along three of the edges of the sheetnbsp;there is a scroll in water-mark. The top, where there is no suchnbsp;scroll, shows signs of having been trimmed, perhaps to the extentnbsp;of 3I/2 cent. The Weimar edition of the letters, at least in thenbsp;volume (9) that would come in for consideration here, volunteersnbsp;practically no information concerning the paper used in Goethe�snbsp;letters, such chance remarks as that the paper of one letter is thenbsp;same as that of another being of little real help.

Johann Heinrich Lips, the subject of the communication, is, of course, well-known as an engraver and painter of the period.nbsp;It was especially for his ability as copper engraver that Goethenbsp;was intent upon drawing him to Weimar. The conditions of this

engagement will be found described in Goethe�s letter to Lips of March 23,1789. For the eight volumes of the Schriften publishednbsp;by G�schen Lips furnished six engravings, the frontispiece of thenbsp;Iphigenie volume, the frontispiece of the Egmont volume, the frontispiece and vignette of the Tasso volume, and the frontispiece andnbsp;vignette of the Faust volume. Of these his engraving after Rembrandt for the frontispiece of the Faust volume is best known.

Lips was apparently not to have regular classes in the Zeichen-schule; see No. 2738 of the Briefe; �Vorerst also soil ich Ihnen 150 rh. jahrlich anbieten, welche Durchl. der Herzog zahlen,nbsp;wogegen nichts von Ihnen gefordert wird, als daB Sie einigennbsp;j ungen Leuten, welche biBher sich im Kupfersfechen ein wenignbsp;geiibt haben . . . Anleitung gaben . . .� Compare also letter No.nbsp;2796: �Von den Stunden Mittwochs und Sonnabends wird ernbsp;wohl zu dispensiren seyn.� This will serve to explain the sentencenbsp;in our letter �Im AdreBkalender k�nnte er bey der Zeichen-Academie aufgefiihrt werden�, which might seem rather singularnbsp;in the case of a man called to Weimar specifically as Professor innbsp;the Zeichenschule. The Pro Memoria that was enclosed, andnbsp;which appears to have been lost, I take to have been a statementnbsp;of the conditions under which Lips took up his residence innbsp;Weimar, a statement that had the signature, or at least thenbsp;approval, of the Duke. Lips had arrived in Weimar on Novembernbsp;13, as we know from Knebel's Tagebuch (WE., Briefe 9, 358).nbsp;Goethe�s letter to the Duke of November 20 has the passage:nbsp;�Lips ist angekommen, seine Gegenwart wird viel gutes undnbsp;erwiinschtes stiften�, with which the phrase from our letter �einnbsp;. . . Kiinstler, der viel Nutzen stiften wird� may be compared.

The addressee of the letter is, I take it, Christian Friedrich SchnauB. For the facts of the life of SchnauB we have the briefnbsp;article of Ruland in the Allgemeine Deutsche Biographic, thenbsp;vivid sketch ^ by Max Hecker in Volume V (1925) of the Jahrhuchnbsp;der Sammlung Kippenberg, and an autobiography 2 publishednbsp;by Beaulieu-Marconnay.

^ Ein neuaufgefundener Brief Goethes von der zwciten Schweizerreise, Jahrhuch V, 5�12.

2 Ein Weimarischer Beamter des achtzehnten Jahrhunderts, von Carl Frei-herrn von Beaulieu-Marconnay, Zcitschrift fiir deutsche Kulturgeschichte, Neue Folge, IV. Jahrgang (Hannover 2875), pp. 649�702. I have used the copy innbsp;the library of the University of Illinois.

almost impossible that a man of the learning and culture of SchnauB, living in so stirring a period and in such close proximitynbsp;to genius, should not feel the need of self-expression in connectionnbsp;with the literary life of the day. And yet the account is absolutelynbsp;barren of anything of this nature. The student of the economicnbsp;life of a Saxon duchy during the second half of the eighteenthnbsp;century will find copious information on the state of the weather,nbsp;on the prices current of grains and victuals, on usages at weddings,nbsp;christenings, and funerals (or, at least, on the expendituresnbsp;incident thereto), but the student of literature will in no waynbsp;find reflected the cultural relations of the most important periodnbsp;of German literary history. When Goethe is mentioned, it is,nbsp;as likely as not, to chronicle his presence at a SchnauB familynbsp;wedding or christening.

SchnauB had since 1772 had a seat and vote in the Weimar Privy Council and became a little later the director of the Ducalnbsp;Library and the Coin Cabinet. Although Ruland does not tell us,nbsp;it also appears that he and Goethe were jointly in charge of thenbsp;Zeicheninstitut. This is shown, among other things, by letternbsp;No. 3232a in Volume 30 of the Weimar Edition, which containsnbsp;the Nachtrage to the earlier volumes. Here SchnauB and Goethenbsp;have together signed an �Unterthanigste Anfrage� to the Dukenbsp;requesting that Heinrich Meyer be given the �Character� ofnbsp;Professor in the �hiesiges F�rstliches Zeichen-Institut�.

Including the present one, we have altogether thirteen letters of Goethe to SchnauB. So far as the Weimar Edition affords anbsp;clue, all of these letters from Goethe to his colleague are in hisnbsp;own handwriting, as was perhaps to be expected. Six of themnbsp;deal directly with some phase of the activities of the Zeichen-schule.

The formulae of salutation and subscription agree with this theory. oGehorsamster Freund und Diener� of the close could,nbsp;in fact, have been used toward only very few persons of Goethe�snbsp;circle. Striking correspondences with other letters to SchnauBnbsp;serve to enforce this argument. It will not be necessary to takenbsp;more than three of these.

On January 2, 1790, about a fortnight after the date of our own letter, Goethe writes to SchnauB as follows:

Mit herzlichem Wunsche zumEintritt in das neue Jahr, sende ich den mir kommunicirten Extracktum Protocolli zur�ck.

Es wire! einen guten Effeckt haben, wenn Ew. Hochwohl-geboren Herrn Lips auf der Akademie einf�hren und vorstellen wollen. Von den Stunden Mittwochs und Sonnabends wird ernbsp;wohl zu dispensiren seyn, da er nur zum Unterrichte der jungennbsp;K�nstler da ist, und solche ihn zu Hause sprechen und seinennbsp;Rath und Lehre einhohlen k�nnen.

erhalten hierbey den von Serenissimo gnadigst genehmigtenVortrag und weiden die G�te haben das weitere gefallig expediren zu lassen.

Die Ackten folgen hierbey mit dem Wunsche, daB Sie solche noch lange ins neue Jahrhundert f�hren mogen.

Unter Anw�nschung eines gl�cklichen Eintritts ins nachste Jahr verharre mit wahrer, lebhafter Hochachtung

Vielleicht fanden Ew. Hochwohlgeboren es in diesem Augen-blicke nicht ungiinstig, wenn wir bei Serenissimo unsers Ab-wesenden, guten Meyers gedachten, demselben den Charackter als Professor erbaten und ihn in dem neuen AdreBkalender unternbsp;die Lehrer an der Zeichenschule, nach Professor Kestner setzten.nbsp;Sie sind mit mir einig, daB er in mehr als Einem Betracht diesennbsp;�ffentlichen Aveu verdient. Ich w�nsche zu horen, daB Sie sichnbsp;recht wohl befinden.nbsp;nbsp;nbsp;nbsp;Ew. Hochwohlgeboren

The AdreBkalender of the above and of our letter is, it may be added, the Hof- und AdreB-Kalender of Weimar.

RHYTHM AND MELODY AS PARODISTIC MEANS IN HEINE�S UNTERWELT

In Heine�s Neuen Gedichten we find a cycle of five poems, entitled Unterwelt, four of which present, transferred into anbsp;bourgeois atmosphere, the melancholy myth of Pluto, Proserpina,nbsp;and Ceres, while the fifth, a stirring lyric complaint, suggests thenbsp;underlying motif of the poet�s own experience. In poem no. inbsp;Pluto, envying the fate of Sisyphus and the Danaides yearns fornbsp;his bachelor days when the strident voice of his termagant wifenbsp;did not drown out the barking of Cerberus. But Proserpina, innbsp;poem no. 2, is in no happier mood; in the cursed rat hole, as shenbsp;calls her consort�s royal residence, Charon, baldheaded andnbsp;spindle-legged, and the judges of the dead with their bored facesnbsp;are her only table guests. The summer vacation which Ceresnbsp;proposes for her daughter is therefore gladly accepted by Plutonbsp;(no. 3), who, drinking punch with Lethe, will be happy to forgetnbsp;his wife, while she is sporting with village clowns at the harvestnbsp;dance on earth. The cruel prosaic directness of the dictionnbsp;combined with the burlesque irony of a consummate stanzaicnbsp;art, the structural importance of highly accented rhymes linkednbsp;in a complex system but of a vulgar connotation the monotonousnbsp;staccato rhythm, the commonplace vocabulary and syntax �nbsp;every means of artful unpoetic expression is combined to createnbsp;the impression of bleak and hopeless distress. And in the middlenbsp;of this cycle, as a climax, we find, prefixed by some verses of

^ Geselle � Holle � Gcbelle; Proserpine � Miene; Ehejoche � Rattenloche; Ceres � gewahr� es; Himmel � Bauernliimmel; verschnaufen � saufen; traurignbsp;� vertraur� ich � schaurig � versaur� ich usw.

Heine�s own make, three stanzas of Schiller�s Die Klage der Ceres, anonymously quoted as �those plaints known by all ofnbsp;you��.

The effect of this interpolation is most startling: �What a platitudinous declamation, what hollow, high-sounding phrasesnbsp;in Schiller�s poem!� is our first reaction. But our puzzlednbsp;astonishment gives way to a second and growing impression,nbsp;a resentment against the parodist and his frivolous play withnbsp;a fellow poet�s artistic creation, which he has somehownbsp;succeded in depreciating. But if, at last, we are able to overcome our offended sense of aesthetic propriety and analyzenbsp;the rhythmic jugglery performed before our ears we cannotnbsp;but marvel at the keenness of acoustic instinct, of parodisticnbsp;roguery involved.

Even then we have not sounded the deepest meaning of this apparent desecration of another creator�s gepragte Form, as wenbsp;shall see later. For the present, we must stop to consider whatnbsp;rhythmic means Heine has used to produce the desired effectnbsp;in the third poem of this cycle

His own introductory stanza consists of eight lines with four crossed, alternating feminine and masculine rhymes (a^b a'b c''dnbsp;c'�d), which, however, have no tectonic function since the strongestnbsp;stop, marked by a period, occurs after line 3 (Erde) instead of,nbsp;as we should expect, after line 4; all other verse-end stops arenbsp;more or less veiled by the syntactic unity of the resulting twonbsp;sentences and the concurrent rise of the voice in at least everynbsp;last syllable of each verse (this end rise, by the way, being characteristic of the whole cycle). The beginning of the line, innbsp;contrast, is marked by a distinct melodic fall, which starts innbsp;the arsis and continues throughout the first trochee; this downward tendency is repeated in every foot and creates the monotonous trot peculiar to the whole stanza, a movement which isnbsp;not relieved but rather reinforced by the arrangement of the fournbsp;arses in two degrees of intensity and on two contrasting tonenbsp;levels. They follow, to use the technical term employed by Eduardnbsp;Sievers �, Type B (light heavy light heavy or low high low high)

3 nbsp;nbsp;nbsp;See Eduard Sievers; Rhythmisch-melodische Studi�n, Heidelberg 1912nbsp;p. 46 and Metrische Studi�n (Abhandhmgen der Kgl. Sachsischen Gesellsch. d.nbsp;Wiss., Phil.-hist. Klasse B. 21, p. 58 �f.),

with the exception of line 6 which accents heavy light light heavy. The diagram of the curve would accordingly look like this:

The tempo is fast. The ratio of the duration of arsis and thesis is about even, for the accented syllables are cut short and level,nbsp;without a pronounced crescendo and decrescendo, not counting,nbsp;of course, the slight circumflection of the word before the pausenbsp;(hauft, lauft usw.). The voice quality which the stanza demandsnbsp;for an adequate rendering is high and hard, the character conversational.

The meter of Schiller�s poem is identical, but in rhytm and melody it differs completely from Heine�s introduction. To benbsp;sure, the stanza, although containing four additional lines, showsnbsp;the same system of crossed, alternating maculine and femininenbsp;rhymes; but while Heine�s quatrains could be strung alongnbsp;indefinitely like beads on a thread, the three component partsnbsp;in Schiller�s poem, again subdivided by two and two, reallynbsp;form a structural unity to which nothing could be added, fromnbsp;which nothing subtracted without destroying the melodic harmony. The experim.ent, for instance, of skipping in an organicnbsp;reading from line 4 (Und des Eises Rinde springt) to line 9 (Innbsp;dem Hain erwachen Lieder) will reveal that the voice drops farnbsp;too low and arrives with the end of line 12 at an impossiblenbsp;guttural level (�.Die Stimme rutscht nach hinten�^). The reasonnbsp;for this fact must be sought in the melodic composition of thenbsp;whole stanza, which seems to move in a preestablished balancenbsp;between the rather high start and the low end cadence.

A detailed analysis shows that the last (third) quatrain of the stanza repeats the melodic movement of the first, only with anbsp;decidedly stronger cadence and on a lower level. Each quatrain

* Concerning �freie und gehemmte Reproduktion� see Sievers, Ziele und Wege der Schallanalyse (in Festschrift f�r Streitberg, Heidelberg 1924, p. 69�70nbsp;and Gunther Ipsen und Fritz Karg, Schallanalytische Versuche, Heidelbergnbsp;1928),

again is split into two chains, the two lines of which form a melodic unit. The first chain prefers a longer rise and short fallnbsp;or a double rise and fall (see stanza I line i and 2), while in thenbsp;second chain the long fall predominates (I, 3 and 4); in bothnbsp;cases the second line serves as a cadence to the first, i. e. i to 2nbsp;and 3 to 4.

This melodic subordination is still more striking in the middle quatrain, where the flattened curve of the first two lines permits anbsp;long unbroken fall, while the second chain keeps up a rather highnbsp;level of sustained interest and thus affords the last quatrain anbsp;sufficiently high start for the range of its cadence.

Needless to say, then, that in contrast to Heine�s dipodic types we have an up and down in scales with at least three virtually

equal dynamic accents to the verse, that in contrast to Heine�s end rise we have a fall at the end of every line and vice versa a risenbsp;at the beginning and were it only in a semivocalic m or 1 or in thenbsp;circumflection of a long vowel (Reis!).

The diversity of tone direction is increased through the fact that the unaccented syllables do not invariably drop to a monotonousnbsp;plane of indifference but have a strong part in shaping the describednbsp;profile of the curve. The tempo is slow, the voice warmer andnbsp;lower and of a singing quality, due to the swelling and ebbingnbsp;of the tone cut of the arses, which have double the quantity ofnbsp;lt;gt; lt;gt; lt;gt;

the theses (erschienen, verjiingt, Zeus, especially noticeable at line ends). If we read the poem in Schiller�s Suabian dialect thisnbsp;latter peculiarity will stand out more strikingly; in that casenbsp;the melody curves, although not changing in shape, would presentnbsp;their symmetrical mirrored reflection, so to say, of high and low. �

Schiller

closed

slow

predominantly melodic (scales)nbsp;monopodicnbsp;legato

long, cescendo-decrescendo following general tone curvenbsp;risingnbsp;falling

Heine

open

fast

predominantly dynamic (types)nbsp;dipodicnbsp;staccato

cut short and even

falling

rising

Stanzaic structure: Tempo:

Accent;

Rhythm:

Rhythmic phrasing: Accented syllables:nbsp;Unaccented syllables:nbsp;Line start:

Line end:

To complete the picture of this radical difference between the work of two poets we must still mention the criterion of thenbsp;Becking curves. It has been established � that there are threenbsp;principal curves, one of which (and only one) may be beatennbsp;(with the index finger of the right hand) to any individual production of prose and poetry while it is being read. The first, innbsp;the shape of a reversed Arabic 6 {cJ), would, for instance, benbsp;called for by the works of Goethe. The second, in the shape of annbsp;horizontal 8 (oo), would apply to Schiller�s works and thus innbsp;our case to Die Klage der Ceres. The third, in the form of an

� Concerning this phenomenon of quot;reversion� see Sievers� Rhythm.-Mel, Stud, p. 63 ff.

A diagram may illustrate the execution of the two curves in question at least for one line of each writer:

Thus far we have intentionally discussed the stanzas of the two authors as if they were two productions reproduced individually and in strict separation. Yet, we have experienced that thenbsp;very effect Heine aimed to create was caused by the law ofnbsp;inertia, so to say, which forced us to carry over the insidiousnbsp;rhythm of the prefatory stanza in every characteristic featurenbsp;and to superimpose it upon the organically different rhythm andnbsp;melody of Schiller�s poem. The most striking result is a changenbsp;to dipodic movement, not, however, as we should expect, anbsp;continuation of type B but a breaking up of the monopodicnbsp;lines into whatever types their gait, if accelerated, would mostnbsp;easily permit: Pegasus assuming the jagged trot of a jadednbsp;carthorse:

and so forth until, completely out of breath and aware of the ridiculousness of our rendering, we stop and consider into whatnbsp;trap we have fallen. It is significant, by the way, that Heine innbsp;his first draft (in 1840) of the cycle and in its first publication �nbsp;quoted only two stanzas of Schiller�s poem and that he addednbsp;the third stanza later, probably in order to facilitate the transitionnbsp;to his own fourth poem (�Meine Schwiegermutter Ceres�) from

7 Zeitung fur die elegante Welt 1842 Nr. ii, 15 Januar. See Walzel�s edition 2,

the last lines of this stanza, which read as follows; �Ihre Trane bringt kein Zeuge Vor der hangen Mutter Blick�. His immediatenbsp;auditory impression of the rhythmic jest was at that time, ofnbsp;course, worn away sufficiently so as to make him unaware ofnbsp;the observation that the psychological compulsion of his prefatorynbsp;lines weakened with the length of the quotation.

What could have moved Heine to such an unprecedented parody of a truly poetic creation of one of his brethren innbsp;Apollo, to such a desecration of an avowedly well knownnbsp;masterpiece of classic German literature? Is it a parallel tonbsp;his unsavory attack on Platen ? Was it born of an innatenbsp;aversion of one poet to another? An examination of all hisnbsp;utterances concerning Friedrich Schiller (facilitated by the indexnbsp;of Walzel�s edition) must convince us that this is not the case,nbsp;that Schiller is one of the German writers whose statue thisnbsp;iconoclast has left unturned. He calls him oden edelsten, wennnbsp;auch nicht den gr�Bten Dichter der Deutschen� (Reisebilder HI,nbsp;Kap. 33).

Let us remember the tone of bitter despair pervading the whole cycle, a cycle which culminates in the poet�s realization of havingnbsp;robbed his wife of youth and happiness:

He who SO fervently wished for a life in Hellenic beauty and innocence after all never tasted the pagan love of Aphrodite butnbsp;remained �� his version of Tannhauser (1836) demonstrates it �nbsp;a Nazarene contenting himself with the love of oFrau Venus, dienbsp;Teufelinne�. Thus his satire of the the Persephoneia legend is anbsp;veiled outcry at the realization of the gulf which separates hisnbsp;petit m�nage with a little Parisian grisette from the world of Greeknbsp;classic lore and love. Schiller�s Klage der Ceres, dragged downnbsp;to the level of platitudinous burlesque, is only the symbol in thenbsp;symbol, the parody a heartrending masterpiece of selfirony.

8 This poem was published later than the other four (Zeitung f�r die Welt 1842, Nr. 104, 31. Mai), perhaps for discretion�s {and Crescentia�s) sake;nbsp;even if it was not written with the others there can be little doubt that itnbsp;presents the quintessence of the cycle.

UNTERWELT.

III.

Die verr�ckte G�ttin lauft Ohne Haube, ohne Kragennbsp;Schlotterbusig durch das Land,nbsp;Deklamierend jene Klagen,

SWEDISH WITCHCRAFT AND THE MATHERS

By September 22, 1692, nineteen persons had been hanged for witchcraft in Massachusetts, besides one who, �in accordancenbsp;with the old criminal law practice, had been pressed to death fornbsp;refusing to plead.� The excitement ran indeed so high, we arenbsp;told, that two dogs accused of witchcraft were put to death. �nbsp;�A certain degree of reaction, however, appeared to be takingnbsp;place,� says an authoritative resum� of the history of the matter,nbsp;�and the magistrates who had conducted the proceedings began tonbsp;be alarmed, and to have some doubts of the wisdom of theirnbsp;proceedings. Cotton Mather was called upon by the governor tonbsp;employ his pen in justifying what had been done; and the resultnbsp;was, the book . . . �The Wonders of the Invisible World,� innbsp;which the author gives an account of seven of the trials at Salem,nbsp;compares the doings of the witches in New England with thosenbsp;in other parts of the world, and adds an elaborate dissertationnbsp;on witchcraft in general. This book was published at Boston,nbsp;Massachusetts, in the month of October, 1692.�^

The governor had selected Cotton Mather, the eminent divine, teacher and writer, to defend or explain the execution of thenbsp;witches (and that of Mr. George Burroughs, �a minister of thenbsp;Gospel, whose principal crime appears to have been a disbeliefnbsp;in witchcraft itself�), because it was he who had been the mostnbsp;ardent believer in witchcraft and the most relentless spirit in thenbsp;prosecution of the victims. He was readily and energeticallynbsp;assisted in his furious campaign by his equally credulous father.nbsp;Increase Mather, the president of Harvard College, who afternbsp;the famous accusation of Mrs. Hale, a clergyman�s wife, publishednbsp;A Further Account of the Tryak of the New England Witches . . .

To Which is added Cases of Conscience^ Concerning Witcher aft and Evil Spirits personating Men. The above-mentioned books bynbsp;Cotton and Increase Mather, now very rare, were both reprintednbsp;in London in the year 1693. For this paper the writer has beennbsp;using the London edition of 1862, which contains both worksnbsp;and an informative introduction�.

But what has the Massachusetts witchcraft mania to do with Swedish witchcraft and the Mathers? Just this: In his cited booknbsp;of defense. The Wonders of the Invisible World, Cotton Mather,nbsp;for instance, makes bold to strengthen his own argument andnbsp;position by quoting a story of the ravages and punishment of thenbsp;witches in the province of Dalecarlia (Dalarne) in Sweden. A fewnbsp;references are made also in Mather�s work to witchcraft visitationsnbsp;in England, Scotland, Denmark, Germany and France; but hisnbsp;argumentative piece de resistance, so far as fortifying himself bynbsp;foreign proof is concerned, appears to be the �Extracts fromnbsp;Dr. Horneck showing the Similarity in the Circumstances attendingnbsp;the Witchcraft in New England and that in Sweedland.� Innbsp;fact, the title-page of the original announces among four specialnbsp;matters to be treated the following item: �A Short Narrativenbsp;of a late outrage committed by a knot of Witches in Swede-Land,nbsp;very much resembling, and so far explaining, that under whichnbsp;New-England has laboured.�� In other words, because of a strongnbsp;alleged resemblance between the cirsumstances and events of thenbsp;witchcraft epidemic in Sweden and that of New England, thenbsp;former became a main prop for the argument of Cotton Mather�s

2 A question of �conscience� was raised, whether or not the devil could assume �the shape of an innocent and pious person [like Mrs. Hale] for thenbsp;purpose of afflicting his victims.� Ibid., ix.

2 An examination of the witchcraft books by the two Mathers reveals a credulity which to-day is unbelievable of course. That two theologians, who havenbsp;always been regarded as able scholars also, could ever descend to such abysmalnbsp;ignorance and superstition as are evidenced in these writings seems indeednbsp;incredible. Yet the horrible truth is all too plain. True, many prominent clergymen, scholars, magistrates and physicians, all over the world were temporarilynbsp;blinded by the witchcraft delusion; but a large percentage of these ultimatelynbsp;realized the error of their ways, and those who did not in despair commit suicidenbsp;repented and used their influence to check the revolting madness. Not so thenbsp;Mathers. Not only Increase and Cotton clung doggedly to their original beliefsnbsp;and assurances, but the latter�s son Samuel appears to have continued far intonbsp;the eighteenth century to champion his father�s claims. The �vanity of superiornbsp;intelligence and knowledge was so great in the two Mathers [Increase and Cotton]nbsp;that they resisted all conviction.� Cf. Ibid., xii,

The Wonders of the Invisible World. And it should be remembered that the author claimed first of all to be quot;an Historian,� and notnbsp;an �Advocate� of witchcrafts His father claimed the samenbsp;distinction. In view of its historical interest we shall here reproducenbsp;in full Cotton Mather�s account of Swedish witchcraft in Mora,nbsp;Dalarne

quot;But is New-England the only Christian Countrey, that hath undergone such Diabolical Molestations? [asks the author] No,nbsp;there are other Good people, that have in this way been harassed;nbsp;but none in circumstances more like to Ours, than the people ofnbsp;God, in Sweedland. The story is a very Famous one; and it comesnbsp;to speak English by the Acute Pen of the Excellent and Renownednbsp;Dr. Horneck. I shall only single out a few of the more Memorablenbsp;passages therein Occuring; and where it agrees with what happenednbsp;among ourselves, my Reader shall understand, by my insertingnbsp;a Word of every such thing in Black Letter.

I. nbsp;nbsp;nbsp;It was in the Year 1669 and 1670, That at Mohr a [Mora] innbsp;Sweedland, the Devils by the help of Witches, committed a mostnbsp;horrible outrage. Among other Instances of Hellish Tyrany therenbsp;exercised. One was, that hundreds of their Children, were usuallynbsp;in the night fetched from their Lodgings, to a Diabolical Rendezvous, at a place they called, Blocktda, where the Monsters thatnbsp;so Spirited them. Tempted them all manner of Ways to Associatenbsp;with them. Yea, such was the perilous Growth of WdsWitchcraft,nbsp;that Persons of Quality began to send their Children into othernbsp;Countries to avoid it. �

II. nbsp;nbsp;nbsp;The Inhabitants had earnestly sought God by Prayer;nbsp;and Yet their Affliction Continued. Whereupon Judges had anbsp;Special Commission to find and root out the Hellish Crew; andnbsp;the rather, because another County in the Kingdom, which hadnbsp;been so molested, was delivered, upon the Execution of thenbsp;Witches.

III. nbsp;nbsp;nbsp;The Examination, was begun with a Day of Humiliation;nbsp;appointed by Authority. Whereupon the Commissioners Consulting, how they might resist such a Dangerous Flood, the Sufferingnbsp;Children, were first Examined; and tho� they were Questioned

� Professor George H. Ryden of the University of Delaware first called my attention to the following account.

� �Blockula� in this paragraph is of course intended to be �Blakulla�; �Elf-dale�, below, the modern �Elvedal�; and �Fahluna�, �Falun.�

One by One apart, yet their Declarations All Agreed. The Witches Accus�d in these Declarations, were then Examined;nbsp;and tho� at first they obstinately Denied, yet at length manynbsp;of them ingeniously Confessed the Truth of what the children hadnbsp;said; owning with Tears, that the Devil, whom they call�d Locyta,nbsp;had Stopt their Mouths; but he being now Gone from them, theynbsp;could No Longer Conceal the Business. The things by themnbsp;Acknowledged, most wonderfully Agreed with what other Witches,nbsp;in other places had confessed.

IV. nbsp;nbsp;nbsp;They confessed, that they did use to Call Upon the Devil,nbsp;who thereupon would Carry them away, over the Tops of Houses,nbsp;to a Green Meadow, where they gave themselves unto him. Onlynbsp;one of them said. That sometimes the Devil only took away hernbsp;Strength, leaving her Body on the ground; but she went at othernbsp;times in Body too.

V. nbsp;nbsp;nbsp;Their manner was to come into the Chambers of people, andnbsp;fetch away their children upon Beasts, of the Devils providing:nbsp;promising Fine Cloaths and other Fine Things unto them, tonbsp;inveagle them. They said, they never had power to do thus, tillnbsp;of late; but now the Devil did Plague and Beat them, if they didnbsp;not gratifie him, in this piece of Mischief. They said, they madenbsp;use of all sorts of Instruments in their Journeys: Of Men, ofnbsp;Beasts, of Posts; the Men they commonly laid asleep at the place,nbsp;whereto they rode them; and if the children mentioned thenbsp;Names of them that stole them away, they were miserablynbsp;Scurged for it, until som.e of them were killed. The Judges foundnbsp;the marks of the Lashes on some of them; but the Witches said.nbsp;They Would Quickly Vanish. Moreover the Children would be innbsp;Strange Fits, after they were brought Home from these Transportations.

VI. nbsp;nbsp;nbsp;The First Thing, they said, they were to do at Blockula,nbsp;was to give themselves unto the Devil, and Vow that they wouldnbsp;serve him. Hereupon, they Cut Their Fingers, and with Bloodnbsp;writ their Names in his Book. And he also caused them to be Baptised by such Priests, as he had, in this Horrid company. In Somenbsp;of them, the Mark of the Cut Finger was to be found; they said,nbsp;that the Devil gave Meat and Drink, as to Them, so to the Childrennbsp;they brought with them: that afterwards their Custom was tonbsp;Dance before him; and swear and curse most horribly; they said,nbsp;that the Devil show�d them a great. Frightful, Cruel Dragon,

3i8 telling them, If They Confessed Any Thing, he would let loosenbsp;that Great Devil upon them; they added, that the Devil had anbsp;Church, and that when the Judges were coming, he told them.nbsp;He Would Kill Them All; and that some of them had Attemptednbsp;To Murder The Judges, but Could Not.

VII. nbsp;nbsp;nbsp;Some of the Children, talked much of a White Angel,nbsp;which did use to Forbid them, what the Devil had bid them tonbsp;do, and Assure them that these doings would Not Last Long;nbsp;but that what had been done was permitted for the wickednessnbsp;of the People. This White Angel, would sometimes rescue thenbsp;Children, from Going In, with the. Witches.

VIII. nbsp;nbsp;nbsp;The Witches confessed many mischiefs done by them,nbsp;declaring with what kind of Enchanted Tools, they did theirnbsp;Mischiefs. They sought especially to Kill The Minister of Elfdale,nbsp;but could not. But some of them said, that such as they wounded,nbsp;would Be Recovered, upon or before their Execution.

IX. nbsp;nbsp;nbsp;The Judges would fain have seen them show some of theirnbsp;Tricks; but they Unanimously declared, that. Since They Hadnbsp;Confessed, all, they found all their Witchcraft gone; and the Devilnbsp;then Appeared Very Terrible unto them, threatning v/ith an Ironnbsp;Fork, to thrust them into a Burning Pit, if they persisted in theirnbsp;Confession.

X. nbsp;nbsp;nbsp;There were discovered no less than threescore and ten Witchesnbsp;in One Village, Three And Twenty of which Freely Confessingnbsp;their Crimes, were condemned to dy. The rest, (One pretendingnbsp;she was with Child) were sent to Fahluna, where most of themnbsp;were afterwards executed. Fifteen Children, which confessednbsp;themselves engaged in this Witchery dyed as the rest. Six andnbsp;Thirty of them between nine and sixteen years of Age, who hadnbsp;been less guilty, were forced to run the Gantlet, and be lashednbsp;on their hands once a Week, for a year together; twenty morenbsp;who had less inclination to these Infernal enterprises were lashednbsp;with Rods upon their Hands for three Sundays together, at thenbsp;Church door; the number of the seduced Children, was aboutnbsp;three hundred. This course, together with Prayers, in all thenbsp;Churches thro� the Kingdom, issued in the deliverance of thenbsp;Country.

XI. nbsp;nbsp;nbsp;The most Accomplished Dr. Horneck inserts a most wisenbsp;caution, in his preface to this Narrative, says, he, there is nonbsp;Public Calamity, hut some ill people, will serve themselves of the

sad 'providence, and make use of it for their own ends; as Thieves when an house or town is on Fire, will steal what they can. And henbsp;mentions a Remarkable Story of a young Woman, at Stockholm,nbsp;in the year 1676, Who accused her own Mother of being a Witch;nbsp;and swore positively, that she had carried her away in the Night;nbsp;the poor Woman was burnt upon it: professing her innocency tonbsp;the last. But tho� she had been an 111 Woman, yet it afterwardsnbsp;prov�d that she was not such an one; for her Daughter came tonbsp;the Judges, with hideous Lamentations, Confessing, That shenbsp;had wronged her Mother, out of a wicked spite against her;nbsp;whereupon the Judges gave order for her Execution too.

But, so much of these things; And, now. Lord, make these Labotirs of thy Servant, Profitable to thy People.�

While these quotations speak for themselves, for the most part, a few comments seem appropriate. In the references to what wenbsp;may call the historical background we find certain statementsnbsp;aiid inferences that are essentially correct. That a particularlynbsp;contagious form of witchcraft raged in a remote part of Dalarne,nbsp;Sweden, in the seventh and eighth decades of the seventeenthnbsp;century is well known. Every Swedish reader will have heard ofnbsp;the witches� flight to Bldkulla and may in Cotton Mather�snbsp;account recognize some other features of demonic tactics so oftennbsp;proclaimed in Swedish legends of the devil and his servants. Butnbsp;the Boston Puritan and his father either accepted everythingnbsp;that they had read at its full face value, or they were the biggestnbsp;scoundrels that Puritanism ever produced. Cotton either did notnbsp;heed or wish to heed the wholesome caution set forth by hisnbsp;source in regard to making false accusations against innocentnbsp;persons. To apply the moral of Dr. Horneck�s counsel to his ownnbsp;actions either did not occur to him or, if it did, he slyly simulatednbsp;innocence and approval of his own doings by calmly and hypocritically reproducing the story of the girl from Stockholm whonbsp;had wrongly accused her mother of being a witch. It is obviouslynbsp;a case either of the most unthinkable blindness or the vilestnbsp;pretense, and a dispassionate judgment forces us to admit thatnbsp;it must have been the former.

But why, in the first place, should an enlightened doctor of theology from Boston seek his heaviest polemic ammunition innbsp;the peasant district of distant Dalarne ? Was it not sufficientnbsp;to note that �Now Hundreds [of witches] are discovered in one

Shire [in England] ?� � Why carry devilish witchcraft coals to Newcastle ? Cotton Mather evidently needed what he considerednbsp;far more tangible and convincing evidence and so turned to anbsp;country beyond Great Britain, the land of his own ancestors. Itnbsp;is an extremely interesting Scandinavian influence with whichnbsp;we are dealing here. It is an unfortunate one perhaps, consideringnbsp;that this is probably the first Scandinavian influence we havenbsp;on a New England book�and Swedes will not be particularlynbsp;proud of it�-but it is one possessing curious interest nevertheless.nbsp;And we shall see later, that unless we consider Cotton Mathernbsp;in all respects an idiot his method of procedure is in realitynbsp;something of a tribute to Sweden.

Cotton Mather mentions a Dr. Horneck as his source. This was Anthony Horneck, a German who had settled in England andnbsp;who had written �An Account of what happened in the Kingdomnbsp;of Sweden in the Years i66g, i6yo and upwards. In relation to somenbsp;Persons that were accused for Witches; and Tried and Executednbsp;by the King�s Command . . . Done into the High-Dutch, by Anthony Horneck D. D.�� He had then translated it into English,nbsp;as we have seen. � This English version had, furthermore, beennbsp;incorporated in Joseph Glanvil�s Sadducismus Triumphatus ornbsp;a full and plain Evidence concerning Witches and Apparatus, ofnbsp;which a second edition appeared in i68i and a third in 1689.nbsp;It seems to have been a very popular work and a fourth edition,nbsp;a copy of which the writer has examined, came out in 1726

Glanvil�s book containing Horneck�s account of witchcraft in Sweden proved a godsend^�if I may use that term-�-to the Mathersnbsp;when the governor of Massachusetts demanded a reckoning.nbsp;�Executed by the King�s Command�� said the official Swedishnbsp;report of the convicted criminals. Under the unbearable pressurenbsp;of the Swedish state clergy, Dalecarlian witchcraft had finally,nbsp;at least nominally, received an official recognition; it had nominally been attested by certain high but frightened personages;

� Ibid., p. 88.

� In his article on quot;Witchcraft� (i868) James Russell Lowell included an account of witchcraft in Sweden. He speaks of the witches� flight to quot;Blockula�nbsp;and of the executions at �Mohra�. Lowell undoubtedly used the same source,nbsp;though I once believed that he might have received an oral account of the persecutions from Longfellow, who had been in Sweden. Cf. my article on quot;Thenbsp;Essays of Fredrika Bremer in the North American Reviewquot;, PMLA, XLI, 754,nbsp;note 12.

and so victims were thereafter dispatched �by the King�s Command.�� This was just what the Mathers needed�-more authority to execute �by the King�s Command�; and if this had been donenbsp;in old civilized Sweden, it should of course be done in the morenbsp;recent settlements of Salem and Boston�according to the lightnbsp;and logic of the Mathers. Here was welcome proof not only of thenbsp;prevalence of witchcraft in general but that the learned divinesnbsp;of Massachusetts had in their prosecution of witches acted justlynbsp;in the past. As we have seen, they m,ade the most of their data.nbsp;It was not brought out, of course, that the most appalling cases ofnbsp;Swedish witchcraft, especially those among Dalecarlian childrennbsp;occurred in the backwoods, so to speak, and that their treatmentnbsp;could hardly be considered a model for an enlightened urban community in any land, not even for the Boston of 1692. But Swedennbsp;was in the seventeenth century everywhere held in high regard;nbsp;the victories of the Protestant champion Gustavus Adolphus hadnbsp;not yet been forgotten; Swedish people had made a permanentnbsp;settlement on the American shore not so many years after thenbsp;Pilgrim Fathers; Scandinavian culture was closely akin to thenbsp;Anglo-Saxon, and therefore any �evidence� from Sweden carriednbsp;weight and was eagerly sought. So Cotton Mather�s methodnbsp;may indirectly be considered a tribute to that country. It is atnbsp;least doubtful, too, whether examples of witchcraft from southernnbsp;Catholic Europe would have been considered valid in a Salemnbsp;court of 1692.

Of course mere chance may have played a part in Cotton Mather�s sources. In his zeal, hurry, and fanatic excitement henbsp;grasped the information nearest at hand and did the best he couldnbsp;with the matter at his immediate disposal. Among these was thenbsp;popular work on witches by Glanvil, containing happily Horneck�snbsp;work on witchcraft in Sweden, perhaps the best and most suitablenbsp;foreign account extant on the subject. TaJes of Swedish witchcraftnbsp;had been very popular in England and Germany, and especiallynbsp;the more flagrant examples of intercourse with Satan, fromnbsp;remote districts, appealed to the morbid interest of high and lownbsp;alike. The fact that devilish things happened in far-off Scandinavianbsp;added a gloomy, romantic and highly effective glamor to thenbsp;stories. Then when Mather found what he could parade as anbsp;governmental stamp of approval of the method of punishment,nbsp;he no longer hesitated to adopt the Swedish material for his chief

.defense.�We shudder to think that the Swedish witchcraft accounts may in the Massachusetts trials have influenced Cottonnbsp;Mather's merciless attitude toward innocent defendands. Butnbsp;this possibility is very remote: no foreign evidence could havenbsp;made the Mathers m^ore adamantly bigoted in their inner convictions regarding witches than they already were. When somenbsp;of their more rational neighbors later began to see the light, thenbsp;Mathers stubbornly commenced to collect material and take suchnbsp;other measures as they deemed imperative to prove their bitterlynbsp;contested ideas and statements, to enforce their opinion uponnbsp;others, and to maintain their old position at all hazards.

We have noted that Increase Mather came to his son�s assistance in defending the powers of witchcraft. His Cases of Consciencenbsp;Concerning Witchcrafts, also, contains striking references tonbsp;Swedish witches and Lapland magic. . . . �It is not usual fornbsp;devils�, says Increase, �to be permitted to come and violentlynbsp;carry away persons through the Air, several miles from theirnbsp;Habitations: Nevertheless, this was done in Sweedland aboutnbsp;twenty years ago by means of a cursed Knot of Witches there�nbsp;The case of the Swedish girl from Stockholm who in 1676 hadnbsp;falsely accused her mother of diabolical relations is told bynbsp;Increase^^ as well as Cotton; and he also tells the story of �A Witchnbsp;in Sweedland [who] confessed, that the Devil gave her a woodennbsp;Knife; and that if she did but touch any living thing with thatnbsp;Knife; it would die immediately�. . . . The fame of the Laplander�s magic, so pronounced in many European traveloguesnbsp;of the time, had reached America also. Increase Mather hadnbsp;absolute faith in the Laplander�s power of �second sight� whichnbsp;was quite independent of any guilty associations with ordinarynbsp;witchcraft� , , . This is common amongst the Laplanders,� assertsnbsp;the learned president of Harvard College, �who are horriblynbsp;addicted to Magical Incantations: They bequeath their Daemonsnbsp;to their Children as Legacy, by whom they are often assistednbsp;(like Bewitched Persons as they are) to see and do things beyondnbsp;the Power of Nature. An Historian who deserves Credit, relates,nbsp;that a certain Laplander gave him a true and particular Accountnbsp;of what had happened to him in his Journey to Lapland; and

Cases of Conscience (bound in one volume with his son�s Wonders of the Invisible World, London, 1862), 242.

further complained to him with Tears, that things at great distance were represented to him, and how much he desired tonbsp;be delivered from that Diabolical Sight, but could not; thisnbsp;doubtless was caused by some Inchantment�

No further comment is necessary. To-day the chief importance of what we have attempted to describe in this paper lies in thenbsp;historical fact that the witchcraft polemics of the Mathers disclosenbsp;definite Swedish influence on two New England writers as earlynbsp;as 1692.

ANGLIST AND ANGLICIST

The terms Anglist and Anglicist are not recorded in the New English Dictionary or the Century. In Webster�s only Anglist isnbsp;given; it is defined as �one versed in English linguistics��. Innbsp;the Standard both terms appear; Anglist is defined as �onenbsp;skilled in English philology��; Anglicist, as �one who is in sympathynbsp;with or advocates anything promoting the interests of Englandnbsp;or the English language or people��. From the following quotationnbsp;it appears that Anglicist has been in the language for nearly anbsp;hundred years:

Macaulay ... at once became the leader of the Anglicists [in the controversy of 1835]

Anglist, on the other hand, seems to be a recent importation into English from the Continent. The term is based on a Latinnbsp;noun, either Anglia �England�� or Anglus �an Englishman��.nbsp;Anglicist, on the other hand, is based on a Latin adjective,nbsp;anglicus �English��. The difference in etymology correspondsnbsp;closely to a difference in meaning, although this difference isnbsp;obscured, or even distorted, by the definitions given in thenbsp;dictionaries. Anglist is a learned or technical term which, so farnbsp;as I am aware, has never had any vogue in popular speech. Onnbsp;the Continent it is applied to an authority on England and thenbsp;English. Such an authority is a man who has devoted himselfnbsp;to the scientific study of the national culture of England, eithernbsp;as a whole (and he is the Anglist par excellence) or in some aspectnbsp;or aspects. The term may therefore be applied to a historiannbsp;of English politics, economics, literature, language, laws, customs,nbsp;what not. This broad connotation did not desert the term whennbsp;it was imported into English, and although Anglist can of coursenbsp;be applied to an authority on English linguistics the definition

quoted above from Webster�s must be condemned as too narrow. The definition given by the Standard is less objectionable, if wenbsp;take philology in the broad sense, as �the scientific study ofnbsp;a national culture�. But since this word is so often used in anbsp;much narrower sense, the definition of Anglist in the Standardnbsp;is misleading and not to be recommended.

Anglicist as an English word has a very different history and background. It was no importation, but a native coinage, andnbsp;was first used as a popular rather than as a learned term. In thenbsp;controversy of 1835, an Anglicist was a man who advocated thenbsp;use of the English language as the medium of instruction innbsp;schools for the natives of India. The term thus had, at first, anbsp;very limited scope, and was closely tied up with a particularnbsp;aspect of English culture, viz., the English language and the literary monuments written in that language. With the settlement of the controversy the word seems to have gone out ofnbsp;popular use, but it has been seized upon as a technical term,nbsp;in the sense �an authority on the English language and Englishnbsp;literature�. This usage fits in admirably with the etymology andnbsp;the previous history of the word alike, and enables us to makenbsp;a neat distinction between Anglist �authority on England�, andnbsp;Anglicist �authority on English�. The definition of Anglicistnbsp;given in the Standard is obviously too broad. The editors arenbsp;to be commended, however, for recording the word, whichnbsp;seems to have eluded all other lexicographers.

But Anglicist is by no means an isolated term; on the contrary, it is only one of a group, and must be considered in its relationsnbsp;to the other members of the group. Latinist is another noun innbsp;-ist based on a Latin adjective. According to the NED it meansnbsp;quot;one who is versed in the Latin language; a Latin scholar�.nbsp;Here we have a definition far from good but highly characteristicnbsp;of the dictionary makers when they come to deal with wordsnbsp;of this type. As a matter of fact, Latinist has two meanings innbsp;English, and these meanings are different enough to deservenbsp;separate treatment. In the learned world, a Latinist is an authoritynbsp;on Latin. In popular use, however, anyone who can read Latinnbsp;with tolerable ease and accuracy may be called a Latinist. Thus,nbsp;Macaulay uses Latinist in the popular sense when he says ofnbsp;Samuel Johnson,

a good Latinist, and he soon acquired ... an extensive knowledge of Latin literature. ^

Other words similarly formed are classicist, medievalist, modernist, orientalist, Gothicist, Italianist, Romanist and Germanist. These words are all recorded in the NED. The first is definednbsp;as (i) �an upholder or imitator of classic style or form;� (2) �onenbsp;who advocates the school study of the Latin and Greek classics�.nbsp;The Century gives, besides, the meaning �one versed in thenbsp;classics�, and Webster�s and the Standard likewise recognize thisnbsp;meaning. In present use, I think, four meanings of classicistnbsp;ought to be distinguished: (i) an authority on Greek and Latin;

(2) nbsp;nbsp;nbsp;one who, without being an authority, is more or less versednbsp;in Greek and Latin; (3) a propagandist for Greek and Latinnbsp;studies in our schools; (4) one who belongs to, follows or upholdsnbsp;any movement called classical, in literature, music or the artsnbsp;generally. The first two of these meanings correspond to the twonbsp;meanings of Latinist and are therefore of interest to us in thisnbsp;connection. Parallel to classicist is medievalist, which the NED.nbsp;thus defines; �one who studies or is skilled in medieval historynbsp;or affairs; one who practises medievalism in art, religion, etc.�nbsp;The other dictionaries hardly improve on this definition. In mynbsp;opinion four separate meanings of the word ought to be given:nbsp;(i) an authority on the Middle Ages, or on one or more aspectsnbsp;of medieval civilization; (2) a student of medieval culture;

(3) nbsp;nbsp;nbsp;a propagandist for medieval studies in our schools; (4) a defender or advocate of anything called medieval. It will be notednbsp;that medievalist in its first or learned sense differs from classicistnbsp;in that it does not have primary reference to language andnbsp;literature. The same must be said of Gothicist, which the NED.nbsp;defines as �one who affects or is conversant with the Gothicnbsp;style, especially in architecture�. Webster�s and the Standardnbsp;give similar meanings; the word is not recorded in the Century.nbsp;On the other hand. Orientalist, according to the NED , meansnbsp;�one versed in Oriental languages and literature�. Here �versed�nbsp;must be taken in a technical or learned, not in a popular sense;nbsp;our word does not seem to have acquired a popular meaningnbsp;distinct from its learned meaning. A better definition for itnbsp;would be �an authority on Oriental languages and literatures�.nbsp;Italianist is defined in the NED. as �one who Italianizes�;

in the Standard, as �one who imitates Italian characteristics�. These meanings are loose and obviously popular; in learnednbsp;circles the word means �an authority on Italian�. I have notnbsp;found a corresponding popular meaning �one who has a knowledgenbsp;of Italy and of the Italian language� but I suspect the word isnbsp;sometimes used in such a sense. Certainly Germanist has a parallelnbsp;use; in the NED. it is defined (i) as �one who has a knowledgenbsp;of Germany and of the German language�, and (2) as �one versednbsp;in Germanic or Teutonic philology�. To these definitions Webster�snbsp;and the Standard add a third: �one learned in the Germannbsp;language�. The last two definitions evidently represent learnednbsp;or technical terminology. They would be better put as (2) �annbsp;authority on Germanic�, and (3) �an authority on German�.nbsp;The restriction to linguistics, implied in the definition of Webster�snbsp;and the Standard, cannot be justified. The first definition reflects popular usage, of course; its learned or technical counterpart, �an authority on German},�, is ignored by all the dictionaries. One definition of Romanist in the NED. is �one who makesnbsp;a special study of Romance languages or philology�. This definition is not recorded in the Century, Webster�s or the Standard.

From this survey, to which other terms might have been added (as humanist, activist, naturalist), we may conclude that nounsnbsp;in -ist based on a Latin adjective are not infrequent in English,nbsp;and that they are given very inadequate treatment by thenbsp;dictionaries. Nouns in -ist based on a Latin substantive are alsonbsp;at home in the English language, but not at home in the dictionaries. I have already mentioned Anglist. Another word of thisnbsp;type is linguist. Here the NED. for once makes a distinctionnbsp;between the popular and technical meanings. It defines linguistnbsp;(i) as �one who is skilled in the use of languages; one who isnbsp;master of other tongues besides his own�, and (2) as �a studentnbsp;of language�. The first of these definitions is happy, but thenbsp;second is unhappy, for �student� is not the right word. Thenbsp;second definition ought to read, �an authority on linguistics�.nbsp;And curiously enough the editor proceeds to mark this meaningnbsp;as obsolete! No one who reads even cursorily our learned publications of today can fail to note that linguist in its technicalnbsp;sense is still very much alive. And this present use is by no meansnbsp;a revival of a meaning long dead. On the contrary, the meaningnbsp;in question can be found in perfectly good nineteenth-century

authors. Thus, Macaulay, in the same essay on Samuel Johnson from which I quoted above, uses linguist in the technical sense.nbsp;He says.

There, too, were Gibbon, the greatest historian, and Jones, the greatest linguist, of the age

The reference is to Sir William Jones, whose researches in the comparative grammar of Sanskrit, Latin and Greek led to thenbsp;greatest single discovery, perhaps, in the history of linguisticnbsp;science and justify the comparison with Gibbon. Here Macaulay�snbsp;meaning is clear. He is speaking of eminent authorities, notnbsp;schoolboys, and he has in mind their scholarly achievements, notnbsp;their mere acquaintance with times and tongues other than theirnbsp;own. In other words, to Macaulay linguist meant �an authoritynbsp;on linguistics�, and the editor of the NED. responsible for markingnbsp;the term obsolete in this sense made a bad blunder. Other wordsnbsp;that belong here are Dantist, Celtist, Slavist, Romancist, Semitistnbsp;and the like. The NED. defines Dantist as �a Dante scholar�,nbsp;and �scholar� evidently means not a mere student but an authoritynbsp;on the subject. Less satisfactory is this dictionary�s definitionnbsp;of Celtist as �one who studies the Celtic languages�. The Centurynbsp;gives a much better definition: �one engaged or versed in thenbsp;study of Celtic language, literature, antiquities, etc.� I shouldnbsp;distinguish between two meanings of this word: (i) an authoiitynbsp;on the Celts, and (2) an authority on Celtic. Both meanings arenbsp;learned, and the word apparently has not developed a distinctnbsp;popular sense. For the second meaning the term Celticist isnbsp;obviously more appropriate, but this term, though actually innbsp;use, is not recorded in the dictionaries. Similarly, Slavist, definednbsp;in the NED. as �one skilled in the Slav languages and literature;nbsp;a Slavonic scholar�, actually has two meanings: (i) an authoritynbsp;on the Slavs and (2) an authority on Slavic, and for the secondnbsp;meaning Slavicist would be a more appropriate term. Romancistnbsp;is explained in the NED. as �a writer or composer of romances;nbsp;a romantic novelist�. So far as I know, this word is never usednbsp;to m.ean an authoiity on Romance philology. We have alreadynbsp;seen that Romanist may have this meaning. Neither Romanistnbsp;nor romancist however strikes one as very suitable for the Romancenbsp;philologist, and the term Romanicist, though unrecorded in thenbsp;dictionaries, seems to be the only word that fits the case. Similarly,

Germanicist is the obviously appropriate term for the authority on Germanic, while Germanist ought to be reserved to the authoritynbsp;on German. The authority on Germany must also be called anbsp;Germanist, since the resources of the English language do notnbsp;permit of a distinction here. Finally, Semitist is defined by thenbsp;NED. as �one versed in Semitic languages, literature, etc.; anbsp;Semitic scholar�. Here again we must distinguish two meanings:nbsp;(i) an authority on the Semites and (2) an authority on Semitic.nbsp;For the latter meaning a term *Semiticist, though in theorynbsp;preferable, is rather heavy from a phonetic point of view, andnbsp;so far as I know it does not occur.

The terms which we have been considering vary more or less in age and historical development, in spite of their similarity ofnbsp;form and meaning. Insofar as they are used in scientific nomenclature, a certain amount of system, and fixity of usage, isnbsp;manifestly desirable. This end is best attained, I think, alongnbsp;the lines which I have indicated. As an example of the systematicnbsp;working-out of the terminology, I present the following, on thenbsp;theme �Iceland�.

Icelandicist i �one well acquainted with the Icelandic language or literature�.

In the case of Egypt, India and China a special terminology exists which has some interest in the present connection. Thenbsp;scientific study of China and Chinese civilization is called Sinology,nbsp;and an authority in this field is known as a Sinologist. Similarlynbsp;we speak of Indology and Egyptology, and of the Indologist andnbsp;the Egyptologist. But this system of terms seems to have littlenbsp;or no tendency to spread, and it is defective in that it makesnbsp;no provision for a distinct term to denote the linguistic andnbsp;philological field. Moreover, such a term as Icelandist has in itsnbsp;comparative brevity a true advantage over *Icelandologist. Mostnbsp;important of all, it has the great advantage of actual existence,nbsp;and this advantage will probably prove decisive.

THE ETYMOLOGY OF STIR 'PRISON�

The word stir, in the sense of prison, has considerable American currency. The �crook� jargon of the �talkies� makes frequentnbsp;use of it, and it appears in newspaper accounts of crime andnbsp;criminals and in fiction dealing with the underworld. Herbertnbsp;Yenne in an article on �Prison .Lingo� in American Speechnbsp;(II, 280) mentions it as used by occupants of prisons in thenbsp;sense of penitentiary. The word is not of American coinage,nbsp;however, but is plainly an importation from the Old World.nbsp;What is its history?

The Century Dictionary enters as a sixth meaning under the standard noun stir, from the verb stir (Old English styrian, rousenbsp;agitate, etc.), the following;

A house of correction; a lockup; a prison. Thieves Slang. Mayhew, London Labour, 1851, I, 421. 'T was in Brummagem and was seven days in the newnbsp;�stir� � (prison).

The New International Dictionary and The Standard also include stir, in the sense of prison, under the familiar noun stir. The Oxford Dictionary enters it as a separate substantive withnbsp;the meaning of prison. The same citation from Mayhew followsnbsp;and no etymology is suggested.

Apparently the history of the term has not been carefully traced. Perhaps, in the absence of links pointing toward the oldernbsp;forms, its ancestry cannot be certainly supplied. It seems to menbsp;far likelier, however, that the word, if coming from the Old English, is to be associated, not with the v�rb styrian, agitate, but withnbsp;the noun steor(stier, styr) defined in Old English dictionaries asnbsp;meaning (i) steering, guidance (2) rule, regulation (3) restraint,nbsp;discipline, check, correction. The phonetic development fromnbsp;either st�or or styr is normal, if we keep in mind the sixteenth-

century coalescence of -er, -ir, -ur, -or. The shortened vowel of a Middle English ster or stir might have arisen out of compoundsnbsp;like st�orness, styrness, correction, discipline, or st�orl�as, withoutnbsp;restraint, steorwirpe, deserving reprobation, etc. Or compare sir,nbsp;shortened from sire. On the semantic side the development of stirnbsp;from Old English steor seems even more probable than on thenbsp;phonetic side.

If the etymology proposed here be sound, lexicographical entries should henceforth associate the underworld word withnbsp;the verb steer rather than, as hitherto, with the verb stir.

106

�1

�.3

:l

STUDIES IN HONOR OF HERMANN COELITZ

TO PROFESSOR COLLITZ.

TO PROFESSOR COLLITZ.

TABLE OF CONTENTS.

XII

BIOGRAPHICAL SKETCH

BIBLIOGRAPHY

OF PROFESSOR COLLITZ� WRITINGS*).

no. 3. 593�601.

NEUTER PRONOUNS REFERRING TO WORDS OF DIFFERENT GENDER OR NUMBER i

DIALECTIC PHONETICS IN THE VEDA: EVIDENCE FROM THE VEDIC VARIANTS

I. Pali and Ardhamagadh�.

2. Asokan Petenika.

3. Girnar traidasa.

4. Additions to Whitney�s Root-Book.

5. The alleged Fox stem anemAdfci- �be cold�.

6. Fox -/a-.

7. A non-Algonquian trait of Arapaho.

41

A MATTER OF SEMANTICS

47

48

ZUR DUENOSINSCHRIFT

54

I. 14 und

65

AENEID, II, 557

THE GERMANIC VOWEL SHIFT AND THE ORIGIN OF MUTATION

75

76

\./

77

SALIC LITUS

84 quered person who has been left on his land or allowed to settlenbsp;on ours�. That a type 1l�ta- had the meaning of a past participlenbsp;is proved by the parallel formation Go fral�ts �apeleutheros�, OHGnbsp;frlldz �libertus�, etc.

93

94

DER GENITIV PLURAL AUF -� IM GOTISCHEN

97

99

GOTHIC SYNTACTICAL NOTES

no

SCANDINAVIAN-POLISH RELATIONS IN THE LATE TENTH CENTURY

BLOTNAUT

142

THE USE OF OLD ENGLISH SWA IN NEGATIVE

CLAUSES!

i6i

2. nbsp;nbsp;nbsp;Ex. 80 ff.:

4. ASHom. 10, 73:

III. Swa in Result Clauses.

1. nbsp;nbsp;nbsp;Finn. 41:nbsp;nbsp;nbsp;nbsp;�

2. nbsp;nbsp;nbsp;Beow. 1506:

3. nbsp;nbsp;nbsp;Az. 60:

4. nbsp;nbsp;nbsp;Or. 198, 8:

3. AGen. 8, 12 b:

TV. Swa pest in Result Clauses.

1. nbsp;nbsp;nbsp;And. 260:

2. nbsp;nbsp;nbsp;EH. 3, 156:

i65

i66

VII. Swa in Modal Clauses.

Here the swa is a conjunctive adverb of manner, and is always found in the initial position.

A. The simple Ste^a-clause.

1. nbsp;nbsp;nbsp;Beow. 2331:

2. nbsp;nbsp;nbsp;Gen. 1565:

3. nbsp;nbsp;nbsp;Rid. 3, i;

B. Swa introducing the second member of anbsp;comparison of inequality.

1. nbsp;nbsp;nbsp;Or. 180, 18:

2. nbsp;nbsp;nbsp;POT. 60�61:

C. Swa in the sense of quot;as if�.

1. nbsp;nbsp;nbsp;WComp. 24:

2. nbsp;nbsp;nbsp;DG. 26, 29:

John Macy: World�s Literature, 165:

CHANGE OF MEANING BY ANALOGY

i85

DER WORTSCHATZ DES AHD. TATIAN UND DIE �BERSETZERFRAGE

84 quered person who has been left on his land or allowed to settlenbsp;on ours�. That a type ¹l�ta- had the meaning of a past participlenbsp;is proved by the parallel formation Go fral�ts �apeleutheros�, OHGnbsp;frlldz �libertus�, etc.