Inheritance
of colours and pod characters
in Phaseolus vulgaris L.

DOOR

R. PRAKKEN

s-gravenhage

MARTINUS NIJHOFF
1934

3IBLI0THEEK DER
RIJKSUNIVERSITEIT
UTRECHT.

lt; r

inheritance of colours and pod characters in
phaseolus vulgaris l.

. 1 ^ 3 R U

Inheritance
of colours and pod characters
in Phaseolus vulgaris L.

PROEFSCHRIFT

ter verkrijging van den graad van doc-
tor in de wis- en natuurkunde aan de
rijks-universiteit te utrecht op gezag
van den rector-magnificus dr. c. w. star
busmann, hoogleeraar in de faculteit
der rechtsgeleerdheid, volgens besluit
van den senaat der universiteit tegen
de bedenkingen van de faculteit te
verdedigen op maandag 2 juli 1934 des
namiddags te vier uur

door

ROELOF PRAKKEN

geboren te enter

AAN DE NAGEDACHTENIS
VAN MIJN OUDERS

't- ƒ

I:-

sl. ^

'nbsp;S

#

Bij het beëindigen van dit proefschrift, hooggeachte Dr. Hoogen-
kaad, gedenk ik met dankbaarheid hoe Gij, als bij zoo velen, ook bij
mij belangstelling hebt weten te wekken voor de levende natuur en
haar problemen.

Hooggeleerde Went, Nierstrasz, Pulle, Jordan en Rutten.
Mijn studiejaren in Utrecht en het onderwijs, dat ik van elk Uwer
mocht ontvangen, blijven mij onvergetelijk. Voor de door U in velerlei
opzicht betoonde tegemoetkoming bij het inrichten der studie en bij
de keuze van de onderwerpen, ben ik ten zeerste erkentelijk.

Aan het werken onder Uw leiding, hooggeachte Dr. Hirsch, zal ik
steeds een aangename herinnering behouden.

Hooggeleerde Honing, hooggeachte promotor, U ben ik ten zeerste
dankbaar voor de gelegenheid, die Gij me hebt geschonken tot het
verrichten van dit onderzoek en voor de grondige critiek, die Gij
tijdens de bewerking van het proefschrift hebt uitgeoefend. Uw be-
langstelling in mijn werk was mij een krachtige steun.

Waarde Van de Peppel, U ben ik dank verschuldigd voor het ver-
vaardigen der foto's. Waarde Jansen en Knoop, U dank ik voor de
verzorging van het materiaal en de nauwgezette hulp bij de verwer-
king daarvan.

TABLE OF CONTENTS

INTRODUCTION................... 178

CHAPTER I. stem, flower and seedcoat colours .... 179
§ 1. Introduction..................

§ 2. Present state of the factorial analysis of seedcoat colour .
§ 3. The cross Fijne tros x Wagenaar.........

a.nbsp;Description of the parent plants and F^.....

b.nbsp;Segregation for stem and flower colours.....

c.nbsp;Seedcoat colour in the backcross F^ with Wagenaar

d.nbsp;Analysis of the Fg..............

e.nbsp;The „shinequot; factor Sh.............

§ 4. The choice of symbols..............

§ 5. Description of the seedcoat colours.........

§ 6. Relation between stem, flower and seedcoat colours. . .
§7. Mottling...................

CHAPTER II. pod characters............

yl. Strength of the string.........

§ 1. Previous investigations............

§ 2. The methods used..............

§ 3. Description of the parent plants and F^......

§ 4. Analysis of the F:...............

§ 5. Fg generation and backcrosses.........

i?. Toughnessofthepodwall.......

§ 1. Previous investigations............

§ 2. Description of the parent plants and F^......

§ 3. Segregation in F^, F^ and backcrosses......

..............................255

LITERATURE....................259

12

INTRODUCTION

The parent plants used in this investigation are the french beans

„Fijne trosquot; and „Wagenaarquot;.

„Fijne trosquot; is a rather late, whiteflowering, white-seeded pole
bean, with non-parchmented, nearly stringless pods.

„Wagenaarquot; is an early bush bean, very pale rose-flowering and
yeiiow-seeded; the pod is semi-parchmented and provided with an

extremely strong string.

The main intention was to investigate the inheritance of the
strength of the string. Besides other characters were investigated. As
the segregation for seedcoat colour was rather intricate, its dis-
cussion takes up most of the space in this work. Length of stem and
linkage relations will not be dealt with in this pubhcation.

In 1930 I had at my disposal 48 F^ plants, of which in 1931 and
1932 I grew the Fg and F^ progeny. Their analysis led to provisional
resuhs. In the same year 1930 I repeated the cross myself. The three
Fl plants (Fijne tros X Wagenaar) resulting from this cross formed
the material for more extensive investigations.

These F^ plants of 1931 (and those of '32) were on a large scale
self-fertilized and backcrossed by the two parent plants. The castra-
tion of Phaseolus flowers is difficult to perform and even with accurate
control of the stigma some pollen grains might be overlooked. I
therefore castrated the flower buds aheady one or two days before
they would have opened; one day after the castration they were
pollinated after renewed close examination of the stigma. Among
hundreds of backcross plants not a single individual has been found
which was apparently due to self-fertilization of the F^ mother planf
The discussion of the inheritance of seedcoat colours is mainly based
upon the backcross with Wagenaar.

CHAPTER I

stem, flower and seedcoat colours

§ 1. Introduction

Many factors for seedcoat colour have as yet been determined by
the various investigators. It is, however, very difficult to find the
connection between their investigations. In the first place the de-
scription of the colour types is often superficial. Secondly the greater
part of the crosses contain rather few of the very numerous colours.
And finally the connection between stem, flower and seedcoat colours
has nearly always been neglected. After the analysis work of
Lamprecht, the situation has become much more favourable,
because we have now the disposal of a very accurately described
material, which has been analysed for many factors.

It is for the following reasons that I venture to add the analysis of
the cross Fijne tros x Wagenaar to the extensive material already
known:

1.nbsp;The analysis concerns seedcoat colour in connection with
stem and flower colour.

2.nbsp;The number of factors involved is rather great.

3.nbsp;Backcrosses serve as a test for the factorial scheme.

4.nbsp;The making of linkage investigations.

Points 1 and 2 make the comparability with earher investigations
fairly great. Therefore I have tried to Hmit the introduction of new
factor symbols as much as possible although I am aware that only
definite crosses could give certainty about supposed identity of
factors.

§ 2. Present state of the factorial analysis of the seedcoat colour.

The factors for seedcoat colour as yet known may be classified as
follows:

a.nbsp;Groundfactor.

b.nbsp;Complementary factors.

c.nbsp;Modifying factors.

a.nbsp;Groundfactor.

This factor must be homozygous dominant or heterozygous for the
seedcoat to be able to show colour. If the groundfactor, which I call
P (Shull, 1907amp;, p. 829) is recessive, the seedcoat and as a rule the
flower too, are white. Kooiman (1931, p. 331) speaks of ground-
or ferment factor, Lamprecht (1933, p. 313) of a f u n d a-
mental gene.

b.nbsp;Complementary factors.

The groundfactor alone does not produce colour. Colour only
appears, if, besides the groundfactor P, there is at least one dominant
„complementaryquot; factor. Each of the complementary factors pro-
duces (in cooperation with the groundfactor P) a definite, mostly
very pale seedcoat colour. More complementary factors together
generally give a darker colour. This was for the first time analysed
by Kooiman (1920), who adopted in the analysis of his cross between
a yellow brown and a white race a groundfactor (A) and three com-
plementary factors, his „c h r o m o g e n o u s f a c t o r squot; B, C and
D. Lamprecht was the first to give complete certainty about the
existence of the two types of factors by crossing two coloured-seeded
races, which gave in the Fa-generation Vie white-seeded plants. As yet
Lamprecht has described six complementary factors or „colour
gene squot;.

The consequences of these relations are:

1.nbsp;White-seeded are all p-plants and also those P-plants in which
all complementary factors are recessive.

2.nbsp;If one or more complementary factors are homozygous domi-
nant, the ratio white-seeded: coloured-seeded can only be I : 3 (Pp-
plants).

3.nbsp;If the groundfactor is homozygous dominant and none of the

complementary factors are homozygous dominant, the proportion of
1

white-seeded plants is — gt; in which formula n isthenumber of hetero-
zygous complementary factors.

4. If the groundfactor is heterozygous and again none of the
complementary factors are homozygous dominant, the proportion

1 1 3
of white-seeded plants is--[■• — x —.

4 4n 4

c. Modifying factors.

They only influence the colours produced by cooperation between
groundfactor and complementary factors. The influence of such
modifying factors (Kooiman 1931, p. 346: „i n t e n s i f i e r squot; E
and F) may be rather general, but in other cases it is restricted to
definite factor combinations.

These relations between ground-, complementary and modifying
factors may be the cause of very intricate cryptomeric, epi- and
hypostatic phenomena.

Here I have to mention some unexpected and as yet unexplained
results of Lamprecht (1934amp;, p. 205). In his cross no. 38 between two
coloured-seeded races, white-seeded plants appeared in the Fg, in
spite of the fact that both parents possessed the complementary
factor C; and in cross 33 a good many more white ones appeared than
was to be expected according to the groundfactor-complementary
factors assumption. One of the parents in both crosses was the partly
coloured („teilfarbigequot;) race L 29. According to Lamprecht „ver-
„bleibt da vorläufig wohl nur die Annahme, dass es eine oder gewisse
„Kombinationen von Genen für Teilfarbigkeit gibt, bei denen die
„Ganze Testa ungefärbt verbleibt.quot;

§ 3. The cross Fijne iros x Wagenaar

The investigation of an Fg (Wagenaar x Fijne tros) consisting of
48 plants and their Fg and F4 offspring (1930—1932) induced me to
the provisional adoption of a factorial scheme. For the analysis to be
given here I will, however, use the more extensive material, viz.
Fl. Fg, Fg and backcrosses with their progeny of the years 1931 —
1933. The provisional scheme was wholly confirmed by this analysis.

In this section colour names and factor symbols are used with-
out ample explanation. The accurate description of the colour types
and the justification of the choice of symbols will be given in the
following sections. The colour numbers refer to the folding scheme
table 31 and to the colour description in § 5.

a. Description of the parent plants and F^

The Fijne tros race is white-seeded and white-flower-
i n g. Hypocotyl, cotyledons and stems are green without any
trace of anthocyanin. Only at the base of the full grown pod very

small violet spots occur.

The Wagenaar race has a yellowish seedcoat, shading
(especially at the ventral side) sometimes to canary yellow; the
hilumring (cf. fig. 1) is yellow brown; around the hilumring is a narrow
violet corona. Cf. the colour

description of no. 3 in § 5. Its
flower colour is a very pale
pink or lilac rose
(Repertoire de Couleurs par
R. Oberthur 130, 1, 132, 1
and paler). At the base of the
standard and wings the colour
is somewhat darker (Rep. de
Couleurs 187, 1, pale light
lilac). Hypocotyl and cotyle-
dons are partly covered with a
rose anthocyanin colour
(Rep. de Couleurs between

150, 3—4 and 118, 3; Ridgway, Color Standards, Plate XIII, 1' c).
This cotyledon colour is brightest immediately after germination
and vanishes about three days later. The anthocyanin colour of the
hypocotyl appears about ten days after germination. The full grown
pod is spotted with the same rose colour.

All Fl plants have a. black mottled seedcoat (fig. 2). Their
flower colour is a light violet; the wings a little more reddish than
Rep. de Couleurs 195, 1 (Violet Mauve); the standard 189, 2 (Bishops
violet). Hypocotyl and cotyledons are partly covered with a very

conspicuous dark blue violet anthocyanin colour (Rep. de
Couleurs between 199, 4 and 190, 4, often much darker; the paler

colours 195, 4; ridgway, Color Standards PI. XLIV, between 69quot;'
i 'and 65quot;' m). The stem is green, except for narrow violet spots
at the nodes; the full grown pod is dark blue violet spotted.
Henceforward I will use the following indications:

b. Segregation for stem and flower colitur

As to stem and flower colour in all generations only the following
three plant types occur:

1.nbsp;Green stem, white flower (white seedcoat).

2.nbsp;Rose stem, pale rose flower (coloured seedcoat).

3.nbsp;Violet stem, violet flower (coloured seedcoat).

As to the violet flower colour, in Fg the violet-1 F^ colour appears,
but also darker types. Moreover some flowers are a little more reddish
violet. The darker colours range between Rep. de Couleurs 189,
2—3 (Bishops violet) and 180,1—2 (Reddish violet). I tried to distin-
guish between the paler F^ colour (violet-1) and darker colours
(violet-2 and 3), but a sharp discrimination was impossible. The F3
and F4 generations, however, have shown that the violet-! Fg plants
nearly always segregate into violet and pale rose, consequently are
heterozygous; violet-2 and 3 plants do not segregate the pale rose
type and therefore are homozygous.

The extension of the violet stem colour in the Fj plants was ex-
tremely variable: sometimes only the cotjdedons and hypocotyl
showed small spots, whereas in other cases nearly the whole plant
was violet.

The rose hypocotyl and cotyledon colour was not always found.
With a view to investigating whether or not this rose colour niay be
totally lacking in pale rose flowering plants, I planted 1600 seeds of
Fl plants in flowerpots and examined the young plants twice closety,
viz. about 2 days (cotyledons) and 10 days (hypocotyl) after their
germination. Then I planted the green seedlings.in the field and
inspected the plants later on as to flower colour. It will be seen
(table 1) that out of 262 rose plants only one had not been recognized
by cotyledon or hypotocyl colour.

The numbers of the three plant types green, rose and violet ap-
proach the bifactorial 4:3:9 ratio, but there is a considerable
shortage of rose. The F^ families in table 2 show the same marked
shortage of rose plants, together with a surplus of violet ones
(D/m = 2.59). The tables 1 and 2 together give:

That is too few rose plants, too many violet ones (D/m still less
than three, but very high!). The two factors involved are the
„groundfactorquot; P (Shull, 1907amp;, p. 829) and a „violet
factorquot; which I call V (Lamprecht, 1932«, p. 177; Johannsen,
1926 p. 443). Both races have a complementary factor J (Lamprecht,
1932«, p. 176) in common, as will be shown later on.

Fijne tros is: pp VV (JJ).

Wagenaar is: PP vv (JJ).

Fiis:nbsp;PpVv(JJ).

Fa consists of 4 white flowering p plants, 3 pale rose flowering Pv
plants and 9 violet flowering PV ones. Of these 9 PV plants 6 are Vv
(violet-1) and 3 are VV (violet-2 and 3); according to table 2 actually
found 567 violet-1 and 283 violet-2 and 3.

The backcrosses of F^ with the parent plants agree with the
bifactorial scheme for stem and flower colour.

Fijne tros x Fj (pp VV x Pp Vv) gave the expected 1 : 1 ratio:

Therefore no trace of certation! Neither white flowering plants nor

violet-2 or 3 ones were found.

The reciprocal cross, F^ x Wagenaar, showed a marked deficiency

of rose plants:

The segregation for the factors P and V in the Fg families is shown
by tables 3—7. Table 3 contains the progeny of homozygous F^ plants.
The flower colour of all PP VV F^ plants was violet-2; in their F3
only violet-2 (and 3) occurred.

Table 4 of Pp vv plants; the agreement with the 1 : 3 ratio is

quite satisfactory:

Table 5 of Pp VV plants. One of the F^ mother plants was classified
as violet-1, three as violet-1 —2 and five as violet-2. In the Fg progeny
the flower colour violet-1 was not found. All families together gave:

Table 6 of PP Vv plants. Nearly all Fg mother plants were noted
as violet-1. The figures point to a clear monofactorial segregation.

There is no evidence of a possibly weaker constitution of rose PP vv
plants as compared with violet PP Vv and PP VV ones.

The progeny of double heterozygous Pp Vv plants (table 7)
showed very surprising results! In most of the families the number

3

of rose plants is considerably lower than the expected —. Only m 7

3

famihesout of 41 it is — or a little more. Summarizing the families,
16

we obtain very high D/m values:

The percentage pale rose flowering plants is 13.67 instead of 18.75.
In order to determine possible differences between Pp Vv plants
(as to the number of rose plants in their progeny), I made (spring 1934)
a second sowing of those famihes, which contained the lowest and the
highest percentages of the rose plant type. Tables 8 and 9 show the
results. The percentages of rose plants in the second sowings of both
groups are nearly the same: 15.32 and 15.42. Therefore I conclude that
the shortage of rose plants is a general characteristic of the Fg and of
probably all Fg families of Pp Vv mother plants.

These „irregularitiesquot; are up to now unexplained. The segregation
for stem and flower colour served as a foundation for the analysis
of the seedcoat colour. The results I arrived at concerning the in-
heritance of seedcoat colour have not given me any indication as to
the cause of the irregularities.

Of the 88 Fa plants of which the progeny has been tabulated, there
appeared to be (tables 3-7) 27 PP plants and 61 Pp ones (the ex-
pected ratio 1 : 2 is 29.3 and 58.7).

c. Seedcoat colour in the backcross F^ with
W a g e n a a r

In all tables the columns for the seedcoat colours are found under
the heading of the three stem and flower colour types, since with each
of them specific seedcoat colours correspond, i.e. the factors P and
V for stem and flower colour are just as well factors for seedcoat

colour.

As the Vv plants can only be distinguished from the VV ones by
the flower colour and not by the seedcoat colour, only the three plant

types

green stem (white flower),
rose stem (pale rose flower) and
violet stem (violet-1, 2 or 3 flower)

are used in the tables.

The backcross F, X Wagenaar (Pp Vv X PP vv) and the recipro-
cal one contain among the rose v plants 4 seedcoat colour types,
which show a strongly marked difference. They are called:
yellowish (the Wagenaar-colour),
orange (yellow brown),
greenish brown and
brown.

Among the violet V plants only 2 seedcoat colour types can be
easily distinguished, viz.:
violet and

The^violets are partly pure violet, partly rather brown violet. As i
could not yet distinguish these two violet types at the time when i
analysed the first backcrosses, i have taken them together as violet.

In each colour class there occur (table 10) about as many plants
with mottled as with selfcoloured seeds; the dark pattern of the
mottled seeds has the same colour as the selfcoloured ones. In each
colour class, except for the pale yellowish, the mottled seeds are

easy to discover.

It appears (bottom rows of table 10) that the numbers of the six

colour types are in accordance with the ratio 1:1:1:1:2:2.

It seems to be (leaving for the present the mottling out of con-
sideration) a trifactorial backcross, the Wagenaar race
(yellowish) being recessive for each of these three factors. One of the
three is of course the factor V. The other two factors must be
responsible for the four seedcoat colour types among rose plants. The
difference between violet and black seedcoat must depend upon one
of these two factors, whereas the influence of the other one is in-
conspicuous in the violet class (brown violet and pure violet) and
hardly or not at all perceptible in the black colour class (the black of
some plants is a very dark chrome green, cf. p. 199 and p. 210).

The progenj^ of the selfcoloured backcross plants exactly confir-
med the trifactorial conception. All plants must be recessive or
heterozygous concerning the three factors involved. As to the
groundfactor, part of the backcross plants appeared to be PP, the
other part Pp. As I have not found any indication of linkage between
P and those three factors, I have counted together the coloured-seeded
offspring of PP and Pp plants and left the white-seeded offspring
oat of consideration. The signification of the homozygous dominant
factors JJ and Sh Sh and of the linked factors CM in the formulae
of the genotypes in tables 11—18 will be explained later.

Selfcoloured yellowish (table 11) does not segregate (except
white).

Selfcoloured orange (table 12) segregates into:

This orange factorquot; I call G (Lamprechx 1932», p. 177;

:o::«trrort;: a;;;, ha^cross p.ants f„r the factors

gg bb vv (no. 3; Wagenaar colour).

Gg bb vv. (no. 9)
gg Bbvv. (no. 15)
Gg Bbvv. (no. 21)
gg bb Vv. (no. 6)
Ggbb Vv. (no. 12)
gg Bb Vv (no. 18) and
Gg Bb Vv. (no. 24)

Selfcoloured r o w n (g. Bb vv; table 14) segregates into:

formula

ob-
served

expected

1:3:3:9

D/m

FamUy 491 does not segregate yellowish and orange, but this

nllw Tquot;-let ,..bbVv; table 15) segregates into:

D/m

0.30

Selfcoloured brown violet (Gg bb Vv; table 16) segregates
into:

These black-seeded mother plants were recessive for the factor g.
Their colour was somewhat greenish black and in family 513 1 found
among the „blackquot; offspring 4 plants with a blackish chrome green
seedcoat colour. In some other families too these greenish black seeds
appeared, but as a rule they did not show a strongly marked difference
with the pure black ones (cf. p. 199 and p. 210).

One selfcoloured black backcross plant (table 18) segregates ac-
cording to all three factors and therefore is Gg Bb Vv:

The colourtypes deah with above could always be nicely discrimi-
nated (except for violet and brown violet). As to the factors ascertained
up to now, the parent plants have the following constitution:
Fijnetros: pp GG BB VV.
Wagenaar: PP gg bb vv.

The analysis of Fj

Cf. the upper half of the folding scheme table 31.
The analysis of the backcross with Wagenaar has been treated
before that of the Fg generation, because in the backcross and its
progeny all the six main colourtypes (or seven, if violet and brown-
violet are separated) could be nicely discriminated and complete
certainty could be obtained as to the influence of the factors G, B and
V. The mottling I left out of consideration.

The mottling I have to deal with is the so-called ever-segregating-
mottling, i.e. mottled plants never breed true, but they always
segregate into mottled and selfcoloured in the ratio 1:1.
In all F2 famiUes together I found:

463 mottled-seeded plants and
465 selfcoloured ones.
It was already shown by Shaw and Norton (1918), Kooiman
(1920) and especially by Lamprecht (1932a) that every mottled
plant of this type always segregates into mottled plants and two
selfcoloured types. One selfcoloured type corresponds with the „back-
ground colourquot; of the mottled seed, the second with its „dark pattern
colour.quot; The ratio between the three colour types is always:

1nbsp;background colour,

2nbsp;mottled,

1 dark pattern colour.
Kooiman and Lamprecht therefore suppose that this motthng
depends upon the heterozygous state of their (complementary) fac-
tor B, resp. C:

background colour type cc (resp. bb);
mottled seedsnbsp;Cc (resp. Bb);

dark pattern colour type CC (resp. BB).
My view of this type of mottling is that it is due to a factor M for
mottling which only works in connection with the dominant factor

C (= B of Kooiman), i.e. M locally suppresses the action of C; these
two factors are absolutely (or nearly absolutely; cf. p. 227) linked.
The outcome of this view will be dealt with in § 7 on mottling.

According to this view one parent must be cM cM, the other
Cm Cm; F^ Cm cM, i.e. mottled. Segregation in Fgi

1nbsp;cM cM: background colour type,

2nbsp;Cm cM: mottled,

1 Cm Cm: dark pattern colour type.

Which parent was Cm Cm, which cMcM?

It has been mentioned above that in the backcross of F^ with
Wagenaar in each of the colourclasses there occur about as many
plants with mottled as with selfcoloured seeds. In all cases the latter
colour is the dark pattern colour of the corresponding mottled seeds.
(In the yellowish class the selfcoloured and mottled seeds could not
be nicely discriminated). The numbers are (table 10):

PP Cm Cm gg bb vv
pp cM c^ GG BB VV.
Pp Gn cM Gg Bb Vv.
And the backcross Fl x Wagenaar is Cm cM x Cm, resulting in

colour type.

1 Cm cM (mottled)

Other colour types than the six (or seven) types of the backcross
Fl with Wagenaar did not occur in the Fg (except for one family
which will be discussed sub e). I may conclude from this that both
races in their genetical constitution only differed as to factors al-
ready discussed.

In the backcross with Wagenaar, however, in each „colour-
classquot; only two types occurred, whereas in Fg I could distingui^
in each „colour classquot; the expected three types: cM cM, Cm cM
and Cm Cm.

In the 4 colour classes among the pale rose flowering v plants
(yellowish, orange, gray-greenish brown and brown) the difference
between the two selfcoloured types, ^ cll and Cm Cm is not
very conspicuous; thecMcM background colours generally
are somewhat paler.

Especially in the yellowish class the difference between the three
types is often so inconspicuous that in Fg (and some other famihes)
all three colour types had to be counted together. Cf. the colour
description of the nos. 1, 2 and 3 in § 5.

The mottled seeds in the orange colour class can always be clearly
distinguished from the two selfcoloured types. Yet the discrimination
between these two types is very difficult in some cases. Again I must
refer to the description in § 5 (nos 7, 8 and 9).

Exactly the same apphes to the brown colour class: mottled seeds
easy to find, discrimination of the two selfcoloured types difficult.
Cf. description of the nos 19, 20 and 21.

In the gray-greenish brown colour class there is a marked differ-
ence between the two selfcoloured types: the background ^lo^r
cMcM is grayish brown, the dark pattern colour Cm Cm
more greenish brown. See description of the nos 13, 14
and 15.

In the violet flowering V plants the motthng of the violet seeds and
of the black ones is often very conspicuous and naturally
the difference between the c'M cM and Cm Cm colour types as well!
The V ^ cM background colour types are only partially tinged
with violet or blue. These tinges are darkest at the
ventral side (hilumring side) near the caruncula and the germ root;
their extension and intensity are extremely variable for
seeds of the same plant! The non-tinged parts have exactly the same

colour as the corresponding cM cM types among v plants: pale
yellowish, pale orange, gray brown and brown.

The violet and bluish tinges of these 4 V c'M cM background colour
types are not equally strong in all plants and in some plants even
totally lacking. In the latter case the difference between the V cM cM
colours and their corresponding v colour is hardly perceptible; the
hilumring colour of the v types, however, is always brighter.

As to the lack of these violet and bluish tinges in VcM cM plants
I have found indications that it depends upon the recessiveness of
one factor, the heterozygous forms being intermediate; there is,
however, no certainty about it, because of the extremely high varia-
bility of the tinges. The influence of this factor should be hardly or
not at all perceptible in the other colour types.

For the description of these four variable background colours I
may refer to the nos. 4, 10, 16 and 22.

The colours discussed so far and the factors involved in the segre-
gation are to be found in the upper half of the folding scheme in
table 31. The lower half of the scheme will be dealt with sub e).

Table 19 shows the numbers actually found in the F2 generations.
For the expected ratio in 1024 plants cf. the scheme opposite the table.
Under each „colour classquot; the three types belonging to it are found,
each indicated by its colour number. It should be remarked that the
nos. 16 and 22, V c'M cM background colours, under the heading
„blackquot;, are not black at all! And that in the „violetquot; V c'M cM back-
ground colour nos 4 and 10 the violet or bluish tinge may be totally
lacking. — There is an entire agreement of all colour types with the
theoretically expected numbers. There seems to be no linkage be-
tween the colour factors P, On (cM), G, B and V.

Table 20, derived from table 19, shows the Fg^segregation (in
coloured seeded plants) concerning the factors Cm (cM), G, B and V.
The ratio mottled: selfcoloured is exactly 1 : 1 (463 and 465). The
numbers in the column of the „total numbers colouredquot; very
nearly approach the theoretically expected ratio 1 :3:3:9:3:9:
36 between the colour classes yellowish, orange, gray greenish
brown, brown, „violetquot;, „brown violetquot; and „blackquot;.

To conclude, I give (table 21) the monofactorial ratios in coloured-
seeded plants for the factors G, B and V, again derived from table 19.

The numbers are:

As to the F3 I may point to table 22. This table and tables 22—27
of the progeny of mottled backcross plants of F^ with Wagenaar
require no further discussion.

The backcross F^ X Fijne tros is in perfect agreement with the

above analysis.nbsp;^ ^

Fijne tros: pp (JJ Sh Sh) cM cM GG BB VV.
Wagenaar: PP ( J J Sh Sh) Cjm gg bb vv.
Flnbsp;Pp (JJ Sh Sh) Cm cM Gg Bb Vv.

The backcross of F^ with Fijne tros therefore must be:
I pp: white.

' i Cm cM : mottled black.
J cM cM : background colour of mottled
black, i.e. brown with (or without) bluish
tinge.

The numbers are :

e. The „shinequot; factor Sh

One of the three large F^ families in 1932 (55-2) showed the same
colour types as the two other famihes (55-4 and 55-6), but also many
additional ones: 55-2 segregated moreover for another factor, for
which I use the new symbol Sh (derived from shine). The F^ plants
55-2 and 55-4 originated from the same cross; probably the Fijne
tros parent plant wiU have been heterozygous for this factor.

All colours discussed above were Sh; they are represented by the
upper half of table 31, the sh types by the lower half.

a.nbsp;All sh sh Cm Cm dark pattern colours (columns III and VI) are
somewhat paler than the corresponding Sh Sh (or Sh sh) dark
pattern colours and especially less shiny, often even dull or
dead. Between the numbers 27 and 33 I could hitherto not sharply
distinguish: both colours probably are of a pale yellow. The other
colours may be indicated by the same colournames as the corre-
sponding Sh types: duU greenish brown (39), dull brown (45), dull
violet (30 and 36) and dull black (42 and 48). In some cases discrimi-
nation between the corresponding Sh and sh types is hardly feasible.

b.nbsp;AU sh sh cM c'M background colour types (columns I and IV)
have a y e 11 o w-b rown hilumring, but for the rest their
seedcoat is nearly colourless. For their indication
I use the name „hilumring typequot;.nbsp;^ ^

So the influence of the factors G, B and V upon these sh sh cM cM
background coulour types is hardly or not at all per-
ceptible. The nos of column I (= v; nos 25, 31, 37 and 43) may
be distinguished by their brighter hilumring colour from the nos 28,
34, 40 and 46 (= V) in column IV. Some plants of the latter have
a gray greenish blue tinge near the caruncula and the germ root,
but the tinged part is always extremely small, never extending
over the greater part of the seedcoat as the violet or bluish tinge in
the corresponding Sh colour nos 4, 10, 16 and 22.

c.nbsp;The mottled sh sh (5n cM types of columns II and V naturally
have a less shiny (dull) dark pattern colour and a
nearly colourless background, by which they are
clearly distinguishable from the corresponding Sh types in the upper
half of the columns (cf. fig. 2, p. 183).

As to the actual numbers found in Fg family 55-2 I have to remark
that in table 29 I combined different colour types, because there were
some difficulties in the classification. As to the factors V and Sh there
occurred among coloured-seeded plants:

The general shortage of pale rose-flowering v plants is high in this
family. There are too many V sh plants.

Perhaps the numbers point to a weak linkage between the factors
V and sh. The constitution of Sh Fg plants suggests the same. Ac-
cording to their F3 there appeared to be:

among 6 v Sh Fj plants: 4 Sh sh and 2 Sh Sh.
among 12 V Sh Fg plants: 11 Sh sh and only 1 Sh Sh.

In F3 all sh sh Fg plants bred true for this factor. The 15 segregating
F3famihes (of Sh sh Fg plants; table 30) gave among coloured-seeded
plants 89 sh and 211 Sh (expected 75 and 225).

I never found clear indications of another segregation into white- and
coloured-seeded but the 1 : 3 ratio. Even the most recessive colour
number (hilumring type no. 25) is not white-seeded. I obtained
this colour number, which is recessive for all the factors discussed
(sh, c, g, b, v) in two F3 famihes. Therefore, at least one
complementary factor is homozygous dominant
in all plants. This „hilumring factorquot; I called J (Lamprecht, 1932a,
p. 176). This factor J is responsible for the fact that coloured seeds
with colourless hilumring never occurred, ahhough the complementa-
ry factor C (= B of Kooiman) produces with P, according to Lamp-
recht and Kooiman, a pale sulphurous or citrine yeUow seedcoat
with a colourless hilumring.

§ 4. The choice of symbols

A long time I have hesitated before I c.ould make up my mind con-
cerning factor names and symbols. In the course of my investigations
I used the following names, some of which indicate in a suitable way
the general or most conspicuous influence of the dominant factor;
other names have been chosen more arbitrarily with a view to their
influence upon one definite recessive genotype.

p

Groundfactor....................^

Hilumring factor (homozygous dominant)........J

Shine factor....................._

Factor pair for mottling...............Cm (cM)

Orange factor....................^

Gray-greenish brown factor..............^

Violet factor....................^

Had I to use new symbols now, derived e.g. from the names given
above ? I have not done so (except for Sh) but I have taken as far as
possible symbols aheady used by other investigators, at the risk of
the same symbol being used for different factors.

As a symbol for the groundfactor the letter P was first used (by
Shull, 1907 h, p. 829).

The letter M for mottling was used by Shull (1908), Emerson
(1909«) and Tschermak (1912). In my opinion the same factor M is
involved in true-breeding and in ever-segregating mottling (cf. § 7).

The colours in the upper half of my scheme are the same or
almost the same as the J colours described by Lamprecht.

His P J C colours:

Schamois.................PJCgbv.

Bister..................PJCGbv.

Mtinzbronze................PJCgBv.

Mineralbraun, dunkel............PJCGBv.

Veilchenviolett, dunkel...........PJCgbV.

Kastanienbraun..............PJCGbV.

ƒ PJCg BV.

Schwarz.................jpjCGBV.

correspond with my P J Sh Cm colours:nbsp;^

Yellowish.............P JShCjng b v (no. 3).

Orange..............P J Sh Cm G b v (no. 9).

Greenish brown..........PJSh(^gBv (no. 15).

Brown..............P J Sh Cm G B v (no. 21).

Violet............... JShCmg b V(no. 6).

Brown-violet............P J ShCm G b V (no. 12).

I P J ShCmg B V(no. 18).
^la^k...............jpjShCiiiGBV (no. 24).

Even Lamprecht's typical colour: Chromgriin-Schwarz: PP CC
JJ gg Bb Vv was found in my material (cf. p. 189, p. 191 and p. 210).

I therefore used the same factor symbols J,
C, G, B and V, supposing that my factors are identical with those of
Lamprecht. In § 5 his colour descriptions are compared with my
own. The main difference concerns the gray greenish brown colours
depending on the factor B: my colours 13 and 15 are somewhat
darker and more greyish than Lamprecht's corres-

ponding colours: „Havannabraunquot; and „Münzbronzequot;. My colour
nos 4 and 16 (V c colours of column IV) are the only two J colours
which up to now have not been described by Lamprecht. Their
character (tinged with violet or blue; extremely variable) fits in very
well with that of the V c colours „Ageratumblauquot; (my no. 10) and
„Graulich Indigoquot; (my no. 22) of Lamprecht.

According to the latter all these factors are complementary
ones. I must emphasize here again that it was not possible for me
to judge about the complementary or modifying character of the
factors involved in my cross, because at least one complementary
factor was homozygous dominant.

About the influence of his complementary factors J, C, G, B, V and
R Lamprecht says (1933 p. 251): „Die verschiedenen Kombinati-
„onen der genannten sechs Gene verursachen meistens mehr oder
„weniger dunklere Töne als einer reinen Mischung der jedem dieser
„Gene (zusammen mit P) entsprechenden Farben zukommen würde.
quot;,Hier bestehen in sofern keine bestimmten Regeln, als etwa dem
„Hinzukommen eines bestimmten Gens zu irgendwelchen anderen
quot;Kombinationen eine bestimmte Wirkung entsprechen sollte. Es
quot;kommt hierbei stets auf die Kombination in ihrer Gänze an, welche

„Farbe erzielt wirdquot;.

I might, however, remark that we may speak of a „generalquot; or
„definitequot; influence of some factors, at least on groups characterized
by definite genotypical constitutions.

Having adopted those symbols for the colours in the upper half of
the scheme, I was obliged to take the new symbol Sh-sh to indicate
the difference between the shiny colours on the one side and the hi-
lumring type and dead colours on the other side. This factor Sh is
supposed to be a modifying factor which is homozygous dominant
in all Lamprecht's colours.

It is noteworthy that nearly all jj CC colours described up to now
by Lamprecht are dead or dull colours. Cf. his description of „Stein-
farbigquot; (1932c. p. 4), „Ambraweiszquot; (1933 p. 255), „Russgrünquot; (1933
p. 256) and „Mattmünzbronzequot; (1933 p. 257). In all these cases he
mentions „mattes aussehenquot; oder „matte Oberflächequot;. At first I was
therefore inclined to ascribe my sh colours to the recessiveness of the
factor J. Then I should be obliged to assume a complementary factor

(other than J) which is homozygous dominant in all my plants and

causes, with P, my hilumring type no. 25, if all other factors involved
in my cross are recessive.

The complementary factor D of Kooiman (1920, 1931) produced,
if no further complementary factors were present, with P beans of
the same appearance as my hilumring type (his „ecruquot;). Together
with other complementary factors (B and C of Kooiman) it „makes the
colours but slightly darker and more greyishquot; (Tjebbes 1931 p. 185).

If we suppose this factor D of Kooiman to be present together
with J in my Wagenaar race, the latter should be of the constitution
PPDD J J Cm Cm gg bb vv. D, J and C are complementary factors,
of which C (= B of Kooiman) without D and J produces (together
with P) a pale sulphurous or citrine yellow seedcoat without
coloured hilumring. DorJ each cause (together with P) a
very pale seedcoat colour with brown or yellow brown
hilumring.

SiRKS (1922«, p. 110) however, crossed a „Wagenaarquot; line with a
„Citroenquot; (lemon-coloured) bean, the latter „without or at most with
a bluish navelringquot; (this „bluish navelringquot; apparently is a corona).
The Wagenaar race is „immediately after harvesting „lemon-
„coloured with a brown navelring, but soon the lemoncolour
„changes into greyish yellow and a year afterwards the seeds are
„entirely yellowish-brownquot;. The colour of „Citroenquot; does not change.
Cf. his col. PI. II nos 18 and 22. The F^ is of the Wagenaar type. The
Fg segregation is unifactorial: 30 „Citroenquot; and 99 „Wagenaarquot;.
(The factor involved is according to Sirks perhaps the same as D of
Kooiman). I therefore left off ascribing to the Wagenaar-race the
constitution PP DD JJ Cm(^ gg bb vv, because in that case in a
cross of my Wagenaar race with „Citroenquot;, the type without a
coloured hilumring would appear in Fg according to the bifactorial
ratio 1:15, instead of the ratio 1 : 3 in the cross of Sirks.
Consequently I assumed all my plants to be JJ and for the uni-
factorial difference between „shinyquot; colours on the one side and
„less shinyquot; colours together with „hilumring typequot; on the other
side, the modifying factor Sh-sh was adopted.

The differences between „Citroenquot; and „Wagenaarquot; of Sirks are
nearly the same as those between the colours

„Geschwefeltes Weiszquot;......P C j g b v and

quot;Schamoisquot;...........PCJgbv

of Lamprecht. „Geschwefeltes weiss ist eine sehr konstante Farbe
und verändert sich auch bei jahrelangem Aufbewahren nur wenigquot;.
And „Schamois verändert sich beim Aufbewahren sehr schnell, es
wird viel dunkler und der oben erwähnte, zuweilen stark kanarien-
gelbe Ton verschwindet vollkommenquot; (Lamprecht 1932«, p. 172).

This makes it all the more probable that we are right in identifying
the factor D of Sirks with the factor J of Lamprecht.

To conclude with the formulae of the races used in my cross are:
Fijne tros: pp JJ Sh Sh c'M cM GG BB VV

(one parent plant: pp_JJ^h sh cM cM GG BB VV)
Wagenaar: PP JJ Sh Sh Cm Cm gg bb vv.

§ 5 Description of the seedcoat colours

These descriptions were made in December 1933 towards the end
of the investigations. At that time I had kept a sample of seeds of
each plant. In each of the colours described here a considerable
variation is found, partly due to individual variation of seeds of one
and the same plant, but no doubt for the other part depending on
genetical differences (not the same as the factors described) betwe^
individuals belonging to the same type. Especially for the PJ Sh cM
c'M colour nos 4, 10, 16 and 22 the variation in the amount of the
bluish or violet tinge is partly individual, partly genetic.

The Sh colours (1—24) of the upper half of the scheme (table 31)
are first described. Between brackets the colour name of the corres-
ponding type of Lamprecht is given. The colour types corresponding
with my nos 4 and 16 have up to now not been described by him.
For the colour description I have used:

1.nbsp;R. Ridgway, Color Standards and Color Nomenclature (1912),

referred to as C. S; and

2.nbsp;R. Oberthür, Repertoire de Couleurs (1905), referred to as R.C.

The order of description of the Sh colours is:

c'M c'M: background colour type

Cm Cm: the corresponding dark pattern colour type.

Cm c'M: the mottled type.

1. PJShcMcMgbv;pale yellowish.

(cf. Lamprecht: Rohseidengelb, 1932«, p. 172).

The common colour is C.S. PL XV M'e (Light Buff — Warm Buff);

R.C. 66, 2—4 (Pale Ecru) or between this colour and 138,3 (Salmon
flesh). The yellower shades are not always easily distinguishable from
no. 3 (cf. Lamprecht 1932«. p. 172); they almost correspond with C.
S. PL XXX 19quot;^ (Cartridge Buff-Cream Buff); R.C. almost 36,2
(Maize Yellow). Hilumring C.S. PI. III13 j (Xanthine Orange-
Amber Brown); R.C. 318,2—328,3 (Rust red-Bistre).

In process of time this colour grows darker and darker. After 1
year: C.S. PL XXIX 16quot; b (Pinkish Cinnamon-Cinnamon Buff);
R.C. 307,1—309,1 (Dark fawn-Buff.) After 2 years: C. S. PL XV 13' j
(Tawny-Russet); R.C. 308,2 (Fawn).

All these colours correspond with those given by Lamprecht.
3. PJShCmCmgbv; yellowish.
The „Wagenaarquot; race,
(cf. Lamprecht: S chamois, 1932«, p. 172).

For the Wagenaar race a difference in colour is characteristic be-
tween the hilum side and the opposite dorsal side of the seed. Dorsal
side C.S. PL XXX 19quot; d (Cream Buff) but a shade paler, or
PL XVI 19' d (Naples Yellow); R.C. 325, 1 (Shamois). Hilum-
side C.S. PL XVI 23' (Strontian Yellow); R.C. 17, 3—4 (Canary
Yellow). The shamois and yellow are not sharply separated; transi-
tion colours occur. The canary yellow colour is very variable in its
extension; sometimes it may be restricted to a very smaU spot close
to the hilumring. Hilumring C.S. PL III between 13 k and 15 i
(Amber Brown-Mars Yellow); R.C. 316, 3—4 (Mars Yellow) or 328,3
(Bistre).

This colour too grows rapidly darker. After 1 year dorsal side C.S.
PI XXIX 17quot; a (Cinnamon Buff-Clay Colour), R.C. 325,4 (Shamois);
hilum side between C.S. PL XXX 21quot; and PL XVI 23'. After 2 years
dorsal side C. S. PL XXIX 15quot; k (Cinnamon-Sayal Brown); R.C. 309,
2—3 (Buff.).

The above description concerns j^hej)ure Wagenaar race,
individuals with the same PJShCmCmgbv constitution often
show a less marked difference between the shamois and the canary
yeUow. — The Wagenaar race has a narrow violet corona (cf. fig.
1, p. 182), which in older seeds is gray brown.

2. P JShOncMghv; yellowish mottled, 3/1.
(cf. Lamprecht: Schamois/Rohseidengelb).

Dark pattern colour as number 3 on groundcolour as number 1;

henceforward indicated as mottled 3/1, analogous to Lamprecht's
mode of indication. The mottling of no. 2 is often very difficult to
distinguish! For this reason I was often obliged to take the colours
1. 2 and 3 together. Many investigators have probably overlooked

this mottled type (Miyake c.s., 1930).

4. PJShcMcMgbV; pale yellowish tinged with

plumbago violet.

(Lamprecht: as yet not described).

Extremely variable colour, just as nos 10, 16 and 22; especially
with respect to the extension of the violet tinge. This tinge may be
more or less clouded and is always deepest at the ventral side near
the caruncula. On the same plant there may occur seeds that are
almost entirely deep gray violet and others almost without any
bluish tinge. The colour withoutblueis C.S. PI. XV 16' (Pale
Ochraceous Buff-Light Buff); R.C. 66, 1-3 (Pale Ecru). The b 1 u e
V i o 1 e t t i n g e is C.S. PL XLIX 53quot;quot; a-c, 57quot;quot; a-c and PI. L
6quot;quot;' a—c (Violet Plumbeous, Light Varley's Gray, Deep Plumbago
Gray); in R.C. less accurately represented; 204, 2-3 (Violet blue)
is the nearest colour, but too bright. Hilumring C.S. Pi. XV
15' i (Ochraceous Tawny) and paler; the darker types 13' i (Tawny);
R.C. 324, 2—4 (Hazel), the darker colours less bright than 321,3

(Dead leaf). ^ _

6. PJShCmCmgbV; violet,
(cf Lamprecht: Veilchenviolet, i932ß, p. 174).

In its brightest shade a little darker than R.C. 192,4 (Violet
purple) • mostly much darker, with transition to 347,4 (Violet black);
CS PL XI 61 n (Fluorite Violet-Black), PL XXV 61'm, 63'm,
65' m (Dark violet colours) and darker. Hilumring about the
same colour, but the corona (cf. fig. 1) is nearly always paler and
more brownish.

Paler types occur as well, especially on badly ripened plants.
Lamprecht (1932a, p. 174) says about this colour: „Schlecht aus-
gereifte Samen zeigen so grosse Unterschiede in der Farbe, dass sie
quot;ohne besondere Kenntnis der Verhältnisse in der Regel nicht zu
quot;,erkennen sindquot;. I could nearly always clearly distinguish the violet

colour no. 6 from the black.

5. P JShCSicMgbV; violet mottled, 6/4.
(Lamprecht: so far not described).

Owing to the great variability in the extension of the violet tinge
of background colour no. 4, the mottling of no. 5 is not always
equally conspicuous.

7. PJShcMc'MGbv; pale orange,
(cf. Lamprecht: M a is g e 1 b, 1933 p. 256).

Palest colour C.S. PI. IV 19 e (Maize Yellow-Buff Yellow); R.C.
326, 1 (and paler) with transition to 36,4 (Maize Yellow). Darker
colours C.S. PL XV 16' (Yellow Ocher-Ochraceous Orange); R.C.
315,2—329,2 (Yell. Tan colour-Raw Siena). Hilumring C.S.
PI. Ill 13 j (Xanthine Orange-Amber Brown); R.C. 329,4 (Raw Siena).

The analogy with no. 9 (see below) is very great. The main differ-
ences are that no. 7 is paler, less deeply orange, less shiny and a little
more reddish (also the hilumring) than no. 9. The intensity of this
pale orange colour in the same plant is rather variable. My pale
colours are a trifle less reddish than the Maize Yellow of Lamprecht.

9. PJShCmCmGbv; orange,
(cf. Lamprecht: Bister, 1932«, p. 173; 1933 p. 256).

C.S. PL III 17 h (Cadmium Yellow-Raw Sienna); the darkest
colour PL III 16 i. R.C. 314,1—329,1 (Ocre de Ru-Raw Sena); the
darker colours 328,2 (Bistre). Hilumring C.S. PL IllilS i (Mars
Yellow) and darker; R.C. 328, 3—4 (Bistre).

It is often very difficult to distinguish between no. 9 and no. 7, cf.
above. It was only after a long time and on comparison with the no. 9
resulting from backcross' F^ X Wagenaar, that I could distinguish

the two colours rather clearly.

This colour (or about the same) has been indicated by different
investigators as brown, yellow brown, orange or yellow.
8. P J Shtoicl Gb v; orange mottled, 9/7.
(cf. Lamprecht: B i s t e r/M a i s g e 1 b).

As a rule the mottling is easy to discover. But sometimes, if the
background-splashes are minute and the colour difference is incon-
spicuous, a very close inspection is necessary.

10. PJShcMcMGbV; pale orange tinged with

ageratum blue.

(cf. Lamprecht: Ageratumblau, 1933 p. 258).

Seeds of the same plant very variable as to the extension and in-
tensity of the bluish tinge. Just as in no. 4 the bluish colour is deepest
near the caruncula and germ root. The colour without bluish

tinge is C.S. PI. XXX 19quot; c (Cream Buff-Chamois); R.C. 325
t in g e ib V.nbsp;r PI XV 16' b (Ochraceous Buff-

1—2 (Shamois). Or more orange. C.S. Pi. XV ib d _ . ,. ^ ^ .

^r 11 N T? rnbsp;1—325 4 The blue violet tinge

quot;pr.nbsp;h^n^^^'^Llf 6r h ,Slate V,o,et-DeepSMe

Violet) RC the paler colours 201. 2-4 and 200,4 (Ageraturn blue,
SLrxZI, tLquot; . KC. 324, 3-. (Ha.el, an. le. .H,«

than 322 3—4 (Brownish terra cotta).nbsp;• -u

lquot; MPKECHT 1933, p. 258) says about this colour: ..Dje
ut^g «e er Farbe Ln diskutiert werden. Ihre A-brldung .st
Xlich rn sehr hohem Grade von den MUieuverhaltmssen abhan-

^rlt-rf^- quot;ï«nbsp;^^^nbsp;der Regel unglcch-

mLt verteüt.... Bei den in Schweden herrschenden Witterungs-
ssen „w das Ageratumblau gewöhnlich mehr oder wem-
g d Utlich ausgebildet, rn sehr warmen und trockenen Sommern,
ITc Twquot; entsprach jedoch ein grösserer Teil der Proben Cu na-

o. bluish tinge m some plants or
ia^i» O, the XenaL X Fijne tros cross is caused genet,cally.
12. PJShCmCmGbV; brown violet.

« aVL, browmsh ring (corona^
quot; xt videt colour in the seedcoat may be totally absen , esp^.a^
in badly ripened or diseased seeds; then the colour .s about R.C. 34,
\ 4 or paL In some cases it was difficult to distmgmsh the nos 6
a'nt 12 (brown violet, clearly, especially in case of very dark
colours Discrimination between 12 (brown violet) aud black (nos

r8a:d24):asalwayspossible,atleastwhenthehgMw.

11. PJShCmc-MGbV; brown violet m o 111 e d, 12/10.

(cf. Lamprecht: Kastanienbraun/Ageratumblau).

The discrimination between 5 (violet mottled) and 11 (brown violet
mottled) is mostly rather clear. The brown violet dark pattern colour
is sometimes much faded.

13. P JShclcUgBv; gray brown,
(cf. Lamprecht: Havannabraun, 1932e, p. 57, 1933 p. 256). '

Greyish brown colour, often with a very faint lilac or violet tinge.
C.S. PL XL 1Tquot; a (WoodBrown-Avellaneous); the browner colours be-
tween the latter and PL XXIX 17quot; i (Tawny Olive); the seeds with
faint violet tinge PL XLVI 15quot;quot; a (Drab-colour). R.C. between 303,1
(Snuff Brown) and 354, 1—2 (Otter brown), with a very faint violet
tinge. Hilumring C.S. PL III 13k (Amber Brown); R.C. 328,
3_4 and 321, 4 (Bistre, Dead leaf) and darker.

The „Havannabraunquot; colour of Lamprecht is somewhat brighter I
The colours described here are distinctly darker and more
greyish, less brownish and greenish.

15. P JShCmOiigBv; greenish brown,
(cf. Lamprecht: Miinzbronze, 1932a, p. 173).

C.S. PL III 17 m (Raw Umber) and lighter; the most greenish
colour PL IV 19 m (Medal Bronze) and lighter; the most brown
colour PL III 16 m and paler. R.C. the darkest colour 343, 3—4
(Chocolate), the greenest 298, 2—3 (Golden bronze green), the more
brownish between 298, 2—3 and 303, 2—3 (Snuff Brown).

This greenish brown colour no. 15 is always markedly different
from the brown colour no. 21 and also from the corresponding cM cM
no. 13, which is more greyish. As to the comparison with Lamp-
recht's Miinzbronze, my no. 15 is less greenish, more „chocolatequot;,

especially the darkest colours.

14. P JShCSicMgB v; greenish brown mottled, 15/13.

(cf. Lamprecht: Miinzbronze/Havannabraun).

Mottling always clear. Not easily confused with any other

type.

16.PJShcMcMgBV;gray brown tinged with slate blue.

(Lamprecht: tiU now not described).

Just as in the nos 4, 10 and 22, the extension of bluish tinge is
extremelyvariablein seeds of one and the same plant. In
some plants or families the bluish tinge is nearly (or totally) wanting.
The colour withoutbluishtinge agrees exactly with the

corresponding gray brown v colour no. 13 (but the hilumring is less
bright): C.S. PL XXIX 17quot; i (Tawny Olive), 15quot; i and k (SayalBrown,
Snuff Brown); R.C. 303,2 (Snuff Brown) and 307, 3—4 (Dark fawn).
The darkest bluish tinge is C.S. PL XLVIII 43quot;quot; 1 (different
Slate colours); in R.C. the exact colour is not to be found; between
231,3 (Indigo) and 348,2 (Bluish black). Paler colours (blended with
the gray brown groundcolour) are R.C. 209, 3—4 (Smalt blue); C.S.
between PL XLVIII 41quot;quot; i (Dark Medici Blue) and PI. LII 35quot;quot;' i (Cas-
tor Gray). The hilumring colour of the dark bluish tinged
seeds is brown, covered with blue.

The black colour nos 17 and 18 will be described below together

with 23 and 24. ^

19. PJShcMcMGBv; (pale) brown,
(cf. Lamprecht: Rhamninbraun, 19326, p. 57).

This colour is rather variable. Brightest colour R.C. 297,2—4
(Brown pink); often more greyish, between R.C. 297,3—4 (Brown
pink) and 303,2—4 (Snuff brown). In C.S. represented less exactly;
the palest colours between PL XV 17' and 15' i (Yellow Ocher-Och-
raceous Tawny); the darker colours between the above mentioned
ones and PL XV 15' j. Hilumring R.C. 328,3 (Bistre)-308,3
(Fawn).

There are two colours which are sometimes difficult to distii^ui£i
from no. 19, viz. 21 and 13. No. 21 is the corresponding brownCmCm
colour (cf. below) and is as a rule of a darker, deeper brown; the hi-
lumring contrasts in number 19 much more strongly with the seed-
coat than in number 21. Concerning the difficulties in discriminating
between 19 (brown) and 13 (gray brown) I believe that the greyish
tints in no. 19 are caused by the same very faint violet as sometimes
appears in beans belonging to number 13.

As to the comparison with Lamprecht's Rhamninbraun, my no.
19 is generally somewhat more greyish, just as is the case with my
no. 13 compared to Lamprecht's Havannabraun.

21. PJShCmamp;nGBv; brown,
(cf. Lamprecht: Mineralbraun, 1932a, p. 173).

C.S. rather exactly PL III 13 m, 15 k, 17 k (Argus Brown, Sudan

Brown, Antique Brown).

R.C. 304,2—3 (Burnt Umber), but often somewhat less reddish,
with transitions to 297,4 Hilumring very little contrasting with

the seedcoat; C.S. PI. Ill 13k and 15 k (Amber Brown, Sudan
Brown); R.C. 308, 2—3(Fawn) and 304,2—4 (Burnt Umber).

For the difference between the brown colour no. 21 and the cor-
responding cM cM colour no. 19, cf. above.

The reddish Mineralbraun (R.C. 339) of Lamprecht does not oc-
cur in my materials; the other colours he refers to, are present.

20. P JShCmc'MGBv; brown mottled, 21/19.
(cf. Lamprecht: Mineralbrau n/R hamninbraun).
Mottling always easy to see.

22. PJShc'Mc'MGBV; (pale) brown tinged with greyish
indigo, (cf. Lamprecht: Graulich Indigo; 1933 p. 258).

This colour type again shows the same variability in the amount of
bluish tinge as the numbers 4, 10 and 16; many mixed colours of
brown and bluish occur. And, because the brown groundcolour is also
rather variable (cf. no. 19), this no. is annoyingly multicoloured! See
the backcross F^ X Fijne tros and reciprocal one (p. 196). Brown and
greenish black seeds (at least near the hilum) sometimes occur in the
same plant; in some plants or families the bluish tinge may be nearly

(or totally) wanting.

The brown groundcolours are C.S. PI. XV 17' i
(Buckthorn Brown), 15'j—14'k (Ochraceous Tawny, Cinnamon
Brown, Russet); sometimes more greyish, C. S. PI. XXIX 16quot; i and
15quot; j. In R.C. 307,4 (Dark fawn), 303,2—3 (Snuff brown), 304,2—3
(Burnt Umber). The darkest bluish tinge (close to the carun-
cula) is C. S. PI. XLVIII 39quot;quot; k (Saccardo's Slate-Dark Greyish Blue
Green), R.C. 232, 4(Greyish Indigo). Sometimes more greenish: C.S.
PI. LI 23quot;quot;'j (Dark Olive Gray-Iron Gray). R.C. 351,1 (Greenish
black). Many paler brown-grey-green-bluish mixed colours occur!
Hilumring R.C. 322,2—3 (Brownish terracotta), 308,3 (Fawn)
and darker, sometimes mixed with a bluish colour. It is not always
possible to distinguish with absolute certainty between the nos 16
and 22. Generally speaking no. 16 is more gray-bluish, 22 rather

brown-greenish blue.

Lamprecht (1933 p. 258) remarks about his colour „Graulich
Indigoquot;: „Diese Testafarbe zeigt ähnlich wie Ageratumblau eine
„sehr ungleichmässige Verteilung und Ausbildung. Gleichwie bei
Ageratumblau ist die Ausbildung von typisch Grauüch Indigo sehr
von den Witterungsverhältnissen Abhängig.quot;

18. P J ShCmCmgB V; black.

24. P J ShCinCmGB V; black.
(Cf. Lamprecht: Schwarz; 1932a, p. 174).

R.C. 349,4 (Black, pure) with transitions to 348,4 (Bluish black),
350,3—4 (Ivory black) and 351,3—4 (Greenish black). C. S.Pl. LII n,
Lin n and black. Hilumring black. It was not possible to make
any sharp discrimination between the different black colours of the
seedcoat.

Lamprecht (1932a, p. 174) describes a colour „(dunkel)
Chromgrii nquot;, which according to his experience always has
the constitution PP CC JJ gg Bb Vv. In my material I also found
this colour with possibly the same constitution, but not always
markedly different from the remaining black colours (cf. p. 189 and
p. 191). Of the self-coloured black-seeded plants out of backcross F^
with Wagenaar one half is (P J Sh Cm Cm) Gg Bb Vv, the other half is
(P J Sh Cm Cm) gg Bb Vv. All plants were black or very dark greenish
black. Especially among the offspring of the latter (gg Bb Vv) some
rather conspicuous greenish black seeds were found. This colour is
C.S. PI. XLII 37quot;' n (Dusky DuU Green-Black) but much more shiny;
R.C. between 236,4 (Chrome green) and 351,3 (Greenish black).

17. PJShCmcMg BV; black mottled, 18/16.

(Lamprecht: till now not described).

23. P JShOncMGBV; black mottled, 24/22.

(cf. = Lamprecht: Schwarz/Graulich Indigo).

Though the background colour types of nos 17 and 23 can be dis-
criminated, the mottled types must be taken together. The mottling
is very conspicuous if the background colour is without any bluish
tinge. In a few cases, when the tinge is extremely strong and dark
the mottling is hardly to be discovered.

As to the sh colours (nos 25—48) I am not yet able to give as accu-
rate a description of them as of the Sh colours (nos 1—24) because of
their great variability.

25. P J sh c'M cM g b v

31.PJshcMcMGb V Hilumring typequot;.

37. P J sh cM cM g B V quot;

43. P J sh cM cM G B v

„Hilumring t y p equot; of pale rose flowering plants. Seedcoat very
slightly coloured with a very pale greyish cream; in most cases a certain

nervationquot; is to be seen. The hilumring types are with or without
coloured „eyequot; around the hilumring. I am not quite sure of it as
yet, but probably the colour of the eye is influenced by the factors
G, B (and V) in an analogous manner as the corresponding totally
coloured Sh numbers. —The pale cream colours are: C.S. paler than
PL XXX 19quot; f (Cartridge Buff), sometimes with a touch of reddish;
R.C. paler than: 66,1 (Pale Ecru), 138,1 (Salmon flesh) and 135,1
(Pale pink). Hilumring: the paler colours C.S. PL IH 15 i
(Mars Yellow); R.C. 316,4 (Mars Yellow); the darker colours C.S.
between PL H 11 k and PL III 13 k; R.C. 333, 3—4 (Indian Chestnut
red) sometimes between this colour and 308,4 (Fawn).

About the same colour as the A b c D type of Kooiman and may
be the pale buff or light ecru of the race Blue Pod Butter used by

Shaw and Norto^

27 PJshCmCmgbv; (dull) yellowish.

C.S. PL XVI 23' b (Citron YeUow) and 22'd—20'd (Barium,
Straw and Naples Yellow). The same series in R.C.: 18, 1—2 (Sul-
phur Yellow), 30,3—4 (Cream Yellow), 29,3—4 (Naples Yellow).

This colour and the corresponding yellow Sh colour (Wagenaar
type) could not always be nicely discriminated.

33. P JshCmCmGbv.

About this colour (corresponding with the orange Sh colour no. 9)
I have as yet no complete certainty. Probably it is yellowish and
difficult to distinguish from the yellow no. 27.

39. P JshOiiCSigB v; dull greenish brown.

About the same colour as the greenish brown Sh colour no. 15,
but paler and less shiny (dull). R.C. between 298,1-2 (Golden bronze
green) and 303,2 (Snuff brown). C.S. PL IV 19 1 (Orange Citrme-
Medal Bronze) and (browner) transitions to PI. XV 16' k. (Cinnamon
Brown-Dresden Brown).

45 P JshOiiC'^iGB v; dull brown.

C S between PI. III13 mand 15 k (Argus Brown-Sudan Brown)
or between PL XV 17' i and PL XXIX 15quot; i. R- C. 304,2 (Burnt
Umber) with transitions to 303,2—3 (Snuff Brown).

(dull) yellowish, 27/25.
dull yellowish (?), 33/31.
dull greenish brown, 39/37.
dull brown, 45/43.

These mottled yellowish, yellowish(?), greenish brown and brown
types by their dull dark pattern colour and especially by their nearly
colourless background are greatly different from the corresponding
Sh colours. The colour of no. 32 (yellowish?) is not yet known for a
certainty.

28. P J sh c'M cM g b V

34. P J sh cM cM G b V

40. P J sh cM cM g B V

46. P J sh cM cM G B V

„Hilumring typequot; of violet flowering plants. Seedcoat very
slightly coloured, often showing a certain „nervationquot;. With or
without coloured „eyequot; around the hilumring. The seedcoat colour is
R.C. 6,1—2 (Purplish tinted white) and 9,2—3 (Fleshy white); the
first of these colours is not accurately represented in C.S.; the second
corresponds with PI. XXX 21quot; f (Ivory Yellow). Sometimes near the
caruncula and germ root a very small spot with a gray-greenish blue
tinge: R.C. 206,1 (Succory blue); C.S. PI. LII 35quot;quot;' d (Dawn Gray).
Hilumring: pale colour: C. S. PI. XV 15' i (Ochraceous
Tawny); R.C. 314,3 (Ru Ochre); darker colour: C.S. between PI.
Ill 13 k and PI. XV 13' k; R.C. 308,3 (Fawn).

In most cases the difference between these V and the corres-
ponding v hilumring types is very slight; but the hilumring colour of
V plants is always less bright than that of v plants.

30. P JshCmCmgb V; dull violet.

36. PJshQnCmGbV; dull brown violet.

Often brown, nearly without violet.

I am not able to describe these colours 30 and 36 accurately, be-
cause of their great variability.

42. PJshCmCmg B V and

48. PJshCmCmGBV; dull gray greenish black.

C.S. PL XLVI 21quot;quot; m and n (Olivaceous Black-Black) and 17quot;quot; n
(Chaetura Black-Black). R.C. 351,2—4 (Greenish black) and 350,
1—2 (Ivory black).

29. P J sh Cm cM g b V
35. P JshtocMGb V
41. P JshCmcMg B V
47. P Jsh(^cMGB V

dull violet, 30/28.
dull brown violet, 36/34.
dull black, 42/40.
dull black, 48/46.

The duU violet, dull brown violet and dull black mottled types
differ on account of their duU dark pattern colour and nearly
colourless background greatly from the corresponding mottled violet
and black Sh colours.

The difference between Sh shiny black mottled and sh duU black
mottled with nearly colourless background may be seen in fig. 2.

§ 6. Relation between stem, flower and seedcoat colours

In my cross a white seedcoat was always accompanied by white
flower and green stem (hypocotyl and cotyledons); a coloured
seedcoat always by coloured flower and stem.

The latter, however, is not always the case. On my reviewing the
literature, the following statements may be made.

a. Coloured-seeded races may have coloured or white flowers.

White-seeded races nearly always have white flowers and a green
stem. According to Fruwirth (1924 p. 179) „finden sich aber auch
Formen, welche Rosa, Violet, Purpur als Blütenfarbe und Weiss als
Samenfarbe zeigenquot;. In several descriptive works on bean varieties
(Von Martensen, Denaiffe, Tracy, Steinmetz) I found recorded
only one white-seeded race with coloured flower (Denaiffe p. 176 :
Haricot nain Prolifique; fleurs blanches, souvent plus ou moins
teintées de rose; grains blancs). In the genetical literature I did not
find any case mentioned.

h. Coloured-flowering races may have a stem (hypocotyl and co-
tyledons) with or without colour.

White-flowering races seem to have in all cases a green stem, at
least green hypocotyl and cotyledons (Miyake c.s. 1930). According
to Fruwirth (1924 p. 179) „ist weisse Blüte mit weissem Samen kor-
relativ verbunden, nach von Tschermak auch mit Fehlen von violet-
ten Flecken auf den Keimlappenquot;. Anthocyanin spots on the full
grown pods of white-flowering plants may occur (Tjebbes and
Kooiman V, 1921amp;; the author; cf. below).

About the appearance of colour in crossing colourless races, the
following facts are known (in most cases a factorial analysis was not
attained) :

1. Seedcoat.

White-seeded X white-seeded, Fj coloured-seeded. The cross of

Davis Wax with Michigan White Wax (Shaw and Norton, 1918 p. 65).

2.nbsp;Flower (in coloured-seeded plants).
White-flowering X white-flowering, F^ with coloured flower.
Mentioned by Shaw (1913 table 9); the parent plants did not

have a totally coloured seedcoat, but an „eyedquot; one.

3.nbsp;S t e m (in coloured-flowering plants).

Green-stemmed X green-stemmed, F^ with coloured stem. Some
crosses mentioned by Miyake c.s. (1930); one parent with „striped-
flower, the other with a totally coloured flower.

Stem and flower colour may either be paler or darker, either
more reddish or more bluish. Up to now, however, only monofactorial
segregations for stem and flower colour have been satisfactorily
analysed.

Shaw pubhshed in 1913 „The inheritance of blossom colour m
beansquot;, without giving a factorial analysis. The difficuhies of classi-
fying must have been rather great, as will be seen e.g. on close ex-
amination of his table 9, which contains many inconceivable results.
The segregation into the flower colours „light pinkquot; and „pinkquot; was
apparently monofactorial. Black-seeded beans seemed always to

have a pink flower colour.

Miyake c.s. (1930) found monofactorial segregation for:
„pinkquot; stem and flower (coloured dilutely) versus „redquot; stem and

flower (coloured intensely).

Tjebbes and Kooiman (V, 1921amp;) reported a spontaneous hybrid
of a light lilac-flowering race with red striped seedcoat and red
striped pod. All the hybrid colours were darker and more bluish:
flower violet, seedcoat bluish black striped, pod dark blue striped.
The Fa segregated according to the ratio 4:3:9.

pod colournbsp;flower colournbsp;seedcoat colour

I 1 pale red 1nbsp;^hite

^ 3 pale blue j

rednbsp;lilacnbsp;without blue

6 light violet 1nbsp;with blue

3 dark violet

The segregation of my cross for the factors P and V shows the
same relations in all respects:

pod colournbsp;flower colournbsp;seedcoat colour

4 little spots whitenbsp;white

(hypocotyl green!)

3 rose pale rose or lilacnbsp;without blue

f 6 violet-1 1nbsp;with blue

9 blue-violet 13 ^ioiet.2 or 3 (nbsp;or violet

The blue-factor B1 of Tjebbes and Kooiman and my factor V are

possibly the same.

In many other investigations analogous relations between flower
and seedcoat colours can be traced, though the required observations
have hardly ever been made.

In this respect I want to refer to Lamprecht's investigations.
The V-plants he used were the black wax varieties „Negerquot; (1932a,
p. 178) and „Merveille du Marchéquot; (p. 190). Their flower colour is
R.C. 189, 1 (Bishops violet) but somewhat brighter. The v parent of
his cross no. II has a yellow-brown (bistre) seedcoat (cf. my no. 9);
the flower colour is R.C. 187,1 (Pale light lilac) but much paler. F^
was violet flowering like one parent plant. Lamprecht does not
mention the connection between flower and seedcoat colours in Fg. Yet
the above-mentioned facts were decisive for me (in connection with
the seedcoat colours) to assume my factor V to be the same as
Lamprecht's.

Many of Lamprecht's v-races (seedcoat Schamois, Bister, Miinz-
bronze, Mineralbraun) are white-flowering. White-flowering races
with black or violet seedcoat have scarcely been described, unless
they are partly coloured, „eyed-seededquot;. The connection between
flowercolour and „eyednessquot; of the seedcoat deserves closer in-
vestigation.

johannsen (1926 p. 443) crossed a white-flowering, yellow-seeded
race with a violet-flowering, black-seeded one. Fj violet-flowering,

black-seeded. F^:

white-flowering (v)nbsp;violet-flowering (V)

yellow-seeded bronze-seeded violet-seeded black-seeded

39 (bv)nbsp;121 (Bv)nbsp;105 (bV)nbsp;293 (BV)

The influence on seedcoat colour of the factors B and V of Lamp-
recht and the author is the same as in this cross of Johannsen.
To wind up with we may say that there is very little known with

certainty about the connection between stem, flower and seedcoat
colours. Unifactorial segregation has been shown (or is at least
probable) for the following flower-colours:
Tjebbes and Kooiman lilacnbsp;violet

Johannsennbsp;whitenbsp;violet

Shaw and Norton pale pinknbsp;pink

Miyake c.s.nbsp;pinknbsp;red

Prakkennbsp;very pale rose (-lilac) violet

In probably all these cases differences in flower- and seedcoat
colours go together. With pale (or white) flower colour generally
correspond yellow and brown seedcoat colours (and red ?). With
darker flower colour violet, brownviolet and black dark pattern
colours (CC); the corresponding background colours (cc) are often
variably tinged with violet or blue.nbsp;^ ^

I may remind of the fact that in my P J Sh cM cM V background
colour types (nos 4, 10, 16 and 22) the violet tinge may be totally

lacking. The corresponding V and v colours are nearly indistinguisha-
ble in that case.

§ 7. Mottling

In § 3 I ascribed the ever-segregating mottling to the influence of
a dominant factor for mottling M, which locally suppresses the influ-
ence of the (complementary) colour factor C, while C and m are abso-
lutely (or nearly absolutely) linked just as c and M.

This hypothesis of the linked factors C and M has consequences
which I will deal with in giving a summary of the various views on
the genetical base of mottling in beans.

As to their inheritance, two types of mottling in beans are known:
a. True-breeding mottling („konstante Marmorie-
rungquot;) of many mottled races. Such a race crossed with a selfcoloured
one gives a mottled F^ and F^ shows a segregation into 3 mottled and 1
selfcoloured (or, if one parent was white-seeded), into 4 white-seeded,
9 mottled and 3 selfcoloured. Part of the mottled F^ plants breeds
true in F3, the other part again segregates into 3 mottled: 1 self-
coloured.

Ever-segregating mottling („Heterozygotmar-
morierungquot;). In our crossing two selfcoloured plants a mottled F^
may appear. In F.^ mottled and selfcoloured plants occur in the 1:1

ratio (if one parent was white-seeded, the F^ ratio is: 4 white-seeded,
6 mottled and 6 selfcoloured). None of the mottled Fg plants are
true-breeding for mottling in Fg or later generations.

The phenotype of the two sorts of mottling is much the
same and it seems doubtful to me, whether they can generally be
distinguished one from the other.

According to Tjebbes and Kooiman (II, 1919) the pattern should
be nearly the same; in the ever-segregating type, however, there
should occur only one background and one dark pattern colour,
whereas the dark pattern in true-breeding mottled plants which they
investigated, consisted of two colours; in a „blackquot; mottled bean e.g.
of parts with very dark blue cells and parts with lighter blue ones.

Kristoeferson (1924) writes: „As to the phenotype both are
similar.quot;

And Lamprecht (1933 p. 260): „Es verdient hier besonders hervor-
„gehoben zu werden, dass zwischen dem gewöhnlichen Typus von
„konstanter Marmorierung, verursacht durch ein besonderes Mar-
„morierungsgen (M), und der durch die Konstellation C c beding-
„ten Heterozygotmarmorierung kein sicherer Unterschied in Bezug
„auf die Zeichnung der Marmorierung hat festgesteUt werden kön-
quot;,nenquot;. In his latest article he says (1934 p. 179): „Sowohl die hetero-
„wie die homozygotmarmorierten Samen zeigen in ihrer Zeichnung
„eine recht beträchthche Variation und scheinen auf Grund dieser
„häufig nicht sicher voneinander unterschieden werden zu können..
„Für die heterozygotmarmorierten Samen ist von mir an einem
quot;.grossen Material nachgewiesen worden, dass die dunkleren Flecken
quot;,der Testa stets der durch einen Genotypus mit CC bedingten Testa-
^}arbe entsprechen, die des helleren Grundes einem im übrigen glei-
quot;,chen Genotypus mit cc. Hier ist die Farbenverteilung also in ihrer
quot;Abhängigkeit von der genotypischen Konstitution vollkommen be-
quot;kannt. Wie die Farbenverteilung bei den homozygotmarmorierten
quot;.Samen durch das Zusammenwirken von M mit den Farbgenen für
quot;die Testafarbe beeinflusst wird, darüber scheint bisher nichts siche-
quot;res bekannt zu sein. Aus oben Angeführtem geht klar hervor, dass
''die Heterozygotmarmorierung nur zweifarbig auftreten kann. Die
,'.homozygotmarmorierten Typen sind diesbezüglich kaum unter-
,'sucht. Soweit mir bekannt, kommen hauptsächlich dreifarbige
„Kombinationen vorquot;.

In the earher investigations (Tschermak, Emerson, Shull a.o.)
differences between the two types are not mentioned.

As to the genetical base of motthng in beans and the
connection between true-breeding and ever-segregating mottling

two contrary views exist:

The two types depend on different factors which are inherited
independently (Emerson's first conception. Kooiman. Lamprecht).

h. The two types depend on the same factors (Emerson-Spill-
man, Tschermak, Shaw and Norton, the author).

As concerns the ever-segregating mottling, it is of importance to
remark that the first investigators (Tschermak. Shull. Emerson)
did not observe that with every mottled type there go two self-
coloured ones. Among mottled and selfcoloured

beans in a cross they distinguished the same

colours.

Shull e.g. (1908) crossed:
Long Yellow Six Weeksquot; (light pink flower, yellow seed as my
Wagenaar race) with „White Flageoletquot; (white flower and seed).
The Fl generation had pink flower and black mottled seed. In F, he
distinguished two main seedcoat colour types:

„brow nquot; (= dark seal brown, dark greenish brown, dark yellow

brown, light yellow) and
blackquot; (= black, weathered black, purple, violet; m some
quot; plants the colour is but p a r 11 y black or violet and the other
part is a brown colour „underlyingquot; the black or violet).
The colours in Fg of this cross were probably t h e s a m e a s m y
Sh colours in the upper half of the scheme, table 31.
My factor B is probably Shull's „dark brown factorquot; D.
My factor V is his „anthocyanin producing factorquot; B.
On my factor G depends the unifactorial difference between
Shull's races „Long YeUow Six Weeksquot; and „Ne Plus Ultraquot;
(orange); this factor is not named by Shull.
The figures in Fg were:

„brownquot; (= my v) „blackquot;(= my V) unclas-

white rii^ttled selfcoloured mottled selfcoloured sified
160nbsp;39nbsp;59nbsp;154nbsp;159nbsp;12

Ratio: 16nbsp;6nbsp;6nbsp;18nbsp;18

Background colour cc and dark pattern colour CC were apparently
taken together! „Brownquot; my columns I and III (v cc and vCC);
„blackquot; my columns IV and VI (V cc and V CC; probably all V cc
colours, my column IV, were strongly tinged with violet; some plants
showed a brown colour „underlyingquot; the black or violet!)

Shull rightly understood that these results were explainable on
the assumption that all individuals heterozygous for a factor M have
the mottled pattern, and that mm and MM are selfcoloured (those
mm and MM types could not be distinguished in his opinion).

Tschermak (1901, 1902, 1904, 1912) just as Shull, distinguished
among (ever-segregating) mottled beans and selfcoloured ones t h e
same colours, viz. his main colour types
blacknbsp;C B

violetnbsp;c B and

brown (with yellow) C b and c b
His cross

„Weisse Wachsschwertquot; (white flower and seed) with
'^Non plus Ultraquot; (pale violet flower, orange seedcoat) gave an
Fl generation with dark violet flower and black mottled seed.
In Fa Tschermak could „leicht 20 verschiedene Farbenklassen
konstatierenquot;. Apparently here too all my Sh colours occurred:
segregation for my factors B (= C of Tschermak), V (= B of
Tschermak) and probably for my factor G (G dominant or re-
cessive does not alter the division into three main colour types
black, violet and brown with yellow).

I discuss the cross Weisse Wachsschwert x Non plus Uhra, be-
cause Tschermak gave (1902 Tab. II) of 98 f^ individuals an accu-
rate description as to their flower and seedcoat colours; later
(1904 Tab. Ill) he gave of the same 98 individuals the division mto
the three main colour groups black, violet and brown.
In different F^ families of this cross he found together:
whitenbsp;mottlednbsp;selfcoloured

167nbsp;(= 4:6:6)

brown violet black brown violet black
38nbsp;33nbsp;92 101 27nbsp;39

3nbsp;9nbsp;9nbsp;3nbsp;4

So he found for the three colours „inversion of the ratioquot; among

mottled beans (4:3:9) and selfcoloured ones (9:3:4). Tschermak
suggested different possibilities as an explanation, but he finished
(1912 p. 187) by saying: „Eine vollbefriedigende Erklärung der Um-
kehr des Spaltungsverhältnisses bleibt noch zu findenquot;.

By accurate comparison of his extensive description of flower and
seedcoat colours of those 98 F^ plants with their later division into
black, violet and blue, I arrived at the apparent solution of this re-
markable „inversion of the ratioquot;.

Among mottled seeds the normal 4:3:9 ratio is found, according to
my factors B and V and possibly G (cf. the colours in the upper
half of my scheme):

1 b V Cm cM 1nbsp;^Q^tigd^ ^ith pale lilac flower.

3 B V Cm cM J quot;
(my nos 2, 8, 14 and 20).
3 b V On c'M, violet mottled, with lilac or dark lilac flower.

(my numbers 5 and 11).
9 B V Cm c'M, black mottled, with lilac or dark lilac flower,
(my numbers 17 and 23).

The „inversionquot; among the selfcoloured seeds is probably caused
by the classification of the background colour
types of black mottled seeds ^n d. v i o 1 e t m o 11-
I e d ones! Of these colours, my P J Sh cM cM V colours (nos 4,
10, 16 and 22), which are pale yellow, pale orange, gray brown and
pale brown, variably tinged with violet or blue,
Tschermak may have classified:

a.nbsp;Not a single one as „blackquot;.

b.nbsp;Those with strong violet or bluish tinge as „violetquot;.

c.nbsp;Those without or with a faint tinge as „brownquot;.

Leaving out of consideration the groundfactor P, the ratio in the
upper half of my scl^m^in table 31 (J and Sh dominant; segregation
for the factor pair Cm-cM and the factors G, B and V) is:

V

cMcM CmcM CmCm cMcM CmcM CmCm

gt;violet gt; violet

5inbsp;9)

jblack black

i 27 J
48

Classified bynbsp;-g ^

Tschermak as: _.nbsp;_. [o g

Mottled: 32 brown, 24 violet, 72 black = 4:3:9.
Selfcoloured:

„brownquot; 16 16 part of 48
violet 12 remaining part of 48
black 36.

The flower colour of those 98 plants was (with 1 or 2 exceptions) in
agreement with my view:

all selfcoloured and mottled black plants: lilac or dark lilac;
aU selfcoloured and mottled violet plants: hlac or dark lilac;
ah mottled brown plants: pale lilac;

selfcoloured brown plants: partly pale lilac, partly lilac or dark lilac.

Emerson (1909«) was the first who tried to analyse the relation
between true-breeding and ever-segregating mottling, after crossing
a great number of white-seeded, mottled and selfcoloured varieties.

In his first hypothesis he assumed that both types of
mottling depend upon different factors. The symbols he used were:
P = groundfactor for colour.
M = factor for true-breeding mottling;

MM and Mm mottled, mm selfcoloured.
X = factor for ever-segregating mottling;
XX and XX selfcoloured, Xx mottled.

(X of Emerson = M of Shull = B of Kooiman = C of Lamp-
recht; Emerson and Shull, however, did not know the difference m
colour between the homozygous types!).

The possible genotypes of the races are:
1 PP MM XX mottled.nbsp;5. pp MM XX

2.nbsp;PP MM XX mottled.nbsp;6. pp MM xx ^^

3.nbsp;PP mm XX selfcoloured.nbsp;7. pp mm XX
4 PP mm xx selfcoloured 8. pp mm xx

According to these views, Emerson expected the following ratios
between white, mottled and selfcoloured beans:

Emerson consiaereu luc pusoiL-^Lx^^j-nbsp;------o-nbsp;-

only in xx and not in XX plants; in this case the ratio in 3« and h

(double heterozygous) would be:

3a. PPMmXxnbsp;0nbsp;H

ft. PpMmXxnbsp;16nbsp;33

Analyzing the resuhs of Tschermak, Shull and himself, Emerson

found no indication of ratios to be expected

incaseofdoubleheterozygosity:MmXx.

Emerson's secondhypothesis (1909amp;) about the con-
nection between true-breeding and ever-segregating motthng was
given on a suggestion of Spillman. The two types of mottling would
depend upon two factors, Y a n d z, w h i c h f a c t o r s
are absoJ.utely linked:

(PP) XZYZ true-breeding mottled race.
(PP) Yz Jzl

(PP) yZ yZ selfcoloured races.
(PP) yz fz

In white-seeded races there may of course occur the same 4
types concerning Y and Z. The various possible crosses and their
(white): mottled: selfcoloured Fa ratio are:

According to Emerson all known facts could be explained by
this Yz-yZ hypothesis as well as by his M X hypothesis of two inde-
pendent factors.

For Emerson the ^z-yZ hypothesis has two advantages:

1.nbsp;The two types of mottling, phenotypically so much alike, de-
pend on the same factors.

2.nbsp;The ever-segregating mottled F^ between two selfcoloured
races is more conceivable than with his monofactorial Xx hypothesis
(or Mm of Shull).

Tschermak (1912) likewise attempted to consider the „konstante
Marmorierungquot; and „Heterozygotmarmorierungquot; from the same
point of view. His „Association-Dissociationquot; hypothesis has some
analogy with the ITz-yZ hypothesis of Emerson-Spillman. Both
allow only of the ratio mottled: selfcoloured being 3:1 or 2:2. The
hypothesis of the absolutely linked factors is, however, more in
harmony with our general genetical conceptions.

The investigations of Shaw and Norton (1918), Kooiman (1920)
and Lamprecht (1931) gave an entirely new aspect to the matter!
They stated that to every class of mottled seeds
there does not belong one selfcoloured type
with the same number, but always two types,
each with half the number:

1nbsp;background colour type,

2nbsp;mottled beans,

1 dark-pattern colour type.

Shaw and Norton and Kooiman tried to explain this fact in
diametrically opposed ways. A third possible way has been chosen
by the author. These three views I will discuss under a, ^n^c.

a. Shaw and Norton (1918) persisted in using the Yz-yZ con-
ception of Emerson-Spillman and assumed a „modifyingquot; f a^t ^r
M which is absolutely linked to the Yz-yZ
factors. One of their coloured-seeded races (Blue Pod Butter)
had a „pale buffquot; seedcoat colour and produced a mottled F^ if
crossed with their yellow, brown and black races.
Blue Pod Butter:nbsp;Y^m Y^m

Yellow, brown and black races: yZM yZM
F^:nbsp;Yzm ^M, mottled.

According to Shaw and Norton the F^ of this mottled F^ always
segregated into:

1nbsp;pale buff coloured:

2nbsp;mottled:

1 yellow, brown or black.

(These relations are apparently the same as in my sh sh colours!) •
Shaw and Norton therefore concluded that a „modifyingquot; factor M
(linked with the Y-Z factor pair) must be dominant for the seedcoat
to be able to show mottling or colour (other than th^ pal^buff of
Blue Pod Butter). True-breeding mottled races are YZM YZM. The
only possible F^ ratios are3: 1 and 1 : 2: 1 (analogous to Emerson-

Spillman's hypothesis).

b. Kooiman on the other hand (1920, 1931) abandoned the linked
factor hypothesis and attributed the ever-segregating mottling to the
heterozygous state of his (complementary) factor B.

Lamprecht (1932a) arrived at the same conclusion as Kooiman,
but used the symbol C for the complementary factor involved.

True-breeding and ever-segregating mottling must now depend
upon different factors and the expected ratios between mottled
and selfcoloured seeds will be the same as in Emerson's first
hypothesis.

According to Kooiman and Lamprecht mottled races (MM) may
be (using Lamprecht's symbol) cc or CC.

The possibilities in crossing a mottled race with a selfcoloured

one are:

1.nbsp;The mottled race is PP MM cc.

a.nbsp;PP MM cc X PP mm cc.

Fl PP Mm cc.

Fa segregates into 3 mottled: 1 selfcoloured.

b.nbsp;PPMMcc X PPmmCC.

Fl PP Mm Cc.

Fa segregates into 14 mottled: 2 selfcoloured.

2.nbsp;The mottled race is PP MM CC.
a. PP MM CC X PP mm CC.

Fl PP Mm CC.

Fa segregates into 3 mottled: 1 selfcoloured.
h. PP MM CC X PP mm cc
Fl PP Mm Cc.

Fa segregates into 14 mottled: 2 selfcoloured.

So each mottled race, if crossed with a cc or with a CC self-coloured
race must give an Fg ratio of 3:1 in the one and of 14:2 in the other
case. Of these 14 mottled beans 12 should be MM or Mm; the remai-
ning 2 mm Cc, i.e. beans of the ever-segregating type must appear!

To my knowledge, the proof that really a 14:2 ratio occurs
■ has up to now not been given.

Some crosses of Shaw and Norton seem to prove the contrary!
They crossed some mottled races with a cc race (Blue Pod Butter;
their Yzm) and with a yellow, brown or black CC race (their yZM
races). The results (derived from their table II) were:

It is impossible to find in each pair of these crosses the 7:1 ratio
for the one and 3:1 for the other cross.nbsp;^

c. My own hypothesis closely approximates the Yz yZ hypothesis
of Emerson-Spillmann and has the same advantages and numerical
results as concerns the possible ratio's mottled: self-coloured (cf. p.
223). In my opinion mottling (true-breeding and ever-segregating) is
due to two factors which are absolutely (or nearly absolutely) linked:
the complementary factor C and

the factor for mottling M, which locally suppresses the action of the
dominant complementary factor C.
CM CM true breeding mottled race.
Cm cM ever segregating mottled type.
Cm Cm

cM selfcoloured races,
cm cm

According to this hypothesis the dark pattern colour of mottled
races is CC. I cannot judge whether this really always holds good

in every case.nbsp;^ ^

Lamprecht (VI; 1933 p. 313) says about the Yz-yZ hypothesis of
Emerson-Spillmann that aU known facts about the two types of
mottling can be explained with it. But „nur müsste dann die sehr
„wenig wahrscheinhche Annahme gemacht werden, dass Y stets mit
quot;c parallel geht und Z mit c oder umgekehrt. Dann erscheint diese
Theorie aber überflüssig und unnötig kompliziert.quot; This objection
does not obtain for my Cm c'M hypothesis, though I must admit that a
cm cm race up to now has not been found.
The possible crosses are:

CcM

cmy

Miyake c.s. (1931) mentioned some cases of crossing over
which with the hypothesis ofKooiMAN andLAMPRECHTare not possible.
The races involved were N7 with red flower and cream seedcoat and
B2a with red flower and black seedcoat. F^ showed a black mottled

seedcoat. Fg consisted of:

111 with dark-pattern colour: black, brown, purple.
273 mottled: black, brown and purple.
121 with background colour: cream.

In Fg they found that:

a. 61 famihes of mottled Fg plants segregated into 322 dark pattern

colour, 616 mottled, 269 cream.
h. 30 families of dark pattern colour Fg plants gave only 577 dark
pattern colour plants.

c.nbsp;27 families of cream Fg plants gave only 609 cream.

These are the normal cases. They found, however, a few ex-
ceptions:

d.nbsp;4 families of mottled F^ plants gave 32 mottled and 17 cream, but
not a single of the dark pattern colour.

e.nbsp;1 family of a dark pattern colour Fg plant segregated into 7
with the dark pattern colour and 4 cream.
Miyake c.s. adopt the Yz yZ hypothesis of Emerson-Spillman

and assume these factors for mottling to be „linked with a factor for

creamquot;. I wiU discuss their results with the aid of my factors C and M.

The genetic constitution of the mottled F^ isCmcM. Its gametes are:

C^ilnbsp;CMl

^ y non-cross-overs.nbsp;— V cross-overs.

cM Jnbsp;cm j

The 4 mottled F^ plants which segregated into mottled and cream

may be considered asCM cM. Among thejr nettled offspring true-

breeding mottled plants (CMCM) must occur. It is a

pity that Miyake c.s. communicate nothing about an F4 generation of

these plants. Now the possibility exists that the lack of dark pattern

plants was merely a chance occurrence.

The Fa plant with dark pattern colour and segregating into 7 dark

pattern colour and 4 cream wiU have been cm Cm. If my hypothesis

is right, these cream Fg plants (cm cm) c a n n o t^ r^ d u c e a

mottled Fl, neither with a cM c'M, nor with a Cm Cm race:

Here the test again fails, while the factorial analysis is not quite
sufficient, so that no definite opinion can be expressed.

Striping and double mottling will be discussed in a later paper in
view of the analysis of another cross. It seems to me possible that
the inheritance of ever-segregating mottling, true-breeding mottUng,
striping and double mottling can be brought under the same point of
view with the aid of a slight completion of the hypothesis given
above.

CHAPTER H

characters of the pod wall

A. Strength of the string

§ 1. Previous investigations

Difficulties in classification according to strength of string always
played an important part in the investigations.

Emerson (1904) reported on crosses between stringy and stringless
podded varieties. The pods of F^ plants were sometimes intermediate
between the parent races, while in other crosses they were very
nearly stringless, so that the difference between them and the pods of
the stringless parent was scarcely discernible. In the Fj generation of
Fl hybrids in which stringlessness was dominant, only stringless and
stringy forms occurred and no intermediate ones (65 stringless and 33
stringy). In the Fs-generation all stringy F^ plants bred true. Part of
the stringless ones also bred true; the other part segregated into 139

stringless and 56 stringy.

Where the Fi was intermediate all three types occurred in the Fa
(114 stringless, 80 intermediate, 78 stringy). As to strength of string
the intermediates, however, varied more in the second than in the
first generation. Part of the stringless and stringy Fa plants bred
true. The remaining stringless and stringy plants and all intermediate
ones segregated. From his table VIII we may derive as follows
about the progeny of segregating Fa plants:

Emerson therefore speaks of „reversal of dominancequot;.

Wellensiek (1922) crossed the stringy race Wagenaar with three
different stringless races. In the F^ generation of the three crosses
stringlessness appeared to be dominant, while in Fg a clear-cut
monohybrid segregation occurred (41 : 18, 46 : 12 and 49 : 25;
together 136 : 55). According to Wellensiek classification into
stringless and stringy did not meet with difficulties.

JoosTEN (1924) examined F3 and F^ generations of Wellensiek's
material. He distinguished 4 degrees of stringiness. The clear mono-
factorial segregation into stringy and stringless seemed not to be

confirmed.

Joosten moreover tested a lot of so-called „stringlessquot; races as to
the degree of their stringlessness. For this purpose he worked out
(according to length and strength of the string of boiled beans) a
scale of „string numbersquot; ranging from 1—10. He pointed out the
high variability of the character and the probable influence of ex-
ternal factors. On one and the same plant of some varieties pods with
both weak or strong strmgs may occur. Really stringless races were

not found.

At the same time he investigated the anatomical structure of the
pod, especiaUy of the sheath of the vascular bundle in the sutures.
He distinguished two principal groups of sheath-types:

1.nbsp;One type (type S) is characterised by the similarity of all the

cells in the sheath; all are s c 1 e r e n c h y m a t i c f i b r e s,

more or less impregnated with lignin.

2.nbsp;The other type (HS) is characterized by a narrow group ot
wood cells, which may sometimes have sclerenchymatic
fibres on the inside, found either alone or joined into larger or
smaUer groups. In varieties with type S not a single case of type
HS occurred. In my investigation I have apparently to deal with

the same two main types.

Currence (1930) in crossing stringy and stringless races found in

two crosses the F^ stringless, in an other one intermediate. He classi-
fied the Fg plants according to the area of sclerenchymatic fibres m
the sheath: stringlessquot;, with an area of less than 0.005 square mm
in each half quot;of the sheath; „stringyquot;, with more than 0.005 square
mm (this division lies between nos 3 and 4 of my fig. 4 p. 236).
Furthermore, Fg plants yielding only progeny with an area of 0.005
square mm or more were considered as homozygous strmgy; Fg

plants yielding only progeny with less than 0,005 square mm as
homozygous stringless; those plants of which the progeny included
both groups were considered to be heterozygous. This method of
classifying is rather arbitrary. According to Fj and F3 of his first
cross (Fj stringless) he arrived at the following results:

9 homozygous stringless,

27 „stringlessquot; Fg plants consisted of

17 heterozygous and
1 homozygous stringy.
4 heterozygous
9 homozygous stringy.

13 „stringyquot; Fg plants consisted of

The numbers of the three types (9:21 : 10) agree very weU with
the 1 : 2 : 1 ratio. For his second cross with stringless F^ Currence
arrived at the same result. The F^ of his third cross was distinctly
stringy (intermediate). According to the Fg progeny he estimated
that there were among 83 Fg plants (classified as 19 stringless and 64
stringy): 8 homozygous stringless, 44 heterozygous and 30 homo-
zygous stringy. These figures agree more or less with the ratio
1 : 8 : 7. In order to explain the data of his three crosses Currence
assumes one dominant factor S producing stringlessness and a second
factor T, which, when present, inhibits the action of the first factor
(i.e. only tt SS and tt Ss are stringless). The first two crosses would
be of the type tt SS (stringless) X tt ss (stringy), F^ stringless, Fg
ratio3stringless: 1 stringy. The third cross would be tt SS (stringless)
X TT ss (stringy), F^ stringy, Fg ratio 3 stringless: 13 stringy; the
stringless plants are tt SS and tt Ss. Currence tried to explain
Emerson's results on the same lines.

The main objection to this hypothesis of Currence is the arbi-
trary way of classifying into the two groups stringy and stringless,
even in the third cross with intermediate F^. The two types of sheaths,
photographed by him, correspond exactly with the two main types
of JoosTEN and my-self (Cf. nos 2 and 8 of fig. 4).

§ 2. The methods used

The partly unclear, partly contradictory results of the investi-
gations discussed above, showed the desirability of an accurate re-
examination. The two main principles in investigating the inheritance
of the string must be:

1.nbsp;To examine the strength of string together with the micro-
scopical anatomy.

2.nbsp;To avoid an a priori division into only two or three types.

Strength of the string.

From each plant 3 nearly full-grown pods were taken. They were
boiled (65 minutes) and afterwards stringed. After Joosten's
example I used a scale of „string numbersquot; ranging from 1 to 10 :

1.nbsp;Without any string.

2.nbsp;Length of string fragments less than 1 cm.

3.nbsp;Length 1—2.5 cm.

4.nbsp;Length 2.5—4 cm.

5.nbsp;Length 4—5.5 cm.

6.nbsp;a. Longer than 5.5 cm, but not continuous.

b. Continuous, but very tender and weak string.

7.nbsp;Continuous, rather weak string.

8.nbsp;Continuous string, rather strong.

9.nbsp;Continuous string, difficult to break; yet ravelling out when

drawn between two finger nails.

10.nbsp;Continuous string; hardly to be broken between the finger
tops; not ravelling out between the nails.

Dorsal and ventral string numbers are calculated by averaging the
numbers of three beans; the string number of each plant is the

average of the six numbers.

In some cases the six numbers differ greatly, the probable error of
the average therefore being high. But in most cases they do not
diverge so much. UsuaUy the ventral string number is somewhat
lower than the corresponding dorsal one.

The microscopical anatomy.

Strength of string depends upon the cell types which compose the
sheath of the vascular bundles in the dorsal and ventral sutures of
the pod. At the beginning of my investigations I described the anato-
my of the sheath in each plant (3 pods; 2 cross sections of each pod)
as accurately as possible. The following points, which refer to all
plants, soon became clear to me:

1.nbsp;The sheath always takes up about the same area.

2.nbsp;It consists of the same three cell types.

3. These cell types have always about the same relative
position.

Only the percentages of each cell type are variable. Therefore the
one thing I had to do was to ascertain (by estimating) the percentages
of the area taken up by each of these three cell types. For this
purpose I studied cross-sections of one nearly full-grown pod from
each plant. After some experiments I chose the double-staining
method with methyl-green and alum-carmine for the differentiation
of the cell types (Strasburger, 1923 p. 231). Besides all investi-
gations have been made with a polarization microscope.

1. T h e area (cf. fig. 3; plates I and II).

The sheath lies outside the dorsal and ventral vascular bundle.

The outer margin is
formed by a crystal-
layer (pl. II, nos 6 and
7). The inner margin
is not so sharply
marked. Yet outside
the phloem portion of
the vascular bundle
we find as a rule some
large parenchymatic
cells (cf. plates I and
II). They are consi-

dered to form the inner margin of the sheath. Tearing off the string
usuaUy takes place along these ceUs. The number of cell-layers
between the crystal-cells and those large parenchymatic cells varies
for the dorsal sheath from 5 to 8, for the ventral one from 4 to 7.
Outside the sheath collenchymatic cells are found (fig. 3; plates I
and II); in other regions of the pod wall only the hypodermic layer is

collenchymatic.

2. The cell types (cf. plates I and II).

a. Parenchymatic cells; non-lignified, with inter-
cellular spaces; red-coloured by the alum-carmine.

W o o d c e 11 s; ceU wall only slightly thickened, strongly
lignified. These ceUs are more or less isodiametrical (cf. pl. II nos 6 and
7). They are stained very dark by the methyl-green and do not show
double refraction. Cf. the nearly black ceU groups in plates I and II.

c Fibres; cell wall strongly thickened. In longitudinal sections
(pi II no. 7) they show the typical characters of fibres: very much
elongated, with narrow, obhque, simple pits. Stained with methyl-
green their colour is a pale bluish green. They show a conspicuous

double refraction.nbsp;t ^ tt\

3 The relative position (cf. fig. 4; plates I and II)

and the percentages of the three cell types are best discussed m the
description of the parent plants and F^.

§ 3. Description of the parent plants and F-^

Thflatomicri structure of the sheath is to be seen in fig. 4 nos 2
and 3 (sometimes between 3 and 4). Plate I no. 1 shows a micro-
photograph of this type. The outer part of the sheath consists of
wood ceUs (between 40 and 75%), the inner part of parenchymatic
ceUs The fibres very often form four more or less distmct cell groups.
This anatomical structure of the sheath corresponds with Joosten s
tvpe HS. In aU Fijne tros plants I grew in the years 1930-1933 the
peLntage of fibres varied between 0 and 25, usually between 5
and l5.ThestrengthofstringinFijne tros lies between2-3 and 5-6.

1930 (35 plants). Highest percentage fibres 10 (1 plant); generally
50/ or lower. Unboiled beans nearly stringless (boiled beans not
examined). Two plants (6-29 and 6-34) were crossed with Wage-
naar; their percentages of fibres were resp. 4 and 2 (average of 3

pods; 2 sections of each pod).nbsp;-r a 90

1931. Strength of string and percentages of fibres m famihe 6-29
are given in the subjoined table.

VENTRAL SHEATH

In aU plants the percentage of fibres was higher than in the mother

plant 6-29 in 1930. External conditions may probably influence the

percentage of fibres and the strength of string. I have tried to
examine these influences by halving seedlings and cultivating the
halves under different conditions of water supply. The results,
however, were not quite decisive. — In family 6-34 there occurred 9
plants with strength 3—4 and 3 with 4-5; these plants have not

been investigated anatomically.

1932. The three famüies belonged to the progeny of family 6-29
1931. Their string numbers and percentages of fibres are given in the
subjoined tables. Famüy 173 (string number mother plant 3.6;
percentage of fibres 10):

Family 178 (string number mother plant 4.4; percentage of
fibres 6):

Family 179 (string number mother plant 5.4; percentage of fibres
25):

The mother plant of family 179 had the strongest string and high-
est percentage of fibres among the members of family 6—29 in 1931.
It appears from the tables that in 1932 family 179 had the highest
strength of string, though not a single plant reached the string
number 6. Slight genetical differences seem probable.

1933. Cf. table 32. Between these families no conspicuous differ-
ences are found. Remarkable is one plant with strength 6.8 in family
327.

2.nbsp;Wagenaar.

The anatomy of the sheath is shown in fig. 4 no. 8 and pi. II no. 5.
Nearly the whole sheath consists of fibres. These fibres are separated
into two parts by a smaU group of wood cells and parenchymatic
cells in the middle. This group of ceUs nearly always has the same
shape as in fig. 4 and pi. II. In the dorsal sheath the wood cells
may be totally lacking in a few cases; the fibres then form one
continuous group. In the ventral sheath the group of wood cells and
parenchymatic ones is extremely narrow and is formed, without
any exception, by the two epidermal ceU layers. The Wagenaar
sheath corresponds with Joosten's type S (cf. p. 231).

Both anatomical structure and strength of string show hardly any
variability in the Wagenaar race. The string number is nearly
always 10, never lower than 9.

3.nbsp;Fl generation.

Its anatomy corresponds with fig. 4 nos 4, 5 or 6; pi. I no. 2 or
between 2 and 3. The percentage of fibres varies (one pod of each

plant has been investigated) between 20 and 65. Fig. 4 and plate I
show clearly that the relative position of the cell types is exactly the
same as in the Fijne tros race. The groups of fibres, however, are
greater and often form (fig. 4 nos 5 and 6) one continuous group in
each half of the sheath. In the ventral sheath the wood ceUs often
lie in three groups, separated by fibres.

The three F^ plants in 1931 showed (5 pods stringed and 5 ana-
tomically investigated) the following figures:

The remaining F^ plants (1932 and 1933) are shown in table 33.
The strength of string always Ues between 6 and 8. The probable
error of these string numbers is very smaU. The pods of some plants
with strength 6—7 received the quahfication „rather youngquot; (based
upon the very young seeds in the examined pods).

The variabiUty in the percentage of fibres in these F^ plants is very
high. Partly this will be due to the fact that but one pod of each
plant has been examined. In the F^ generation no differences between
the progenies of F^ plants with 20 and with 50% fibres have been
found (cf. next § and table 37).

§ 4. Analysis of the F^

In the correlation tables 34, 35 and 36 the three F^ families of 1932
(55-2, 55-4 and 55-6) are shown, arranged according to strength of
string and percentage of fibres. The correlation between these two
characters is rather strong and it might have been even stronger, if
the anatomy of more than one bean could have been investigated.
In a sheath without or nearly without fibres the percentage of wood
ceUs probably slightly influences the strength of string.

In some cases of very bad agreement between string number and
percentage of fibres I find the probable cause in my notes, viz. that

only one very bad pod was available for the string investigation.
Another remarkable cause of discrepancies between the two charac-
ters was the following. In Fg some plants have pods with strong
ventral curvation. When I gathered these pods it might happen that
the ventral string broke in one or more places without my noticing
it. The same breaking may occur in the boiled strongly curved beans.
This breaking of the string in curved beans I have not sufficiently
realized in the beginning of the Fg investigation. After becoming
aware of it, I reexamined the records, and found indeed that in
some cases of strong discrepancies (high percentage of fibres, low
string number) in curved beans the ventral string number was
extremely low; sometimes the unusual high strength of the short
ventral string fragments was mentioned. A third cause of inaccuracy
lies in the difficulty of always using the same standard for the
strength of string. In spite of these and possibly other causes of in-
accuracy the Fg and F3 results lead to very definite conclusions.

The undermost rows of the three tables 34, 35 and 36 show that a
division into three groups, Fijne tros, F^ and Wagenaar (according
to strength of string) is impossible. Nor can we discern a definite
Wagenaar type from a not-Wagenaar type, though the number of
plants with strength 8—9 is not very high.

With regard the anatomy of the sheath aU types of fig. 4 or plates I
and II occur. From this continuous series it becomes entirely clear that
the relative position of the three cell types is always the same. In this
respect I with to mention a remarkable analogy. Figure 4 no. 1 shows
those parts of the wood cells of no. 2 which are lignified first; these
parts are exactly the same as the narrow groups of wood cells in
the sheath types nos. 6 and 7.

Though the anatomical structures showed a continuous series,
the pure stringy type (i.e. the Wagenaar-type)
always could be nicely discriminated. Its
characteristics are (cf. fig. 4 no. 8; pi. II no. 5): sheath entirely
consisting of fibres, except for the narrow rectangular part in the
middle, which in the ventral sheath is strictly limited to the con-
tinuation of the two epidermal ceU layers.

This pure stringy (Wagenaar) type is to be found in the correlation
tables 34, 35 and 36 under the group with 80% fibres or more. It
will be seen that nearly aU Fg Wagenaar-types have the string num-

ber 9—10 and likewise that nearly all Fg plants with string number
9—10 are of the pure stringy Wagenaar type.

All Fg families investigated are found in table 37 (in table 36 only
part of fam. 55-6 is recorded). Each family shows a clear cut mono-
factorial segregation (1:3) into a pure stringy type (Wagenaar) and
a not pure stringy one. All families together give:

In table 37 there is a difference between the three Fg families of
1932 and those of 1933. In the first three families many not pure
stringy plants occur with string number 8—9 and even a few with
9—10; their average strength is 6.30. In the families of 1933 on the
other hand, the pure stringy (Wagenaar) type is by its strength of
string markedly different from the remaining not pure stringy plants.
Not any of the latter has a string number 9—10 and only a few 8—9;
their average strength is 5.76.

I am not certain what this difference between the Fg families of
1932 and 1933 means. Some F^ mother plants of the latter famüies
indeed had a somewhat lower string number and a lower percentage
of fibres than the three F^ plants in 1931. It must, however, be
remarked that of each F^ plant in 1932 only one pod was investigated
anatomicaUy (in 1931 five) and that the F^ families 393—400 (1933)
in table 37 do not show düferences between the progeny of F^ plants
with a high and that of F^ plants with a low percentage of fibres.
Moreover strength of string and percentages of fibres in the Wagenaar
and Fijne tros parent plants (and their progeny; compare the table
on p. 235 with the first table on p. 237) of the two groups of F^ fami-
lies were the same. Possibly the differences between the F^ families of
1932 and those of 1933 may be (partly) due to weather influences
or to differences in classification caused by the subjective method of
determining the strength of string.

r. prakken, inheritance of colours
§ 5. Fg generation and backcrosses

As might be expected all pure stringy F^ plants (Wagenaar type)
breed true in the Fg (table 38). The strength of string is nearly always
9—10. Many hundreds of these Fg plants have been investigated
anatomically; their sheath always shows the characteristics of the
pure stringy Wagenaar type.

In tables 39 and 40 the Fg progeny of not pure stringy F^ plants is
found. The families in the two tables have been arranged according
to increasing strength of string in the F^ mother plants. Those in
table 39 segregate the pure stringy (Wagenaar) type, while those

in table 40 do not.

As to the segregating families in table 39 the numbers of the two
types always fit in very weU with the 3 : 1 ratio. In aU these fami-
lies together there occur:

As to the not pure stringy type there are, however, striking differ-
ences between the families! If the F^ mother plant has a low string
number and percentage of fibres, the two types may be easily
discerned in the Fg family; not pure stringy types with a higher string

number than 6—7 or 7—8 do not occur. In families of F^ plants
with high string number and percentage of fibres the two types on
the contrary can not be discriminated by strength of string at all:
many plants with string numbers 8—9 and 9—10 are not of the pure
stringy Wagenaar type. Discrimination, however, is always possible
by means of anatomical investigation.

Table 40 shows the Fg famihes which do not segregate the pure
stringy type (progeny of 55-2 kept apart from 55-4; in table 39 the
two progenies have been arranged in one series). Possibly some fami-
lies might have segregated the pure stringy type, if a greater number
of Fg plants could have been examined. Here as well as in the
segregating famihes, there is a correlation between strength of string
in the Fa mother plants and in their Fg progeny, though there is some
regression. A comparison of table 40 with table 32 (Fijne tros families
in 1933) shows that no Fg families with lower string numbers than
those of table 32 occur among the famihes of table 40.

In table 41 the Fg mother plants of all Fg families are represented;
they are arranged according to strength of string, percentage of
fibres and their Fg progeny (non-segregating, segregating and pure
stringy). 29 families segregate the pure stringy type, 24 families do
not. This serious divergence from the expected 2 : 1 ratio may be
partly due to the very low number of plants which has been investi-
gated in some of the non-segregating families.

From these F^, Fa and Fg results we may conclude that one main
factor is responsible for the differences between the sheath types of
Wagenaar and Fijne tros. The F^ is intermediate, though approaching
the Fijne tros parent. In F^ only one fourth, the homozygous pure
stringy plants (Wagenaar type) may be discerned from the remaining
not pure stringy ones; the pure stringy type therefore might be mdi-

cated as „recessivequot;.

The stringiness of the not pure stringy types (homozygous and
heterozygous) is influenced by other factors. For me it is impossible
to say with certainty something about their number and character.
Especially the heterozygous type seems to be greatly influenced by
these factors. As a result some families with a nearly „stringlessquot;
mother plant (e.g. string number 4) segregate into „stringlessquot; and
„pure stringyquot; without (or nearly without) „intermediatequot; forms with
string jiumbers 6, 7, 8 or 9 (cf. table 43, families 496, 507, 503).

The backcrosses of F^ with the two parent types confirmed the
monofactorial scheme of the inheritance of stringiness. In the back-
cross with Fijne tros not a single pure stringy plant has been found.
The string numbers in this backcross were:

1—2 2—3 3—4 4—5 5—6 6—7 7—8 8—9 9—10 total.
0 4 16 32 22 32 42 3 0 151

In the backcross with Wagenaar (table 42) there occurred:

The not pure stringy backcross plants of 1933 had lower string
numbers than those of 1932 (cf. table 42). Exactly the same differ-
ence as had been found between the Fg famihes of those years (cf.
p. 241). I am not certain about the cause of it. It may be due to
genetical differences in the Wagenaar race. Possibly, however, the
differences are caused by weather influences or by the subjective
method of determining the strength of string. AU not pure stringy
backcross plants are heterozygous for the main factor and therefore
in all their progenies the pure stringy type must appear. These
progenies are given in table 43. No. 492 (12 individuals) is the only
family which does not segregate the expected pure stringy type.

As concerns the symbolizing of the main factor for stringiness
Currence used the symbol S for the monofactorial difference between
stringy and stringless types. The symbol S, however, is generaUy used
(Tschermak, Tjebbes and Kooiman) for the striped character of
the seedcoat. I, therefore, wiU use the abbreviation St to symbohze the
monofactorial main difference between the two races.

Fijne tros (nearly stringless) St St.

Wagenaar (pure stringy) st st.

B. Toughness of the pod wall
§ 1. Previous investigations

When the pod wall is characterized, two main types are generally

distinguished:

1.nbsp;Tough, parchmented or fibrous.

2.nbsp;Tender, non-parchmented or fibreless.

Emerson (1904, p. 54) found a „strong tendency toward domi-
nance of tenderness, the pods of first generation hybrids being
quot;almost as tender as the tender podded parentquot;. In the second
generation (different crosses combined) he found 22 tender-podded
plants, 38 intermediates an d 17 tough-podded ones.

Tjebbes and Kooiman (VII, 1922) on the other hand observed in
Fl dominance of the parchmented type. In F^ they found segregation
into 174 parchmented and 38 non-parchmented; 130 segregating Fg
families segregated into 1572 and 512. They therefore suggest a

monofactorial segregation.

Tschermak (1901 and 1902) mentioned dominance of „gewölbtquot;
over eingezogenquot;. He found (1922 p. 40) in the cross „Tausend für
Einequot; x „Ankerquot; among 94 F^ plants „ 16 mit Schnür-oder Perlhülsen,
78 mit glattgewölbten Hülsen (vermutlich 3 : 13), was auf bifak-

torielle Grundlage hinweistquot;.

currence (1930) reported on F^ F^ and Fg of crosses between
fibrous and fibreless pods, of which the F^ was always fibrous. The
difficulties of properly classifying the F^ plants appeared to be very
great. Of 23 F^ plants e.g., which had been classified as fibrous, there
appeared to be according to their Fg progeny: 7 homozygous fibre-
less 11 heterozygous and 5 homozygous fibrous (Cf. his table VI).

Lamprecht (19326, p. 306ff.) distinguished the two mam types

einfach gewölbtquot; and „eingeschnürtquot;. He obtained entirely new
Results, which may be shortly reviewed. Cross XVII: Both parents
and Fl „einfach gewölbtquot; (one of the parents less typical). Fg segre-
gated into 563 einfach gewölbtquot; and 34 „eingeschnürtquot; (= Vie)-
Among the firstquot; group typical and less typical forms occur in the
ratio 1 : 2. Dies lässt die Annahme wahrscheinlich erscheinen, dass

teüs die beiden für die Ausbüdung der einfach gewölbten Hülse
quot;verantwortUchen Faktoren je für sich eine weniger typisch ge-

„wölbte Hülse bedingen und dass dies teils auch für die in diesen
„beiden Faktoren heterozygoten Typen der FaU ist; bei dieser
„Annahme haben wir mit folgendem theoretischen Verhähnis zu
„rechnen: 5 typisch einfach gewölbt: 10 weniger typisch gewölbt:

„1 eingeschnürt____ Damit soU nicht behauptet werden, dass es

„sich so verhalten muss. Die weniger typisch gewölbten Hülsen
„könnten auch durch den Effekt anderer in den in Rede stehenden
„Biotypen vorkommenden Faktoren bedingt werdenquot;. Cross XVIII:
Again both parents and F^ „einfach gewölbtquot;. In Fg segregation of
V64 „eingeschnürtquot; (1635 and 30). Cross XII: Between an „einfach
gewölbtquot; and an „eingeschnürtquot; race; F^ „einfach gewölbtquot;. In Fg
the ratio of the two types is 9 : 7 (348 and 311).

The results of these three (and other) crosses Lamprecht explains
by the adoption of the four factors Fa, Fb, Fc and Fd. „Fa ist ein
„Faktor, dessen Anwesenheit erforderhch ist, wenn es überhaupt
„zur Ausbildung einer einfach gewölbten reifen Hülsenwand kom-
„men soh. Fr wirkt demnach als Grundfaktor. Fb, Fc und Fd sind
„drei Faktoren, die je für sich allein, aber nur bei Anwesenheit von
„Fa zur Ausbildung einer einfach gewölbten, reifen Hülse führenquot;.

Further on I shaU have to revert to his view about the anato-
mical base of the pod wall character.

The cross Fijne tros X Wagenaar is of a type up to now not de-
scribed as far as I know.

§ 2. Description of the parent plants and F^

Figures 5 and 6 show the full grown and the dry pods of Wagenaar,
Fl and Fijne tros; on fig. 7 halves of dry pods are to be seen.

Fijne tros is a race with non-parchmented, tender pod waU, which
in the ripe pods is greatly constricted between the seeds and then
has a much shriveUed surface. A membrane (parchment layer) is
never found in boiled beans.

The Wagenaar race has a „semi-parchmentedquot; pod waU, which in
the ripe pod is only slightly constricted and shrivelled in a rather
variable degree. Wagenaar can be threshed. (Fijne tros cannot). In
boiled beans a thin, weak membrane can be shown by scraping off
the outer, soft layers.

The Fl pod waU is strongly parchmented, neither constricted nor
shriveUed. The dry pods sometimes dehisce of their own accord

•a

o
ft
t»

and roll up spirally. In boiled beans the membrane is extremely
thick and tough.

As regards the anatomy (cf. fig. 3, p. 234) the pod wall of each of the
three types consists of an outer parenchyma (with the vascu-
lar bundles) and an inner parenchyma. In the Wagenaar race

outer and inner parenchyma are separated by a layer of fibrous cells,
which run obliquely across the pod wall. Only the cells in the outer
part of the layer are somewhat thickened and lignified. This Ugnified
part is rather variable; sometimes true fibres are not found at all, in
other cases the greater part of the layer is lignified.

In the Fijne tros race a fibrous layer does not occur. Yet the
separation between inner and outer parenchyma is usually clear, on
account of some differences in cell width and shape, the occurrence
of crystal-cells and sometimes of a few scattered fibres.

In all Fl plants the fibrous layer is as thick as (or even somewhat
thicker than) in the Wagenaar race; it is totally lignified. The cell
walls, however, are never very much thickened.

So the three types Fijne tros, Wagenaar and F^ seemed to be quite
different. In the second generation therefore the bifactorial ratio
4:3:9 might be expected, analogous to results obtained with
Pisum sativum (Wellensiek, Rasmusson) and Vicia Faha (Sirks
1932, p. 319: „der Faktor Q verursacht eine lederartige Hülsen-
„struktur, welche vom S-Faktor verstärkt werden kann; der
„S-Faktor als solcher hat bei Abwesenheit des Q-Faktors keine
„Wirkungquot;).

§ 3. Segregation in F^, F^ and backcrosses

In all Fg plants I studied the character of the pod wall, making use
of three different characteristics:

1.nbsp;Toughness of the parchment-layer in three nearly full grown
boiled beans of each plant.

2.nbsp;Microscopical structure of the fibrous layer on cross-section.

3.nbsp;Degree of shriveUing and constriction of the ripe pods.

By none of these methods (nor by combining two or all three) I
arrived at a perfectly sharp division into two or three types. The
relation between toughness, microscopical structure, and constriction
and shrivelling of the dry pod wall generally was the same as was
expected on the ground of parent types and F^, though some ex-
ceptions occurred. Intermediate types were usually intermediate in
all respects.

The only fairly marked division was into n o n-p archmented
(Fijne tros type) and p a r c h m e n t e d (in all degrees). In some
intermediates between Fijne tros and Wagenaar the anatomical
structure was decisive for discriminating the non-parchmented type:
total lack of the fibrous layer; some scattered fibres or groups
of fibres may occur. In a very few cases the existence of a layer
was doubtful and could not be ascertained with certainty from cross-
sections.

In the Fg-families the number of this purely non-parchmented
type was always about (highest D/m value 0.88). AU famihes
together (table 44) gave:

Among the 302 backcross plants F^ x Wagenaar and the re-
ciprocal cross not a single non-parchmented plant occurred. Dis-
crimination between Wagenaar and F^ type was not possible.

The backcross plants Fijne tros X F^ could easily be divided into
87 non-parchmented (3 less typical) and 56 parchmented (most of
them of the F^ type; 6 more or less Wagenaar type). The deviation
of the expected monofactorial backcross numbers 71.5 : 71.5 is very
great, D/m = 2.58. This backcrossing, however, was chiefly done
with a view to seedcoat colour of coloured flowering plants;
as the castration of the Fijne tros mother plant was performed less
accurately than in other crosses, there may be among the white
flowering, non-parchmented plants some pure Fijne tros individuals.
The figures in the backcross are:

In the Fg generation all non-parchmented Fj plants bred true;
fibres may occur, but only in scattered groups between outer and
inner parenchyma, never forming a more or less continuous layer.

Of parchmented F^ plants 38 Fg families segregated into both
types; 25 families did not segregate the non-parchmented Fijne tros
type (expected according to the 2 : 1 ratio 42 and 21). The segre-
gating families are to be found in table 45. The agreement with the
unifactorial ratio is quite satisfactory:

(The 7 doubtful plants have not been investigated anatomically).

Fg plants of the strongly parchmented F^-type sometimes segre-
gated (except for non-parchmented plants) into strongly and semi-
parchmented, in other cases the latter type was not found or only
in a few plants. Semi-parchmented Fg plants generally gave none or
only few of the strongly parchmented type. It was, however, on
the very ground of the F3 segregations impossible to conclude to a
bifactorial segregation.

Figure 8 shows a plant of the non-parchmented „Perfectquot; race,
of which one branch bears three pods of the parchmented type. It
may be of some importance with a view to the inheritance of this pod
character to examine the offspring of this plant. Woycicki (1930)
mentioned a probable case of mutation from non-parchmented into
parchmented in the variety „Japonaisequot;.

Now I will revert to Lamprecht's investigations. He writes (19Z2b,
p. 309): „Die Steifigkeit der Bauch- und Rückennaht beruht auf
„einer Einlagerung von Sklerenchymzellenbändern, was u.a. von
„Joosten (1927) untersucht worden ist. Die Steifigkeit der übrigen
„Hülse wird durch Einlagerung einer Kollenchymzellenschicht unter
„der Epidermis verursachtquot;. And (p. 314): „Mit ziemlicher Wahr-
„scheinlichkeit kann angenommen werden, dass diese Faktoren (i.e.
„Fb, Fc and Fd) für einen Verholzungsprozess der unter der Epider-
„mis gelegenen Kollenchymzellenschicht verantwortlich zu machen
„sindquot;.

If Lamprecht's supposition about the collenchymatic layer as the
anatomical base of his types „einfach gewölbtquot; and „eingeschnürtquot; is
confirmed by further anatomical investigations, there would be
two quite different types of toughness of the pod wall.

About the connection between his pod wall characters „einfach
gewölbtquot; and „eingeschnürtquot; and stringiness Lamprecht writes

;i932ö, p 308): „Ein weiteres Charakteristikum (der eingeschnürten
„Hülsen) gegenüber der ersten Gruppe ist ferner, dass hier auch
„Rücken- und Bauchnaht der Hülsen beim Reifen in hohem Grade
„nachgeben und eingezogen werden. In der ersten Gruppe gewahrt
man auch bei den weniger typisch einfach gewölbten Hülsen stets

„eine gewisse Steifigkeit der Rücken- und Bauchnaht. Unter den
„Typen mit eingeschnürten Hülsen scheinen demnach keine solchen
„„mit stärkerer Fädigkeit vorzukommenquot;. And about the possible
influence of the „groundfactorquot; Fa (p. 314): „„Es verbleibt zu unter-
„,suchen ob es bei Anwesenheit von Fa bzw. fa allein zur Ausbildung
„nur eines unbedeutenden Sklerenchymstreifens längs der Rücken-

„naht (und vielleicht auch längs der Bauchnaht), der sogenannten
„Fädigkeit der Bohnenhülsen, kommt oder nicht, und inwiefern
„sich Fa hierin von fa unterscheiden lässtquot;.

In the cross Fijne tros with Wagenaar toughness of the pod wall
and stringiness are quite independent of each other, both characters
depending on a different main factor. Only one Fg family (55-4 1932)
has so far been investigated with a view to linkage relations. In
this family the four possible types occurred in the expected 1:3:3
: 9 ratio of independent inheritance.