Digitized by the Internet Archive

in 2011 with funding from

LYRASIS members and Sloan Foundation

http://www.archive.org/details/inheritanceinmaiOOeast

THE CONNECTICUT

AGRICULTURAL EXPERIMENT STATION

NEW HAVEN, CONN.

BULLETIN 167, APRIL, 1911.

INHERITANCE IN MAIZE.

By E. M, East and H. K. Hayes.

The Bulletins of this Station are mailed free to citizens of Connecticut
who apply for them, and to others as far as the limited edition permit.

Comecticnt Apicjltiral Eiperlient Station.

OKF^ICERS AND STA.KF

BOARD OF CONTROL.

His Excellency, Simeon E. Baldwin, Ex Officio, President.

Prof. H. W. Conn, Vice President Middletown.

George A. Hopson, Secretary Wallingford.

E. H. Jenkins, Director and Treasurer New Haven.

J. W. Alsop ... Avon.

Charles M. Jarvis . . . . • Berlin.

Frank H. Stadtmueller Elmwood.

James H. Webb Hamden.

Administration

STATION STAFF.
E. H. Jenkins, Ph. D', Director and Treasurer.
Miss V. E. Cole, Librarian and Stenographer.
Miss L. M. Brautlecht, Bookkeeper and Stenographer.

Chemistry.

Analytical Laboratory,

John Philips Street, M. S., Chemist in Charge.
E. Monroe Bailey, Ph. D., C. B. Morrison, B. S.
R. B. Roe, A. B., C. E. Shepard, Assistants.
Hugo Lange, Laboratory Helper.
V. L. Churchill, Sampling Agent.

Proteid Research,

T. B. Osborne, Ph. D., Chemist in Charge.
Miss E. L. Ferry, A. B., Assistant.
Miss Luva Francis, Stenographer.

Botany.

G. P. Clinton, S. Ti., Botanist.

E. M. Stoddard, B. Agr., Assistant.

Miss M. H. Jagger, Seed Analyst.

Miss E. B. Whittlesey, Herbarium Assistant.

Entomology.

W. E. Britton, Ph, D., Entomologist; also State

Entomologist
B. H. Walden, B. Agr., D. J. Caffrey, B. Agr.,
A. B. Ch.amplain, Assistants.
Miss E. B. Whittlesey, Stenographer.

Forestry.

Samuel N. Spring, M. F., Forester; also State

Forester and State Forest Fire Warden
W. O. FiLLEY, Assistant.
Miss E. L. Avery, Stenographer.

Plant Breeding.

H. K. Hayes, B. S., Plant Breeder.
C. D. HUBBELL, Assistant.

Buildings and Grounds.

William Veitch, In Charge.

TABLE OF CONTENTS.

Introduction,

Page.
5

PART I. — The Material and the Problem.

The Plant and its Origin,

The Varieties of Maize, . .

The Problem and its Treatment,

Previous Work on Inheritance in Maize,

The Material used.

Methods used, ....

Experimental Error,

10
14
18
21
25
28
31

PART II. — Endosperm Characters.

Inheritance of Starchiness,

32

Conclusions, . . . .

43

Inheritance of Yellow Endosperm,

46

Conclusions,

56

Inheritance of Aleurone Color, .

57

Family (24 x 54),

59

Family (8 x 54),

67

Family (15 x 54),

74

Family (18 x 54),

76

Family (7 x 54),

80

Family (17 x 54),

80

Family (19 x 54),

81

Family (60 x 54),

81

Conclusions,

90

PART III.— Xenia.

General Discussion of Zenia,

101

PART IV.— Plant Characters.

Page.

Podded and Podless Maize ,

105

Pericarp Color, ....

105

Cob Color, ....

108

Silk Color, ....

108

Glume Color, ....

109

General Discussion of Red Sap Color, .

109

Physical Transformations of Starchiness,

110

Size Characters, ....

117

Number of Rows of Seeds, .

120

Heights of Plants,

123

Length of Ears,

124

Weight of Seeds,

124

PART V. — Plant Abnormalities.

Dwarf Forms,

Regularity of Rows of Seeds,

Bifurcated Ears, .

Ears with Lateral Branches,

Plants with Striped Leaves,

Hermaphrodite Flowers,

129
131
.132
133-
134
134

GpQeral Conclusions,
Literature Cited,

136

13&

INHERITANCE IN MAIZE.

E. M. East and H. K. Hayes.

INTRODUCTION.

The investigations reported in this paper were begun in the
spring of the year 1906. During the first four years the work
was conducted at the Connecticut Agricultural Experiment
Station. Since the fall of 1909, it has been carried on both
there and at the Bussey Institution of Harvard University.
Strictly speaking the researches comprise more than five years'
work, for several of the pure varieties used as parent stocks
had been selfed for the two previous years, so that a number
of crosses were made in 1905 with full assurance that as far as
most of the visible characters were concerned, the parent
strains were pure. There was some further advantage gained
in that the writers have been interested in experimental maize
breeding since 1900, for without this experience the probable
error of the results would be greatly increased.

Genetic research during the past decade has been very fruit-
ful of results ; nevertheless one could scarcely say that the field
has been thoroughly surveyed, much less that any part of it
has been completely investigated. The rediscovery of Mendel's
law in 1900 opened up a new era in the search for the principles

*" Contribution from the Laboratory of Genetics, Bussey Institution
of Harvard University, No. 9."

6 INHERITANCE IN MAIZE.

of heredity. Mendel's chief discovery — segregation of poten-
tial characters in the germ-cells of hybrids and their fortuitous
recombination — was one of the really great achievements in
biology, but even so, it may be questioned whether his chief
legacy is not his method of work. As has already been stated
by Bateson, previous investigators even including the biometri-
cians dealt with facts en masse, and the seeming order of the
mathematical formulas deduced served rather to conceal than
to reveal the individual facts. Mendel's method of individual
analysis by the study of simple characters in carefully con-
trolled pedigree cultures, however, has yielded and will continue
to yield results of great value to science. Still, since we are
dealing, as yet, with the simplest elements of the science of
genetics, the subject matter of an investigation may be expected
to yield results (other things being equal) somewhat in the
proportion in which it fulfills the following technical require-
ments.

1. The genus or species under investigation should be
variable. There should be a goodly list of types which are
differentiated by definite characters easy of determination.
That is, the differences should be largely qualitative and not
quantitative.

2. The different types should be freely fertile inter se, unless
an investigation of the causes of sterility is contemplated.

3. The flower structure should be such that the technique
of crossing and selfing is simple and accurate.

4. Since the accuracy of an analysis of the manner in which
characters are inherited increases — generally speaking — as
the square root of the number involved, the subjects should
return a large number of seed per operation (selfing or crossing).

5. The flowering branches should be numerous. This is
necessary for three reasons. If one is dealing with flower
characters he must be able to determine the character of the
plant from a mature flower while immature blossoms still
remain for the production of the controlled seed. Obviously,
it is also an advantage when dealing with plant characters,

INTRODUCTION. 7

to have more than one opportunity to secure a desired union.
Further, it is advantageous to be able to make several different
crosses upon one plant.

6. Seed should be viable for several years in order that
different generations may be compared at the same time. This
enables one to reduce to a minimum the physiological fluctua-
tion due to varying environment which many characters undergo
in a marked degree.

7. The subject material should be "workable" cytologically
in order that it may be attacked from both standpoints.

It might be remarked here that some botanists consider that
genetic research can throw no new light upon evolution and upon
the meaning of species, unless the subject material is an unculti-
vated genus or species. This criticism is apparently no more
pertinent than the one the chemists had to meet years ago
when they were told that synthetic compounds could not
possibly be the same as those produced by nature. The fact
of physiological fluctuation due to varying environment is
admitted, but it is not admitted that the mechanism of hereditary
transmission of the character in question is affected by these
fluctuations.

Some idea as to the effectiveness of an inquiry concerning
inheritance in maize from the standpoint of science may be
gained then by examining the degree in which the plant fulfils
the above requirements.

Although the forms of maize are regarded by botanists as
belonging to the one species Zea mays L., there is probably no
species of the flowering plants — if horticultural color varieties
are excepted — that appears under such varied forms. These
forms are perfectly fertile inter se, moreover, so that the first
and second of our requirements are fulfilled perfectly. The
third requirement, that of an easy technique and accurate
control of desired matings, is met very imperfectly. The
plant is monoecious. Ordinarily, this type of flowering habit
is desirable in pedigree culture work because accidental selflng
is usually much more easy to prevent than it is in hermaphroditic
plants. In the case of maize, however, there is such an enor-

8 INHERITANCE IN MAIZE.

mous production of pollen that it is continually present in the
air of the maize field. In spite of all precautions, therefore,
seeds of unknown paternal ancestry do creep into the cultures.
The general error due to this source has been determined in
cases which are described later, but the determination of a
probable error in a mass of data is not sufficient in genetic
work. An actual error in which a single seed of unknown
paternity becomes the ancestor of a pedigreed line, is sufficient
to upset all inductions drawn from the data. For this reason
the cultures have had to be larger than would otherwise have
been necessary.

The requirement of a large number of seeds from one union
to reduce the probable error of chance fertilization among
gametes differing in character is quite satisfactory in maize as
from two hundred and fifty to twelve hundred seeds are pro-
duced on the cobs of the various races. The small number of
flowering branches, however, is a serious objection. In some
cases there are two or even three and four ears upon each plant ;
but in most cases, especially in the large races, there is but one
ear. And even where there is an extra ear one gains but little
advantage. The ears mature about the same time and it is
impossible to find out what seed characters the plant possesses
before pollinating the ear which is to have its place in the con-
trolled culture. The disadvantage of this fact to the investi-
gator is apparent if one remembers that when studying ear
abnormalities sometimes twenty to twenty-five cobs must be
selfed by hand to be reasonably certain that one selfed ear
with the desired characters is obtained.

Maize seed is rather delicate and must be carefully dried in
a place where there is a good circulation of air. When dried
until the moisture content is only about ten per cent, it remains
in fairly good condition for three seasons. Seed older than
this is almost worthless. In fact, there is a possibility of
obtaining distorted results even in the second year. Ninety-
eight to one hundred per cent of properly dried seed should
germinate the next spring after harvesting, but this percentage
falls to about ninety the year following. If, therefore, seeds
of any particular gametic structure should lose their vitality
first, incorrect results would be obtained.

The chromosomes of maize are small and difficult to studv

INTRODUCTION. g

and scarcely anything is known of their behavior during the
maturation divisions.

This discussion should give some idea of the advantages and
disadvantages that maize presents as subject material for a
genetic investigation from the standpoint of pure science. The
plant, however, even if not as perfect as some others from this
point of view, has another claim which ought not to be "dis-
regarded. The fact that maize is the basis of the agricultural
wealth of the country makes it eminently desirable that every
fact about the inheritance of its characters, should be learned
as soon as possible. It is only through the application of such
knowledge that the present arbitrary, and, in a way, unscien-
tific methods of its improvement as an economic crop will be
placed upon a definite and orderly basis.

INHERITANCE IN MAIZE.

PART I.

THE MATERIAL AND THE PROBLEM.

The Plant and its Origin.

Although there is no absolute information as to the origin
of maize, most botanists agree that its original home is Mexico
(Harshberger '93) or the region to the south of there. As to
how it originated there has been much speculation, and various
views are held by different writers. We think it not out of
place to give here a synopsis of the most important theories,
because in our opinion, the results from the pedigree culture
work on the inheritance of plant characters described in Parts
IV and V throw considerable light on the subject.

The Tribe Mayde^ of the order GramineEe comprises but
seven genera and only sixteen or seventeen species. The two
genera which interest the maize student are Zea and Euchlsena
both of which are monotypic. The especial distinctions be-
tween the two are given by Lamson-Scribner ( : 00) in his key
to the genera of Maydese as follows :

"Euchlaena, pistillate spikes axillary fasciculate, distinct,
axis of each articulate."

"Zea, pistillate spikes axillary, grown together, forming a
compound spike with a much thickened, continuous axis."

His complete descriptions are:

''EuchlcBua Schrad. Ind. Sem. Hort. Goett. 1832. Spikelets
unisexual, monoecious; the staminate 2-fiowered, in pairs,
one sessile the other pedicellate, arranged in terminal paniculate
racemes; the pistillate 1-fiowered, sessile and solitary at each
joint of an obliquely articulate rhacis of a simple spike; the
spikes fasciculate in the leaf axils and each more or less enveloped
by a foliaceous bract. Glumes in the staminate spikelets 4,
acute, the first two membranaceous, empty; flowering glumes
smaller and like their paleas, hyaline. Stamens 3. Glumes of
the pistillate spikelets 4, the outer one broad and boat-shaped,
smooth, soon becoming hard, surrounding the inner glumes and

ORIGIN OF THE PLANT. ii

narrow rhacis, second glume empty coriaceous, third glume
hyaline with a palea but no flower; fourth or flowering glume
with its palea hyaline. Styles very long, filiform, shortly bifid
at the apex.

Tall annuals with long and broad leaves, closely resembling
Indian corn in habit. Species one with several varieties in
Mexico and Central America."

''Zea Linn. Sp. PI. 971. 1753. Spikelets unisexual, mon-
oecious; the staminate 2-fiowered in pairs, one sessile the other
pedicellate, along the numerous branches of a terminal panicle;
the pistillate 1-fiowered, sessile, crowded in several rows, along
a much thickened continuous axis arising from the lower leaf-
axils and closely enveloped by numerous large foliaceous bracts.
Glumes four, awnless; those of the staminate spikelet acute;
those of the pistillate very broad and obt.use or emarginate.
Grain hard, only partially inclosed by the fruiting glumes.
A well-known tall and striking annual grass with erect stems
and broad leaves. The terminal staminate inflorescence forms
the "spindle" [tassel], and the long projecting styles of the
pistillate flowers constitute the "silk." The cob is formed by
the union of the axes of several female spikes into a much
thickened body. Species one or two, of American origin,
presenting many varieties in cultivation known as corn, Indian
corn or maize."

From these descriptions of the two monotypic genera, it is
seen that Euchlczna mexicana Schrad., the common teosinte,
is not greatly different from Zea mays L., our ordinary maize.
Indeed to one who has grown and followed the extraordinary
variability of both, it does not seem a greater step from teosinte
to the maize most similar to it — the short many branched
pop or flint types — than it does from the small dwarf pop
maize to the giant dent forms. Teosinte is perfectly fertile
with maize, which fact has led to some confusion, for Watson
('91) thinking that hybrids between the two constituted a pure
wild species, named it Zea canina Watson. Segura (Harsh-
berger, '96), however, by remaking the crosses and growing
them near the region where the "Zea canina" was found, clearly
proved the true nature of the latter. Harshberger ('93) first
fell into the same error as Watson but later (Harshberger '96)

12 INHERITANCE IN MAIZE.

discovering the true state of affairs suggested that maize origi-
nated either from (1) a cross between teosinte and some extinct
but closely related plant, which by variation under a better
environment finally produced a plant with larger maize-like
ears; or that it came from (2) a cross between teosinte and a
race of the latter that had varied under long continued culti-
vation. The basis of Harshberger's argument that teo inte
must have been crossed by another form is his idea that only
in the progeny from a cross would sufficient variability have
appeared to have produced the more vigorous plant — the
aboriginal maize.

More recently Montgomery [ : 06] has advanced the theory
that teosinte and maize are both descended from an unknown
many-branched grass whose branches terminated in a panicle
of spikelets bearing hermaphrodite flowers. He says: "As
evolution progressed, the central tassel came to produce only
staminate flowers, these being higher and in a better position
to fertilize the flowers on the lower branches. At the same
time, the lateral branches came to produce only pistillate
flowers, their position not being favorable as pollen producers,
while, on the contrary, they were favorably placed to receive
pollen. This difl^erentiation in the flowers was accompanied
by a shortening of the internodes of the lateral branches until
they were entirely enclosed in the leaf sheaths [the husks]."
The especial difference between the evolution of teosinte and
of maize was thought to have been in the development of the
ear of the first from the lateral branches of the tassel-like panicle
and the ear of the second from its central spike. This argument
of Montgomery is directly opposed to the old theory that the
cob of modern maize is the result of a fusion of a number of
two-rowed pistillate spikelets such as are found upon teosinte.
His theory then, emphasizes the nature of the changes that
took place; Harshberger's theory, the way they were trans-
mitted. >

In addition to these views it seems only necessary to mention
that since maize is the only grass with a naked seed, the podded
variety Z. mays tunicata Sturt. is by many considered to be an
earlier stage in maize development.

Our own views on the subject have resulted from a considera-
tion of the behavior of the characters of the various races of

PLATE I.

a. Ear with hermaphrodite flowers from the dwarf plant which appeared
in Stoweh's Everg-reen suRar maize.

b. Mature seeds on male spike of plant heterozygous for starchiness,
showing segregation. A common physiological fluctuation.

Abnormalities

ORIGIN OF THE PLANT. 13

maize when crossed, the data on which they are based being^
given later. The matter is largely speculation and should be
considered as such. It is merely the simplest manner of inter-
preting the known facts, by connecting maize with the othe-i
Maydese by a short series of changes that involve characters
that mendelize. On the whole it does not differ greatly from
Montgomery's theory.

Since we now believe that the essential role of hybridization
is to recombine the characters possessed by the parent plants
in definite ratios without actually producing anything new,
[new combinations may produce characters formerly unknown],
there is no necessity of postulating hybridization of teosinte
with a more maize-like variety. It is known that when teosinte
is cultivated in rich soil it sometimes produces ears having
an irregular development of four rows. This is only an
expression of one of the commonest modes of variation, repeti-
tion of parts or meristic variation. This variation in the ear
has taken place again and again in maize and is inherited
although sometimes obscured by physiological fluctuation.
The ear of maize then is a meristic variation produced from the
central spike of the tassel of the lateral branches of teosinte
or of a teosinte-like plant, and not a fusion of the lateral spike-
lets. Lateral spikelets still appear in maize, apparently as if
variation ran in grooves or paths of least resistance. This
character has been found to segregate in a manner essentially
Mendelian. The podded character also mendelizes and is
allelomorphic to its absence. If then progressive meristic
variations occurred in the central spikes of the side branches
of the teosinte-like ancestor, followed by retrogressive varia-
tions affecting both the lateral spikes of the lateral branches
and the pod character, a plant would have originated bearing
naked hermophroditic ears. Further change might easily
have come about, as Montgomery suggests, by a shortening
of the side branches producing the modern husk, and finally
the origination of the monoecious character. The latter occur-
rence is not at all hard to picture for the change of the staminate
inflorescence to an hermaphroditic or even a pistillate one, is
something which is exceedingly common in all or almost all
strains of maize. It is a physiological fluctuation produced
by excessive rainfall and fertile soil. The appearance of stamens

14 INHERITANCE IN MAIZE.

on the modern maize ear is much more rare but that it does
occur is shown by the ear pictured in Plate I. In fact one of
our sterile dwarf mutations had nothing but hermaphroditic
flowers.

The Varieties of Maize.

Although all of the varied forms of maize are regarded by
modern taxonomists as sub-divisions of the species Zea mays
L., many varieties have at various times been given specific
rank. The Index Kewensis gives the six following types as
species. The original sources have been consulted but the
descriptions have been shortened to include only essential facts.

Z. Curagua, Molina, J. I. Saggio sulla storia naturale del
Chili, pp. 306, Bologna, 1810. = Z. mays.

This variety is distinguished by its serrate leaf-edge. It
has never been cultivated in the United States, but appears
to be a flint type, Z. mays indurata. Syn. Z. Caragua, Stend.
Nom. ed II, ii. 797.

Z. erythrolepsis, Bonafous, M. Histoire naturelle, agricole
et economique du Mais. Folio, pp. 181, Plates 19, Paris, 1836,
= Z. m,ays.

"Glumis rubris, seminibus compressis." "Le Mais a rafle
rouge cultive sur les rives du Missouri, se distinque par I'aplatis-
sement de ses grains, et surtout par le coleur rouge des ecailles
et corallines de I'epi femelle. La Constance de ce caractere
tend a lui meriter le titre d'espece."

This form could hardly be considered a variety as it is a
common variation in all of the commonly recognized varieties.

Z. hirta, Bonafous, M. Note surune nouvelle espece de Mais.
Ann, Sci. Nat. Ser. I v. 17; 156-158. 1829. = Z. mays. '' Foliis
hirtis et dependentibus; spiculis m,asculis sessilibus, diandris
triandrisve; antheris subaureis.''

A good variety, originally sent to Bonafous from Balbis of
the Jardin des Plantes de Lyon. It varies into a series of flint,
pop and dent types.

Z. japonica, Van Houtte, Fl. des Serres, XVI (1865-67), 121.
t. 1673-74. 1867. =Z. mays. Syn. Z. vittata, Hort. and Z.
variegata, Hort.

A small variety with leaves variously striped with white.
A small flint type is the one chiefly cultivated for ornament,

VARIETIES OF MAIZE. 15

but the variety occurs again and again in fields of all of our
common maize strains. It could undoubtedly be isolated pure
by careful selection of these individuals.

Zea macrosperma Klotzsch. Bot. Zeitung 9; 718. 1851.
In der Sitzung des Ges. naturf. Freunde zu Berlin. =Z. mays.
Seed received by Humboldt from Cuzco. It is simply a large
seeded dented starchy type.

Zea rostrata, Bonafous, M. Ann. Soc. Agr. Lyon. v. (1842),
197. =Z. mays. Simply a hook-seeded form of pop maize,
somewhat similar to our common rice pop.

Of these types Z. mays Curagua Molina and Z. mays hirta
Bonafous might be considered as good varieties. The four
remaining names and also the varieties of Z. mays listed in the
Index Kewensis might well be placed under the classification
porposed by Sturtevant ('99), leaving out his Zea mays amy-
leasaccharata because the latter is a type which is probably
identical with the "flinty" sweet corns with which canners
have so much trouble. The three ears from the San Padro
Indian collection sent to Sturtevant by Palmer, and upon
which he based the variety, failed to yield a mature crop in
Geneva, New York. It is therefore unknown whether this
type would really prove true. Sturtevant's classification
follows although I have added the word mays and have listed
them as varieties instead of species. It is not strictly correct
to give him as the authority for the names, as he used them
specifically, but since they have come into general use in the
United States it seems more convenient to keep them. Sturte-
vant himself based his claim for separate species principally
upon the fact that intermediates were either absent or rare.
This fact comes about, as will be shown later, from the alter-
native manner in which the distinguishing characters are
inherited. In reality many other characters are inherited in
the same manner, and it is only because the chief characters
of these five varieties are striking to the eye that it is advanta-
geous to keep them in use.

Zea mays tunicata, the pod corns. Sturtevant, Bui. Torr.
Bot. Club 1894, p. 355 (Also St. Hil., Ann. Sci. Nat. 16; p.
143, fide De Candolle). A form in which each kernel is enclosed
in husks (usually four) besides the foliaceous bracts that enclose
the ear.

i6 INHERITANCE IN MAIZE.

This form was first described by C. Bauhin in 1623, and has
been the basis of a long list of synonyms since that time. It
is probable that the prototype of the species possessed this
character for it would thus be linked closer to the other May-
dese. The strains now obtainable under this name have been
hybridized until ears can be found whose kernels run the whole
gamut of the other four kinds. The tendency of plants to
form anew characters once possessed that have been lost, is
well illustrated here. We have come across several ears of our
ordinary varieties in which a few of the kernels at the base
have been podded. Sturtevant gives two authentic cases
where fully podded ears have appeared in other varieties under
such conditions that it is very improbable that it was the result
of hybridization.

Zea mays everta, the pop corns. Sturtevant, Bui. Torr. Bot.
Club, 1894, p. 324.

"This [species] group is characterized by the excessive pro-
portion of the corneous [starch in the] endosperm and the small
size of the kernels and ear. The best varieties have a corneous
endosperm throughout. This gives the property of popping,
which is the complete eversion or turning inside out of the
kernel through the explosion of the contained moisture on the
application of heat."

Strains of the pop maizes are the smallest of our cultivated
corns, and although there are varieties that reach a height of
nine feet when cultivated on fertile soil, plants comparable in
size to the average dent or starchy maize are never found.
There appears to be a distinct correlation between size of seed
and size of plant; therefore, since one never obtains large size
seeds without a development of soft starchy matter, pop ker-
nels much larger than those now grown are not likely to be
produced through selection or hybridization.

Two forms of seed are known in the pop corns; one is simply
a small seed with rounded crown similar in shape to the small
flints; the other, characteristic only of pop corns, is peaked
at the point where the style or "silk" was attached.

Other variations such as purple colored aleurone cells, yellow
endosperm, red silks, and red and variegated pericarps charac-
terize the pop maizes in common with the flint, sweet, dent
and starchy corns. The modal number of rows also varies

VARIETIES OF MAIZE. 17

in different varieties from eight to sixteen. In this, pop maize
is similar to dent, sweet and starchy, but different from flint
maize. It is doubtful whether any true flint maize exists with
a mode for number of rows higher than twelve.

Zea mays indurata, the flint corns. Sturtevant, Bui. Torr.
Bot. Club, 1894, p. 327.

This group is characterized by the seeds having a corneous
starchy endosperm, surrounding a soft starchy center immedi-
ately behind or partially surrounding the embryo. The strains
in common cultivation are considerably larger than the pop
corns, but varieties do exist which form a definite series from
pop to dent differing only by the amount and extent to which
soft starch replaces corneous starch in the endosperm. The
same color varieties that were described for pop corns occur.

Zea' mays indentata, the dent corns. Sturtevant, Bui. Torr.
Bot. Club, 1894, p. 329.

A group characterized by the extension of the soft starch
until it completely covers the summit of the seed. Corneous
starch, however remains at the sides of the kernel and acts as
a frame work to support the drying seed. The soft starchy
portion shrinking in drying to a much greater extent than the
other forms a characteristic indentation. Dent varieties occur
averaging from five feet to twenty feet (reported) in height,
with from eight to twenty-four rows as the modes (extremes
to thirty-six rows) . The usual color varieties occur.

Zea mays amylacea, the soft or flour corns. Sturtevant,
Bui. Torr. Bot. Club, 1894, p. 331.

A group characterized by entire absence of corneous starch
in the endosperm. Uniform shrinkage in drying usually gives
a seed with no indentation. The mummy corns of Peru, Mexico
and the southern United States appear to belong to this group,
but this is not absolutely certain. The specimens that we have
examined belonging to the New York Botanical Garden might
have been flint corns which owe their floury appearance to
partial decomposition.

This group marks the final disappearance of corneous starch
in the endosperm. It is the end of a series beginning with the
pop corns and coming up through the flints and dents. For
this reason one might expect them to possess the largest seeds, as
the length of time necessary for maturing the seed undoubtedly

i8 INHERITANCE IN MAIZE.

has something to do with the amount of soft starch formed.
The plants are indeed large, but seeds occur varying from the
size of the smaller flints to that of the larger dents. The origin
of the starchy corns is not due simply to their correlation with
the general plant structure and therefore a simultaneous origin
with large varieties, but is dependent upon a separate character
or group of characters. The usual color varieties occur.

Zea mays saccharata, the sweet corns. Sturtevant, Bui.
Torr. Bot. Club, 1894, p. 333.

"A well defined group characterized by the translucent horny
appearance of the kernels and their more or less crinkled,
wrinkled or shriveled condition." The sweet corns are simply
pop, flint and dent varieties (East : 09) that have lost their
ability to mature starch normally. Some few starch grains
are produced but they are generally small, angular and abortive.
The reserve material of the endosperm seems to undergo a
decomposition to cane sugar and the various hexoses. There
is apparently something more than a simple non-development
of starch, for the sweet corns in the "milk" state contain greater
percentages of sugar than do the starchy varieties in a similar
stage of ripeness. The same color varieties occur.

The Prohlein and its Treatment.

It is apparent that maize furnishes an admirable series of
types which are perfectly fertile among themselves. The
primary object of our work is to obtain pure forms of these
diverse types by inbreeding, then to test the mechanism of
inheritance of each separate character by controlled matings
and an analysis of the resulting progeny. In doing this we
simply follow Mendel's method of the individual analysis of
pedigree cultures.

The specific questions attacked are numerous. Our principal
object is to find whether the different characters under obser-
vation all obey the same law of heredity or whether separate
principles are involved, and whether characters apparently
inherited independently are not sometimes correlated with each
other. The question of dominance of a character in the first
generation of a cross has also interested us. Some characters
are perfectly dominant, other characters imperfectly dominant,
while still others form heterozygous combinations differing from

THE PROBLEM. 19

either of the parents. There are even cases in which dominance
appears to be reversible. If such a thing is possible, an explana-
tion is desirable. The study of the phenomenon of Xenia,
which has already formed the basal object of Correns' ( : 01)
fine monograph, throws gome light on these questions.

Another object we have kept in mind is the problem of the
purity of extracted homozygotes. It is a matter of common
knowledge that characters that have been lost through retro-
gressive variations — characters that behave as mono-hybrids
in inheritance — often reappear. The reverse of this pheno-
menon is also true. Is it because there is a phylogenetic "path"
in which these changes run, so that the same variation appears
again and again, or is there no absolute purity of the germ-cells
but only a comparative purity as indicated by Morgan ( : 10) ?

The idea of prepotency has been held with great tenacity
up to the present time. We hope these researches will throw
some light upon this subject of so much importance to practical
breeders. If there are individuals whose constitution is such
that chance production of zygotes is interfered with, the fact
brings many complications into the study of inheritance; but
such complications must not interfere with the facts. Various
other questions will be discussed in the proper places and for
this reason will not be considered further here.

It may be well to mention that although these questions
smack of the technical, it is maintained that in just so far as
one contributes toward their solution, that far is the broad
practical problem of better methods for the production of new
economic maize types solved. The questions of purity of
homozygotes, inheritance of size and number of ear-rows in
the different sub-species are easily seen to be of practical agri-
cultural importance. Other questions may seem of less import-
ance or even of no importance from this point of view, but this
is fallacious as is easily shown by illustrations from the science
of chemistry where abstruse theoretical researches have con-
tinually proved to be the most practical in the end.

In certain quarters there has been a marked reaction against
the continued Mendelian interpretation which has been given
to every paper published since the year nineteen hundred in
which actual experimental studies concerning the mechanism
of inheritance have been reported. This reaction has taken

20 INHERITANCE IN MAIZE.

the form of a philosophical query as to whether the characters
of the organic complex of which living organisms are composed
can in any sense be dissected and analyzed into the "miits"
of heredity which are the basis of Mendelian inheritance.
Further it has been questioned whether there is any justifica-
tion for the increasing complexity with which Mendelian
formulse are involved. It has been argued that with a mul-
tiplicity of "factors" any particular case can be interpreted
as segregating according to the Mendelian hypothesis. For
these reasons the writers wish to have their position in report-
ing the following investigations distinctly understood at the
beginning.

It is fully understood that there is danger in improper analysis
of complex ratios from pedigree cultures. This is inevitable.
Yet it is not a more pertinent criticism to condemn complexity
in biological facts than it is to frown upon the intricacies of
modern organic chemistry because it is so different from the
simple chemistry of Liebig. The answer is that the facts of
heredity are complex.

In regard to the question of the ultimate nature of unit
characters or the possibility of absolute segregation of characters
in the germ-cells so that in the recessive there is actual absence
of the character (gene) in question, we must await more results
from the different points of view of the breeder, the cytologist,
the physiologist and the physiological chemist. The facts
reported in genetic investigations remain indelible. The
interpretation of these facts may or may not be correct; they
simply arbitrarily represent the facts in a convenient system
of notation much as the facts of chemistry are represented by
structural formulae. This is the idea in the minds of the
authors in the following report. It is thought moreover to
represent the attitude of most genetic investigators and the
excuse for making the above statements lies in the fact that
unfortunately one often finds no appreciation of this attitude
by biologists not actually engaged in genetic research.

We have, then, used the ordinary Mendelian notation, with
allelomorphic pairs interpreted as presence and absence of
characters not because we know that there is actual absence
but because this interpretation fits our present knowledge.
We have interpreted complex characters such as height which

PREVIOUS INVESTIGATIONS. 21

we are not able to analyze completely, as segregating characters.
The conclusion that there is a segregation in the second hybrid
generation very much in excess of the sum of the non-inherited
fluctuation and of other variation due to the heterozygous con-
dition of the pure (?) forms used and also of their combination
in the first hybrid generation, is justified by the data. This is
the essence of Mendelian theory; and, whether absolutely
correct or not, it is an interpretation that cannot fail to be
valuable to the commercial plant breeder. It gives him some
knowledge of what may be expected in his endeavors to produce
new types of maize by hybridization.

It might also be mentioned that following Johannsen, the
word "gene" has been used to signify that substance present
in the germ-cell which represents potentially the "unit char-
acter" or whatever it may be called that acts as an entity in
heredity.

Previous Work on Inheritance in Maize.

Before describing in detail the material used in these experi-
ments it may be well to give a short summation of the previous
work in the field.

The early hybridists, Camerarius, Logan, Pontedera and
Henschel, each made a desultory study of maize crosses, but
obtained no results of present interest. Hardly more satis-
factory are the papers of Dudley (1724), Sageret ('26), Puvis
('37), Gartner ('49), Naudin ('63), Hildebrand ('67, '68),
Vilmorin ('67) and Focke ('81), although these researches —
representing work of the principal students of hybridization
of the period — each gives several observations concerning the
immediate effect of pollen upon the endosperm, — that
phenomenon called Xenia by Focke ('81). These observations,
however, can hardly be compared with those made since the
cause of Xenia was discovered for the obvious reason that the
facts concerning the changes in the endosperm were almost
lost to sight in the search for effect of cross-pollination on the
tissues of the maternal parent.

In the work of a slightly later period particularly in the
United States (Kellerman and Swingle '89, '90; McCluer '92;
Morrow and Gardner '92) a great improvement was made in

22 INHERITANCE IN MAIZE.

the methods of investigation employed. The parental stock
was often inbred to establish its purity, crosses were made by
hand upon protected flowers, and the resulting progeny were
studied with great care. Many facts of inheritance are uncon-
sciously reported in their papers which are confirmed in the
post-Mendelian work which gives them a meaning For exam-
ple one finds these data in Kellerman and Swingle ('91). A
*chance hybrid evidently produced by the pollination of a white
maize with pollen from a variety with purple aleurone cells was
inbred. A hand-pollinated ear contained 370 seeds, of which
206 were blue, 71 pink, 71 orange-yellow and 22 pure white.
One wonders how the essential facts of dominance and segre-
gation remained unnoticed in the face of such ratios as this.
But even if it is interesting to reread these papers and consider
them from a more modern viewpoint, it is hardly profitable
to discuss them further here. The work previous to 1900 was
in the wrong epoch, and since that time three valuable con-
tributions to the subject in hand have been made. (De Vries
1899 and 1900, Correns 1,899, 1900 and 1901, and Lock 1906.)

It is interesting at least, to note that in the cases of both
De Vries and of Correns the studies of maize hybrids in which
presence and absence of yellow and .presence and absence of
starch in the endosperm were concerned, contributed largely
to their independent discoveries of dominance and segrega-
tion in hybrids, which they both unselfishly credited entirely
to Mendel after their discovery of his previous paper. Thus
Zea mays shares with Pisum sativum the honor of being the
subject material in the establishment of Mendel's laws.

Correns' (:01) beautiful monograph was written with the
especial idea of furnishing an explanation of the phenomenon of
Xenia, but it naturally contributed a large amount of data
upon the mechanism of inheritance of the characters with which
he worked.

Correns' technique was as follows. The seeds were planted
first in pots, allowed to attain a healthy start, and finally trans-
planted to the field. In the first year (1894) the plants to be
used as "mothers" were planted together in his experiment
field, castrated at the right time and the silks protected between

* The immediate parents were thought to have been white, but
this was probably an error.

PREVIOUS INVESTIGATIONS. 23

pollinations with paper bags. The individuals that furnished
the pollen were planted together in places apart from the pro-
posed mother plants and from them pieces of the male panicles
(tassels) were carried in glass bottles to the mother plants.
Slight changes in the plan of planting were made in 1895 and
1896, but I cannot find in any case that either the male flowers
were protected from foreign pollen during their maturation
or that special care was taken to have pollen for a cross fur-
nished by an individual plant. Furthermore in handling the
hybrids individuals were not selfed but bred inter se. Some
of the families were given to gardeners who were growing no other
maize, while others were detasseled and naturally pollinated
en masse with the pollen of a pure race. The first method
predominated. We can see then that the methods in use
furnished correct results only when the characters in question
were simple and of such nature that the races could be kept
pure by inspection. Complex ratios such as are furnished
when maize with purple aleurone cells is crossed with various
white maizes differing in gametic structure, could not possibly
be analyzed correctly.

Correns reached conclusions regarding the mode of inheri-
tance of the following characters but it must be borne in mind
that these results came from the study of data more or less
massed, and not the study of individual crosses in as precise
a manner as that outlined by Mendel.

Yellow endosperm was found to be dominant to its absence,
and starchiness dominant to absence of starchiness (sweet). Both
of these characters behaved as Mendelian mono-hybrids. It
cannot be definitely stated, however, that crosses between all
races of maize where presence and absence of these characters
are concerned would give the same results. Long aleurone
cells also proved dominant to short aleurone cells, and red
pericarp to absence of red, but Correns was not entirely satis-
fied that these characters behaved as simple Mendelian mono-
hybrids although he supposed this to be the case.

Purple aleurone cells appeared to form an allellomorphic
pair with absence of purple, but he found that the heterozy-
gotes when bred inter se did not give the normal number of
whites. What he took to be heterozygotes of the same char-
acter were either pure purple, partial purple, or pure white

24 INHERITANCE IN MAIZE.

when the purple was used as the male parent. In the reverse
cross the purple appeared to be fully dominant. Correns
(:01) endeavors to explain this phenomenon by the fact that
in the formation of the hybrid endosperm two nuclei come
from the female and but one from the male parent. He sup-
poses that in some cases this may cause a dominance of the
female characters. This purple character seemed to interfere
with normal inheritance in still another case (Correns :02),
where a blue sweet corn was crossed with a non blue pop. Here
the second generation yielded only about 153^% of sweet
kernels out of a total of over 8,000. Pollinated with the reces-
sive parent there appeared nearly 50% of sweet kernels so that
the female germ-cells seemed to have segregated normally.
Correns suggested that in this case the four possible combinations
of characters in the germ-cells did not take place with equal
facility.

Our own data shows the error in the first case to be due to
the fact that white races differ in their gametic structure in
characters which affect the purple color. The observations
in the second case have not been confirmed, but were probably
due to improper classification of the heterozygous dominants
and the recessives. (See starchy and non-starchy crosses.)

The shape and size of seed and relative weight of embryo
and endosperm Correns thought behaved in a non-Mendelian
manner, although he was not prepared to say in exactly what
manner these characters were inherited.

Lock ( : 06) carried out a much more extended series of
maize crosses at Peradeniya, Ceylon, from 1902 to 1906. His
technique in certain cases was a considerable improvement
on that of Correns in that both the male and female infior-
escences were enclosed in bags and thus crosses were made
between single individuals. Unfortunately his method was
later changed and cross pollination was accomplished by plant-
ing the two races in alternate rows on an isolated plot of ground,
and detasseling all plants of the race which it was proposed to
use as the female parent. This method of course gave no
chance to make a proper analysis of complex characters for
could not be known just what gametic composition was pos-
sessed by the male parent. This criticism was anticipated
by Lock himself but the method was used because he desired

THE MATERIAL. 25

to have a large amount of data from which to establish the
mathematical accuracy of Mendel's hypothesis of definite
segregation and chance mating. In the cases of starchy and
non-starchy, and yellow and non-yellow endosperm Lock's
results were in accord with those of Correns. Furthermore
he showed definitely that red pericarp behaved as a Mendelian
character, allelomorphic to absence of red. Lock also crossed
indented and non-indented races and remarks that in the F2
generation a high degree of variability appeared, but without
making crosses between individual plants and studying the
progeny he could not decide whether or not Mendel's Law was
followed. No data is reported on inheritance of height of
plants but a number of crosses were made between Fi plants
and the shorter of the parental races, and he states that no
segregation into short and intermediate plants took place. The
plants on the contrary were remarkably uniform in height and
he believed blended inheritance to be the rule for this character.

Lock's results in crossing races with purple aleurone cells
with races with non-purple aleurone cells is so seriously com-
plicated from the fact that he followed out no individual crosses
that it is impossible to criticize his data. From the fact that
individual ears showed such different ratios as 3 : 1, 9 : 7 and
1 ; 3 we may suspect that he was dealing with white races of
varying composition such as are described in our work on this
character.

These short abstracts from the work of Correns and of Lock
do not give an adequate idea of the large amount of painstaking
investigation for which each should be credited however, and
anyone interested in the subject should therefore consult the
original papers.

The Material Used.

The types of maize which furnished the parental stock with
which crosses were made for this series of studies were in most
cases inbred by hand for at least two generations before any
hybrids were actually made. When this procedure was im-
possible the parental ears were obtained from various commer-
cial growers who made a specialty of the types which they
furnished. From the maize obtained from them single ears
were selected and planted. The plants forming the immediate

26 INHERITANCE IN MAIZE.

progeny of these ears were used in part as the parents of crosses
and in part to inbreed. When any of the seeds from the origi-
nal ear were found to be heterozygous in any characters the
fact is noted when the crosses are described. In this manner
we were able to determine the purity of the parental stock
used, for all of the grosser characters. Of course new varia-
tions were continually isolated and these were given numbers
which show their origin. For example, the original stock of
Longfellow corn is No. 15; if, however, new variations appeared
in the Longfellow progeny they were numbered 15-1, 15-2, etc.
The following descriptions, then, comprise only original
material; that is, single ears of various commercial varieties.

Zea mays tunicafa. The podded corns.
21. Podded maize.

A fourteen-rowed ear with four husks around each kernel in addition
to ths usual paleas. The seeds looked like rice pop; they were small
but showed a considerable amount of white starchy matter.

Zea mays everta. The pop corns.
20. A flint-like 8-row purple pop.

A pop with purple aleurone cells, showing a small amount of white
starchy matter immediately behind the embryo, sufficient to keep the
seeds from popping well. Ear 1.5 cm. long, 11 cm. in cir. Seeds .9 x .9
cm., white endosperm. Cob white.

60. Tom Thumb pop.

A dwarf true pop. Ear 7.5 cm. long, 8 cm. in cir., 12-rowed; pericarp
colorless. Seeds rounded, true pop, .5 x .4 cm., endosperm yellow. Cob
white.

23. White rice pop.

A white true pop. Ear 15.5 cm. long, 10 cm. in cir., 16-rowed. Seeds
white, .9 X .5 cm., hooked. Cob white.

26. A white, flint-like pop.

A white true pop with rounded flint-like seeds. Ear 17 cm. long,
9 cm. in cir., 8-rowed. Seeds .8 x .9 cm. rounded. Cob white.

27. Red rice pop.

A true rice pop with red pericarp. Used only for inheritance of
pericarp color.

28. White rice pop.

A true rice pop with white or colorless pericarp. Used only in cross

with No. 27.

Zea mays indurata. The flint corns.
4. Benton maize.

An eight-rowed race intermediate between the flint and the dent
corns. Ear 34 cm. long, 14 cm. in cir., 8-rowed, pericarp red becoming
colorless at summit. Seeds 1.1 x 1.4 cm., some very slightly dented;
endosperm yellow, slightly more starchy than a true flint. Cob white.

THE MATERIAL. 27

5. Watson flint.

A true flint with a pericarp rose red when developing in full sunlight,
the seeds at the tip usually showing simply red striations beginning
at point of attachment of the silk. Ear 27 cm. long, 13 cm. in cir., 8-
rowed. Seeds 1.0 x 1.2 cm., endosperm colorless. Cob white.

11. Sturges' flint.

A twelve-rowed yellow flint race. Ear 20 cm. long, 14 cm. in cir.,
12-rowed, pericarp colorless. Seeds 1.0x1.0 cm., endosperm yellow.
Cob white.

13. Sanford flint.

An eight-rowed race. Ear 30 cm. long, 13 cm. in cir., 8-r6wed;
pericarp colorless. Seeds 1.0x1.3 cm.; endosperm colorless. Cob
white.

15. Longfellow yellow flint.

An eight-rowed yellow race. Ear 27 cm. long, 11.5 cm. in cir.; 8-
rowed; pericarp colorless. Seeds .9x1.2 cm.; endosperm bright
yellow. Cob white.

17. Palmer's red-nosed yellow flint.

An eight-rowed yellow race. Ear 22 cm. long, 12 cm. in cir.; 8-rowed;
pericarp a dirty red at the sides of seed becoming almost colorless at
summit. Color not deep as in common red maize. Seeds 1.0x1.4
cm.; endosperm yellow. Cob white.

24. Rhode Island white cap.

An eight-rowed flint race. Ear 29 cm. long, 12 cm. in cir.; 8-rowed;
pericarp colorless except for a slight pink tinge of rose similar to No. 5
but less in amount. Seeds .9x1.2 cm.; endosperm colorless. Cob
white.

25. Brindle flint.

A common flint race not breeding true to the character from which
it derives its name, — a mosaic pericarp formed by slashes of dark red
extending irregularly from the point of the attachment of the silk. Eight-
rowed true flint.

Zea mays indentata. The dent corns.

2. Illinois low protein dent.

A white dent selected for low proteid content at the Illinois Agri-
cultural Experiment Station for eight generations. Protein content
8.30 per cent. Ear 19 cm. long, 18 cm. in cir.; 16-rowed; pericarp
colorless. Seeds 1.5 x .8 cm.; endosperm colorless; white starchy
matter largely increased in summit over usual dent type. Cob white.

8. Illinois high protein dent.

A white dent selected for high proteid content at the Illinois Agri-
cultural Experiment Station for eight generations. Proteid content
15.46 per cent. Ear 20 cm. long, 14 cm. in cir.; 14-rowed; pericarp
colorless. Seed 1.1 x .9 cm.; endosperm colorless; white starchy matter
decreased from amount usual in dent types but summit still well dented.
Cob white.

3. Leaming dent.

A yellow dent race. Ear 21 cm. long, 16 cm. in cir.; 20-rowed;
pericarp colorless but sometimes very slightly tinted with dirty brick
red at sides of seeds. Seeds 1.3 x .7 cm.; endosperm dark yellow;,
considerable soft starch at summit; well dented. Cob dark red.

28 INHERITANCE IN MAIZE.

6. Learning dent.

Same race as No. 3 but of different ancestry. Ear, 19.5 cm. long,
18.5 cm. in cir. ; 18-rowed.

7. Learning dent.

Same race as No. 3 but of different ancestry. Ear 18 cm. long, 17
cm. in cir.; 20-rowed.

9. Learning dent.

Same race as No. 3 but of different ancestry. Ear 18.5 cm. long,
16.5 cm. in cir.; 16-rowed.

12. Leaming dent.

Same race as No. 3 but of different ancestry. Ear 19 cm. long, 17
cm. in cir.; 18-rowed.

16. Leaming dent.

Same race as No. 3 but of different ancestry. This ear was 18-rowed
and perfectly formed. It was surrounded by five lateral branches each
having either four or eight rows of seeds.

1. Missouri cob pipe dent.

A very large dent race characterized by large cob. Ear 28 cm. long'
22.5 cm. in cir. ; cob 14 cm. in cir. ; 20-rowed; pericarp colorless. Seeds
1.5 X .9 cm.; endosperm white. Red cob.

22. Mosaic red dent.

A dent characterized by dark intense red pericarp. Used only for
study of that character.

Zea mays amylacea. The flour corns.

10. White floury.

A thoroughly floury race, showing absolutely no corneous starch.
Ear 22 cm. long, 14.5 in cir.; 14-rowed; pericarp colorless. Seeds 1.2 x
1.0 cm.; endosperm colorless. Cob white.

Zea mays saccharata. The sweet corns.
19. Stowell's evergreen.

A large-eared extremely wrinkled-seeded late sugar corn. Ear 16
cm. long, 15.5 cm. in cir.; 14-rowed; pericarp colorless. Seeds 1.4 x .7
cm. ; endosperm colorless. Cob white.

18. Early Crosby.

A twelve-rowed sugar corn. Ear 14.5 cm. long, 14 cm. in cir.; 12-
rowed; pericarp colorless. Seeds .9 x .9 cm.; decidedly wrinkled but
thick full seeds; endosperm colorless. Cob white.

54. Black Mexican.

An eight-rowed sugar corn characterized by purple aleurone cells.
Ear 13.5 cm. long, 12 cm. in cir.; 8-rowed; pericarp colorless. Seeds
.9 X 1.1 cm.; aleurone cells purple; endosperm colorless.

Methods Used.

In carrying out the large amount of tedious routine work
necessary in the collection of data from the crosses of the above
types, a great effort was made to reduce experimental errors

EXPERIMENTAL METHODS. 29

to a minimuin. No part of the work was left to farm workmen
except the preparation of the breeding plots and their culti-
vation. The planting, labeling of families, crossing, selfing,
harvesting, filing of seed and collection and reduction of data
were done by the authors. The senior author alone is respon-
sible for the details of the work until 1909. In 1909 and 1910
the senior and junior authors both shared in the labor. Since
1907 efficient aid in harvesting, filing seed, etc., has been given
by Mr. C. D. Hubbell of the Conn. Agr. Exp. Station. In
1910 Mr. D. W. Davis and Mr. 0. E. White, graduate students
at Harvard University, aided in selfing ears of various selections.

The ears have always been shelled and seeds classified and
filed in seed envelopes. Where there has been the least question
about classification the work has been duplicated by two
observers. If then there has been a doubt concerning the
characters borne by particular seeds, those in question have
been grown for another generation. The planting has been
done from the seed envelopes directly to the field. There
they were planted in hills three and one-half feet apart each
way, four seeds to the hill. It was not considered necessary
to start the seeds in the greenhouse in sterilized soil as is done
with smaller seeds. Maize seed ver}^ seldom germinates after
remaining in the ground over the winter in this climate. Fur-
thermore the corn which was not hand pollinated was not
husked directly on the field so that there was but little chance
that any seeds should remain upon the ground. Great care
was taken not to drop seeds at planting time in other than the
hills marked out. These were covered carefully and after the
young plants appeared above the surface, any individuals
not exactly in the hill were removed. No plants have ever
given evidence that they were misplaced and there is every
reason to believe that the work is accurate in this regard.

The different families were marked in the field by heavy
stakes to which wired tree labels were attached. As an addi-
tional precaution against mis-labeling or misplacement of
labels, however, a planting plan was always kept on file showing
the exact location of every plant in the field. With this safe-
guard every field stake might have been removed without
making the least confusion.

All crossing and selfing were done by hand. Individual

30 INHERITANCE IN MAIZE.

plants were used as the male parent in nearly every case. If
possible the male parent of a cross was also selfed with its own
pollen so that selfed seed of that individual was accessible if
necessary. If for any reason it was particularly desirable to
have the progeny of a plant where through an accident none
of its own pollen was available, it was pollinated from a sister
plant. This fact was always noted, however, and the male
parent selfed if possible.

Heavy manila paper bags were used to protect both male and
female inflorescences from foreign pollen. These were found
much more desirable than paraffined bags as the latter were
likely to become inverted and filled with water during a rain
storm. The manila bags stood up well in the rain, dried out
quickly, and seldom failed to furnish dry viable pollen. The
tassels were bagged about three days before any pollen was
ripe. Of course here there was a slight chance of enclosing
foreign poUen. This pollen, however, would have been three
or more days older than the pollen coming from the bagged
flowers, and therefore much less viable. Even disregarding
this fact, however, the immense amount of pollen furnished
by the bagged infloresence would so dilute any foreign pollen
that the ratio would be at least 10,000 to. 1 in favor of the former.

The female flowers were always bagged of course before any
of the silks were showing, and any bracts or leaves showing
foreign pollen were carefully removed. Here again, however,
is a slight chance of enclosing foreign pollen. This error has
been determined by bagging 53 ears and allowing them to re-
main in the bags. Forty-four ears formed no seeds, six ears
formed one seed each, two ears formed two seeds each and one
ear formed four seeds. There are over five chances to one
then that no viable foreign pollen enters in this way.

The pollination is accomplished by removing the bag from
the tassel, shaking out the empty anthers and dusting the pollen
over the silks of the proposed mother plant. The bag covering
the silks is not entirely removed but is held so that its opening
is horizontal with the silks resting inside. The pollen is then
shaken in at the opening as quickly as possible, taking care
not to let the silks touch the hands or clothing of the operator
or the leaves or stem of the plant. It is sometimes impossible
to keep from touching the silks with the fingers, as it may be

EXPERIMENTAL ERROR. 31

necessary to rearrange them in the bag. To guard against
contamination from this source the hands are carefully cleaned
with 95 per cent, alcohol after each pollination.

The silks at the base of the ear mature first ; those at the tip
of the ear last. For this reason, if one is to be absolutely
certain of a well filled ear, it is necessary to pollinate two or
three times with fresh pollen. This procedure has the disad-
vantages of increasing the error, however, not to speak of the
difficulty of obtaining pollen, so that in this work but one
pollination was made in each case. When pollinated about
five days after bagging, fairly well filled ears were generally
obtained, particularly with the small races.

Immediately after pollination the ear is rebagged and tagged.
From this time until the ears are mature they are inspected
every little while to see that the bags are not too tight for the
maturing seeds. The bags remain on until the ears are har-
vested. They are then picked, husked, tagged with wired
tree labels and dried. Boards through which wire nails have
been driven are hung from the ceiling of the drying room to
prevent the depredations of mice. The ears are impaled upon
these nails and thus dry surrounded by a current of air.

Experimental Error.

The manipulation during pollination is undoubtedly pro-
ductive of an experimental error which even the most careful
work cannot entirely prevent. This error was determined as
follows. Twenty-five ears were bagged and allowed to remain
in this condition for five days. The bags were then opened
and given the manipulation that was necessary for hand-
pollination, except that no pollen was applied. The ears were
then rebagged and remained so until harvest time. No seed
were formed on sixteen ears; three ears produced one seed
each ; four ears produced two seeds each ; while one ear produced
four seeds and one ear produced five seeds.

There is a possibility then of an experimental error of five or
six seeds out of the two hundred to eight hundred produced
per ear. This is to be considered as a maximum error and
not the probable error, the latter being less than one seed per
ear.

INHERITANCE IN MAIZE.

PART II.

ENDOSPERM CHARACTERS.

These hybridization studies are all reported under the head-
ings of the different characters investigated, as this seems to be
the method calculated to show the data with the least con-
fusion. The female parent is written first, using the variety
number given under the description of the material under
investigation. For example 7 x 54 represents a cross of Learning
yellow dent female with Black Mexican sweet male. When
this cross is grown and hand-pollinated selfed ears are obtained,
they are numbered (7 x 54)-l, (7 x 54)-2, (7 x 54)-3, etc. Should
ear number 2 be grown for still another generation, the crop
obtained is numbered (7 x 54)-2-l, (7 x 54)-2-2, etc., thus the
exact generation of a particular ear is always shown. The
characters under consideration are known by letters; '5', for
example means presence of starchy character and 's', absence
of starchy character: 'P' , represents presence of purple aleurone
cells; 'p' , its absence. An ear numbered (7 x o4)-2-l P S
represents an ear of the third or Fs generation from Avhich
purple starchy seeds have been selected for planting.

Starchiness and Non-starchiness.

Starchiness is the condition of the endosperm of all of Sturte-
vant's maize varieties except Zea mays saccharata, regardless
of the physical condition — corneous starch or soft starch —
in which it appears. The starch grains are fully developed
and possess a shape characteristic of the species Zea mays.
The sugar maize does not have the ability to develop these
starch grains to maturity. Some starch is formed but it remains
small, angular and abortive, hence the seeds ripen from the
stage of maturity called the "milk" without much change,
giving the seed a wrinkled translucent appearance. The dif-
ference in size of the starch grains in the two races is shown in
Table 10. This difference in the size of starch grains however,

PLATE II.

5-

.->r%

No. 2'4 Rhode Island white cap (starchy parent) ; b. No. 53 Crosby
non-starchy parent) ; c. result of cross 24x53 showing heterozygous
seeds in which starchiness is completelj^ dominant, d, an ear with
F2 seeds showing mono-hybrid segregation. Lower row daughters
of d. E, f and g, results from planting" starchy seeds. One ear out
of three is pure starchy, li, result from planting non-starchy seeds.

Segregation of Starchines.s and Non-starchiness.

INHERITANCE OF STARCHINESS.

33

is not the whole difference between starchy and non-starchy
races. As the starchy races ripen, starch formation goes on
at a steady rate, while in the non-starchy races there is an
actual breaking down of endosperm materials into cane sugar
and various hexoses. This is shown by determinations we have
made of reducing sugars in both starchy and non-starchy
races when both were at the "milk" stage of maturity. The
non-starchy races contained from one and one-fourth to two
and one-half as much reducing sugar as the starchy races.

Correns (:01) has already shown that starchiness behaves
as a Mendelian dominant allelomorphic to its absence. Domi-
nance was complete, and segregation generally * exact and
inheritance discontinuous. It is not to confirm his work that
the matter is taken up here, but to consider other questions
to which the data are relevant. These questions relate chiefly
to the mathematical hypothesis of Mendelism, to prepotency
of individuals, and to gametic purity. The data from which
the problems are discussed are not selected, but the figures

The one exception was the pop and sugar cross mentioned later.

TABLE 1.

NO. 15 FLINT STARCHY X NO. 54 NON-STARCHY.

Ear

No.

S

s

Total

Ratio

per 4

Dev.

P. E.

(15x54

-1

135

48

183

2.9508

1.0492

0.0492

0.0864

( "

-2

253

85

338

2.9944

1 . 0056

0.0056

0.0635

( " ^

-3

150

42

192

3.1248

0.8752

0.1248

0.0843

( " ^

-4

238

96

334

2.8504

1.1496

0.1496

0.0639

( " ^

-6

190

72

262

2.9008

1.0992

0.0992

0.0722

( "

-8

302

96

398

3.0352

0.9648

0 . 0352

0.0586

( " ^

-11

242

105

347

2.7896

1.2104

0.2104

0.0627

/ If

-15

236

79

315

2.9968

1 . 0032

0.0032

0.0658

(15x54

-2-1

235

70

305

3.0820

0.9180

0.0820

0.0669

( " ^

-2-2

242

79

321

3.0156

0.9844

0.0156

0.0652

/ « '

-2-3

248

66

314

3.1592

0.8408

0.1592

0.0659

( " ^

-2-4

227

68

295

3.0780

0.9220

0.0780

0.0680

( "

-2-5

200

59

2.59

3.0888

0.9112

0.0888

0.0726

( "

-2-6

182

74

256

2.8436

1.1564

0.1564

0.0730

^ / 11

-2-7

238

91

329

2.8936

1.1064

0.1064

0.0644

( "

-2-8

195

58

253

3.0832

0.9168

0.0832

0.0734

( "

)-2-9

162

38

200

3 . 2400

0.7600

0 . 2400

0.0826

( "

1-2-10

131

53

184

2.8476

1.1524

0.1524

0.0861

( "

)-2-ll

132

40

172

3.0696

0.9304

0.0696

0.0891

( "

)-2-12

101

32

133

3.0376

0.9624

0.0376

0.1013

34 INHERITANCE IN MAIZE.

include only about one-fourth of the hand-pollinated ears at
our disposal, belonging to the starchy and non-starchy cross.
This number seemed sufficient for our purpose, and the segre-
gating kernels were not counted on the remaining ears. It
should be mentioned however, that any wide departures from
the normal on any of the four hundred selfed heterozygous
ears of this cross would have been noted and reported if such
had occurred.

Dominance was found to be complete. In no case was there
the slightest difference between the homozygous and the heter-
ozygous seeds in either outward appearance or in the character
of the starch cells when examined microscopically. Whatever
it is that is brought in by the starchy parent to cause starch
formation is sufficiently active to bring about complete change
when present in one "dose" (that is from one parent). As in
all endosperm characters, when S is the male parent the starchi-
ness appears in the current generation so called, giving the
most perfect illustration of Xenia there is known. As a matter
of fact, one is not dealing with the current generation but with
the Fi generation, the endosperm being a younger generation
than the plant which bears the ear. In no case, in an experience
with several thousand seeds, did an Fi seed showing Xenia
fail to show a heterozygous condition; nor did extracted reces-
sives (sugar seeds) of the F2 generation ever show a heter-
zoygous condition. From this, one may conclude that the second
male nucleus that fertilizes the endosperm nucleus always
bears the same characters as the first male nucleus that ferti-
lizes the embryo nucleus or egg. Several heterozygous seeds
have been found, however, that were not completely starchy,
but had developed bilaterally into half starchy and half non-
starchy. There was not a gradual change from the one condi-
tion to the other, but a distinct line of demarkation, with one
side as absolutely distinct from the other as are the pure races
of each kind. None of these seeds were homozygous starchy,
and Correns' interpretation of similar phenomena as cases in
which the second male nucleus did not fuse with the endosperm
nucleus but each developed separately, seems well founded.
Attention is called to the matter for this reason. It is an
hypothesis generally received with quiescence if not with
acquiescence, that starchiness (and other "presence" characters)

INHERITANCE OF STARCHINESS.

35

is due to presence of an enzyme not possessed by the allelo-
morph. Now if this is true, the enzyme must be a colloid with
such large molecules that there is absolutely no dialysis, other-
wise it seems as if it would diffuse through the unripe seed
sufficiently to act as a catalyser throughout the entire endo-
sperm. No matter what is the correct interpretation, there
is certainly a definite chain of hereditary transmission of char-
acters from cell to cell during development, and each original
cell follows an inertia of its own with little influence on others.

TABLE 2.

NO. 24, FLINT STARCHY X NO. 54, NON-STARCHY.

Ear No.

S

5

Total

Ratio per 4

Dev.

P. E.

(24 X 54)-l

274

94

368

2.9784 : 1.0216

0.0216

0.0609

( " )-2

219

73

292

3 . 0000 : 1 . 0000

0.0000

0.0684

( " )-6

256

89

345

2.9680 : 1.0320

0.0320

0.0629

( " )-8

200

64

264

3.0304 : 0.9696

0.0304

0.0719

( " )-9

155

69

224

2.7680 : 1.2320

0.2320

0.0781

( " )-io

212

59

271

3.1292 : 0.8708

0.1292

0.0710

( " )-ll

213

77

290

2.9380 : 1.0620

0.0620

0.0686

( " )-12

268

80

348

3.0804 : 0.9196

0.0804

0.0626

( " )-13

264

106

370

2.8540 : 1.1460

0.1460

0.0607

( " )-14

227

90

317

2.8644 : 1.1356

0.1356

0.0656

(24 X 54)-l-2

207

68

275

3.0108 : 0.9892

0.0108

0.0704

( " )-l-6

223

75

298

2.9932 : 1.0068

0.0068

0.0677

( " )-l-8

235

90

325

2.8924 : 1.1076

0.1076

0.0648

( " )-l-9

106

36

142

2.9860 : 1.0140

0.0140

0.0980

We have said that dominance appears to be complete; segre-
gation also appears to be complete. It is seldom necessary to
subject extracted recessives to proof by growing them a further
generation. Some strains of non-starchy maize, No. 18 for
example, are much less wrinkled than others; and when such
a strain is crossed with a flint type there is less difference
between dominants and recessives in appearance than when
certain other types are crossed. But in no case is there the
least difficulty in separating the segregates correctly. Whether
this apparent segregation is as complete as it appears^ we shall
discuss presently. It should further be mentioned that the
seeds can also be classified with absolute exactness by micro-
scopical examination.

36

INHERITANCE IN MAIZE.

Tables 1-9 contain the proportion of the starchy and non-
starchy seeds obtained as progeny when heterozygous seeds
were planted; although, as was stated before, only a few ears
from each family were counted. One object in view is to show
the behavior of starchy and non-starchy in several races.

TABLE 3.

NO. 5, FLINT STARCHY X NO. 18, NON-STARCHY.

Ear No.

5 X 18)-4

" )-8

" )-10

" )-16

" )-18

" )-21

" )-25

" )-30

s

5

Total

181

68

249

172

66

238

215

68

283

225

69

294

186

61

247

136

42

178

176

50

226

218

68

286

Ratio per 4

2.9176 : 1.0924
2.9908 : 1.1092
3.0388

3.0612
3.0120
3 . 0560
3.1152
3.0488

0.9612
0.9388
0.9880
0 . 9440
0.8848
0.9512

Dev.

0 . 0924
0.1092
0.0388
0.0612
0.0120
0.0560
0.1152
0.0488

P. E.

0 . 0740
0.0757
0.0694
0.0681
0.0743
0.0876
0.0777
0.0691

There is reason to believe that different races can be identical
in appearance, but may have such different gametic composition
that they may affect a character possessed by a race with which
they may be crossed, in very different manners. (See purple
aleurone cells and non-purple.) Examination of the tables
shows this not to be the case with starchy and non-starchy.
All of the starchy and non-starchy races with which crosses
have been made behave in exactly the same manner. There
is no difference in appearance in heterozygotes from different
races that is not accounted for by the different shaped seeds
possessed by the parents, and wide variations in shape occur
only in generations later than Fi. In the Fi generation the
shape is intermediate between that of the two parents.

TABLE 4.

NO. 11, FLINT STARCHY X NO. 18, NON-STARCHY.

Ear No.

S

J

Total

Ratio per 4

Dev.

P. E.

(11 X 18)-7
( " )-14
( « )-15
( « )-22

220
218
200
235

74
81

77
87

294
299

277
322

2.9932 : 1.0068
2.9164 : 1.0836
2.8880 : 1.1120
2.9192 : 1.0808

0.0068
0.0836
0.1120
0.0808

0.0681
0.0676
0.0702
0.0651

INHERITANCE OF STARCHINESS. , 37

If one examines the tables carefully however, he sees at once
that there is quite a difference in the ratios obtained. They
vary from a ratio of 2.7896 : 1.2104 in ear (15 x 54)-ll, Table
1 to a ratio of 3.2020 : 0.7980 in ear (19 x 7) -2, Table 7. This
brings up an important question. Does this discrepancy
represent an expected probable error in chance matings; or, is
there a prepotency in certain families through which excessive
numbers of dominants or of recessives tend constantly to
reappear? Correns did indeed find in one family such an excess
of starchy seeds, but it is not known whether this apparent
prepotency was transmitted in further generations.

Tables 1-9 show several cases where ears with a ratio deviat-
ing from the expected 3 : 1 of Mendelian hypothesis have been
grown for another generation. For example, ear (8 x 54)-l
of Table 6 has a ratio of 2.9420 : 1.0580 while ear (8 x 54)-5
of the same table has a ratio of 3.0780 : 0.9220; yet the progeny
of these ears average just about the 3 : 1 ratio of theory. There
are even more ears with an excess of recessives from the ear
that had the excess of dominants and vice versa. Other deviants
have been grown for several generations, and while the exact
ratios have not been recorded it may be stated with confidence
that wide deviations occurring in considerable numbers would
have been noticed while making other records. It may be
concluded then that no prepotency or tendency to aberrant
ratios is a constant characteristic of any of our families. How
then can the discrepancies from theoretical ratios be explained?

To study this question the probable errors of all of the ratios
have been calculated. The method used has been that of
Johannsen (: 09, p. 405), except that the mean error has been
reduced to the probable error by multiplying by the factor
0.6745. The standard deviation of a Mendelian proportion is

VpXq
=t where p and q are the Mendelian terms, in this case

p+q

-\/3Xl 1.7321

case 3 and 1. Thens.D.= ± =± =±0.4330. The

4 4

S. D.

probable error, E= ±0.6745 where n is the total number

of variates.

38 INHERITANCE IN MAIZE.

To find out whether the different ratios given in Tables 1-9
are what should reasonably be expected if the Mendelian
theory of chance matings of equal numbers of gametes 5 and 5
in both male and female germ-cells is correct, it should be under-
stood just what is meant by probable error in the law of error.
Plus errors and minus errors should occur with equal frequency,
small errors should occur more frequently than large errors,
and very large errors should not occur. Determined as above
the probable error means that the chances are :

1 to 1 that the true value lies within =>= E

4 . 5 to 1 that the true value lies within ± 2E

21 to 1 that the true value lies within ± 3E

142 to 1 that the true value lies within ± 4E

The theory of error also provides for errors of any size in their
proper frequency or rather infrequency, but as a matter of
fact in practice if errors greater than ± 4E occur they are
probably due to experimental errors or avoidable mistakes.

We may consider each ear given in Tables 1-9 as a determi-
nation and its probable error as the probable error of a single
determination. With this in mind we find that in the 94 ears
tabled there are 49 plus errors and 45 minus errors. Further
we find that the theoretical mode or 0 error is almost 3:1,
being in fact very slightly greater. The errors are distributed
as follows :

Within ± E 47 . 8% — Theory 50 . 0%
Within ± 2 E 83 . 0% — Theory 82 . 3%
Within ± 3 E 96 . 8% — Theory 95 . 7%
Within ± 4 E 100 . 0% — Theory 99 . 3%

The sum total of these segregates is 23529 to 7811, a ratio of
3.0031 : 0.9969 ± .0066.

It should be mentioned simply in order to suppress no data
that one ear was found with a probable error somewhat in
excess of ± 4 E. It was found by growing this ear for another
generation however that this was due to an experimental error.
There was a great excess of sugar seeds, but in the starchy
seeds there proved to be about 4 heterozygotes to 1 homozy-
gote. Since no other ear like this has ever been obtained, and
since it is known that during the progress of this experiment

INHERITANCE OF STARCHINESS.

39

several ears were first selfed with pollen killed by rain and
afterward pollinated with pollen from a sugar plant to get
material for another purpose, it seems highly probable that this
ear was of a similar mixed parentage and that its explanatory
label had been lost.

As to whether these data support the Mendelian h3qDothesis
or not there may be slight grounds for a difference in opinion.
Our own opinion is that when we take into consideration the
chance for experimental error, the ratios are well within the
limits of probable error. One thing at least is brought out clearly,
the behavior of segregates in more than one generation and a
variety of matings are necessary, if one is to draw conclusions
as to the exact mode of inheritance of character pairs from
small numbers.

One further point remains for discussion. Do the extracted
homozygotes breed true ? In other words, is segregation an
absolute separation of a gene from its absence ? or, is there only
a relative segregation? Morgan (; 10) has suggested that
relative segregation may explain Mendelian facts, if one pre-
supposes that when the amount of the gene falls below a certain
limit the dominant fails to develop. This idea while interpret-
ing the facts in the F2 generation is inadequate to explain the
apparent purity of further generations of extracted recessives,
for if this hypothesis were true many recessives would show the
dominant character when crossed.

In Table 9 is shown the segregation of extracted dominant
starchy seeds. The ratio is as nearly the expected 2 hetero-
zygotes to 1 homozygote as could well be expected. Several

TABLE 5.

NO. 17, FLINT STARCHY X NO. 54, NON-STARCHY; NO. 18, NON-STARCHY X
NO. 58, FLINT starchy; and NO. 7, DENT STARCHY X NO. 54,

NON-STARCHY.

Ear No.

S

5

Total ; Ratio per 4

1

Dev.

P. E.

(17 X 54)-l
(18 X 58)-l
( 7 X 54)-l
( " )-2

328
332
379
493

102
102
137
131

430
434
516

624

3.0512 : 0.9488
3.0600 : 0.9400
2.9380 : 1.0620
3.1604 : 0.8396

0.0512
0 . 0600
0.0620
0.1604

0.0563
0.0561
0.0514
0.0467

40 INHERITANCE IN MAIZE.

thousand dominant homozygotes have been bred for further
generations and these have all bred true to the starchy character.
This is, in general, the case with the extracted non-starchy
seeds. Furthermore, there are in commercial use many sugar
corns that are extracted recessives. Golden Bantam, Late
Egyptian and many others are examples of races that have
originated from crosses with starchy varieties. The wrinkled
seeds have been selected and have bred true. Out of the many
million seeds that are annually grown for the canning factories,
however, there does appear an occasional ear with semi-starchy
seeds. These ears transmit the character and give the canners
no end of trouble. There is no way to find out whether these
ears appear only on varieties which somewhere in their ancestry
had a starchy parent. One can only say that they do appear
in ratios not exceeding one ear in ten thousand. By some
lucky chance some of these ears made their appearance in our
controlled cultures. All of our extracted recessives have
proved true to non-starchiness (*) except from the progeny of
ear (8 x 54) -1-6. The majority of the progeny of this ear were
also non-starchy, but three ears appeared which were decidedly
semi-starchy, one of which is shown in Plate III, fig. b. There
was no possibility that these ears could have grown from a
normal heterozygous seed. They were not plump seeds like
a true heterozygote nor did they segregate into starchy and
non-starchy in the next generation. The entire ear was rather
uniformly semi-starchy and quite different from the true starchy
ears. Microscopical examination showed definitely that starch
grains had been developed normally to a size intermediate
between the true starchy and the true sweet seeds of the same
family. This fact is shown in Table 10.

* There are other cases where some apparent starchiness is always
developed, namely when pop races are crossed with non-starchy races.
We interpret this as being due to the small size of the resulting F2 seeds
borne on intermediate Fi ears. When the seeds are small the endosperm
material more nearly fills the pericarp than when they are large. The
wrinkled condition is therefore less apparent. If one has had consider-
able experience in classifying starchy and non-starchy seeds, such crosses
are seen to show normal segregation. If, however, careful classification
is not made and the seeds are not tested in further generations pop and
non-starchy crosses always appear to show an excess of starchy seeds.
It is suggested that this is the explanation of Corren's failure to obtain
normal ratios in a similar cross. These cases are not real exceptions
to the statement made above, however, for recessives extracted from pop
crosses are never grown commercially as sugar corns.

PLATE III.

.^'w'^j.'Ct-*^^x^j.\'ii-yi-

a. P2, Fi, and F2 seeds from cross between No. 19 Stowell's Evergreen
sugar and No. 2 Illinois low protein dent maize.

b. Middle ear is a semi-starchy ear No. (8x54) -1-6, progeny of an ex-
tracted recessive (wrinkled) seed. On the left is an extracted dom-
inant (starchy) ear of the same cross. On the right is a well
wrinkled ear, sister of No. (8x54) -1-6.

Segregation of Starchiness and Non-Starchiness.

PLATE IV.

a Random sample of progeny of starchiest seeds of semi-starch}^ ear
shown in Plate III.

b. Random sample of progeny of most wrinkled seeds of semi-starch}'
ear shown in Plate III.

Gametic Purity.

INHERITANCE OF STARCHINESS.

41

TABLE 6.

NO. 8, DENT STARCHY X NO. 54, NON-STARCHY.

Ear No.

S

.J

Total

Ratio per 4

Dev. P. E.

(8 X 54)-l

( " )-2
( " )-3
( " )-5

381
340
400

384

137
137
141
115

518
477
541
499

2.9420 : 1.0580
2.8512 : 1.1488
2.9576 : 1.0424
3.0780 : 0.9220

0.0580 0.0513
0.1488 i 0.0535
0.0424 0.0502
0.0780 0.0523

(8 X 54)-l-3PC
( " )-l-4PC
( " )-l-6PC
( " )-l-8PC
( " )-l-14PC
( " )-l-6P
( " )-l-llP
( " )-l-13P
( " )-l-15P
( " )-l-l
( " )-l-2
( " )-l-4
( " )-l-10
( " )-l-29

243
302
231
321
238
145
268
320
293
237
236
176
242
272

236
294
147
277
357
324
306
249

64

105

81

117

66

40

78

107

96

88

88

60

80

93

307
407
312
438
304
185
346
427
389
325
324
236
322
365

3.1660 : 0.8340
2.9680 : 1.0320
2.9616 : 1.0384
2.9316 : 1.0684
3.1316 : 0.8684
3.1356 : 0.8644
3.0984 : 0.9016
2.9976 : 1.0024
3.0128 : 0.9872
2.9168 : 1.0832
2.9136 : 1.0864
2.9832 : 1.0168
3.0064 : 0.9936
2.9808 : 1.0192
2.8956 : 1.1044
2.9924 : 1.0076
2.8268 : 1.1732
3.1748 : 0.8252
2.9748 : 1.0252
2.9932 : 1.0068
3.1304 : 0.8696
2.9732 : 1.0268

0.1660 0.0667
0.0320 0.0579
0.0384 , 0.0672
0.0684 0.0558
0.1316 0.0670
0.1356 0.0859
0.0984 i 0.0628
0.0024 0.0565
0.0128 0.0592
0.0832 0.0648
0.0864 0.0649
0.0168 0.0760
0.0064 0.0651
0.0192 0.0611

(8 X 54)-5-2
( " )-5-3
( " )-5-4
( " )-5-5 •
( " )-5-6
( " )-5-8
( " )-5-10
( " )-5-ll

90
99
61
72
123
109
85
86

326
393
208
349
480
433
391
355

0.1044 0.0647
0.0076 0.0589
0.1732 0.0810
0.1748 0.0625
0.0252 0.0533
0.0068 0.0651
0.1304 0.0591
0.0268 0.0638

Seeds from the ear shown in Plate III, fig. b, were divided
into two classes, those most nearly starchy and those most
nearly non-starchy, and planted. A number of selfed ears
were obtained from each class. Those resulting from the seeds
most nearly non-starchy were in part what would immediately
be classified as non-starchy and in part as starchy in appearance
as the parent ear. The ears resulting from the seeds most
nearly starchy were all as starchy as the parents and certain
of them even more so. This fact is shown in Plate IV. Micro-
scopical examination of the most starchy" seeds of this genera-
tion showed that the starch grains were most of them developed
to normal size. The ears were not uniform nor was there
uniform starchiness among the seeds of a single ear. Seeds could
be selected which formed a series running from true sweet to
true starchy, yet those most nearly starchy had a rough appear-

42

INHERITANCE IN MAIZE.

TABLE 7.

NO. 19, NON-STARCHY X NO. 7, DENT STARCHY AND NO. 19, NON-STARCHY
X NO. 8, DENT STARCHY.

Ear No.

S

s

Total

Ratio per 4

Dev.

P. E.

(19x7)-2.

( " )-5

297
486

74
156

371
642

3.2020 : 0.7980
3.0280 : 0.9720

0.2020
0.0280

0.0607
0.0461
0.0575

(19 x 7)-5-l

304

109

413

2 . 9444 : 1 . 0556

0.0556

(19x8)-l
( " )-2
( " )-3

( "■ )-4
( " )-5

183
464
449
303
414

64
152
151

96
139

247
616
600
399
553

2.9636 : 1.0364
3.0128 : 0.9872
2.9932 : 1.0068
3.0376 : 0.9624
2.9948 : 1.0052

0.0364
0.0128
0.0068
0.0376
0.0052

0.0743
0.0471
0.0477
0.0585
0.0507

ance very different from the well-filled pericarp of the true
starchy seeds of the same family. These seeds will be selected
for Starchiness and if uniform ears are finally obtained, will
be crossed with non-starchy again to see if their behavior
is the same as normal starchy maize. Provisionally one is
forced to one of two conclusions. Either homozygous reces-
sive s {and likewise dominants) are not complete segregates, but
products of a partial quantitative separation of genes allowing
traces of the dominant character to remain, traces which may
sometimes accumulate sufficiently to bring out the dominant
character: or, progressive variations are constantly taking place
in small numbers, 'most often along paths that have been passed
before.

TABLE 8.

NO. 60, POP STARCHY X NO. 54, NON-STARCHY.

Ear No.

S

s

Total

Ratio per 4 i Dev. P. E.

(60-5 X 54)-2
( " )-6
( " )-8
( " )-ll
( " )-12

274
273
163
191
249
296
260
243

82

102

53

58
84

356
375
216
249
333

3.0788 : 0.9212 0.0788 0.0619
2.9120:1.0880 0.0880 0.0603
3.0184 : 0.9816 0.0184 0.0795
3.0684:0.9316 0.0684 0.0740
2.9908: 1.0092 1.0092 0.0640

(60-3 X 54)-l
( " )-5
( " )-6

87

107

73

383 1 3.0912:0.9088 0.0912 0.0597 \
367 2.8336 : 1.1664 0.1664 0.0610 !
316 1 3.0760:0.9240 0.0760 0.0657

(60-8 X 54)-l
( " )-8

227
224

67
71

294 3.0884 : 0.9116

295 3.0372:0.9628

0.0884 0.0681
0.0372 0.0680

INHERITANCE OF STARCHINESS. 43

It is our opinion that dominant starchiness — if it is the
same dominant starchiness — has been formed anew. It
occurs too rarely to support a partial segregation theory such
as Morgan's (: 10). If it is asked why starchiness is the char-
acter that arises anew rather than another variation, it is sug-
gested that the peculiar chemical structure of the germ cell of
maize may be such that a molecular readjustment is much more
likely to bring about starchiness than any other variation.
Such a path of least resistance for variations might account for
the many cases in animals and plants where the same variation
has apparently occurred again and again.

Conchision.

These starchy and non-starchy crosses represent a much
larger number of individuals than have ever before been studied
in accurately controlled pedigree cultures. Taking them as a
whole they show that the mechanism by which the members
of an allelomorphic pair are distributed among the gametes, is
accurate. The aberrant ratios sometimes obtained are what
should be expected by the Law of Error. They are not
inherited, and we believe this to show that there is no such
thing as prepotency per se which would cause abnormal ratios.
We might extend this conclusion further and say that there is
no conclusive evidence of a failure of segregation of male gametes
or of selective fertilization (Lock : 06) , or of partial gametic
coupling that presupposes gametes bearing opposite genes to
be formed in unequal numbers (Bateson and Punnett : 08).
Disbelief in prepotency of the kind described above does not
indicate disbelief in different "potencies" as described by
Davenport (: 10). Different potencies, that is various degrees
of manifestation of the same character due to its modification
during development by the action of other developing genes
possessed by the individual, is a different thing and is entirely
logical. In prepotency or potency of this kind segregations
are perfectly normal, and modifications which occur in characters
are due to the gametic constitution of the individual.

The aberrant ratios obtained by Correns in the pop-sugar
cross referred to above, may have been due to modification by
other unknown characters possessed by the parents, but it

44 INHERITANCE IN MAIZE.

seems more likely that they were due to improper classification
of dominants and recessives for the reason that recessives in
such crosses although hyaline and easily classified microscopically
often do fill the pericarp with endosperm material owing to the
small size of the seed.

If then, in cases of simple mono-hybrids where there are no
complications, a ratio of 3.0031 : 0.9969 ± .0066 is obtained;
are we not compelled to take the view that segregation occurs
at the reduction division? Could any less exact division give
the distribution of genes necessary for such exact recombi-
nations? Of course it has long been suspected that this was
the time of segregation, but Bateson (: 09 p. 271) has felt that
obstacles were in the way of interpretiiig the chromosomes
as such important bearers of hereditary qualities. These
obstacles were three in number; first, it is objected that no
correspondence has been shown between visible differences
of type (except sex) and chromosome differences; second, that
no correspondence between complexity of type and chromosome
numbers has been shown; and third, that bud sports are somatic
segregates. There are, it seems to us, no real obstacles here.
One should expect that the quality of the chromosome and not
shape or number, is the important fact. It is even likely that
most of the important morphological characters are carried
by all of the chromosomes, hence a doubling of chromosome
number as has occurred in Oenothera gigas may be relatively
unimportant. The case of bud sports is also fairly clear since
Winkler ( : 09) has shown that a graft hybrid between the
black night shade and the tomato proved to have the sum of
the haploid numbers of the two parents and not the sum of the
diploid numbers. The somatic cell then has a regulatory appa-
ratus of its own. What might be called the normal bud sport
(other sports probably occur from abnormal cell divisions) is
probably due to the fusion of two somatic cells of a hetero-
zygote, followed by a reduction, in which one of the homozygote
forms appears. It must be not understood however that be-
cause Bateson 's objections are considered surmountable, we
therefore believe it to be proved that the chromosomes are the
sole bearers of hereditary characters and that the reduction
division is the time of Mendelian segregation. Judgment
must still be suspended on these matters.

INHERITANCE OF STARCHINESS.

45.

TABLE 9.

EARS FROM F2 GENERATION PLANTS OF STARCHY AND NON-STARCHY

CROSSES.

Starchy Seeds Planted.

Selection

Heterozygous S

Homozygous S

( 8 X 54)-l

6

4

( " )-2

4

5

( " )-3

30

17

( " )-5

75

31

( " )-l-l

67

32

( " )-l-2

44

25

( " )-4

71

38

( " )-io

48

28

(15 X 54)-2

46

13

( " )-3

25

17

(24 X 54)-l

. 28

14

Total

444

224

Ratio

1.93

1

TABLE 10.

RANDOM COMPARISON OF DIAMETER OF STARCH GRAINS.

Extracted Starchy Seeds from {8 x 54)-i o,nd Semi-Starchy and
Non-Starchy from (8 x 54)— 1—6.

Diam. in mm.

.009

.017

1

.034

9

52

.052

.069

.086

.103

.12

.138

.155

Total

No. variates

from starchy

seeds

23
57

34

48

66

36

16

12

3

198

No. variates

from semi-starchy

seeds

17

17

11

5

227

No. variates

from non-starchy

seeds

34

94

52

13

193

46 INHERITANCE IN MAIZE.

Yellow and Non-yellow Endosperm.

Correns (:01) and Lock (:06) each found a yellow color in
the endosperm which behaved with its absence as a single alle-
lomorphic pair. We have found two * yellow colors in the
endosperm each behaving when crossed with its absence, as
an independent allelomorphic pair. A part of the experiments
with these characters has been described in a previous paper
(East :10). In this paper some further data are presented.

Both of these yellow colors, although they behave in inheri-
tance as separate entities are either identical or very similar
in composition. They are insoluble in water, somewhat soluble
in alcohol and easily soluble in ether, chloroform, benzine,
benzol and carbon bisulphide. They occur in rhombic plates
in the starch cells and possibly also in the chromoplasts although
this is not certain. From these facts it might be supposed that
they were hydrocarbons with compositions similar to carotin.
They do not give the general reactions however which the fatty
pigments or lipochromes - — of which carotin is an example —
give with sulphuric acid or iodine dissolved in aqueous potas-
sium iodide. Independent of their solubility reactions, this
would class them with tlie anthochlorins (Courchet '88).

Considering the importance to Mendelian theory of the
discovery that two similar and possibly identical characters
may each act with its own absence as an independent allelomor-
phic pair, further chemical investigations are being made which
will be reported in a separate paper. It may simply be stated
here that as far as is known these colors are indistinguishable,
but as they behave differently in crosses they will be known as
Yi and Y^.

~ A number of crosses were made between yellow and non-
yellow which gave only 3 : 1 ratios. The remaining crosses
shown in Tables 11-16 each showed one or more ears with
dihybrid ratios.

* Lock mentioned that light yellow seeds appeared in his crosses, but
he classed them as whites which vitiates his study of Mendelian numerical
proportions.

INHERITANCE OF YELLOW ENDOSPERM. 47

TABLE 11.

F2 SEEDS FROM CROSS OF NO. 1 WHITE DENT X NO. 7 YELLOW DENT.

Ear No.

Y

y

Ratio Approx.

(1 X 7)-l
(1 X 7)-2

587
127

212
30

3 : 1
3 : 1

Table 11 gives the results from two selfed ears of No. 1 white
dent crossed with No. 7 yellow dent. They approximate 3 to 1
ratios although ear No. 1 has an excess of non-yellow and ear
No. 2 an excess of yellow seeds. This cross proved to be too
late for the Connecticut climate and the resulting Fa seeds were
immature and difficult to classify. Yellow was dominant and
appeared as Xenia in the Fi seeds but the F2 seeds varied in
different ears in a peculiar manner. Where there was suffi-
cient soft starchy matter in the caps of the seeds the hetero-
zygotes were considerably lighter colored at the cap than when
the seeds possessed more corneous starch. The same phenome-
non occurred in reciprocal crosses; so that when there was
sufficient soft starchy matter the heterozygotes could be dis-
tinguished from the homozygotes either way the cross was made.
(See cross of floury yellow with non-yellow.)

Ear (1 X 7)-lY gave only one selfed ear with 126 yellows of
various shades, 14 white, and 3 doubtful seeds. The mother
seed was probably Yi Y2 yi y2. Several open field ears from
yellows with white caps all proved to be heterozygous, thus
proving the above statement regarding Xenia. The crop
from the white seeds proved pure for non-yellow.

INHERITANCE IN MAIZE.

TABLE llA.

F3 SEEDS OF EAR NO. 2 OF CROSS SHOWN IN TABLE 11.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(1 X 7)-2-3

423

108

3 : 1

' )-2-7

351

108

3 : 1

' )-2-8

375

120

3 : 1

' )-2-12

343

111

3 : 1

' )-2-14

577

Pure yellow

' )-2-17

286

'58

3 : 1

' )-2-18

209

87

3 : 1

' )-2-19

360

135

3 : 1

' )-2-20

341

105

3 : 1

' )-2-22

319

92

3 : 1

' )-2-23

408

168

3 : 1

' )-2-25

633

40

15 : 1

Table 11a shows the results from planting (1 x 7) -2 Y seeds.
Ears Nos. 3 and 17 have an excess of yellow seeds. Possibly
they were 15 : 1 ratios in which the yellows were very light and
could only have been classified with certainty by growing the
supposed whites another generation. The remaining ears all
showed 3 : 1 ratios except ear No. 25. This ear was clearly
a 15:1 ratio. The crop from (1 x 7)-2y (extracted whites)
gave 12 pure white ears, showing that the classification of the
F2 seeds was correct.

A cross between No. 5 white flint and No. 6 yellow dent
(Tables 12 and 12a) showed in all cases complete dominance
of yellow. In the Fi seeds which were of course flinty like the
mother, there was no soft starch in the cap and the heterozygotes
were exactly like pure yellow flint seeds. In the Fi plants the
F2 homogygous and heterozygous yellow seeds were also indis-
tinguishable. It was necessary to grow them to distinguish
heterozygous yellow from homozygous yellow. In the F2
plants with F3 seeds, however, there was a considerable segre-
gation of dented ears from flint ears. Here as in the cross of
(1x7) it was fairly easy to distinguish heterozygous yellows
from homozygous yellows when the seeds of the former had a
well developed soft starchy zone in the cap.

Although as has been stated the Fi seeds were all exactly

INHERITANCE OF YELLOW ENDOSPERM.

49

like pure yellow flint seeds, they nevertheless belonged to two
classes. The Xenia seeds (Fi seeds) of the hybrid ear con-
tained 159 seeds which were dark yellow and 145 seeds which
were a considerably lighter yellow. This striking phenomenon
was not understood until another generation was grown from
the seeds. Table 12 showing the selfed ears resulting from the
dark seeds, and Table 12a showing the selfed ears resulting
from the light seeds make this matter plain. Excluding ear
(5 X 6)-9 from Table 12 because it evidently came from a pure
yellow seed grown in this family through an error, and ear
(5 X 6)-lla from Table 12a which evidently grew from a self-
pollinated seed of the mother No. 5, it is clear that the No. 6
plant furnishing the pollen for the cross was homozygous for
one yellow and heterozygous for the other. The classification
into light and dark yellows was somewhat arbitrary and there-
fore some ears in Table 12 gave ratios of 3 : 1 and some ears
in Table 12a gave ratios of 15 : 1 but the fact that about one-
half of the Fi seeds had a gametic formula of Yi yi Y2 ji and

TABLE 12.

F2 SEEDS FROM CROSS OF NO. 5 WHITE FLINT X NO. 6 YELLOW DENT.

Dark Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(5 X 6)-l

326

29

15 : 1

' )-2

316

27

15 : 1

' )-3

313

28

15 : 1

' )-7

354

122

3 : 1

' )-8

331

109

3 : 1

' )-9

307

Pure yellow

' )-10

475

25

15 : 1

' )-ll

298

113

3 : 1

' )-12

368

19

15 : 1

' )-13

363

35

15 : 1

' )-14

489

29

15 : 1

' )-15

427

118

3 : 1

one-half the formula Yi yi, or Y2 y2, is certain. Ear (5 x 6)-7a
is the only ratio in doubt. It is probably 15 : 1 as the yellows
were very light and difficult to classify, and some were probably
placed with the non-yellows.

50

INHERITANCE IN MAIZE.

TABLE 12A.

F-i SEEDS FROM SAME CROSS AS SHOWN IN TABLE 12.

Light Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(5x6

)-la

359

117

3 : 1

( "

-2a

144

54

3 : 1

( "

-3a

173

63

3 : 1

( "

-4a

433

136

3 : 1

( "

)-5a

557

35

15 : 1

( "

)-6a

316

120

3 : 1

( "

-7a

450

49

10 : 1

( "

-8a

229

86

3 : 1

( "

-9a

325

115

3 : 1

( "

-10a

227

87

3 : 1

( "

-11a

434

Pure white

( "

-12a

318

118

3 : 1

( "

-13a

256

93

3 : 1

Tables 13, 13a, b, c, d, show results from an opposite cross.
No. 11, yellow flint was the female parent and No. 8, white
dent was the male parent. There was no effect of Xenia, as
the Fi hybrid seeds were as yellow as the pure No. 11. Table
13 shows the results from the Fi hybrid seeds. Every ear
approximates a 3 : 1 ratio except ears (11 x 8)-7 and (11 x 8)-8.
Ear (11 X 8)-7 is shown afterwards by Tables 13b and c to have
been in reality a 15 : 1 ratio. In other words it was a Yx yi Y2 yj

TABLE 13.

F3 SEEDS FROM CROSS OF NO. 11, YELLOW FLINT X NO. 8, WHITE DENT

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(Ilx8)-1

358

154

3 : 1

( " )-2

124

41

3

( " )-3

389

127

3

( " )-4

340

96

3

( " )-6

252

83

3

( " )-7

454

145*

3

( " )-8

204

70**

3

* ** Proved to be a mixture of Y y and y, with preponderance of Y y.

INHERITANCE OF YELLOW ENDOSPERM.

51

ear. Ear (11 x 8)-8 probably was also of the same character
as about half of the seeds classed as white proved to be heterozy-
gous. Table 13d shows only two ears out of eight to have been
other than white but an inspection of the open field crop showed
such a large proportion of apparently heterozygous ears, that
this ratio is probably not the real one.

Ear (11 X 8) -2 proved to be Yi yi or Y2 y2 as is shown in Table
13a. There is a ratio of about 2 heterozygous to 1 homozy-
gous ears.

TABLE 13A.

Fs SEEDS OF EAR NO. 2 OF CROSS SHOWN IN TABLE 13.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(llx8)-2-l

275

Pure yellow

( " )-2-3

237

75

3 : 1

( " )-2-5

244

71

3 : 1

( " )-2-7

374

Pure yellow

( « )-2-8

344

113

3 : 1

( " )-2-9

280

Pure yellow

( " )-2-10

99

3i

3 : 1

( " )-2-ll

173

38

3 : 1

( " )-2-15

274

75

3 : 1

Ear (11 X 8)-7 was evidently wrongly classified as is shown in
Tables 13b and 13c. Ear (llx8)-7-l is probably a 15 : 1
ratio. If this is true then there were 2 ears with gametic formula
Yi yi Y2 ya, 2 ears with gametic formula Yi yi or Y2 y2, 1 ear
with formula Yi Yi Y2 Y, [Ear (11 x 8)-7-9], and 3 ears with
formula yi y2. The apparently white seeds from this ear were
not all non-yellow, but partly pure and partly heterozygous
light yellows. That is, they were Yi Yi or Y2 Y2 or Yi yi or
Y2 y2.

52

INHERITANCE IN MAIZE.

TABLE 13B.

F3 SEEDS OF EAR NO. 7 OF CROSS SHOWN IN TABLE 13.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(11 x8)-7-l

207

25

8 : 1

( " )-7-4

253

68

3 : 1

( " )-7-6

193

73

3 : 1

( " )-7-8

163

79

3 : 1

( " )-7-9

456

Pure yellow

( " )-7-ll

108

35

3 : 1

( « )-7-14

88

5

15 : 1

TABLE 13C.

F3 SEEDS OF EAR NO. 7 OF CROSS SHOWN IN TABLE 13.

Apparently White Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(11 x8)-7-la
( " )-7-2a
( " )-7-3a
( " )-7-4a
( « )-7-5a

271
323

504

330
117
300

Pure non-yellow
Pure light yellow
Pure non-yellow

3 : 1
Pure non-yellow

TABLE 13D.

F3 SEEDS OF EAR NO. 8 OF CROSS SHOWN IN TABLE 13.

White Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(11 x8)-8-l

406

194

3 : 1

( " )-8-3

394

Pure non-yellow

( " )-8-6

560

( " )-8-7

348

((

( " )-8-9

490

it

( " )-8-ll

360

u

( " )-8-12

360

u

( " )-8-13

442

Pure yellow

INHERITANCE OF YELLOW ENDOSPERM.

53

TABLE 14.

f2 seeds from cross of no. 11 sturges' yellow flint x no. 24
sanford's white flint.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(11 X 24)-3
( " )-4

( " )-5
( ■ " )-6

467
320
499
356

164
137
142
116

3 : 1
3 : 1
3 : 1
3 : 1

Table 14 shows the results from selfing the Fi seeds of a cross
between No. 11, yellow flint and No. 24 white flint. There
was no effect of Xenia. The ears gave 3 : 1 ratios and the
extracted non-yellows proved to be pure in the Fa generation.

TABLE 15.

F2 SEEDS FROM CROSS OF NO. 15 LONGFELLOW FLINT X NO. 8
WHITE DENT.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(15x8)-l

305

73

3 : 1

( " )-2

166

12

15 : 1

( " )-3

246

85

3 : 1

( " )-4

428

142

3 : 1

( " )-5

393

124

3 : 1

( " )-6

353

106

3 : 1

( " )-7

480

140

3 : 1

Table 15 gives the results from selfing the Fi seeds of a cross
between No. 15, Longfellow yellow flint and No. 8, white dent.
There was no appearance of Xenia in the Fi seeds. The Fj
seeds segregated in 3 : 1 ratios with the exception of ear (15x8)-
2. This ear was originally classified as bearing 128 yellow and
50 non-yellow seeds. The Fs seeds produced by the supposed
whites, however, showed the correct ratio to have been 166
yellow and 12 non-yellow. The whites proved true in three
other ears. The white seeds from ear (15 x 8)-l were not

54

INHERITANCE IN MAIZE.

grown, and therefore the large excess of yellow seeds cannot be
explained. It is possible of course that this ear as well as one
or two others that were not planted really had light yellows
classified as whites. If this were true one might consider that
the original mother plant was homozygous for one yellow and
heterozygous for the second. It seems not improbable that
this was the case, for the same results were obtained in two
other instances.

• TABLE 16.

F2 SEEDS FROM CROSS OF NO. 19 WHITE SWEET X NO. 7
YELLOW DENT.

Yellow Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(19 X 7)-2
( " )-5

277
599

77
43

3 : 1
15 : 1

One other cross. No. 19 non-yellow sweet and No. 7 yellow
dent (Table 16), gave di-hybrid ratios. The hybrid seeds were
yellow starchy varying somewhat in shade. Only two selfed
ears were obtained from the Fi seeds. As shown in Table 16
one is a 3 : 1 ratio and one is a 15 : 1 ratio. Here again is
evidence that the male parent was homozygous for one yellow
and heterozygous for the second yellow. To be sure there is a
slight excess of non-yellows in ear (19 x 7)-5, but this is accounted
for in the F3 generation. The supposed non-yellows gave one
heterozygous yellow to seventeen non-yellows. The true ratio

TABLE 16A.

Fj SEEDS OF EAR 5 OF CROSS SHOWN IN TABLE 16.

Dark Yellow Starchy Seeds Planted.

Ear No.

Y

y

Ratio Approx.

(19 X 7)-5-l
( " )-5-6
( " )-5-9
( « )-5-12
( « )-5-13

315
320
19
203
440

98
97
1
14
25

3 : 1
3 : 1

15 : 1
15 : 1
15 : 1

INHERITANCE OF YELLOW ENDOSPERM. 55

then is 601 : 41 which is very close to theoretical expectancy.
The results from planting the yellow starchy seeds of (19 x 7)-5
are shown in Table 16a. Unfortunately the admixture of
segregates with wrinkled endosperm made these a little difficult
to classify, but there is scarcely a doubt that 2 ears were mono-
hybrids and three ears di-hybrids, although no dependence
can be placed on ear (19 x 7)-5-9 with only 20 seeds. No pure
yellows were obtained from these seeds unless ear (19 x 7)-5-9
were of this class. The deficiency of these data was supplied
by the crop of the yellow sweet F2 seeds of the same ear.
Twelve selfed ears were obtained. They are not given in a
table because we were not able to prove the classification by
growing for another generation, and it is difficult to make
exact visible classifications of yellow and non-yellow sugar
seeds. There is scarcely any doubt however that two ears
were pure for both yellows (seeds all dark yellow), two pure
for light yellow, (seeds all light yellow) three heterozygous for
one yellow (seeds light yellow and white), one at least and
probably two heterozygous for two yellows (seeds dark yellow,
light yellow and white) and the rest homozyous for one yellow
and heterozygous for one yellow (seeds dark yellow and light
yellow) .

This family gave b y far the best demonstration of two yellows
as far as the eye is concerned. The ears homozygous for two
yellows would never have been classed as the same variety with
those homozygous for one yellow. Nearly all the seeds were
absolutely distinct, and yet when they were arranged in a series
there woidd always be a number that were difficult to place.

Table 17 gives the F2 segregates of a mono-hybrid cross be-
tween No. 10 white flour and No. 6 yellow dent. There seems
to be no question of a di-hybrid ratio, but the cross is interest-
ing for another reason. The heterozygous seeds are lighter
than the homozygous so that the effect of Xenia is shown
either way the cross is made; that is, Xenia is shown both
where white flour is crossed with yellow, and where yellow
Hour is crossed with white. The effect is the same as that
shown when light starchy caps are formed when a starchy
yellow dent is pollinated by a non-yellow, but as in this case
the whole seed is floury, therefore it is all changed to lighter
yellow.

56

INHERITANCE IN MAIZE.

It might be mentioned that No. 60 yellow pop crossed with
No. 2 white, and No. 9 yellow dent crossed with No. 10 flour
also show Xenia. The hybrid seeds become so much whiter
that there is no difficulty in distinguishing the greater part of
them from homozygous yellows.

TABLE 17.

F2 SEEDS FROM CROSS OF NO. 10 WHITE FLOUR AND NO. 6
YELLOW DENT.

Yellow Seeds Planted.

Ear No.

Dark Y

Light Y

Total Y

y

Ratio Approx.

(10x6)-l

( " )-2
( " )-3
( " )-4

162
141
175
131

357
242
301
243

519
383
476
374

187
119
156
127

3 : 1
3 : 1
3 : 1
3 : 1

Conclusions.

This completes the list of crosses in which new facts have
been observed in regard to yellow endosperm. Other crosses
might be described where simple mono-hybrid ratios were
obtained, but these have already been described by Lock. The
di-hybrid ratios have been described in greater detail because
they belong to a class of facts having a very important theo-
retical bearing on the Mendelian hypothesis, which is discussed
later in the paper.

It should perhaps be stated that Correns' other general facts
have been corroborated. The pure extracted dominants of
the Fs generation have appeared in about the general ratio of
1 homozygote to 2 heterozygotes when dealing with mono-
hybrids. There have been insufficient numbers to determine
the exact ratio of extracted dominants when dealing with
di-hybrids, but in both cases the F4 generations have in every
case bred true. This fact we hold to be more important than
the ratio. It may look somewhat queer to say that the extracted
F2 non-yellows have always bred true, when a number of cases
have been described in which the seeds that were thought to
be non-yellows, proved to be heterozygous yellows. This

PLATE V.

At left, No. 24 Rhode Island white cap (white endosperm), at right.
No. 15 Longfellow (j-ellow endosperm). In center, hybrid showing
dominance of vellow. Below. F2 seeds showing segregation.

b. An ear showing dominance of red pericarp in F^. The pericarp has been
removed from two rows of seeds, showing mono-hybrid segrega-
tion of F2 endosperms beneath it into yellow and non-yellow.

Segregation of Yellow and Non-Yellow Endosperm.

PLATE VI.

Cross 24x54. T. Ear ( 24x54 ) -12-5 : a ])uvv extracted purple. 2. Ear

(24x54)-i2-6 ; purples 208, non-purples 65, a 3:1 ratio. 3. Ear

(24x5b)-i2-4; purples 147, non-purples 117, a 9:7 ratio. 4. Ear

(24x54)-i2-3 ; a pure extracted non-purple.

b. Purple seeds produced by random crossing of non-purple seeds of ear
(24x54)-i2 shown in Table i8g. i. Ear (24x54) -12-9x12x8; ratio
I purple: 3 non-purple. 2. Ear (24x54) 12-11x12x10; ratio i pur-
ple : I non-purple.

Inheritance of Aleurone Color.

INHERITANCE OF ALEURONE COLOR. 57

is due to the simple fact that the seeds with the gametic formula
Yi yi or Y2 72 vary in color intensity so that it is generally
impossible to classify correctly from 1 to 5 seeds per ear. These
Fo seeds prove their gametic structure in the F3 generation;
and those that have behaved as pure extracted non-yellows
in F3 have never given anything but pure non-yellows in the
F4 generation.

The occurrence of the two yellow colors casts a further doubt
upon the correctness of Lock's work since his main object was
to show the truth of Mendel's mathematical conclusions when
dealing with large numbers. Our results both here and in the
case of the purple aleurone cells show the futility of not making
crosses between individuals and of not selfing individual Fi
plants. This is a further excuse for presenting in detail the
individual crosses between starchy and non-starchy races with
the same object as Lock.

Purple and Non-purple aleurone cells.

The consideration of the inheritance of this character includes
also that of a hypostatic red color which appears in crosses
between the various purple and non-purple families. The
pigments are both fairly easily soluble in water. They are
seen first in the aleurone cells of the maturing seeds a few days
after fertilization. When the seed is mature the red color
becomes an intense dark rose madder, and the purple becomes
almost black. Several tests of each pigment were made by
macerating the aleurone cells in 50% alcohol and testing the
filtrate. With lead subacetate both turned green and a green
precipitate separated. The precipitate from the red seeds was
somewhat darker and turned greenish brown on evaporation
while that from the purple seeds remained a lighter green.
Ferrous sulphate added to the red pigment produced but little
if any change in color although a dirty precipitate separated on
shaking. When added to the blue pigment, however, a dark blue
precipitate separated leaving the liquid colorless. This pre-
cipitate left a blue residue on evaporation, while the residue
from the red pigment was simply a slight discoloration, dark,
but with no distinct color except possibly a redness at the
edges. Ferric chloride however gave markedly different
reactions in the two cases. Added to the red pigment an orange

58 INHERITANCE IN MAIZE.

color was produced which became somewhat darker on evapora-
tion. The precipitation was sHght. Added to the blue pigment
the color was first greenish with blue edges. This turned dark
blue and a bluish precipitate separated which later turned
green and remained so on evaporation. With ferric alum there
was no change except that each pigment became more intense
in color. Sodium hydroxide formed brownish green precipi-
tates, darker with the purple. Acids gave red colorations
which were lighter with the red pigment. The acid and alkali
tests are evidently the usual reactions with vegetable color
"indicators" and differ only through the various amounts of
pigment present.

It is recognized that tests such as these are arbitrary in
nature and cannot form the basis of conclusions as to the chemi-
cal composition of the pigments. It seems certain however
that they differ somewhat in composition, although they are
probably different stagfes of oxidation of the same color base.

It will be seen in the following pages that purple crossed with
different strains of non-purple gave different results. This
is clearly due to the various gametic formulse possessed by the
different whites. It may also be that the purples differ some-
what among themselves in unseen characters even though
they were pure for purple when selfed. Our analysis of the
large amount of data which follows shows that there is simple
Mendelian segregation and recombination of several factors
and that there is really no confusion of results such as led
Correns and Lock to advance various supplementary hypotheses
to account for the facts. The use of the color factor C, shows
how Lock obtained his purples by crossing white seeds sup-
posedly heterozygous for purple, with white; but it is impos-
sible to analyze his data since individual pollinations were not
made. A supplementary hypothesis of Correns should also'
be mentioned because, if it were true it would necessitate a
very different conception of the interpretation of the inheritance
of all endosperm characters. Correns supposed that purple
X non-purple always gave purple while non-purple X purple
sometimes gave non-purple and sometimes gave purple. He
accounted for this by the supposition that since the endosperm
nucleus is formed by the union of two maternal nuclei with one
paternal nucleus, therefore the maternal endosperm characters

INHERITANCE OF ALEURONE COLOR. 59

would often dominate the paternal characters through the
effect of the greater amount of maternal nuclear material.
This is never the case and the fact is quite important. If Cor-
rens' supposition were true and the amount of nuclear matter
determined the characters to be formed, no Mendelian segre-
gation of endosperm characters and their recombination by
chance matings could be demonstrated. Since all of our data
shows it to be untrue, it follows that the quality and not the
quantity of nuclear material is the important thing. The
nucleus evidently regenerates or throws off material to come
to its proper adjustment for the performance of its functions,
and always in accordance with the quality of its structure.

In order to facilitate a consideration of the data, it will be
presented in families. Each family comprises the progeny
resulting from a particular cross. They are taken up in the
order of increasing complexity.

Family (24 and 54)

This family includes all of the progeny of the cross of No. 24
white flint with No. 54 Black Mexican sweet, this being the
variety with purple aleurone cells. The Black Mexican which
furnished the pollen for this cross had proved true to the purple
color for three generations, but pollen for the crosses of the
different hybrid families came from several different ears. For
this reason there is no certainty that the purple aleurone parent
had the same gametic structure in each family. The data for
the above family are reported in the sub-divisions of Tables
18 to 20. In these tables there is no correlation of the purple
and starchy characters, there being a simple 3 : 1 relationship
of starchy and non-starchy seeds in both the case where purples
and non-purples were obtained in F2 in the ratio of 3 : 1 and
where they are obtained in the ratio of 9 : 7. We may therefore
leave this character out of consideration and consider only the
purple character.

The Fi seeds formed in the hybrid ear were all purple. Upon
growing these seeds nine selfed ears were obtained with the
ratios of purples to non-purples shown in Table 18. The purple
color of these segregates was of full depth and covered the
entire seed with one or two exceptions. These exceptions
were zygotic variations due to heterozygosis and were quite

6o INHERITANCE IN MAIZE.

different from the partial or light purples obtained in other
families. In the latter case it was due to a transmissible gametic
factor which will be explained later. Table 18 shows the
ratio of purples to non-purples to be 3 : 1 in the case of seven
ears and 9 : 7 approximately in the case of two ears. This
immediately suggests the mono-hybrid ratio in the first case
and a di-hybrid ratio in the second case. That this is the true
state of affairs is shown by the behavior of the seeds of these
ears in later generations. The progeny of the purple seeds of
ear (24 x 54)-l (Table 18a) were either pure purples or heter-
ozygous purples segregating in the ratio of 3:1. The non-
purple seeds of the same ear (Table 18b) produced only non-
purples. The same ratio was obtained from purple seeds of
ear No. (24 x 54)-ll shown in Table 18c.

The fact that Fs extracted non-purple seeds continued to
breed true is shown by the results of the F4 generation shown
in Table 18d. Extracted purple starchy seeds were also planted
from Ear No. (24 x 54)-l-4 and ten selfed ears proved pure.
Twenty-six ears were also obtained from the open field crop
which were also pure purple, six being pure starchy and twenty
heterozygous starchy.

These continued 3 : 1 ratios with purity of the extracted
homozygote are what should be expected from the progeny of
the mono-hybrid ears of Table 18. If the 9 : 7 ratios given by
ears No. 9 and No. 12 of Table 18 are true di-hybrid ratios
resulting from the interaction of two factors both of which are
necessary for the production of the purple color, one should
expect in the Fs generation but one pure purple out of nine
to occur and the remaining ears to be about 50% monohybrids
with a 3 : 1 ratio and 50% di-hybrids with a 9 : 7 ratio. The
progeny of ear No. (24 x 54)-12 (Tables 18e, 18f) shows how
nearly these expectations are confirmed. Out of a total of
nineteen selfed ears two were pure purple, ten were mono-
hybrids and seven were di-hybrids. It must be concluded
therefore that the purple color is due to the action of the factor
P upon another color factor C, which is probably similar in
nature to that which Bateson found in sweet peas. The gametic
structure of No. 24, the non-purple variety, evidently differed
in the ovules of the seeds of the original hybrid ear. Part of
them lacked both P and C and gave a 9 : 7 ratio when crossed

INHERITANCE OF ALEURONE COLOR. 6i

with the purple (C P), and part of them contained either P or
C and therefore gave a mono-hybrid ratio when crossed with
C P. If one supposes C to be contained by the non-purple
in the first case then the result is as follows, Cp xCP = CCPp.
The gametes formed differ only in presence or absence of P and
a simple mono-hybrid ratio is obtained in the F2 generation.
In the second case the cross is cpxCP = CcPp, and the
F2 populations have the formulae and ratios 9CP:3C:3P:1
c p, the first nine being purple and the last seven being white.
This being the case the various non-purple seeds of F2 should
prove true non-purples when selfed but should sometimes
give purples when crossed. The non-purples exist in the
following ratios:

1

C

C

p

p

2

C

c

p

p

1

c

c

p

p

2

c

c

p

p

1

c

c

p

p

When crossed at random there are 7 x 6 = 42 possible combi-
nations of which 24 should give all non-purple and 18 some
purples. Of these eighteen ears 2 should be pure purples, 8
purples and non-purples in the ratio 1:1, and 8 purples and
non-purples in the ratio of 1 : 3. In Tables 18g and 18h besides
the selfed non-purples seven combinations of different non-
purples are shown, besides several reciprocal crosses. Of these
one combination and its reciprocal gives a 1 : 1 ratio and one
combination and its reciprocal gives a 1 : 3 ratio.

None of the F2 seeds of the selfed ears of this cross showed any
seeds with red aleurone cells. Among the open field ears
containing F-. seeds however, were noticed several seeds with
aleurone cells of a peculiar blue color and several of the red
color. Five selfed ears were obtained from the blue aleurone
seeds (Table 19). Four of these ears gave 9 colored (P and R)
seeds to 7 non-colored and one gave a simple mono-hybrid
ratio in which no reds were found. The red seeds varied in
shade until the darkest seemed to the eye to be purple. They
could be separated accurately only by a microscopic examina-
tion of sections of the aleurone cells. The purples (the blue
seeds proved to be exactly like ordinary purple seeds) occurred

62 INHERITANCE IN MAIZE.

in greater numbers than the reds but the exact ratios were not
determined in this family, because their parentage was not
certain.

The red seeds found in the open field ears also proved to be
heterozygous for red as shown by Table 20. They gave simple
3 : 1 ratios except ear No. 1 which proved to be pure red although
heterozygous for starchiness. Fa seeds were obtained from
the red seeds of ear No. 8 as is shown in Tables 20a and 20b.
It happened that in this small number five pure red ears were
obtained and only three ears that were heterozygous and seg-
regated in the ratio of 3 : 1.

Besides the ears shown in Table 20a, two ears from extracted
red seeds were crossed with pure extracted non-purples (whites)
of the Fs generation of cross (24x54). Ear No. 1 gave 125
purples and 123 non-purples. Ear No. 2 gave 108 purples
and 124 non-purples. The red ears, the maternal parents of
the crosses, were evidently heterozygous and therefore a 1 : 1
ratio was obtained. The non-purple which furnished the
pollen must have carried the P factor which oxidized the seeds
which otherwise would have become red to the purple color.
This fact proves the epistatic nature of P over R and is a further
proof of the di-hybrid nature of the purple color. Another
ear crossed with non-purples of the same family as above gave
all purple seeds. This ear evidently was homozygous for red
and all of its seeds were oxidized to purple. Two other of these
red ears were crossed with extracted purples of the same cross
from which came the extracted whites used above. The seeds
of the resulting ears were all purple. (See Plate 8a.)

Several red non-starchy seeds from ear (24 x R)-16-8 (Table
20b) were also planted. Three selfed ears resulted in two pure
for red and one giving 248 reds to 60 non-reds, a 3 : 1 ratio.
One ear of this lot was crossed with the same extracted purples
used in crossing the starchy^ red seeds resulting in an ear with
all purple seeds. Another ear was crossed with one of the
extracted non-purples used in crossing the red starchy seeds
and resulted in an ear with 119 purple starchy and 124 non-
purple starchy seeds. The results from the non-starchy seeds
of this family were therefore the same as those from the starchy
seeds.

The non-red seeds from (24 x R)-16-8 both starchy and non-

INHERITANCE OF ALEURONE COLOR.

63

starchy bred true to the non-red character. Four crosses
between individual ears of this lot were made and the resulting
seeds were all non-red. This is the result which should be
expected from an ear giving a mono-hybrid ratio as did ear
(24 X R)-8 and shows that the purples resulting from the crosses
between the non-purples coming from the 9 : 7 ratios were not
accidental.

TABLE 18.

F2 SEEDS FROM CROSS OF NO. 24 WHITE FLINT X NO. 54 PURPLE
ALEURONE NON-STARCHY.

Purple Aleurone Starchy {PS) Seeds Planted.

Ear No.

PS

Ps

pS

ps

Total P

Total p

Ratio
Approx.

(24

X 54)-l

207

67

67

27

274

94

3 : 1

" )-2

170

54

49

19

224

68

3

" )-6

197

65

59

24

262

83

3

(

" )-9

83

44

72

25

127

97

9

(

" )-10

166

40

46

19

206

65

3

(

" )-12

153

40

115

40

193

155

9

" )-8

159

41

41

23

200

64

3

" )-ll

166

55

47

22

221

84

3

*(

" )-13

205

81

59

25

286

84

3

* All purple seeds were full dark purples except a few splashed purples
from this ear.

TABLE 18A.

Fs SEEDS OF EAR 1 OF SAME CROSS AS TABLE 18.

Purple Aleurone Starchy {PS) Seeds Planted.

Ear No.

PS

Ps

pS

ps

Total P

Total p

Ratio
Approx.

(24 X 54)-l-3

144

Pure P

( " )-l-4

384

Pure P

( " )-l-5

96

PureP

( " )-l-ll

320

Pure P

(24 X 54)-l-2

161

55

46

13

216

59

3 : 1

( " )-l-6

171

56

52

19

227

71

3 : 1

( " )-l-8

180

71

55

19

251

74

3 : 1

( " )-l-9

79

29

27

7

108

34

3 : 1

( " )-l-10

255

91

255

91

3 : 1

( " )-l-14

195

80

195

80

3 : 1

Total

1251

410

INHERITANCE IN MAIZE.

TABLE 18B.

Ft SEEDS OF EAR 1 OF SAME CROSS AS TABLE 18.

Non-Purple Aleurone Starchy (ps) Seeds Planted.

Ear No.

P

P

Ratio Approx.

24 X 54)-l-4a

. " )-l-5a
" )-l-6a
" )-l-7a
" )-l-15a

208
312
362
320
296

Pure white

a
u
U
u

TABLE 18C.

F3 SEEDS OF EAR 11 (TABLE 18a) OF SAME CROSS AS TABLE 18.

Purple Aleurone Non-Starchy (Ps) Seeds Planted.

Ear No.

Ps

ps

Ratio Approx.

(24x54)-ll-2

312

PureP

( " )-ll-3

368

a

( " )-ll-4

280

li

( " )-ll-l

240

82

3 : 1

( " )-ll-5

197

78

3 : 1

( " )-ll-6

205

52

3 : 1

( " )-ll-ll

40

12

3 : 1

Total Het.

682

224

TABLE 18D.
r4 SEEDS OF EAR (24 X 54)-l-6 (extracted PS. TABLE 18b) of

SAME CROSS AS TABLE 18.

Non-Purple Starchy (ps) Seeds Planted.

Ear No.

ps

PS

(24 X 54)-l-6-l

All

( « )- 1-6-2

u

( " )-l-6-5

u

( " )-l-6-8

u

( " )-l-6-l x(24x54)-8-5 RS

All

( " )- 1-6-6 X ( " )-8-8 RS

a

( " )-l-6-12 X ( " )-8-3 RS

i<

( « )-l-6-9 X ( " )-8-10 RS

a

( " )-l-6-10 X ( " )-8-l RS

u

Open-field crop all white.

PLATE VII.

a. Flint and dent segregates from F2 of cross 8x54. Flint character car-
ried by No. 54.

f ^

r '

Z?. F3 types from cross 8x54. i. Pure extracted purple (PPCC). 2. Pure
extracted parti-colored ((PPcc). 3. Pure extracted non-purple
(ppCC or ppcc).

Inheritance of Aleurone Color.

INHERITANCE OF ALEURONE COLOR.

65

TABLE 18E.

F» SEEDS OF EAR (24 X 54)-12 OF SAME CROSS AS TABLE 18.

Purple Starchy {PS) Seeds Planted.

Ear No.

PS

P

SS or Ss

Ratio Approx.

(24 X 54)-12-l

280

Ss

Pure

( « )-12-2

147

40

Ss

3 : 1

( « )-12-3

190

60

Ss

3 : 1

( « )-12-4

147

117

Ss

9 : 7

( « )-12-5

288

Ss

Pure

( « )-12-6

208

65

Ss

3 : 1

( " )-12-7

188

115

Ss

9 : 7

( « )-12-8

237

72

SS

3 : 1

( " )-12-8i

212

72

Ss

3 : 1

( « )-12-9

159

120

SS

9 : 7

( « )-12-10

145

56

Ss

3 : 1

( " )-12-ll

95

30

Ss

3 : 1

( « )_i2-12

179

59

Ss

3 : 1

TABLE 18F.

Fj SEEDS OF EAR (24 X 54)-12 OF SAME CROSS AS TABLE 18.

Purple Non-Starchy (Ps) Seeds Planted.

Ear No.

P

P

Ratio Approx.

(24 X 54)-12-la

( « )-12-2a
( « )-12-3a
( « )-12-4a
( " )-12-6a
( " )-12-9a

160
186
137
97
109
123

53
64

115
65
80

120

3 : 1
3 : 1
9:7

9 : 7
9 : 7
9 : 7

TABLE 18G.

Fi SEEDS OF EAR (24 X 54)-12 OF SAME CROSS AS TABLE 18.

Non-purple Starchy (pS) Seeds Planted.

Ear No.

P p

Ratio Approx.

(24 X 54;

-12-3b

All

( " ^

-12-4b

«

( " "

-12-6b

((

C " '

-12-7b

u

( " ^

-12-12b

a

( " ^

-12-2b X 12-4b

u

( " ^

-12-4b X 12-2b

u

( " '

-12-5b X 12-lb

11

( "

-12-8b X 12-9b ]

L3 62

i : 3

( " ^

-12-9b X 12-8b i

n 226

1 : 3

( " '

-12-10bxl2-llb

19 86

1 : 1

( " '

-12-llb X 12-lOb <

)3 99

1 : 1

66

INHERITANCE IN MAIZE.

TABLE 18H.

Fj SEEDS OF EAR (24 X 54)-12 OF SAME CROSS AS TABLE 18.

Non-Purple Non-Starchy (ps) Seeds Planted.

Ear No.

P

P

(24 X 54)-12-2c

All

(■ « )-12-3c

u

( " )-12-5c

a

( « )-12-9c

u

( " )-12-10c

u

( " )-12-12c

u

( " )-12-13c

u

( " )-12-lc X 12-4C

u

,

( " )-12-4c X 12-lc

u

( " )-12-7c X 12-3c

u

( " )-12-3c X 12-7c

u

( " )-12-6c X 12-8c

a

( " )-12-8c X 12-6C

tt

TABLE 19.

Fa SEEDS FROM CROSS BETWEEN NO. 24, WHITE FLINT AND NO. XP
PURPLE ALEURONE.

Purple Aleurone Starchy {PS) Seeds Planted.

Ratio

Ear No.

P + R

P

Approx.

Notes

(24 X P)-16-2

287

192

9 : 7

SS: some seeds
red

( " )-16-5

141

117

9 : 7

ss: few P's
strongly colored

( « )-16-6

165

115

9 : 7

ss: few P's
strongly colored

( " )-16-7

278

89

3 : 1

_ Ss: 84 P's
lighter at cap

( " )-16-8

253

193

9 : 7

ss: 69 P's
lighter at cap

TABLE 20.

Fj SEEDS FROM CROSS BETWEEN NO. 24 WHITE FLINT AND NO. XR
RED ALEURONE.

Red Aleurone Starchy (RS) Seeds Planted.

Ear No.

RS

Rs

rS

rs

Total R

Total r

Ratio
Approx.

(24 X R)-16-l
( " )-16-4
( " )-16-6
( " )-16-8

160

26

140

195

52
12
43
73

is

53
41

3

22
19

212

38
183
268

ie

75
60

Pure red
3: 1
3 : 1
3 : 1

INHERITANCE OF ALEURONE COLOR. (
TABLE 20A.

Fs SEEDS OF EAR (24 X R)-16-8 OF SAME CROSS AS TABLE 20.

Red Aleurone Starchy {RS) Seeds Planted.

Ear No.

R

r

Ratio Approx.

(24 X R)-16-8-3
( " )-l 6-8-4
( " )-16-8-5
( " )-16-8-6

( " )-16-8-8

360
161
60
172
320

65
53

Pure red
3 : 1

Pure red
3 : 1

Pure red

TABLE 20B.

Fs SEEDS OF EAR (24 X R)-16-8 OF SAME CROSS AS TABLE 20.

Red Aleurone Non-Starchy (Rs) Seeds Planted.

Ear No.

R

r

Ratio Approx.

(24 X R)-16-8-la
( " )-16-8-2a
( " )-16-8-3a

160

248
280

60

Pure red

3 : 1
Pure red

Family (8 x 54)

The Fi Xenia seeds of the cross between No. 8 non-purple
dent starchy and No. 54 purple non-starchy were all purple in
color. Four selfed ears were obtained when these hybrid seeds
were planted. The segregation of the F2 seeds is shown in
Table 21. A new phenomenon of peculiar interest appeared
in this family. A certain number of seeds were solid dark
purple, others were splashed dark purple, others were a very
faint purple and have been called particolored, while still
others were without the purple color. The splashed dark
purples were seeds that had a break in the purple color ; that is
the purple color was dark but appeared in patches. These
splashed purples are found in all of the purple-non-purple
crosses except the family * just described. It seems evident
then that they are due to the interaction of characters which

Only one or two splashed purples were ever found in family (24 x 54).

68 INHERITANCE IN MAIZE.

happen to be absent from the (24 x 54) family ; but at the same
time they are zygotic variations which are not inherited, for
their progeny are exactly like the progeny of the dark purple
seeds. Further, these patches are not in a regular pattern
nor does the selection of seeds of this nature have the slightest
tendency to fix the phenomenon as a separate character. There
is reason for believing however that no homozygous purples
(C C P P) are ever of this nature, and that the splashing is
simply due to incomplete dominance, but caused by a factor
or factors brought in by the non-purple parent.

The fact that particolored or very light purples which trans-
mitted the character also appeared in this family made it seem
probable that a new character had appeared, making the family
a tri-hybrid. But this is not the simplest interpretation.
We have seen in the other family that the behavior of purple
is best interpreted as the interaction of two factors C and P.
In this family the hypothesis that either Cp or cP seeds are not
pure whites but very light purples is supported by all of the data.
At first sight it seems more reasonable that they should have
the formula Cp. If in accordance with older interpretations
of color inheritance, the purple color is formed by an enzyme,
P acting upon a chromogen C it is more reasonable to suppose
that in the presence of the chromogen an exceedingly small
amount of the enzyme might give rise to the particolored seeds,
than it is to believe that the normal amount of enzyme would
form the purple color with a trace of chromogen. The reason
for this statement rests upon the well known fact that enzymes
are organic catalysers and can accelerate reactions involving
quantities very disproportionate to their own amount. There
is an objection to this interpretation, however, for when parti-
colored seeds are crossed with those having red aleurone cells
and which therefore have the gametic formula R C, they invari-
ably give purples. This proves that the gametic formula of
the particolored seeds is c P and they are so designated in the
tables.

The suggestive work of Miss Wheldale (: 09, : 09a, : 10) in
correlating the results of biological chemistry with those of
genetics, has made it very probable that a basic chromogen is
present in all flowers which are able to form a sap color, and
that the complexities of color inheritance may be referred to

PLATE VIII.

Pure purple aleurone resulting from crossing pure extracted red
aleurone with pure purple. 2. Same result from crossing pure ex-
tracted red aleurone with colorless aleurone. 3. Seeds half purple
resulting from crossing heterozygous red aleurone with colorless
aleurone. 4. Result from selfing the male parent of 3.

I. (24x54)8-3 pure extracted red aleurone. 2. (24x54) -8-6 heterozy-
gous red aleurone. Cut does not show color value when compared
with Fig a.

Inheritance of Aleurone Color.

INHERITANCE OF ALEURONE COLOR. 69

the dual nature of the oxydases necessary for the formation
of the color compounds. It is quite likely that the color in the
aleurone cells of maize is similar in nature to flower color ; and,
as we fully agree with Miss Whedale's conclusions, none of our
factors C, R and P are to be regarded as chromogens. The
argument above is in agreement with this viewpoint. If one
wishes to denote a chromogen, the addition of an X to represent
it, common to both families, makes no difference in the inter-
pretation of the results.

If we are dealing with a di-hybrid ratio, one pure purple ear
out of every nine should be expected in the F2 generation.
Tables 21a and 21b show that one such ear was obtained out
of seventeen ears. If the total purple seeds and the sum of the
particolored and white seeds is considered in Tables 21, 21a
and 21b a close approximation to a 9 : 7 ratio is obtained. If
the particolored seeds could in every instance be distinguished
from whites the ratio of purples to particoloreds to whites
should be 9 : 3 : 4. It will be noticed however that in the ears
from which this ratio should be expected there is generally an
excess of whites. This is explained by the fact that parti-
coloreds especially when non-starchy are not always distin-
guishable from whites. The last two ears shown in Table 21d
are in fact ears grown from seeds which were originally classed
as whites. If this hypothesis in regard to the particolored is
true, one should expect the purple F2 seeds to give in the F3
generation, one ear pure purple, two ears showing segregates
of purple and particolored in the ratio of 3 : 1, two ears showing
segregates of purple and non-purple in the ratio of 3 : 1, and
four ears showing purples, particolored and non-purples in the
ratio 9:3:4. Among the ears received (Tables 21a, 21b)
there were one of the first class, six of the second class, three
of the third class and seven of the fourth class.

Tables 21c and 21d show the results from growing the parti-
colored seeds of the same ear. No. (8 x 54) -1. One ear should be
pure particolored to two showing segregates of particolored and
non-purple in the ratio of 3 : 1. Out of the fifteen ears obtained
three were evidently of the first class and twelve of the second
class.

In all of these tables the progeny of hybrid starchy seeds
segregated normally.

Seeds classified as non-purples were also planted from this

70 INHERITANCE IN MAIZE.

same ear No. (8 x 54) -1. The thirteen selfed ears resulting
as progeny of starchy seeds all proved to be non-purple. Two
particolored ears, however, appeared in the eight selfed ears
resulting from planting non-purple, non-starchy seeds. This
showed that there was more difficulty in classifying the non-
starchy non-purples than in classifying starchy non-purples.
Non-purple seeds planted from ear No. (8 x 54)-5 also gave a
few particolored progeny.

Four thousand seeds from tested whites of the Fs seeds were
planted in an isolated plot the next season and were allowed to
inter-cross naturally. If we were dealing with di-hybrid non-
purples in this case, such inter-crossing should give some purples,
such as were obtained in the (24 x 54) family. The resulting
crop of this large number of plants however were all true non-
purples, proving that we were dealing with non-purples with
formulas either CC, Cc or cc. Further proof of the constitution
of the particolored is shown in the following facts. No parti-
coloreds ever gave full purples. Furthermore, pure extracted
particoloreds (c c P P) from ear No. (8 x 54)-l-13b of Table 21c
were grown for another generation and their gametic structure
tested by various crosses. Several of these ears were selfed and
all proved to be pure particoloreds (ccPP). Three different
ears were crossed with pure extracted purples from progeny of
ear No. (24 x 54) -1-4. As would be expected all of the seeds
were purple. Two of the ears however had a decided reddish
purple color while one was dark purple without the reddish
tint. Four ears were crossed with extracted red seeds (RRCC)-
All produced purple seeds. Nine ears were crossed with plants
of the progeny of the non-purples of ear No. (24 x 54) -12. It
will be remembered that this ear gave a ratio of nine purples to
seven non-purples. The seven non-purples would have the
following formulse: 1 PPcc, 2 Ppcc, 1 ppCC, 2 ppCc, Ippcc.
Crossing the particoloreds at random with pollen of individual
plants of this lot should give on the average one ear with all
purple seeds when pollinated by ppCC, two ears with 50%
purple and 50% particolored when pollinated by ppCc, four
ears pure particolored when pollinated by PPcc, Ppcc or ppcc.
Nine ears were obtained of which one had all purple seeds,
three had 50% purple and 50% particolored with a total of
308 purple seeds to 294 particolored seeds and five were all

INHERITANCE OF ALEURONE COLOR.

n

particolored. It should be mentioned, however, that the
particolored seeds obtained by crosses with the whites of this
(24 X 54) family in which the (ccPP) seeds were not particolored,
gave seeds which averaged much lighter in appearance than
the pure particolored. In other words particoloreds crossed
with whites of other families show imperfect dominance of
particolored. Some gene common to both parents of the
(8 X 54) family, therefore, accounts for the production of the
color.

These two families differed in no other endosperm character
except presence and absence of starchiness. No correlation
of any kind was observed between these two allelomorphic
pairs.

TABLE 21*.

Fj SEEDS FROM CROSS BETWEEN NO. 8 NON-PURPLE DENT STARCHY
AND NO. 54 PURPLE NON-STARCHY.

Purple Seeds Planted.

Total

Total

Ear No.

CP

cP

Cp or cp

Purple

Non-Purple

(8 X 54)-l

297

75

146

297

221

( " )-2

230

75

172

230

247

( " )-3

302

239

( " )-5

270

229

Total

1099

936

* There were 1,514 starchy and 521 non-starchy seeds.

72

INHERITANCE IN MAIZE.
TABLE 21A*.

Fa SEEDS OF EAR NO. (8 X 54)-l OF TABLE 21.

Purple Starchy (PCS) Seeds Planted.

Total

Total

Starch-

Ear No.

CP

cP

Cp or cp

Purple

Non-Pur.

mess

(8 X 54)-l-l

233

70

233

70

SS

( " )-l-2

16

6

7

16

13

Ss

( " )-l-3

238

69

238

69

Ss

( " )-l-4

321

86

321

86

Ss

( " )-l-6

312

Ss

( " )-l-8

239

106

93

239

199

Ss

( " )-l-9

223

65

95

223

160

SS

( " )-l-12

285

66

285

66

SS

( " )-l-14

160

33

111

160

144

. Ss

( " )-l-20

126

54

126

54

SS

* There were 1,362 starchy and 435 non-starchy seeds in the Ss ears.

TABLE 21B.

F3 SEEDS OF EAR NO. (8 X 54)-l OF TABLE 21.

Purple Non-Starchy (PCs) Seeds Planted.

Ear No.

CP

cP

Cp or cp

Total
Purple

Total
Non-Purple

(8 X 54)-l-la

229

79

229

79

( " )-l-3a

236

44

116

236

160

( " )-l-4a

295

93

295

93

( " )-l-6a

260

86

260

86

( « )-l-7a

86

20

38

86

58

( " )-l-10a

239

89

239

89

( " )-l-lla

223

55

88

223

143

INHERITANCE OF ALEURONE COLOR.
TABLE 210*.

Fj SEEDS OF EAR NO. (8 X 54)-l OF TABLE 21.

Particolored Starchy (cPS) Seeds Planted.

73

Ear No.

cP

Cp or cp

Starchiness

(8 X 54)-l-2b

322

99

SS

( " )-l-5b

402

ss

( « )-l-6b

115

70

Ss

( « )-l-7b

150

64

SS

( « )-l-10b

386

SS

( " )-l-llb

254

92

Ss

( " )-l-13b

427

Ss

( " )-l-14b

262

112

SS

( " )-l-15b

256

133

Ss

* There were 1,026 starchy and 321 non-starchy seeds in the Ss ears.

TABLE 21D.

Fs SEEDS OF EAR NO. (8 X 54)-l OF TABLE 21.

Particolored Non-Starchy (,cPs) Seeds Planted.

Ear No.

cP

Cp or cp

(8 X 54)-l-2c
( " )-l-3c
( " )-l-4c
( « )-l-5c
( " )-l-lw
( « )-l-12w

149
364
168
123
230
131

110

'89
59

115
99

74 INHERITANCE IN MAIZE.

Family (15 x 54).

This family brings in a third allelomorphic pair namely
presence and absence of yellow in the endosperm. No. 15 is
Longfellow pure for the presence of starchiness and for a single
yellow factor. This cross was made to find out whether there
were further differences in the behavior of the purple factor
when crossed with other non-purples, and it was thought that
the yellow endosperm might prove a disturbing factor. This
is not the case for the Fi seeds were all purple with the exception
that a few splashed purples, which behaved like the normal
hybrid purples, also occurred in this family. Eight ears were
obtained by growing the Fi seeds, and starchiness and yellow-
ness were found to segregate in a normal manner. There was
a total of 1765 Y to 604 y and 1746 S to 623 s seeds.

There is only one fact of particular interest in this family.
Table 22 shows the eight selfed Fi ears grown from the purple-
starchy hybrid seeds, containing the F2 seeds. It will be
noticed that in the table, six of the ears appear to show mono-
hybrid segregation and two of them di-hybrid segregation.
This is not really the case. All of the ears giving the 3 : 1 ratio
were also di-hybrids. The figures in the column headed "Purple"
contained purples, splashed purples and particoloreds. Some
unknown cause produces many seeds in this cross that are
heavily splashed with purple. These always behaved as
heterozygous purples, although the heterozygous purples were
not always splashed, but were generally full colored purples.
The particoloreds are seeds containing the P factor but lacking
the C factor as in cross (8 x 54) . The difficulty here was to
distinguish by sight all of the splashed purples (P C) from the
particoloreds (Pc). They were all included in the table there-
fore as "Purples."

The ears (15 x 54) -2 and (15 x 54) -3 did not show parti-
colored seeds, but that the same gametes were concerned is
shown by the following data. Theoretically, if ear (15 x 54)-2
is a di-hybrid the purple seeds when selfed should give 1 ear
with all purple seeds, 4 ears with 3 purple seeds to 1 non-
purple seed and 4 ears with 9 purple seeds to 7 non-purple seeds.
Twelve selfed ears were obtained in the next generation. One
had all purple seeds. Seven had purple and non-purple seeds

INHERITANCE OF ALEURONE COLOR. 75

at the ratio of 9 : 7, there being a total of 1035 purple and 763
non-purple seeds. Four had purple and non-purple seeds at
the ratio of 3 : 1, there being a total of 480 purple to 162 non-
purple seeds. It should be remarked that in two of these ears
a few very light particolored seeds were found, showing that
the seemingly aberrant ear (15 x 54) -2 had a slight tendency
to throw particoloreds like the other ears of the family. There
is also some evidence that microscopical examination of the
embryo stem would show particoloreds in the ordinary ratio.

The non-purples from this ear were also grown. Eighteen
selfed ears were obtained. All of them were true to non-purple.
Two of them had a few particolored seeds (6 in one case and 14 in
another). These seeds might possibly have been produced by
the contamination of a few grains of foreign pollen, but they
might very well be white seeds of the formula Pc which were
showing the racial tendency to a slight production of pigment
(i. e. particoloreds).

Non-purples from the other aberrant ear No. (15 x 54) -3 were
also grown and when selfed gave only non-purples. Three inter-
crosses and their reciprocals were made between different plants
from this lot. It happened that no purple seeds were obtained
as should be expected in a portion of the cases, as explained
before. That the non-purples did differ in composition among
themselves was shown however by crossing a pure particolored
(PPcc) of the (8 X 54) family with pollen from one of our non-
purples, ear No. (15 x 54)-3-10. The ear resulting from the
cross had 179 purple seeds and 168 particolored seeds. This
1:1 ratio could only have been obtained from a non-purple
heterozygous for C (Cc). As a non-purple with the formula
Co could only have been obtained from a di-hybrid cross, it is
proved that all of the ears of this family were di-hybrids. The
complete gametic structure of the hybrid seeds, speaking of
endosperm characters only, is YySsCcPp.

76

INHERITANCE IN MAIZE.

TABLE 22.

Fi SEEDS FROM CROSS BETWEEN NO. 15 NON-PURPLE YELLOW
STARCHY AND NO. 54 PURPLE NON-STARCHY.

Purple Seeds Planted.

Ear No.

Purple*

Non-Purple

Ratio Approx,

(15 X 54)-l

138

45

9-f3 : 3 + 1

( " )-2

203

135

9 : 7

( " )-3

109

83

9 : 7

( " )-4

250

84

9-t-3 : 3-fl

( " )-6

201

61

9-1-3 : 3-hl

( " )-8

307

91

9+3 : 3 + 1

( " )-ll

254

93

9+3 : 3 + 1

( " )-15

239

76

9+3 : 3 + 1

* Every ear except ears 2 and 3 contained splashed purples ■which
act as heterozygous purples in inheritance and particoloreds which act
as if they had the gametic formula (cP), but the intergradations were so
gradual that it was impossible to make an accurate classification. The
matter is not worth mentioning here except for the reason that persons
who had not had experience with the behavior of purple and non-purple
crosses in other families would be utterly at loss for a classification and
it would be necessary for them to grow each individual seed for another
generation to determine its gametic formula.

Family (18 x 58)

No. 18, the female parent of this cross is a small non-purple
sugar maize which usually has twelve rows. The purple parent
is a small eight-rowed flint. P The Fi seeds were purple. Only
one selfed ear was obtained from the Fi plants through an
unfortunate loss of pollen. The segregation of F2 seeds is shown
in Table 23. The hybrid' 'seeds have the gametic formula Pp
RrCc'. The seeds with the formula PR and probably also
with the formula P give particoloreds or very light purples as
they did in family 8 x 54. They were very light however and
the 138 seeds classed as whites' or non-purples contained same
particoloreds as is shown in thelFs generation. Theoretically
in the F2 generation there should be 36 purples (27 PRC+9 PC)',
9 reds (RC), 12 particoloreds (9 PR +3 P) and 7 whites (3 C+3

PLATE IX.

'-¥f<**i(^

a. F3 color segregates from cross (18x58). i. Pure extracted purple.
2, 3. Ears from heterozygous plants. 4. Pure extracted hon-purple.

: — ^ -
^T r

b. F3 color segregates from cross (18x58). i. Pure extracted red.
2, 3. Ears from heterozj'gous plants. 4. Pure extracted non-red.
Proper color values are not shown.

Inheritance of Aleurone Color.

INHERITANCE OF ALEURONE COLOR. n

R+1 pre). There is an excess of whites because the parti-
coloreds could not be classified easily, so it might be said that
there should be 36 purples: 9 reds: 19 particoloreds and whites.
In the one ear obtained there is still an excess of the last class,
but the behavior of the seeds in the Fs generation proves the
gametic constitution of the parents. Tables 23a and 23b give
the results from planting purple F2 seeds. The last four ears
shown in Table 23a were planted from splashed purples but
they gave the same result as the full purples. We may con-
clude therefore that splashed purples behave the same as self-
colored purples in inheritance. Theoretically the entire lot
of purples should have the following gametic constitutions and
proportions :

Class 1. 1 P P R R C C = Pure purple.

" 2. 2PpRRCC= 3 purple: 1 red.

" 3. 2PPRrCC= Pure purple.

" 4. 2PPRRCc= 3 purple: 1 white.

" 5. 4 P p R r C C = 12 purple: 3 red: 1 white.

" 6. 4PpRRCc= 9 purple: 3 red: 4 white. 3

being particolored.

" 7. 4 P P R r C c = 12 purples: 4 white. 3 being part-
icolored.

" 8. 8 P p R r C c = 36 purples: 9 reds: 19 whites. 12

being particolored.

« 9. 1 P P C C = Pure purple.

" 10. 2 P p C C\ o 1 1 u-^

" 11 2PPCcJ ^ purples: 1 white.

" 12. 4PpCc =9 purples: 7 whites and parti-

coloreds.

These ears when selfed should give the proportions shown at
the right of the above column. An examination of Tables 23a
and 23b show that out of the 23 selfed ears obtained the expected
ratios were followed rather well. There were two pure purple
ears, Classes 1, 3 and 9; 2 ears of Class 2; 3 ears of Class 12; 4
ears of Classes 4, 7, 10 and 11 which collectively give 3 purples:
1 white; 3 ears of Class 8; 9 ears of Classes 5 and 6. The parti-
colored seeds are very light in color and although they are
classified as nearly as possible in the tables this classification
should be considered only an approximation and not a reality.

78

INHERITANCE IN MAIZE.

Particoloreds and whites are considered together in determining
the gametic constitution of the ears. The particolored seeds
proved to be true particoloreds of the same nature as those of
family (8 x 54) . The selfed ears resulting from such seeds of
ear No. (18 x 58) -1 of Table 23, gave no purples. Pure parti-
coloreds ears and heterozygous particolored ears were obtained
but no exact visual classification of the latter could be made
and it was not considered worth while to determine their precise
constitution by breeding.

The red segregates occurring in ear No. (18 x 58) -1 were also
tested in the F3 generation. Fifteen selfed ears were obtained
and are shown in Table 23c. Among them were five pure red
ears, six which threw reds and whites in the ratio of 9 : 7 and
four which threw reds and whites in the ratio 3:1. The
number of pure red ears obtained was slightly greater than
should generally be expected but such a deviation should occur
about once out of five times when dealing with lots of only
fifteen ears. The selfed white segregates of ear No. (18 x 58)-l
of Table 23 yielded about one particolored ear either homo-
zygous or heterozygous out of every four. This shows the error
in trying to classify particolored and white seeds. There is no
doubt however that when pure white segregates are planted
they always breed true.

TABLE 23.

Fa SEEDS OF CROSS BETWEEN NO. 18 NON-PURPLE NON-STARCHY
AND NO. 58 PURPLE STARCHY.

Ear No.

Purple
PCR + PC

Red
RC

Particolored
PR-fP

Non-Purple
+ some P

(18x58)-!

191

56

42

138

INHERITANCE OF ALEURONE COLOR.

79

TABLE 23A.

F3 SEEDS OF EAR (18 X 58)-l OF TABLE 23.

Purple Starchy Seeds Planted.

Ear No.

Purple

Red

* Parti-
colored

Non-Purple

(18 X 58)-l-l

167

49

84

( " )-l-2

18

4

( " )-l-3

41

13

( " )-l-6

211

72

92

( " )-l-8

221

66

96

( « )-l-8a

138

65

83

( " )-l-ll

80

( " )-l-12

66

i7

i2

17

( " )-l-ls

240

78

( " )-l-2s

141

48

72

65

( " )-l-3s

121

38

113

( « )-l-4s

93

21

15

60

s Planted splashed purples.

Particolored classification is only approximate.

TABLE 23B.

F3 SEEDS OF EAR (18 X 58)-l OF TABLE 23.

Purple Non-Starchy Seeds Planted.

Ear No.

Purple

Red

* Parti-
colored

Non-Purple

(18 X 5S)-l-2

49

25

26

( " )-l-3

68

20

5

66

( " )-l-4

183

61

73

( " )-l-5

240

61

( " )-l-6

22

7

8

( " )-l-7

207

147

( " )-l-8

184

24

140

( " )-l-9

360

( " )-l-10

186

68

( " )-l-ll

99

22

41

( " )-l-12

84

34

* Particolored classification is only approximate.

8o

INHERITANCE IN MAIZE.

TABLE 23C.

F, SEEDS OF EAR (18 X 58)-l OF TABLE 23.

Red Starchy Seeds Planted.

Ear No.

Purple

Red

Non-Purple

Ratio
Approx.

(18 X 58)-l-l

222

162

9 : 7

( " )-l-2

350

Pure

( " )-l-3

222

'so

3 : 1

( " )-l-5

212

171

9 : 7

( " )-l-6

195

115

9 : 7

( " )-l-7

148

74

3 : 1

( " )-l-12 !

187

102

9 : 7

( « )-l-2s

300

Pure

( " )-l-3s

350

Pure

( « )-l-4s

276

Pure

( " )-l-6s

209

'63

3 : 1

( " )-l-7s

44

35

9 : 7

( " )-l-9s

237

141

9 : 7

( « )-l-10s

361

Pure

( " )-l-lls

206

"ei

3:1

s Red sugar (s) seeds planted.

Family (7 x 54)

This cross introduces a combination of yellow endosperm
and a dent character, the Learning parent having long dented
yellow seeds usually with formulae Y1Y1Y2Y2. The current
efifect of the cross gave purple seeds some of which were splashed
as in the previous case where yellow flint was the non-purple
parent. There was nothing of special interest in the F2 gene-
ration as the ears segregated purple and non-purple seeds in
di-hybrid ratios without the appearance of particolored (cP)
seeds. The characters in which these parent varieties differed
segregated absolutely independently of each other.

Family (17 x 54)

The yellow flint which is the non-purple parent in this family
is similar to No. 15. The ear is shorter, however, and has
present a red pericarp color described under pericarp color as R4.
The Fi seeds were purple. They were sometimes splashed
purples but more rarely than in the other crosses. The F2 seeds
gave a simple mono-hybrid ratio but they were not followed
into further generations. The plant of No. 17 used as the
parent was homozygous therefore 'for either C. or P.

PLATE X.

No. 54. Black Mexican sugar and No. 60 Tom Thumb pop above. Be-
low F J ears with F2 seeds. At left ear from the famity without
factor inhibiting the formation of color in aleurone cells. Other
ears contain this inhibiting factor (heterozygous in mother plants).

Inheritance of Aleurone Color.

INHERITANCE OE ALEURONE COLOR. 8i

Family (19 x 54)

No. 19 the female parent of this cross is a large sugar corn
comparable in size with the large dent varieties. The Fi seeds
were deep purple and the F2 seeds segregated in ratios exceed-
ingly close to the theoretic number for mono-hybrids.

Family (60 x 54)

No. 60 is a dwarf pop maize with a yellow endosperm, known
as Tom Thumb. The individuals used as parents in the various
crosses were pure Tom Thumbs but it is not certain that they
were the product of a single selfed ear. They were grown from
an ear which was self -pollinated, but because the silks appear
in this variety while the young ear is entirely hidden in the
axil of the leaf, it is less certain that foreign pollen was excluded
from the bagged ears than it is in the case of our other crosses.
The bags were slipped down into the leaf axil as firmly as pos-
sible but there was still some chance for cross pollination.
This chance existed only among plants of the same variety,
however, for no other pollen was mature at the same time.
As several of these crosses were made upon different plants
of variety No. 60 it is not strange therefore that one or two
of the crosses acted as if parents with different gametic formulae
had been used. It does not follow that this was the case,
for the Tom Thumb or the Black Mexican or both might have
been heterozygous in some non-visible character.

The result of the immediate cross was different from our other
crosses in which the purple aleurone cells were concerned; some
of the seeds were dark purple, some were varying shades of
light purple and some were white (i. e. non-purple). The
behavior of the purple and non-purple hybrid seeds in the next
generation showed conclusively that we were dealing neither
with a reversal of dominance nor with a character in which the
female gametes segregated normally and the male gametes
abnormally as suggested by Correns, but with an entirely new
dominant factor in which the Tom Thumb variety was probably
heterozygous. This factor we take to be an actual inhibiting
factor similar in action to the dominant white found in poultry.
It is also analogous to the latter in that it does not always
completely inhibit the development of color, in which case

82 INHERITANCE IN MAIZE.

light purples similar in appearance to the particoloreds of
earlier crosses develop. They are not like the particoloreds
of family (8 x 54), however, for in the cross under consideration
seeds with the gametic formula cP do not develop color. The
light purples behave as if the inhibiting factor could vary
zygotically so that in some cases light purples are developed
while in others the color is completely inhibited, and also as if
various amounts of color are developed in the presence of the
inhibiting factor due to different combinations of other gametes.
For example, it seems probable that more color may be developed
in the presence of the inhibiting factor when the zygote is homo-
zygous for the purple factor than when it is heterozygous.
Further it seems less likely that any purple color develops when
the inhibiting factor is homozygous than when it is hetero-
zygous. This makes the segregating seeds of Fi or F2 ears
very difficult to classify visually. The only accurate determi-
nation of the gametic structure of a seed is through its own
further breeding.

Fifteen ears of Tom Thumb were crossed with the purple
sugar corn, but only five crosses were selected from which to
grow the Fo generation. One of these, No. 60-5 x 54, had dark
purple and non-purple seeds, while the other four crosses had
only non-purple or very light purple seeds. It was a little un-
fortunate perhaps that this selection was made. The white seeds
in cross 60-5 x 54 proved to be selfed Tom Thumbs, and the be-
havior of the purple hybrids showed that no inhibiting factor had
been present in ear 60-5. The behavior of the crosses made on
ears 60-2, 60-3, 60-8 and 60-11 showed that they had been
homozygous in the inhibiting factor. No doubt a number of
the other crosses would have shown that the maternal plants
were heterozygous in the inhibiting factor.

For these reasons the data from cross 60-5 x 54, which may be
called the purple family (without the inhibiting factor), have
been listed in Table 24, while the other four crosses containing
the inhibiting factor are shown in Table 25.

The resulting F2 seeds obtained by selfing the purple Fi seeds
of cross 60-5 x 54 shown in Table 24 were purples, reds and nqp.-
purples. No light purples (particoloreds) appeared in this
family. Splashed purples occurred as in other families, but
as in other families all splashed purples were heterozygotes

INHERITANCE OF ALEURONE COLOR. 83

and not all heterozygotes were splashed purples, showing the
phenomenon to be due — as before — to incomplete dominance
caused by other factors. The only peculiar thing about this
cross was the appearance and inheritance of the red color.
The extracted reds bred true apparently and were hypostatic
to purple as in other families, but they were purplish red (dark
magenta) in appearance and not clear reds such as appear in
other crosses. It is conceivable that this red is not the same
red that appeared in the other crosses. It may be caused by
something similar to what Wheldale (: 10) has suggested may
occur in stocks; viz. that the blue oxygenase may act in con-
junction with the red peroxydase or vice versa. The only
difficulty in alining the results obtained with the ordinary
behavior of the known factors, is the fact that almost none of
the ears of the F3 generation show the same ratios as the F2
generation.

The ratio obtained in this generation, 1843 purples: 188 reds:
545 non-purples immediately suggests 12 : 1 : 3, which could
be obtained from Fi seeds with a formula Pp Xx CC RR where
X is an inhibiting factor from the Tom Thumb which affects
R but not P. Our failure to obtain ears in F3 with segregates
of 9 purple; 3 red; 1 non-purple caused this hypothesis to be
discarded. The same ratio could be obtained by supposing
that there is a partial gametic coupling between P and R similar
to that obtained by Bateson and Punnet (: 08) between purple
color and long pollen in the sweet pea. These authors suppose
gametes to be produced after the general formula 7 AB : 1 Ab :
1 aB: 7 ab, from which result zygotes 3n2-(2n-l) AB : 2n-l
Ab: 2n-l aB: n2-(2n-l)ab. Such an interpretation, while it
may represent Bateson's and Punnett's facts, throws no light
on the mechanics of heredity for there is no reasonable way
known at present for such a segregation to come about. In
our own case * no such excess of purples was obtained in the
Fs generation. It seems better therefore to consider the results
of the F2 generation in the light of the breeding records of the
F3 generation. If this is done, the following interpretation
fits the facts best. Tom Thumb, the female parent has the
gametic formula pcR, and Black Mexican the male parent has

* Bateson and Punnett have never reported their F3 generation of
sweet peas, although they state that it gave conflicting results.

84 INHERITANCE IN MAIZE.

the formula PCR. The Fi generation is therefore PpCcRR.
If this is the true formula there is the following theoretical
expectation in the F2 and F3 generations :

F2 gives in F3

9
Purple

fl P P C C R R
•2 P p C C R R
!2P PCc RR
[4 P p C C R R

Pure Purple.
3 Purple : 1 Red.
3 Purple : 1 Non-purple.
9 Purple : 3 Red : 4 Non-purple

3

Red

/I CCRR
\2 Cc RR

Pure Red.
3 Red : 1 Non-purple.

4

Non-purple

fl PPRR
2 P p RR

U P P c c R R

Pure Non-purple.
Pure Non-purple.
Pure Non-purple.

An examination of the F3 segregates given in Tables 24a-e
show how nearly the experimental results accord with the
theory. First notice that out of 55 ears obtained by selfing
purples of the F2 generation, 20 segregated purples and non-
purples without reds. This is more than our own theory calls
for (theoreticall}^ 12 out of 55), so that here is clear evidence
that we do not have to deal with partial gametic coupling of
the kind described by Bateson and Punnett. But, since the
deficiency of reds in the Fo generation is too great to be due
to chance and since there is a certain excess of purples in the F3
generation, we must say frankly that we are dealing with some-
thing that we cannot yet explain.

The entire data from the 55 purple F2 seeds from which selfed
ears were obtained may be classified as follows: 8 ears pure
purple; 20 ears segregating purples and non-purples in the
ratio of 3 : 1 ; 9 ears segregating purples and reds in the ratio of
3:1; 1 ear each segregating purples and reds in ratios of 5 : 1,
6 : 1 and 12 : 1; 11 ears segregating purples, reds and non-
purples in the ratio of 9 : 3 : 4; 3 ears segregating purples,
reds and non-purples in the ratio of 48 : 3 : 13 or thereabouts.

From the red Fo seeds 13 selfed ears were obtained. Out of
these, 3 were pure red and 10 segregated reds and non-reds in
about the ratio of 3 : 1. It- should be remarked, however, that
in three cases the heterozygous reds gave a greater excess of
reds than usually should be expected with chance mating.

INHERITANCE OF ALEURONE COLOR. 85

From the non-purple F2 seeds 16 selfed ears were obtained: 15
were pure non-purple while one gave 12 purples to 49 non-purples.
As this is a poor ear, the 12 seeds may be due to foreign pollen,
or a chance pollen grain possessing the inhibiting factor may
have produced the F2 white seed from which the ear resulted.

With the plants resulting from non-purple F2 seeds random
intercrosses were also made. Of these 13 gave ears with all
non-purple seeds and one gave an ear with 49 purples and reds
and 140 non-purples.

These results generally follow our theory pretty closely,
but there are abnormalities difficult to explain. We seem to be
dealing with only two heterozygous factors — since 8 pure
purples are obtained from 55 ears — yet tri-hybrids and tetra-
hybrids are possible which give such results. By our theory
no whites should give purples when crossed at random. One
such ear occurred. Was it an error? It is difficult to say.
But if we were dealing with heterozygous red (Rr) we should
expect more than one ear out of 14 to give purples on random
crossing. Furthermore, it can be seen by inspection that
there are many reasons why we cannot be dealing with simply
a heterozygous red factor. It is not denied however that
several other unknown factors with a heterozygous red factor
might interpret the facts. It does not seem possible to explain
the results by any reasonable system of gametic coupling or
by selective mating. P and C certainly are present in an heter-
ozygous condition. R is probably homozygous although it was
not found in the Black Mexican in other crosses. But this is not
peculiar since the Black Mexican used in the cross can only be
said to be pure for purple. On the other hand, the red does not
appear to the eye to be exactly the same red which appeared in
the other crosses. It is more purplish in color, as if it were a
modified purple. Nevertheless it always bred true after ex-
traction.

Let us now turn to what may be called the white side of this
family. As was stated before ears 60-2, 60-3, 60-8 and 60-11
gave no dark purples when crossed with No. 54. (Some seeds
were afterwards found to be very slightly purple.) One may
conclude therefore that they (the maternal plants) were either
homozygous for a factor that inhibits the development of the
purple color; or, that there is a reversal of dominance, which is

86 INHERITANCE IN MAIZE.

improbable. There were other ears that gave both purple and
non-purple seeds in crosses. These were either heterozygous
for an inhibiting factor or exhibited dominance of both purple
and non-purple on the same ear which is still more improbable.
None of these ears were followed into the F2 generation, but
progeny of all four of the ears of the first type were grown.

The results of the F2 generation from these ears are shown in
Table 25. There is no reason why some of these families might
not differ from others in invisible factors, for different plants
of No. 60 were crossed with pollen from different plants of No. 54.
They are placed in one table here but certain differences in their
behavior in F3 leads us to consider them separately. There
is a total of 662 purples, 94 reds and 2838 light colored purples
and non-purples. The reason for classing the light purples
and non-purples together will be seen later.

The results of the F3 generation as well as our experience with
other crosses are such as exclude the possibility of a reversal
of dominance. The purples did not breed true nor did the
behavior of any of the classes indicate anything other than a
normal Mendelian segregation involving several characters.
Furthermore, a belief in reversal of dominance in our opinion
strikes at the foundation stone of Mendelism. Not that
dominance is an important part of Mendelism. It is not. Yet
no analysis can be made of breeding records without following
every individual for several generations if dominance is reversible.
Of the thousands of extracted recessives that have bred true,
many would have proved to be heterozygous dominants if
dominance is reversible.

Taking the same Fi gametic formula that served for the
purple side of the family and adding an inhibiting factor I which
comes from No. 60, gives the best interpretation of the data.
This makes the Fi formula PpCcIiRR. In F2 the following
classes would be expected :

Color non-purple.

" non-purple.

" purple.

" non-purple.

" non-purple.

" non-purple.

« red.

" non-purple.

J7 P I C R

9PI R

9PCR

91 CR

3PR

31 R

3CR

1 p c i r

INHERITANCE OF ALEURONE COLOR. Sy

The ratio is 9 purple : 3 red : 52 non-purple. The experimental
results given in Table 25 show that here also there is a deficiency
of reds. Many light purples also appeared, but these were
classed as non-purples. This was done because in Fs the light
purples behaved as if they possessed the factor I in a heter-
ozygous condition, the variation in color being due to the dif-
ferent combinations in which the factors P and C appeared.

With this theory the expectation in Fa is 52 non-purples and
light purples giving :

28 Producing all Non-purple seeds.

2 " 1 Purple : 3 Non-purple.

4 " 3 Purple : 13 Non-purple.

4 " 3 Purple : 1 Red : 12 Non-purple.

8 " 9 Purple : 3 Red : 52 Non-purple.

2 " 1 Red : 3 Non-red.

4 " 3 Red : 13 Non-red.

9 Purples giving :

1 Producing all Purple seeds.

2 " 3 Purple : 1 Red.

2 " 3 Purple : 1 Non-purple.

4 " 9 Purple : 3 Red : 4 Non-purple.

3 Reds giving :

1 Producing all Red seeds.

2 " 3 Reds : 1 Non-red.

Let us now examine Tables 25a-e which give the results from selfing
the seeds of certain of the F2 ears. Table 25a shows the progeny
of ear (60-3x54)-l. This ear gave the smallest proportion
of purple seeds in F2, and such purples as were produced in F2
were lighter in color than normal full purples. In F3 the purples
are again light in color. They are classed in with the non-
purples in the last column, those showing some color being
given first. The first two ears are progeny of the darkest
purples; one has purple and non-purple seeds in the ratio of
3 : 1 and one is pure purple. The next two ears planted from
lighter purples show a difference between themselves. One
gives 3 light purples : 1 non-purple, the other gives 1 purple : 2
non-purple. The latter probably came from an ear heter-
ozygous for the inhibiting factor and the former from a real

88 INHERITANCE IN MAIZE.

purple. Of those ears resulting from white seeds, one has 33
red seeds dark enough to be classed as real reds and a number of
very light reds classed with the non-purples — a total of prob-
ably near 25% reds, while another gives light purples (and
possibly light reds) and non-purples. The remaining ears are
non-purples. Five plants from non-purple F2 seeds were also
crossed, and gave all non-purple seeds. In reality, however,
only two random crosses can be said to have been made, since
the pollen of No. (60-3 x 54)-l-2 ES was used three times, while
once the same plant was used as the mother. The progeny of
ear (60-3 x 54)-l, therefore, behave like those of other ears of
this family except that all of the progeny of purples are light in
color. They give pure purples and purples segregating into
3 purples : 1 non-purple, but none are dark like normal purple
ears. Some geneticists would probably interpret this as pre-
potency of the non-purple or rather lack of prepotency of the
purple. But when one talks of prepotency he really confesses
ignorance of the gametic constitution of his cultures. Is it
not much more likely that the true reason for the production
of these light purples lies in a fact more in keeping with what
is known of hereditary phenomena? May not one say that
here is a dominant purple character coming from the individual
of unknown character of variety No. 54 which was used as the
male parent ? If the purple gene from the male parent was such
as to give always a lighter purple in zygotic combinations where
purple is visible then no dark purples would occur in the segre-
gates resulting from crosees. Such results were obtained from
four selfed plants. Two ears resulted from planting purples
which were only slightly lighter than normal dark purples, such
as the parents of ears (60-3 x 54)-l-l and (60-3 x 54)-l-2, and
two ears resulted from planting seeds quite light in color. (Table
25a).

Similar results were obtained from cross (60-8 x 54) , of which
the Fs generation from ear (60-8 x 54) -8 are shown in Table 25e.
Here eleven ears resulted from selfing seeds with the modified
color if two red seeds are included. None of these ears had
seeds dark in color, but the ratios are no doubt the same as those
given in Tables 25 b-d. The general reduction of the amount
of purple color, however, makes the error of classification too
great for safe conclusions. There is even some doubt about

INHERITANCE OF ALEURONE COLOR. 89

the classification of the seeds from the F2 generation of these
two crosses (Table 25), but the results of the F3 generation are
such as to give us considerable faith in them.

Tables 25 b-d give a considerable number of F3 progeny from
F2 seeds of three other ears. There seems to be no reason why
they should not be considered together. From the purple F2
seeds planted, twelve selfed ears were obtained. Three ears
were pure purples of the normal shade. One ear gave a ratio
of 3 purples : 1 red and two ears a ratio of 3 purples : 1 non-
purple. The other six ears gave purple, red and non-purple
segregates. Four of these ears were clearly of the ratio 9 : 3 : 4,
but in the remaining two there was a considerable deficiency
of red seeds. From the F2 red seeds planted, only one selfed
ear was obtained. This ear gave red and non-red segregates in
the ratio of 3 : 1.

A large number of selfed ears were obtained from the F2 light
purple and non-purple seeds. Ears of each of the classes
expected by the proposed theory were obtained, as will be seen
by an examination of the Tables 25 b-d ; but as the visual classi-
fication is arbitrary owing to the light color of most of the seeds,
it could not be depended upon without further breeding. The
light colored seeds are given first in the last column of the
tables, followed by the seeds which were apparently non-purple.
If one is a little charitable about the exactness of the classifi-
cation the following conclusions can be drawn.

Both seeds which were apparently non-purple and seeds which
were light purple in color in the F2 generation gave light purple
seeds among the F3 segregates. This fact proves both the impos-
sibility of exact classification and the gametic identity of seeds
slightly different in their appearance.

Two plants from light red seeds (Table 25b) were selfed.
One resultant ear showed a ratio of 1 dark red : 3 light red and
non-red; the other ear showed only light red and non-red seeds
which were classed together. Thirty-six plants from light purple
and from non-purple F2 seeds were selfed. Of these, fifteen
ears resulted from planting seeds classified as non-purple in F2.
Only four of them threw dark purple segregates in F2. On the
other hand only two of the ears resulting from selfed plants
which were progeny of seeds classified as light purples, threw
no dark purple segregates. It seems to us that this shows a

90 INHERITANCE IN MAIZE.

fair classification of seeds heterozygous for the inhibiting factor'
A few seeds, however, were wrongly classified in the F2 genera-
tion and proved their proper status in the F3 generation.

Out of the total of 36 selfed ears from light purple and non-
purple F2 seeds, 23 threw dark purple segregates and 13 produced
only light purple and non-purple seeds. Of those ears which
threw purple segregates, none of them had ratios of purple to
light purple plus non-purple greater than might reasonably be
expected by chance mating. The different ratios expected in
Fs were followed rather well, although it is recognized that these
ratios could not be determined accurately with such small
numbers.

Conclusions.

There can be but little doubt that the factors I, C, P and R
are concerned in this cross. Whether there is another factor
which modifies the purple color or not, is a question that cannot
yet be settled, because we have no data concerning the indivi-
dual plant of No. 54 that formed the male parent; yet there
seems to be no other way to account for the light purples in
Table 25a and Table 25e. The ultimate analysis of the behav-
ior of the R factor in this cross must also be left in abeyance.
These unsettled questions however have no bearing upon two
important conclusions which the evidence forces upon us. The
first is that one should be exceedingly careful before he decides
that the transmission of certain characters is an exception to
the general law of Mendel. When a collection of white or non-
purple aleurone strains are promiscuously crossed with a purple
aleurone maize, the results seem almost impossible to bring
into conformity with simple Mendelian results, yet this con-
fusion is brought about simply by the gametic differences of
the non-purple races. If such confusion can result in the
case of a simple color inheritance, much more care must be
taken to analyze the transmission of more complex characters
before subsidiary hypotheses are submitted.

The second important fact is in regard to prepotency. It
has been shown that certain families of purple and non-purple
hybrids produce very light purples when P exists alone without
C, while other families produce no color. No modification

INHERITANCE OF ALEURONE COLOR. 91 j

of the Mendelian ratio occurs, yet some transmissible difference j

in the two families gives this different result. Here is a probable |

explanation of prepotency. If these white families were mixed 'i

together, a mixture more easily imagined in the case of bisexual |

individuals, there would appear to be a difference in prepotency \

of the purple character. It -therefore seems probable that i

prepotency is due only to a difference in gametic character ;

which modifies somatic appearances and not to an actual modi- 1

fication of Mendelian chance ratios as others have suggested. \

The behavior of the other families is so simple that we think l

there can now be no question but that the purple aleurone color ■

behaves as a normal Mendelian character in inheritance. 1

92

INHERITANCE IN MAIZE.

TABLE 24.

F2 SEEDS OF CROSS BETWEEN NO. 60-5 NON-PURPLE POP AND
NO. 54 PURPLE SWEET.

Purple Seeds Planted.

Ear No.

Purple

Red

Non-Purple

(60-5 X 54)-2
( " )-3
( " )-4
( " )-5
( " )-6
( " )-8

( " )-io

( " )-ll
( " )-12

271
236
92
203
272
144
198
190
237

28
21
11
36
33
14
21
4
20

57
71
33
69
71
58
55
55
76

Total

1843

188

545

TABLE 24A.

F3 SEEDS OF EAR NO. (60-5 X 54)-2 OF TABLE 24.

E

ar No.

Planted from

Purple

Red

Non-
Purple

(60-5 X 54

-2-1

Purple S

277

23

/ (f

-2-2

S

16

5

C "

-2-3

S

176

45

69

C "

-2-4

s

233

48

92

( "

-2-1

s

209

49

72 ,

( " ^

-2-4

s

396

( "

-2-1

Red S

219

91

( "

-2-5

" S

All

/ u

-2-1

L. Purple S

194

27

71

( "

-2-2

S

167

56

80

( "

-2-1

Non-Pur. S

Pure

/ (f

-2-2

S

((

/ u

-2-3

S

u

( "

-2-4

s

u

( " ^

-2-2 BS X 2-1

Red X Red S

380

( "

-2-3 AO X 2-2

Pur. X Pur. S

280

98

( "

)-2-3 CS X 2-1

L. Pur. X L.

Pur. S

152

37

54

( "

)-2-5 ES X 2-4

Non-Pur. x
Non-Pur. S

Pure

( "

)-2-2 EO X 2-1

Non-Pur. x
Non-Pur. s

u

INHERITANCE OF ALEURONE COLOR.

93

TABLE 24B.

F3 SEEDS OF EAR NO. (60-5 X 54)-6 OF TABLE 24.

Ear No.

Planted from

Purple

Red

Non-
Purple

(60-5 X 54

)-6-3

Purple S

87

24

38

/ <f

-6-4

S

204

60

( "

-6-6

S

262

80

/ «

-6-7

S

318

58

/ u

-6-4

" s

265

83

( "

-6-1

Red S

135

29

/ « ^

-6-2

" S

287

76

( "

-6-3

" S

164

48

( " ^

-6-1

" s

240

71

( " ^

-6-1

Non-Pur. S

Pure

/• «

-6-2 AS

x4

Pur. x Pur. S

384

/• K ^

-6-5 AS

x7

S

420

( "

-6-2 AG

x3

s

200

( "

-6-2 ES

xl

Non-Pur. x
Non-Pur. S

Pure

( " ;

-6-4 ES

x5

((

u

TABLE 240.

F3 SEEDS OF EAR NO. (60-5 X 54)-8 OF TABLE 24.

Ear No.

Planted from

Purple

Red

Non-
Purple

(60-5 X

54)-8-l

Purple S

125

( "

)-8-2

S

176

55

( "

)-8-3

S

212

60

( "

)-8-4

s

170

/ ti

)-8-6

s

183

60

( "

■ )-8-l

s

180

28

65

( "

)-8-2

s

182

35

/ «

)-8-5

" s

153

35

/ «

)-8-6

s

217

71

/ ((

)-8-7

s

150

( "

)-8-l

Red S

176

34

/ u

)-8-2

" S

182

( "

)-8-2

" s

156

57

( "

)-8-l

Non-Pur. S

180

/ u

)-8-4

S

12

40

( "

)-8-5

S

250

/ «

)-8-l

s

250

( "

)-8-2

s

220

C "

)-8-3

" s
f Non-Pur. x \
1 Non-Pur. x ]

240

(60-5 X

54)-8-2 ES x4

400

/ Non-Pur. x 1
1 Non-Pur. x J

/ «

)-8-7 ES x6

49

140

[ Non-Pur. x )

( "

)-8-4 EG X 5

i Non-Pur. x [

220

94

INHERITANCE IN MAIZE.

TABLE 24D.

F3 SEEDS OF EAR NO. (60-5 X 54)-ll OF TABLE 24.

Ear No.

Planted from

Purple

Red

Non-
Purple

[60-5 X 54

)-ll-3

Purple S

115

42

47

1 u >

-11-4

S

175

49

c u >

-11-5

S

175

65

(( >

-11-6

S

180

62

« ^

-11-8

S

209

69

' If >

-11-10

S

210

67

r li >

-11-11

S

101

38

(( (

-11-12

« S

112

39

r If >

-11-2

« s

46

21

r (( ^

-11-3

s

144

46

If ^

-11-4

s

178

64

« ^

-11-7

s

180

. . .

* <

-11-1

L. Purple S

116

25

49

' l( ^

-11-3

S

204

41

75

' II ^

-11-5

S

124

52

' II ^

-11-1

s

218

29

75

' « 1

-11-2

L. Red S

163

38

If

-11-1

Non-Pur. S

Pure

' II >

-11-2

" S

II

11 ^

-11-5

S

II

II

-11-7

S

(f

II ■

-11-1

s

II

11

-11-2

s

(f

60 X 54-5:

-11-7 AS

xll-6

Pur. X Pur. S

142 Pur.

and Red

56

" )

-11-3 ES

xll-5

Non-Pur. x
Non-Pur. S

Pure

" )

-11-6 ES

xll-4

Non-Pur. x
Non-Pur. S

(f

" ^

-11-2 EO

xll-3

II

(f

INHERITANCE OF ALEURONE COLOR.

95

TABLE 24E.

F3 SEEDS OF EAR NO. (60-5 X 54)-12 OF TABLE 24.

E

ar No.

Planted from

Purple

Red

Non-
Purple

(60-5 X 54'

-12-3

Purple S

206

78

80

( " ^

-12-4

S

201

44

67

( " ^

-12-4a

S

350

c " ^

-12-5

S

169

59

/ « ^

-12-6

S

245

65

/ «

-12-7

s

204

67

/ li

-12-8

s

191

49

88

( "

-12-9

s

187

66

( "

-12-10

s

239

66

/ «

-12-13

s

248

77

( "

-12-14

s

217

77

( "

-12-1

s

240

72

/ «

-12-3

s

350

/ »

-12-4

s

184

43

56

( "

-12-5

s

300

/ (1

-12-7

s

147

53

/ (1

-12-1

Red S

229

76

(" "

-12-3

" S

280

( "

-12-4

" s

172

56

( "

-12-1

Non-Pur. S

Pure

( "

-12-8

s

a

(60-5 X 54

)-12-2 AS X 3

Purple S

lei

56

53

( "

-12-11 As x6

S

175

63

( "

-12-2 AG X 1

S

229

74

( "

-12-3 ES x5

Non-Pur. S

Pure

( "

-12-4 ES x5

S

(t

/ «

-12-5 ES x7

S

u

/ «

-12-7 ES X 5

S

u

/- «

-12-9 ES x8

S

u

o6

INHERITANCE IN MAIZE.

TABLE 25.

F2 SEEDS OF CROSS BETWEEN NO. 60 NON-PURPLE POP AND NO. 54
PURPLE SWEET.

Very Light Colored and White Seeds Planted.

Ear No.

Purple

Red

L. Pur. + Non-Pur.

(60-8 X 54)-l

83

5

66 -f 135 =201

( " )-7

19

7

44 + 150 = 194

( " )-8

35

4

41+215=256

(60-11 x54)-2

22

4

22+ 40= 62

(60-2 X 54)-l

68

15

96 + 159=255

( " )-7

86

7

99 + 150=249

( " )-io

89

14

69 + 148=217

(60-3 X 54)-l

26

76+282=358

( " )-3

46

7

87+140=227

( " )-.5

54

12

102 + 159=261

( " )-6

65

6

113 + 144=257

( " )-7

69

13

117 + 184 = 301

Total

662

94

2838

TABLE 25A.

Fa SEEDS OF EAR NO. (60-3 X 54)-l OF TABLE 25.

Ear No.

Planted from

Pur.

Red

L. Pur. +
Non-Pur.

(60-3

X 54)-l-l

Purple S

235+ 76=311

)-l-2

S

245 =245

)-l-l

L. Purple S

217+ 69=286

)-l-2

S

39+ 72 = 111

)-l-5

Non-Pur. S

0+384=384

)-l-6

S

105+204=309

)-l-7

S

33L

380*

)-l-9

S

0+390 = 390

)-l-10

S

0+280=280 1

( '

)-l-l

" s

0+448=448

( t

)-l-2

" s

0+200=200

)-l-3

" s

0 + 152 = 152

' )-l-4

" s

0+280=280

(60-3

X 54)-l-l ES X 1-2

Non-Pur. x
Non-Pur.

0+250=250

/ <

)-l-2 ES X 1-1

«

0+258=258

)-l-3 ES X 1-2

a

0 + 110 = 110

)-l-4 ES X 1-2

u

0+308=308

/ I

)-l-8 ES X 1-6

u

0+352=352

Light reds and non-reds.

INHERITANCE OF ALEURONE COLOR.

97

TABLE 25B.

F3 SEEDS OF EAR NO. (60-3 X 54)-5 OF TABLE 25.

E

ar No.

Planted from

Pu

r. Red

L. Colored +
Non-Colored

(60-3 X 54

)-5-l

Purple S

16.

5 47

0+66 =66

( "
( "
C "
( "

)-5-2
-5-3
)-5-4
)-5-6

" S
S
S
s

25(
20:
30(

27:

)

5 62

)

5

"6+ 95= 95

/ «

-5-1

s

17(

3 13

0+ 69= 69

( "

-5-1

Red S

190

0+ 71= 71

( "

-5-1

L. Purple S

'hi

i

136 + 110=246

/ «

-5-2

S

9(

) 49

115+ 96=211

/ u

-5-2a

S

10^

t

78 + 152=230

( "

)-5-3

S

55

)

91+145=236

( "

)-5-5

S

8(

)

58+ 97 = 155

( "

)-5-5a

S

7{

)

65+ 54 = 119

I "

-5-6

S

88+ 26 = 114

/ «

-5-4

L. Red S

(

52

78+ 95 = 173

( "

-5-2

s

75+ 51=126

( "

-5-2

Non-Pur. S

0+352=352

/ «

-5-4

S

0+ 30= 30

/ «

-5-7

S

'6'

I

138 + 193=331

C "

)-5-8

S

0 + 380=380

( "

)-5-9

S

0+345=345

/ «

-5-2

s

0+390=390

(60-3 X 54

)-5-l ESx5-7

Non-Pur. x
Non-Pur.

125+313=438

/ « ^

-5-2 ESx5-4

4'

I

14 + 126 = 140

( "

-5-5 ESx5-4

0+320=320

/ «

-5-10 ES X 5-9

90 + 154 = 244

/ u

-5-11 ESx5-9

43+ 87 = 130

( "

-5-17 ES X 5-7

109 + 144=253

/ «

)-5-l ESx5-5

0 + 360=360

/ u

)-5-3 ESx5-4

0 + 104 = 104

( "

)-5-4 ESx5-3

0+254=254

r "

)-5-7 ESx5-5

0+400=400

INHERITANCE IN MAIZE.

TABLE 25C.

F3 SEEDS OF EAR NO. (60-3 X 54)-6 OF TABLE 25.

E

ar No.

Planted from

p

ur. Red

L. Colored +
Non-Colored

(60-3 X 54^

-6-2

Purple S

1^

iO

0-1- 71= 71

( "

)-6-2a

S

1'

75 9

0+ 64= 64

( "

-6-5

S

1:

38 20

0-f 40= 40

( "

-6-1

L. Purple S

31 5

36-1- 38= 74

( "

-6-la

S

-:

19

68+ 49 = 117

( "

-6-3

S

27

53+ 48 = 101

( "

-6-4

s

(

36 13

63+ 61=124

( "

-6-6

s

{

36

144 + 130=274

( "

-6-1

s

'

30

108 + 142=250

( "

-6-1

Non-Pur. S

0+250=250

( "

-6-3

S

25 '.'.

70 + 106 = 176

( "

-6-5

s

i

21 3

45+ 62 = 107

( "

-6-6

s

0 + 152 = 152

( "

-6-7

s

(

59 '.'.

53 + 105 = 158

(60-3 X 54

-6-3 AS

x6-2

Pur. x Pur.

^

51

88+ 80 = 168

( "
( "

-6-1 AO

)-6-2 ES

x6-2
x6-3

Non-Pur. x

1^

]

52 57

L4

'i5 + '3r='46

Non-Pur.

( "

-6-3 ES

x6-5

«

0+380 = 380

( "

-6-5 ES

x6-6

li

'. *4

21+117=138

( "

)-6-6 ES

x6-5

(I

17

78 + 122=200

( "

-6-7 ES

x6-3

11

37 + 170=207

( "

-6-8 ES

x6-6

u

; '4

65+ 61=126

( "

)-6-2 EO

x6-l

11

84+222=306

( "

-6-4 EO

x6-3

il

33+ 26= 59

( "

)-6-3 EO

x6-4

u

t

^6 25

24 + 171=195

PLATE XI.

fsn lili ^ /f /,i^ Ji ^^B

\ I- M n > I

_K

<u

,,1,

w

r^

U

<U

i-t

rt

ct:

*-'

'a

"Tj

+_)

jz!

c/5

K

CO

O

-1-t

fe

M-l

p

'O

3

<v

Oh

•

•^

^

m

n

OJ

tn

QJ

<u

^

r^

u

in

OJ

1-

,

OT

c

w

,5

S

rt

o

en

rt

t3

cu

o

, ,

cu

en

O

rt

cu

U

OJ

a

cu

CJ

i-l

H

en

a

SJ

p;

c/i

Q_

O

OJ

'— .

en

M-l

'a

rt

TD

;_,

cu

be

OJ

cu

OJ

rt

'uJ

en

biD

.^

cu

rt

h-^

(U

r^;

CJ

'^

u

o

tH

I3J

OJ

t-H

■S

c

0^

be

o

^

.

"'3

OJ

iT

p

M-H

rt

'Hh

u

;_,

c

OJ

p

3

^

Y^

-o

o

^

a

G

^

o

rt

•^

n

_5

o

b/]

cu

cu

rt

tj

-i-)

lU

!^

im

41]

OJ

S

rs

-;-»

OJ

jQ

■3^

^^

be

rt

^'

1^

Tj

p

p

i::^

n

rt

rt

->-<

en

^

^-^

A

lo

cu

0

X

c

jr!

r:J

1J^

en

n

cu

en

bo

o

OJ

o

Q

S

u

Inheritance of Aleurone Color.

i

INHERITANCE OF ALEURONE COLOR.

991

TABLE 25D.

F3 SEEDS OF EAR NO. (60-8 X 54)-l OF TABLE 25.

Ear No.

Planted from

Pur.

Red

L. Colored -|-
Non- Colored

(60-8 X 54)-l-6

Purple S

135

36

0+ 56= 56

( " )-l-7

S

67

16

0+ 20= 20

( " )-l-3
( " )-l-l

s
L. Purple S

280
20

■66 + '76 = i42

( " )-l-la

S

51

i7

63 + 164=227

( " )-l-2

S

35

1

31+ 84 = 115

( " )-l-3

S

40

20

67 + 124=291

( " )-l-5

S

20

16

27+ 21= 48

( " )-l-l

s

74

101+115=216

( " )-l-l

Very L. Pur. S

92+121=213

( " )-l-2

S

67

5

83 + 109 = 192

( " )-l-4

Non-Pur. S

0+250=250

( " )-l-5

s

0+250 = 250

( " )-l-6

s

0+230=230

'( " )-l-7

s

0+240=240

(60-8 X 54)-l-6 ES x 1-7

Non-Pur. x
Non-Pur. S

0 + 100 = 100

( " )-l-7 ES X 1-6

11

0+2.50 = 250

( " )-l-8 ES X 1-9

11

43+ 66 = 109

( " )-l-9 ES X 1-8

u

151+158=309

( " )-l-10 ES X 1-9

u

68

0 + 161=161

( " )-l-l EG X 1-2

Non-Pur. x
Non-Pur. s

0+240=240

( " )-l-3 EG X 1-7

li

K* + K =1

Approximated.

.lOO

INHERITANCE IN MAIZE.

TABLE 25E.

F3 SEEDS OF EAR .NO. (60-8 X 54)-8 OF TABLE 25.

Ear No.

Planted from

Pur.

Red

L. Colored +
Non-Colored

(60-8 X 54

)-8-2

Purple S

t54L

236 + 100=336

( "

)-8-5

S

212+ 0 = 212

( "
( "

)-8-l
)-8-2

Red S
" S

266L
5L

■26 + "8 = '34

( "

)-8-l

L. Purple S

110 + 104 = 214

/ «

)-8-2

S

224 + 123=347

( "

)-8-5

S

50 + 148 = 198

( "

)-8-7

S

62L

0+248=248

( "

)-8-8

S

7bL

84 + 164=248

/ a

)-8-l

s

90+202=292

( "

)-8-l lower ear

" s

97+243=340

( "

)-8-3

Non-Pur. S

80 + 294 = 374

( "

)-8-9

S

80+216 = 296

("

)-8-2

' " s

37L

0+263=263

(60-8 X 54

)-8-l ES X 8-7

Non-Pur. X
Non-Pur. S

0+300=300

/ a

)-8-2 ES X 8-3

S

39+260=299

( "

)-8-4 ES X 8-9

S

49+214 = 263

/ «

)-8-5 ES X 8-17

S

0+200=200

/ «

)-8-6 ES X 8-9

S

0+300=300

/ «

)-8-7 ES X 8-17

S

24+270=294

( "

)-8-l EO X 8-2

s

88 + 174=262

/ <(

)-8-5 EO X 8-3

s

48* + 153=201

* Several seeds rather dark purple.

t Those marked L are light in color but not nearly as light as those
given in the last column.

PLATE XII.

Xo. 21. Pod.k-il maize. b. No. 7, non-podded maize.

f ^-*,-4.''iN.vV^

qK-

■c. Cru-..- Jix;. I- ^ aljoxt'i pud cliaractcr tull\- domniant. Fj below: com-
plete segregation in monohybrid ratio.

Inheritance of "Podded" Character.

XENIA. loi

PART III.

XENIA.

The appearance of the endosperm in the Fi generation in the
crosses discussed in Part II really include almost all of our
observations of true Xenia, but the subject is sufficiently import-
ant to warrant a more systematic arrangement of the facts.

The word Xenia was proposed by Focke to denote the effect,
if any, produced by the action of pollen upon the maternal
tissue of the seed plant. The classical example of such effect
was the endosperm of maize. After the discovery by Guignard
('99) and Nawaschin ('99) that the endosperm is in reality a
part of the filial generation formed by the development of the
endosperm nucleus after fusion with the second male nucleus
of the pollen cell, De Vries ('99), Correns ('99) and Webber ( : 00)
saw in this the explanation of the phenomenon in maize. These
facts took away the only authentic illustration of Xenia in its
original use — the effect of foreign pollen on matern^,! tissue.
In this older sense the word is therefore of no value, and it may
be used solely to describe the visible effect of the second male
nucleus on the endosperm. Unfortunately, botanists have not
been so prompt in discarding belief in the original meaning of
Xenia as the zoologists in discarding telegony. In the experi-
ence of Correns, of Lock and of ourselves the effect of the second
male nucleus has never extended to maternal tissue. One
of the present authors has made several experiments in which
pollination without fertilization (between infertile species) has
had an effect on maternal tissue, (parthenocarpie) , but this
effect was simply that of a chemical stimulus or irritant produc-
ing cell division in the carpels.

The visible effects of double fertilization have been found
in the following cases, in all of which the parents have been
selfed strains that precluded errors in the observations. Non-
starchy seeded plants crossed with starchy seeded plants always

102 INHERITANCE IN MAIZE.

show starchiness. Starchiness is completely dominant, there-
fore the reciprocal cross, bringing in the "opposing" character,
never shows Xenia.

Yellow endosperm is also completely dominant in most cases.
Non-yellow crossed with yellow endosperm therefore shows
Xenia while the reciprocal shows no Xenia. Three exceptions
to this rule were found, however. In the large races of dent
maize where the zone of soft starch at the summit of the seed
is extensive, the heterozygous yellow is somewhat lighter in
color than the homozygous yellow, and Xenia appears when
the cross is made either way. It shows as a cap of lighter color
than the homozygous yellow. When floury yellow races are
crossed with floury white races this lighter color of the heter-
ozygote extends throughout the seed. In this case difference
in color is always great enough to be noticed by a careful observer
in either cross, but where the cap only is floury the color inter-
grades to that of the homozygous yellow. When dealing
with races with corneous endosperms, such as the flint and pop
varieties, there is so little difference in color that the homo-
zygous yellow is generally indistinguishable from the heter-
ozygous yellow; therefore Xenia occurs only when the
white is the female parent. Even here, however, we have
found two different cases where a few heterozygous yellows
were distinguishable from homozygous yellows when the latter
were used as the female.

Both the red and the purple colors in the aleurone cells behave
in the same way as regards Xenia. When the two parents
differ only in these characters, they are completely dominant
and Xenia occurs only when they are possessed by the male
parent. Even in the race in which a slight purple color appeared
when the color factor was absent (P c instead of P C) the same
slight color appeared when it was used as the male upon a race
in which P and C were both absent. Furthermore when this
race was crossed either way with a white race bearing the color
factor (PcxpCorp CxPc), the full purple developed and
appeared as Xenia. The red color undoubtedly behaves in
the same way although we have made no original crosses deal-
ing with these conditions. Again, two pure white races (P c
and p C) which show not the slightest color may bring together

XENIA. 103

the two factors P and C necessary for full development of the
purple color, and Xenia results when either is the female parent.

The next and last case in which we have observed Xenia is
that in which the white parent possesses a character that inhibits
the development- of red or purple aleurone cells. Correns and
Lock probably used races containing this character but they
did not distinguish it from a recessive white or simple absence
of the purple character. They therefore concluded that when
a white race was crossed with a purple, Xenia sometimes results
and sometimes does not result, and that no change occurs when
the purple is the female parent. The true state of affairs is
just the opposite of this. When a white containing the inhibi-
tor is the male parent, a white seed results, and while the same
result is obtained in the reciprocal cross it is of course unnoticed
when the white is the female parent. Sometimes the purple
is not fully inhibited and then a light purple results no matter
which parent is the mother.

If one considers these observations as a whole, the following
law regarding Xenia may be formulated :

When two races differ in a single visible endosperm character in
which dominance is complete, Xenia occurs only when the dominant
parent is the male; when they differ in a single visible endosperm
character in which dominance is incomplete or in two characters
both of which are necessary for the development of the visible dif-
ference, Xenia occurs when either is the male.

Correns observed that in every case where Xenia may be
expected to occur, the seeds showing Xenia were always hybrids.
This fact was assumed to prove that the second male nucleus
always bears the same characters as the one that fuses with
the egg cell to form the embryo. For this reason Mendelian
segregation of the gametes must have occurred previous to the
division of the pollen nucleus. Our observations are entirely
in accord with those of Correns. The latter author and also
Webber observed several cases where Xenia occurred in only
one-half of the endosperm. These rare phenomena which are
probably similar in nature to the gynandromorphs occurring
in insects, they both interpreted as the independent develop-
ment of the endosperm nucleus and the second male nucleus.
We have observed many instances of this phenomenon and
have grown a number of them to see if the tendency was inherited

T04 INHERITANCE IN MAIZE.

but without positive results. Correns' and Webber's expla-
nation of the cause of these seeds is probably correct, yet
the suspicion cannot be avoided that if the two nuclei can
develop independently then the female nucleus ought some-
times to develop to the total exclusion of th-e male. If this
were true a seed showing no Xenia where it is to be expected,
should sometimes prove to be a hybrid. This has never occurred
in our work, a fact in disagreement with the work of Webber.
It may be possible then that the cause of these seeds is Mende-
lian segregation in somatic tissue, such as often occurs in bud
sports. This could be proved if there occurred among the Fi
seeds of a cross in which the parents differed in two characters,
an individual in which the characters were segregated dif-
ferently: for example, if a white sweet maize were pollinated
with a yellow starchy race, and a seed developed having one
half yellow sweet and the other half white starchy. The matter
is simply mentioned because it is important to biological theory,
and it was thought that some experimentalist might happen
upon such an individual.

It is thought that Webber's idea that seeds with splashed
purple aleurone cells are due to mosaic development of cell
descendants of the endosperm nucleus and of the second male
nucleus, is incorrect. If this idea were true, in cases where the
endosperm is heterozygous yellow, this character also should
be mosaic. Such cases have never been reported. It therefore
seems better to consider the splashed purples as cases of incom-
plete dominance caused by other factors as was explained in
greater detail earlier in the paper.

PLATE XIII.

a. Podded maize. The four husks successively removed showing naked
seed at right. The double rowed condition characteristic of all
maize varieties is seen most clearlv.

Male spikes (tassels) showing development of seeds, b. a dominant F2
plant; c, a recessive F2 plant. Segregation is persistent in this cross,
21x7.

Inheritance of "Podded" Character.

I

PLATE XIV.

t left, the color which develops in sunlight — R4; in center variegated
or mosaic seeds — R2; at right, common red pericarp — Ri.

fl. At left, the cok

b. Segregation of pericarp color R4 in F2 of cross 5x11. Amount of color
developed is variable depending on light conditions during
maturation.

Pericarp Colors.

INHERITANCE OF POD CHARACTER. 105

PART IV.

PLANT CHARACTERS.

In this part of the paper the inheritance of normal plant
characters is considered. These characters in general have no
effect upon the endosperm — the new generation — and there-
fore do not show as Xenia in the daughter seeds of the ear that
has been crossed.

Podded and Podless Maize.

The inheritance of the podded character is interesting because
it is a shining example of a case where a gross morphological
character behaves as a simple Mendelian mono-hybrid. No.
21 a podded maize was crossed with a common Learning dent
like No. 7, but not of the same stock. The Fi generation was
as perfectly podded as the podded parent. There was of course
some variation in the length of the husks of the seeds, a varia-
tion apparently physiological in character depending upon the
vigor of the mother plant, but this variation was no greater
in the Fi generation than it was in the pure podded maize. The
F2 generation yielded 64 podded and 21 non-podded individuals.
The latter were without any trace of husk and were not dis-
tinguishable from ordinary non-podded corn which had never
been crossed with a podded variety. (See Plate XII.)

The Fi generation was also crossed back with the recessive —
the non-podded variety — and in the next generation yielded
41 podded ears and 50 non-podded ears. In other words
Hh X h gave 50% Hh and 50% hh as was to be expected.
The character was again strictly discontinuous. The extracted
recessives proved absolutely true.

Pericarp Color.

There are various red sap colors appearing in the pericarp,
the cob, the husks, the silks, the glumes and the anthers of maize.
We have not been able to make a chemical study of them and

io6 INHERITANCE IN MAIZE.

so cannot say if they are due to the same compound, but the
comparatively small amount of data regarding their inheritance
that we have obtained is particularly interesting on account of
the number of different organs in which color occurs. It has
long been thought that such colors that manifest themselves
in different parts of a plant, are single units as regards heredity,
but are produced in visible quantities only when developmental
conditions are favorable or when certain transmissible limiting
factors or extension factors which effect their development,
are present or absent. Our especial problem was to find out
whether these red colors occurred and were transmitted separately
or whether they were linked together in genetic or in chemical
relationships. This work is therefore simply a report of progress.

The first red pericarp, which we will call Ri was found in
No. 27, a rice pop maize. It was the ordinary dark red color
of the varieties commonly known as red corns. It did not
have a red cob or red silks, although the glumes of the male
flowers were sometimes reddish. Crossed with number 28, a
rice pop with white pericarp, white cob and silks, it gave 75 red
and 22 white ears in F2. The color was inherited absolutely
discontinuously, the reds being all dark and the whites showing
no trace of color.

The only other cross with apparently this same dark pericarp
color, was a peculiar ear found in a field of dent maize of unknown
parentage. This ear, as shown in Plate XV, fig. a, had seeds
with red pericarp on one side and seeds which were sometimes
white and sometimes striped with red on the other side. The
ear appeared in a field of white maize in which only white maize
was planted. It must have been produced therefore by a
hybrid seed Ri ri. Furthermore since it was the only ear in
the field showing red pericarp, it is likely that it was nearly all
pollinated by white. One would expect its seeds therefore to
be half Riri and half riri, and that they would give in the next
generation 50% red ears, 50% white ears. In order to test
any possible transmission of the variation which appeared in
this ear however, both the red seeds and the seeds from the
side which had white and striped kernels were planted. From
the dark red seeds were obtained 22 dark red ears and 22 white
ears; from the white and striped seeds were obtained 15 ears
showing a few red striped seeds and 15 ears with only white

INHERITANCE OF PERICARP COLOR. 107

seeds. No difference was observed between the progeny of
white and of striped seeds. Both kinds of seeds from this side
of the ear gave striped ears and white ears. A selfed red ear of
this generation gave a simple mono-hybrid ratio in the next
generation — 75 red ears and 26 white ears. The explanation
of this phenomenon evidently is the same as that of the bud
variations that sometimes occur in perennials. They occur in
annuals but are usually unnoticed. The plant due to produce a
red ear varied somatically so that one-half of the ears was red and
one-half striped. This variation was transmitted by seeds,
but at the same time the hybrid character of its seeds was
unchanged as shown by their segregation into reds and whites
in the next generation and the normal segregation of the hybrid
dark reds in a further generation. This strain had red cobs,
and there was perfect coupling between the two characters in
the next generation.

Two other red pericarp colors seemingly independent of red
in other parts of the plant have been found, which may be
called R2 and R3. R2 is a dark red that occurs as irregular red
stripes radiating from the point where the silk was attached;
R3 is a dirty red color more abundant at the base of the seed
and almost wanting at the summit. The dye occurs in small
amounts. The latter red, which occurs in Palmer's red-nosed
yellow appears to be completely coupled * with red silks. It
is almost certain that this red forms an allelomorphic pair with
its absence that is entirely independent of Ri, R2 and R4, but
our numbers are too small to make a full report on the matter.
The mosaic red (R2) is also one that has not been subjected to
sufficient genetic study. Thus far (2 generations) it has not
bred true but has thrown a percentage of non-reds.

Two other red pericarps have occurred, however, which are
interesting because they are the same in appearance but are not
allelomorphic to each other. The first is a rose red (R4) charac-
teristic of No. 5. It develops only in presence of light, hence
the ears with thick husks show the color but faintly. When
the husks are stripped away and the ear matures in full sun-
light, however, the color appears over the entire ear as a bright

* Coupling is proved by the fact that red silks occur without red peri-
carp in other combinations.

io8 INHERITANCE IN MAIZE.

rose red. In numbers 2, 8 and 18 there appeared another red
which we at first thought was the same as the above. It
occurs in less amounts and on thick-husked ears can only be
detected by careful examination. Since these two reds behave
as separate allelomorphic pairs they are called R4 and Rg.

The transmission of these two reds was shown by crossing
No. 5 (R4) with No. 18 (Rs). In Fi all of the ears were red. In
F2 there were 131 red ears and 7 white ears. No. 5 (R4) was also
crossed with No. 2 (Rs) and gave similar results although the
number of plants was small. In F2 there were 52 red ears and
2 white ears.

It may be asked whether the red in No. 5 (R4) acts as a simple
mono-hybrid in crosses with strains having no red in the peri-
carp. We have only one cross of this kind in which data for
pericarp color were taken. No. 11-2 (r4) was crossed with No.
5 (R4) and yielded 251 red ears and 91 white ears in F2.

None of these varieties had the red color in other organs.

Cob Color.

Several crosses were made in which one parent had a red cob
and one a white cob. None of the parents had dark red peri-
carps (Ri) but in one case R4 was present (the light red pericarp
developing in presence of light). In a cross between No. 5
and No. 6, F2 yielded 277 ears, of which 212 had red cobs and
65 white cobs. It was strictly a mono-hybrid cross, and the
character red-cob seemed not to be coupled with the pericarp
color. This red we may call Re.

The parents in this case were tested for purity although there
are strains of No. 6 in our possession that do not have red cobs.
The results of the other crosses were similar and space will not
be taken to report them in full. It must be noted however, that
although no di-hybrid reds were found, it is not beyond prob-
ability that such might be found in an extensive series of
crosses.

Silk Color.

Varieties are also obtained which have red silks although the
red color. is not manifested in other parts of the plant. In fact,
No. 19, which has the darkest red silks of any variety in our pos-

PLATE XV

a. Somatic or bud variation from dark red seeds to slightly variegated
seeds in ear whose seeds were supposed to be half R^, ri, and half

ri, rj.

b. Progeny of red seeds of a. Half
dark red, half white.

Progeny of sliglulx \ariegated
seeds of a. Half slightly var-
iegated, half white.

d. Similar bud variation in which R2 is concerned.
Pericarp Colors and Somatic Segregation.

DISCUSSION OF SAP COLOR. 109

session, has white cob and pericarp. It is not quite clear, how-
ever, how this character is transmitted. The facts are obscured
by the action of the bag over the ear to be hand-poUinated,
which prevents the full development of the red color by shutting
out the light. For this reason one cannot be certain whether the
Fi plants which are selfed are full reds or only red-haired silks.
An illustration of what is obtained in a cross between red
silk and non-red silk varieties is as follows. No. 12-2 which is
pure fof non-red silks was crossed with No. 9-2 which is pure
for red silks. In Fi there were 110 plants with red silks and 27
with greenish- white silks with red hairs. In F2 the progeny of
3 Fi plants were grown. The first ear gave 123 plants with red
silks and 40 with white silks. The progeny of the other two ears
were of three classes; reds, greenish-whites with red hairs and
greenish-whites in the numbers 198 : 29 : 94. We will not
attempt to analyze this ratio. It is simply mentioned to show
that the silk color does mendelize without the production of
color in other parts of the plant.

Glume Color.

No plant has yet been obtained which has red glumes and yet
shows no red color in other parts of the plant. One has been
found however that is pure for red glumes and shows no red in
other parts with the exception of the silks.

General Consideration of Red Sap Color.

It is difficult to put aside the thought that all of these red
colors are localizations of the same general pigment. If this
were true, there should be a series of varieties in which increas-
ing extension of color is found, until red appears in all the
organs in which it ever occurs. This is not true. Varieties
exist, for example, with red pericarp and red cob, with red peri-
carp and white cob and with white pericarp and red cob. If
these formed a series with increasing extension of red one might
find the color localized in the cob and not found in the pericarp,
but the theory could not account for the existence of varieties
with red pericarp and white cob. It seems as if these facts
would drive us to one of two conclusions. We are dealing.

no INHERITANCE IN MAIZE.

either with different color compounds each of which manifests
itself in only one organ, or with identical genes held in the germ
cells in such different combinations that they may be mani-
fested differently. The latter interpretation is more probable,
and the natural assumption is that identical genes held by
different chromosomes in some way accounts for the different
manifestations. Yet there is an obstacle to this assumption
which though not necessarily insurmountable, is at least impor-
tant. One cannot quite understand why a red color should be
manifested in different organs simply because its gene is held by
different chromosomes.

Since the first draft of this paper was written Emerson * has
reported important data from many crosses where certain of these
red colors of maize are sometimes absolutely coupled in their
inheritance and sometimes show spurious allellomorphism. For
example, if a plant with a red cob and a red pericarp is crossed
with one in which these colors are absent, there is segregation
in F2, but the colors remain together. On the other hand, if
a cross is made between a plant having a red cob and a white
pericarp and one having a white cob and a red pericarp, the
colors show spurious allelomorphism. The spurious allelo-
morphism is shown by the F2 generation, in which is produced 1
red pericarp-white cob : 2 red pericarp-red cob : 1 white pericarp-
red cob. His idea is that in the case first mentioned the colors
are both carried in the same chromosome while in the second
case they are carried in different but homologous chromosomes.
As Emerson himself has stated, this theory assumes the inevi-
table pairing of the two chromosomes carrying the colors, which
is probable but unproved. Our own data show no facts diamet-
rically opposed to this hypothesis but the criticism regarding
genes held by different chromosomes that was made above would
also apply here.

Physical Transformations of Starchiness.

Although presence and absence of starchiness behaves as a
Mendelian allelomorphic pair in heredity, the physical condition
of the starch is a different matter. Starchiness acts as a filial
or endosperm character and shows as Xenia in individual seeds.

At meeting of Amer. Soc. Nat., Ithaca, N. Y., Dec. 29, 1910.

PHYSICAL FORM OF STARCHY CHARACTER. in

The physical condition of the starch behaves as a plant character
affecting the entire ear. One may have ears which show a
tendency towards the dented character in some seeds and a
tendency towards the flint character in other seeds. Such ears
are probably always heterozygous dent -flint combinations, and
simply show zygotic variations. The different kinds of seeds
give the same results in the next generation and show no tendency
toward a real segregation of dent and flint characters in the
individual seeds. .

The difference in the appearance of the starch in the different
races of maize has been described earlier in this paper. The
immediate cause of these differences is the amount and location
of the soft starch formed in proportion to that of corneous
or translucent starch. In the pop corns there is total absence of
soft starch or at most only a small amount immediately sur-
rounding the top and back of the embryo. As this amount of
soft starch increases, the starch cells of the seeds lose their
ability to hold the steam formed by the moisture they contain
when heated, and can no longer evert their entire contents as
cooked starch. They may pop slightly but they can no longer
be considered commercial pop corns. They have passed into
the flint corn class. This class includes varieties with varying
amounts of soft starch up to those in which it covers the cap.
The latter are dent corns, for the dent is simply formed by the
greater percentage of contraction which the soft starch under-
goes in drying. The amount of dentness is in direct proportion
to the thickness of the soft starchy layer at the cap. A few
varieties are known in which the soft starch has replaced almost
all the corneous starch. They are known as semi-starchy corns.
They are not so well known however, as the floury corns in
which the corneous starch is absent.

As all of these varieties are known in Z. mays curagua, Z. mays
hirta and Z. mays tunicata, it is obvious that the proportions
these two kinds of starch (in appearance at least) plays a great
part in the commercial classification of maize. Also, since so
many varieties are known in which every possible ratio of corne-
ous starch to soft starch occurs, it is evident that the trans-
missible characters which cause these differences are relatively
numerous and their interactions complex. For these reasons,
it is perhaps too much to expect that the inheritance of this

112 INHERITANCE IN MAIZE.

complex of characters will be cleared up until all possible com-
binations of these varieties have been made and studied. Our
data serve only to establish certain general facts.

The first bit of evidence in the matter comes from a con-
sideration of the behavior of the only class of maize varieties that
apparently are beyond the scope of the subject in hand — the
sugar varieties. When the latter are crossed with starchy
varieties it is perfectly clear that starchiness is a separate
character independent of the physical form in which it exists.
Sugar varieties are found that are simply dents and flints
which lack starchiness. We have also produced by crossing,
sugar varieties that -are characteristic pop corns lacking starch.
No sugar varieties are known which would be soft starch varie-
ties (Z. mays amylacea) if they contained the S factor, but it
can hardly be doubted that such could be produced. The
experimental evidence is as follows. When Black Mexican,
Early Crosby and Golden Bantam are crossed with dent varie-
ties, the Xenia starchy seeds, or Fi generation are all flint-like
in character. These when grown produce Fi ears which have
an appearance intermediate between dents and flints and give
in F2 ears which are characteristically flint in character. In
the case of the cross between Black Mexican sugar No. 54 and
Illinois High Protein dent No. 8, these flint segregates of F2
were carried to the F3 generation and bred true. Since pure
dent varieties were the male parents of these crosses, the occur-
rence of flints in F2 can only be accounted for by supposing that
the sugar varieties that were used as the female parents of
the crosses were latent flints. In the same way Stowell's Ever-
green sugar and Late Egyptian sugar were proved to be latent
dents by crossing them with starchy flint varieties. The Xenia
seeds were dented and pure dents appeared in the F2 generation.
One peculiar thing occurred in the cross between Black Mexican,
sugar. No. 54 and Illinois High Protein dent, No. 8. In Fi all
of the ears were intermediate between dent and fiint with a
tendency toward dentness, except one. This ear was a pure
flint in appearance. Only one of the intermediate ears was
grown in the F2 generation and it produced 91 dents and inter-
mediates and 6 flints. The pure dents could not be separated
from the intermediates but flints occurred in the ratio of one
out of sixteen. The ear which was apparently flint in Fi proved

PLATE XVI.

■H

sS^

At left, No. 15, Longfellow flint. At right, No. 8 Illinois high protein
dent. In center, Fi ears of cross 15x8, showing interinediate char-
acter of physical condition in which the starch is stored.

Dent-Flint Crosses.

PHYSICAL FORM OF STARCHY CHARACTER. 113

to be an intermediate in F2. Thirty-four ears were obtained,
of which three were clearly dented, a number were intermediate,
while from ten to twenty would ordinarily be classed as flints.
Thirteen of the latter were grown in the F3 generation and
produced from 50 to 175 ears apiece. Nine out of the thirteen
gave only flint ears in a total of 947 individuals. The other four
ears produced a total of 264 ears of which between 10 and 20
were flints (i. e. ten were certainly flints and ten others were
questionable). Therefore, since 9 out of 13 of the 20 ears
classified as "probable flints" in F2 proved to be true flints in
F3, we have 14 ears pure flint to 20 dents and intermediates in F2.
We do not know enough about this cross to say just what occur-
red here, but it is probable that one factor for dentness was miss-
ing in the pollen which produced the hybrid seed from which
this lot F2 ears came. In the other case a di-hybrid ratio
appears.

Several other crosses were made between true dent and true
flint races, that is, races in which the parents both were starchy.
No. 15 Longfellow flint was crossed with No. 8 Illinois High
Protein dent. The Fi generation was intermediate in character.
Through an unfortunate oversight data regarding the segrega-
tion in F2 were taken on the progeny of only one ear of the
three Fi ears planted. This ear gave 33 dents and intermediates
to 3 flints. About 200 ears were obtained from the other
two Fi ears planted and from our general fleld notes we can
say that not less than 15 dents and intermediates to each flint
ear were obtained. One flint ear gave a crop of 94 ears in F3,
all of which were flint. One dent ear grown in F3 also proved
to be pure. A better idea of these results is given by the photo-
graphs on Plates XVI and XVII, however, than can be given by
written description.

Two crosses were made between No. 11, Sturgis' flint and
No. 8, Illinois High Protein dent. Both were intermediate in
Fi. In F2, progeny of one Fi ear of the first cross gave 44 dents
and intermediates to 3 flints. In F3, one ear from an inter-
mediate of F2 gave 23 dents and intermediates and 2 flints.
Five Fi ears of the other cross were grown in F2 resulting in
175 dent and intermediate ears, and 17 flint ears. The ratio
here is about 10 : 1, but if any error was made in the classifica-
tion it certainly occurred by placing intermediates in the flint
class.

114 INHERITANCE IN MAIZE.

Another cross of this kind was that of No. 5-5, Watson's
fiint with No. 2, Illinois Low Protein. The ears were inter-
mediate in Fi. In Fo there was segregation, for ears exactly
like No. 2 were obtained. Out of the 101 ears obtained, how-
ever, no ears were produced that could be classed definitely
as flints. One or two flint-like ears occurred which will be tested
for purity this coming season. It is quite likely that we have
here a tri-hybrid or possibly a tetra-hybrid.

The female parent of this cross. No. 5, was also crossed with
No. 6, Learning dent. Fi generation was intermediate as before.
Five Fi ears were grown with the following results:

Dents and Inter.

Flints

98

16

71

17

51

5

42

7

Total, 262

45

These ears gave different ratios. Probably more ears were
classed as flints than would prove to be such in the Fa generation,
yet they were classified similarly in each case and Fs tests
would probably only reduce the proportion of flints from each
ear. Paradoxical as it may seem, however, different ratios
are to be expected in F2 if the general hypothesis concerning
the applicability of Mendelian principles to cases where varia-
tion is apparently continuous, is true (East : 10). This is
explained in the following paragraphs.

In the crosses described above three facts stand out definitely.
The characters which give the flint or the dent appearance to
maize are transmitted as plant characters to the entire ear and
not as endosperm characters to the individual seed. They
conform to the essential feature of Mendelism by showing
segregation; and they, are due to the action of more than one
transmissible character. The question remains, can any or all
of these characters be named ?

Our experience suggests that the proportion of corneous
starch to soft starch depends partially upon size and shape of
the pericarp and upon the number of rows per ear. All of the

PHYSICAL FORM OF STARCHY CHARACTER. 115

races (pop corns) in which soft starch is absent have small seeds,
and the full corneous starch character cannot be transferred to
large seeds by recombination through hybridization. On the
other hand, by crossing a pop maize with a dent maize dent
seeds may be obtained which are much smaller than many races
with flint seeds. Further, dent races are known which have
much larger seeds than some races in which the corneous starch
is entirely absent (the flour corns). There is also some rela-
tion between the size of the plant and the amount of soft starch
in their seeds. The floury or semi-floury corns are in general
larger than the corneous starchy corns. Here again, however,
there is an overlapping, for we have produced dent races by
crossing with dwarf pop races, which are much smaller in size
than the large pop and flint races.

Relationship between the physical character of the starch and
shape of pericarp is much more intimate than it is between the
former and size characters. In the rice pops the pericarp is
drawn to a point at the place where the silk is attached. This
makes the rice pop races have rather long slender seeds, but it
is probably due to a separate character or characters. Leaving
this complication out of consideration one may say that the
pop corns have small seeds which are almost as broad as they
are long. As the seeds become larger, if the ratio of length to
breadth remains about unity or less, flint races are formed.
If, instead, the ratio of length to breadth increases, dent races
are formed. On the other hand, medium large to large seeded
races may have almost any ratio of length to breadth and be
either flint, dent or floury varieties.

Of course the shape of the pericarp depends somewhat on the
number of rows, as the greater this number the more the seeds
are crowded together and thus lengthened. Small-seeded pop
and flint races exist with as high as 20 rows, but when the seeds
are medium in size flint races are usually 8-rowed and 12-rowed,
and never — in our experience — over 16 rowed. Dent races,
on the other hand, seldom occur with less than 12 rows, but
when large seeded they do exist with as few as eight rows.
Floury races we have never seen with less than 10 rows, but
they reach as high as 24 rows.

These relationships may simply be correlations and not
direct causes of the proportion of corneous starch to soft starch

ii6 INHERITANCE IN MAIZE.

that exists in various strains of corn. But even if the\^ were
directly concerned, they could not account for the large number
of differences in varieties, for none of the correlations are suf-
ficiently high. Many other characters, the exact nature of
which is unknown, must be concerned in the matter. The
simplest interpretation of the matter seems to be the interaction
of independent allalomorphic pairs of the nature reported by
Nillson-Ehle (: 10) and East (: 10) in earlier papers. If this
interpretation be granted, one should expect that greatest
difference in character pairs would exist in the case of pop and
starchy races. Flint and dent races with about the same size
seeds and small differences in number of rows should differ
by fewer pairs of characters.

We have seen that in two of such crosses the evidence points
to the existence of two allelomorphic pairs giving pure flints and
pure dents in the F2 generation once in every sixteen individuals.
In another cross (5-5 x 2) at least three character pairs are con-
cerned. It happened that in two of these cases the male parents
were Illinois High Protein and Illinois Low Protein dent races,
which gives us some idea as to why there was a di-hybrid ratio
in one case and a higher ratio in the other case. These two
strains were both isolated by selection from a commercial
variety known as Burr's White. This variety, as are most
commercial varieties, is a mixture of complex hybrids. By con-
tinued selection of ears high in protein and of ears low in protein
with close interbreeding of the progeny these two strains were
isolated. The high proteid race is characterized by a high
percentage of corneous starch, bringing it into closer relationship
to the flint corns. The low proteid race is characterized by
a high percentage of soft starch, bringing it into closer relation-
ship with the flour corns. It was the high proteid strain, that
is, the one nearer the flint varieties, that gave the di-hybrid
ratio when crossed with a flint race; while the low proteid
strain, — the one nearer the flour corns, — gave the higher
ratio.

This result is what one should expect, but can the 6 : 1
ratio obtained in the cross between No. 5 and No. 6 be explained
so easily? We believe it presents no difficulties if the complex
gametic constitution of No. 6 is properly appreciated. The
individual which furnished the No. 6 pollen came from a selfed

SIZE CHARACTERS. 117

daughter ear of the original No. 6. Its sister ears varied in
number of rows from 12 to 20 with the mode at 16. The
individual furnishing the pollen in cross 5x6 was in all prob-
ability therefore a complex hybrid itself, and the cross instead
of being simple was really a collection of crosses. There is
no doubt that many intermediate ears were classed as flint
in the table given above. If they could all be grown for another
generation it is quite likely that a series of mono-di-tri and
higher hybrids would be found. It may be asked why, if this
is the case, were not the other crosses complex? The answer
is that they undoubtedly were more complex than they seemed.
For example, if a large number of Fi ears were grown it is likely
that some would give ratios other than those found. It was
simply chance that gave us fairly good di-hybrid ratios from
a few Fi ears in two instances. The most important reason
why the cross with No. 6 was likely to be more variable than
the others, however, lies in the fact that all of the other strains
had been inbred for much longer periods.

Size Characters.

The remainder of Part IV will be devoted to a discussion of
the inheritance of size characters, — variations that have been
considered to be and to casual observation are, continuous.
Our studies have been concerned with the number of rows per
ear, height of plant, length of ear and size of seed.

It is perfectly obvious to one familiar with the maize plant
that it is almost impossible to work out in detail the inheritance
of the complex factors that interact to cause the transmissible
differences in the size of its organs. That size characters are
complex in themselves is shown by the numerous varieties
grown commercially. They each vary from their own means,
but different variety means in height are found all the way
from two and one-half to fourteen feet with but little actual
difference between the most similar strains. Further to com-
plicate matters, all size characters respond to environmental
stimuli, and these non-inherited fluctuations obscure the
analysis of pedigree cultures in a still greater degree. For
these reasons we do not attempt to analyze our results further

ii8

INHERITANCE IN MAIZE.

than to say that they do show segregation in every case* And
segregation is held to he the important and essential feature of
Mendelism. Therefore we believe that size characters mendelize.
Let us now consider the hypothesis by which segregation in
characters apparently continuous in their variation, could come
about. Nillson-Ehle (: 09) has shown that black glumes in oats
when crossed with their absence behave sometimes as mono-
hybrids and sometimes as di-hybrids, and that presence and
absence of red pericarp color in wheat sometim.es behaves as a
tri-hybrid. He further showed, although not quite so con-
clusively, that presence and absence of ligule in oats behaves
as a tetra-hybrid. In this and in a former paper (East : 10)
it has been shown that yellow endosperm, red pericarp and

TABLE 26.

INHERITANCE OF ROWS IN CROSS (5 X 6).

No.

Gen.

Rows

of

Parents

Row Classes

8 • 10

12

14

16

18

20

22

24

No. 5 flint
(2 yrs.)
No. 6 dent
No. 5 X 6
(5 X 6)-l
(5 X 6)-2
(5 X 6)-22
(5 X 6)-23

P
P

F2

F2
F2

8

18
8
12
10
10
12

289

13
12

7
8
4

2

36

48
22
45
25

2

6
53
35
15
31
60

31
10

9
2

1

18

51

1

4

18
2

4
1

dented seeds (as opposed to flinty seeds) behave as di-hybrids,
with so many data that the facts can hardly be questioned.
We have also shown although less conclusively that other red
pericarped varieties and other varieties which differ in their
ratios of soft to corneous starch behave as higher hybrids.

It should be clearly understood what this means to Mendelian
theory. Several genes for the same character may exist in the
germ cells of one organism, the number being limited possibly by
the number of chromosomes. The limited number of cases thus

* It is probable that the number of internodes per plant is one of the
factors directly concerned in the inheritance of height of plant.

PLATE XVII.

a. F2 dent segregate above (frequency about i in 10). Random sample
of its F3 progenj' below.

- -A

^-vS^

.Vf' i

b. F2 flint segregate above (frequency about i in 16). Random sample
of its F3 progeny below.

Dent-Flint Crosses.

SIZE CHARACTERS.

ug

far found presumably is due to the fact that few size characters
have been investigated, for nowhere would these phenomena be
so likely to occur as in quantitative characters.

It is fortunate for us that it has been possible to prove the
presence of several independent allelomorphic pairs due to
produce the same somatic character, for characters like color
where dominance is relatively perfect. Beginning with this
as a basis, one can extend the theoretical possibilities of such
facts to other cases and thus be better prepared for the paradox-
ical complexities that occur in actual pedigree cultures. When
in a cross there is simple presence dominant to absence of one
gene for a certain character, the ratio in F2 is 3 dominant to
1 recessive ; when two independent allelomorphic pairs producing
the same character are concerned, the ratio in F2is 15 dominants

TABLE 27.

INHERITANCE OF ROWS IN CROSS (5 X 2).

No.

Gen.

Rows

of

Parents

Row Classes

8

10

12 14

16

18

20

22

24

No. 5 flint

(2 yrs.) ^
No. 2 dent
No. 5 X 2
(5 X 2)-6

P

P

Fi
F2

8

16

8

10

289

i

4

2
'9

18

2

2 14
20 1 4
61 14

56
3

42
1

20

1

1

to 1 recessive. In general then if n allelomorphic pairs are
concerned, in F2 there will be a ratio of 4''-! dominants to 1
recessive. It is not likely however that dominance is ever
perfect in these complex hybrids. For example, in the case of
the two yellow colors in the maize endosperm, the intensity of
the yellow decreases in the following order Y1Y1Y2Y2, Yiyi
Y2Y2 or YiYiY2y2, YiYi or Y2Y2, Yiyi or Y2y2 and yiyiy2y2.
In size characters dominance is probably very incomplete or
absent. A heterozygous combination presumably produces
half the effect of a homozygous combination. Then as domi-
nance becomes less and less evident the Mendelian classes vary
more and more from the formula (3 + 1)" and approach the
normal curve of error (3^ + 3^)." When there is no dominance

I20 INHERITANCE IN MAIZE.

and open fertilization, a state is reached in which the curve of
variation simulates the fluctuation curve, with the difference
that the gradations are heritable. The heritable variations
are always more or less obscured, however, by the ever present
fluctuation.

The experimental results may now be considered — remembering
only that fluctuations are present and that in maize many
genotypes are often present in one parent. In Table 26 are
shown the results from a cross between a race practically pure
to the eight rowed type. No. 5, and a dent No. 6, which varies
from twelve to twenty rows with the mode at sixteen rows. The
Fi generation is intermediate and furnished four inbred ears
that were grown in the F2 generation. Now three of these
four F2 families show no greater range of variation than Fi,

TABLE 28.

INHERITANCE OF ROWS IN CROSS (11x5).

No.

Gen.

Rows

of

Parents

Row Classes

8

10

12

14

16

18

20

22

24

No. 11 flint
No. 5 flint
No. 11 X 5
(11 X 5)-8
(11 X 5)-18

P
P

Fi
F2
F2

12
8
12
12
10

1

289

2

10

19

4

2
11^
38
33

387

2

26

107

100

7

2

23

5

1

8

yet it is a noticeable fact that they vary in different ways. Ear
(5 X 6)-l shows a modal condition at ten rows. It may be
considered that the crossed seed from which the Fi ear that
produced this crop came, contained the genes for lower numbers
of rows from the varying parent. No. 6. Ear (5 x 6)-23, on the
other hand, evidently contains genes from No. 6 that were due
to produce higher numbers of rows.

Table 27 shows a slightly higher variability in F2 than in
Fi.

Table 28 is interesting because it shows the results of a cross
between two varieties that have been selected for many years
until they are relatively true to the 12-rowed and 8-rowed

PLATE XVIII.

a. No. 60, Tom Thumb maize, showing variation in length of ear. Class
centers are even centimeters {}q).

/»j^/ /3 14 /s J6 n n n 20 I)

Ky.^ 3 n n ts W rf 10 i 1

b. No. 54. Black Mexican sugar maize, showing variation in length of
ear {}{).

Inheritance of Length of Ear.

PLATE XIX.

Length 7

16 17 18 19

11 11 H n

Variation in length of ear of Fi, generation of cross between No. 60
and No. 54 (ig).

^ '^ 10 n 12 1^ 14 15 10 J7 lb 19

-V,. K., f s II S(. do ns izf 7; q n n i> 1

b. Variation in length of ear of F2 generation of cross between No. 60
and No. 54. Family (60-5x54) CY).

Inheritance of Length of Ears.

PLATE XX.

a. Variation in length of ear of F2 generation of cross between No. 60
and No. 54. Family (60-3x54) (i,;)-

7

8

9

10

11

12

13

14

15

IG

17

18

0

20

2)

2

5

n

33

53

33

n

It

13

10

II

\i

1

^

/

b. Variation in length of ear of ¥2 generation of cross between No. 60
and No. 54. Family (60-8x54) {}'^).

Inheritance of Length of Ears.

PLATE XXI.

^

V

K

^% ^%

K

^WQ9^^

r.

'^€^'

a. Average size of seeds of No. 60 (upper left) and No. 54 (lower left)
and the Fi generation of the cross between them. E.xtremes of the
F2 generation at right.

>r

h. Average ears of No. 60 (left) and No. 58 (right) with average of Fi
generation in center. Extremes of F2 generation shown.

Size Inheritance.

SIZE CHARACTERS.

121

TABLE 29.

INHERITANCE OF ROWS IN CROSS (11 X 18).

No.

Gen.

Rows

of

Parents

Row Classes

8

1
13

2
1
8

10

12

14

16

18

20

22

24

No. 11 flint
No. 18 sugar

(2 yrs.)
No. 11 xl8
(11 X 18)-4
(11 X 1S)-10

P
P

Fi

F2
F2

12

12

12
12
10

10

9

13

387
51

24

78
62

7
4

1
10
13

1
1

conditions, respectively. F2 shows a distinctly higher varia-
bility than Fi. It is expected that 8-rowed F2 plants may
breed relatively true.

Table 29 is given simply to show that a cross between two
12-rowed varieties does not show an extension of the row classes.
Such a condition should sometimes be possible if our general
hypothesis is true, yet it might not occur in more than one cross
in hundreds.

Table 30 shows the results from a cross between another variety
true to the eight rowed condition and a variety which varies from
ten to eighteen rows with the modal condition at twelve. Un-
fortunately only a few plants matured in the Fi generation and
no conclusions can be drawn regarding its variability. The F2
generation apparently shows a marked segregation. The

TABLE 30.

INHERITANCE OF ROWS IN CROSS (15 X

Rows

Row Classes

No.

Gen.

of
Parents

8
100

10

12

14

16

18

20

22

24

No. 15 flint

P

8

1

No. 8 dent

P

14

3

54

36

12

2

No. 15 X 18

F,

8

2

5

(15 X 8)-2

F,

10

14

15

28

9

1

(15 X 8)-3

Fo

12

4

13

25

6

3

(15 X 8)-2-10

F,

14

1

8

14

6

1

1

(15 X 8)-2-l

: F,

8

32

35

23

4

(15x 8)-2-5

1

Fs

12

4

. 41

116

15

3

1

122 INHERITANCE IN MAIZE.

results in the Fs generation are the most interesting, however,
for the progeny of an eight rowed F2 show a distinct tendency
toward an 8-rowed condition, while progeny of F2 ears having
twelve and fourteen rows respectively, though highly variable,
show a transmission of their parental qualities.

Our largest pedigree series for number of rows is shown in
Table 31. The male parent is the same as was used in the
previous cross. The female parent is an eight-rowed type but is
not so pure for this condition as the eight-rowed varieties prev-
iously used. The general crop in Fi was discarded before the

TABLE 31.

INHERITANCE OF ROWS IN CROSS (8x54).

Rows

Row Classes

No.

Gen.

of
Parents

8

10

12

14

16

18

20

22

24

No. 8 dent

P

12

3

54

36

12

2

No. 54 sugar

P

8

89

25

7

No. 8 X 54

El

12

1

6

14

(8 X 54)-l

F,

12

9

22

16

1

( " )-5

F,

12

1

3

16

1

( " )-!-!

F,,

10

15

87

4

( " )-l-2

F,,

8

20

38

50

( " )-l-2a

F,,

10

61

48

54

( " )-l-3

F,,

10

32

75

15

( " )-l-3a

F,

8

5

20

27

1

( " )-l-5

F,,

12

33

158

26

3

( " )-l-6

F,

12

4

36

109

8

2

( " )-l-10

Fa

8

Very irre

gular

, mostly 8-rowed

( " )-l-13

Fa

10

96

43

8

data was taken upon the number of rows. This oversight is
partially rectified by the records from 21 hand-pollinated ears,
but the true variability is presumably somewhat greater. Two
Fi ears were grown in the F2 generation, one having the moda
condition at ten rows and the other at twelve rows. Nine ears
from the Fj progeny from (8 x 54)-l produced Fa crops. This
table should be examined in order to appreciate the significance
of the results of this generation. There is a marked tendency
in different ears to segregate into twelve-row and eight-row
types. Two of the ears have modal conditions at ten rows.

SIZE CHARACTERS. 123

but their variability is so great that the presumption is that
this represents simply the continuance of the heterozygous
condition. In our opinion there is no question about segre-
gation of number of rows but we are perfectly aware that the
believer in selection would be justified in the criticism that that
is the cause of the results obtained.

Table 32 shows the frequency distribution of the heights of
two varieties Nos. 5 and 6, and the Fi and F2 generations of the
resulting cross. A good idea of the possible segregation in the
F2 generation of such crosses as this, is obtained by the compara-
tive size of the coefficient of variation of the Fi and the F2
generations. In eveyy case it is at least 50% higher in the Fj
generation than in the Fi generation. The Fi generation is
not intermediate between the two parents but is nearly as high
as the taller parent. This fact is not to be regarded as in any
way connected with dominance. It is due to the increased
vigor which comes from crossing in maize as shown in a previous
paper (East : 09) . The total results of the F2 generations
show segregation from the lowest class range of the shorter
parent to the highest class range of the taller parent. It must
not be thought however that these segregates are regarded as
pure types. Their behavior in further generations is still
problematical. Continued selection of shorter or taller segre-
gates presumably will give an approach toward the selected
condition. The criticism that any such results would be due
to selection and not segregation is not valid in this case, however,
for segregates of extreme types that never appear in either of
the parents alone have occurred here in the F2 generation.

Table 33 shows similar segregation in heights of plants in
another cross, No. (54x60). The frequency distribution of
the heights in No. 54 was obtained from plants grown during
the season and on the same soil upon which the F2 generation
was grown. The exact distribution of heights of No. 60 and
of the Fi ears was not taken because at that time another
object was in view. The range of distribution as shown by
the black lines, is correct. From notes recorded at the time
we know that the Fi generation was quite uniform, the measure-
ments being distributed around classes 67 to 73. Here again
the effect of crossing is observed in the relatively tall plants of
this generation. The lowest plants in the F2 generation reach

124 INHERITANCE IN MAIZE.

the upper range of No. 60 while the highest plants are practically
the height of the highest plants of No. 54. The reason that
no plants were obtained in the lower range of No. 60 is due no
doubt to their continued heterozygous condition in some of
their characters and therefore an increksed vigor.

Table 34 shows that the lengths of ears in the above cross
segregate in a similar manner. The Fi generation is not forced
toward the long-eared parent as it is in the heights of the plants.
In other words ear length does not show the increased vigor
due to heterozygosis that is seen in the heights of plants. There
can be scarcely a doubt that the greatly increased variability
in Fi is the direct .result of segregation.

The segregations of weights of seeds in the above cross is
shown in Table 35. The Black Mexican parent No. 54 shows
somewhat distorted variation in this character as there are
four classes of large sized seeds containing only six ears in all.
No F2 segregates occurred of this size. The reason is that the
ears of No. 54 which produced this crop were commercial seed
of which only three individuals were used in crossing. The
Fi generation in both Tables 34 and 35 were recorded from only
one cross although three crosses were made. To be strictly
fair, therefore, the F2 generation of cross No. (60-5 x 54) is the only
one that can be directly compared with the Fi generation given.
We have records, however, of a sufficient number of ears of the
other two crosses to know that they differ but slightly if any
from the one recorded in the tables. But even if we should
be conservative and leave out of consideration the Fj genera-
tions of crosses (60-8 x 54) and (60-3 x 54) , there is still no
question but that segregation has occurred.

>

d

rt OS CO ■* >Q OS O
C<l t^ IC 00 1^ lO lO

T)H (N lO CD CO lO ■*

•h' 41 -h"-h'4|' •«'•«'

1-1 -H 00 m Tt< rq lO
mocDi--cooo

OS >o 00 lo CO c<i cq

S. D.

■* CO OS CD t^ 00 ■*

IN 00 ■-! 00 CD CO CO
CO O] lO r^l lO Tf CO

41414141414141

OS t^ — 1 CO 00 lO 00

>!»< o cq i> i> ic c<i

CD IC 00 IN 1-1 OS O

<

CD O O lO N 1-1 CO

0 C Tt< 00 O IN 1-1
Tt< r}( 1> CD 00 CO lO

41 41 41 41 41 41 4l'

01 00 CO o CO CO Tti
<N 1-1 in o CO Tji csi

00 i-< ^ ^ 00 OS T-4

CD C OS 00 00 t- 00

ji
K

a

o

a>

o

C

a
O

c8
O

T-H

!« : :(N : ;

00

o

loot^c^cq : :

o

• rflTjH^OO -IN

IN

P

•COr)H-*OrtTj<

03
03

■ -rf "t 00 CO th in

CO

•COCDOOOi-lO
•tH rH 1-1

CO

OS

• IN OS CO lO 00 1-1

o

OS

• IN OS CO CO 00 OS

1>

■ 00

• i-l CO 1-1 IN O CD

00

• • t i> 00 lo th

rHiN

00

^ -lOOlNINt^

. i-HrtrHIN

00

CO • -COlOCOiH

lO

O • -iNCOOSi-l
1-1 • -IN <N

s

00 • • O IN 1> !>•

OS

CO

CD
CO

t> • -C^TiKr^OO

t^ • • lO IN ■* t-

CO
CO

rt . .CO -rf C^

o

CD

lo : icoiN :■*

iti ! !co ; ;c<i

s

IN :;■-<: ii-H

lO

CO ! In iiN ; 1

d

El, rH M 00 -H

^^^xxxx

CO CO CO CO CO
X X t< X ^
IC lO IC lO »o

H £

r26

INHERITANCE IN MAIZE.

in o lo CD o ro 00 05

o

TfOiOlMCO^CJO

l-Hi-H COIMTtI

^

03

'"'

00

00

CO : : A :-■ :

00

- :aa

:- :

IN

Tf< ■

CO-* •

00

03

■* •

coco ■

t>

CO

^ .

(M03 •

1>

<M •

cq

CO

CO •

i-ioin

t>

C-1 •

C0O4

o

TJ< •

O5rHC0

^

l>

IM •

incqi-H

a

ca

l>

1> •

CC1I>TJH

Oi

o

osoqrt

^

O •

cocOt>

o

C^CO-cH

■a

,_4

CO ■ V

t-^OJ-*

CD

.-^iNco

T-H

00

cq • ■

inmiN

lO

oscoi>

in

■\|rTt<C0O

* TfT-HCO

lO

<N

• • -V -CD CO 00

e

lO

+j

03

OOCOrH

0)

O

'^

rt CO

M

o

oomcq

CO

o

1-1 .-HCO

-*

^

o

:a : : :^ :S

1

■*

l>

• • -(Nrtin

CO

Tf

CO

^

CO

00

<M

in

(M

<N
(N

:v : : : : : :

fefefefri'fefe

^^^^^-^^^^^y^

TfOTt'-!l<Tt<Tt<^Ttl

d

in o in in in lo lo lo

►^'^inoocoinooco

■^■^ 1 1 1 1 [ J

oooooo

CD CO CO CO CO CO

s °

.S "5

^ "^

t- 1-

hn

ctf

bsi

B

G

m

bo

&

q S

« s S

u, u "^

?S S '3

^ s ?^

CR +j +J

see

o o o

^« & & &

* n

o o o

SIZE CHARACTERS.

127

CO "-H (M CO CO ■*

00 CO CQ 05 r-( ■^

•H'-H-H-tl-tl'-H

oq 1-1 -* CD Ti< CO

rH 00 CO t^ CO t^

10 00 00 CO lO 00
000000

-H-D-H-H-H-n'

■-H t^ i-H a> 10 ^

00 001OO5IM00

rH,-lr-l(N<M

CO rH ,-1 00 CO 00

•t] -H -H ii 41 -H

O00rHt>02 CO

(N

:iM ! ! I"-!

0

:t^ ! ! !iM

0

■0 •-l.^r-l

00

•lO -COOStN

m

i>

•CO -t^lCi-l

'0

CO

.10 -t^iOO

■ T-i -cqiMi-i

10

•(MTt<C0O:C0
• rt CDCO.-1

■*

•^O3-H00^
•1-1 050(N

"2

CO

■COt>O5 00t>
i-HNcDiM

0

IN

• •■* mcoco

• -^TtltvCO

1
0

0

-

• -(MOI^CO
■ -rt 00 TfICO

0

• -c^cococo

• • rt in (>) CO

03

00

00 • -"OO lO

l^

IN • •

CD

C^

"3

■*;;;!;

. . >< X X X

O O in LO CC CO

2:2: I AAA

0000
OCOCO CD

01

ol

lU

(1)

0

>

i*

.a

fl<

T)

s

s

g

"^ i & ^

o o o o
2; O O O

OiCOIN (NffiCO

O CD i-t "* Oi CO

>

00O5 05 -^t^ lO

^^^l^^^^

o

rt<TjH^tD^O

rt< in 05 lOOO lO

-#^C0003 05

rH^rt C^r-lrt

■* ^ 1-1 C<l t^ CO

IM t-- Tt< (N Tt CO

oooooo

Q

41-H^^fl^

M

o --1 Tf C! CO t>

C0IMCDOC5 1-H

rt rHr-<,-l

Tt< l> 03 OJ t- t^

cooincoOTfi

Or-OOOO

<

-H-H-H-H-H-H

t> CO CD CO (M O

iM 00 TtH in CO CD

lO

^

IN : : : :

rH

o

f_(

tH

in

.(M . . . .

o

o

o

lO

.■* .

05

o

:"= : --KM

>o

•O -.-10500

00

• 1-1 .

o

•-* -000005

00

lO

-i> -cocqo

■o

- T— 1 1— 1 1— )

w

o

-in -t~oOT-i

lO

• oqrt(N

O

in

-eg -COCON

ifl

-T)<!NiO

JS

bs

•IN CO 03 COO

<D

CD(M-#

^

Ih

• -ThCOOCO

*2

in

ooi-iin

o

t-^ CO CD 1-1

oj

rHrt rHTh

(h

1—1

o

in

CO 00 in Tf

■*

iHOii-llN

o

cqo-.coo

1^

■*

rtTt< i-<

<0

O

in

Tt<

int>i-i TtH

m

CO

1-1

o

00

• t^ . .

CO

(N

in

<N

-i-H • -

(M

IM

O

1>

cq

fCfeC^fe

Tj<T(<-*^

in in in in

5

Ot)H X ►< X ►<

inmooco

iAii

CO CD CD CO

n1

cS

(U

lU

(U

o

0)

o

>

is

s

S

S

^ ^ ^

o o o

Vh U Ih

O O O

PLATE XXII.

a. Fasciated ears. F2 generation of cross with normal showing domi-
nants and heterozvgotes.

b. F2 cob of heterozygote above; sample of F3 progeny below. From
left to right first six show the abnormality in different degrees. Last
two ears are normal.

Inheritance of Ear Fasciations.

o,

DWARF FORMS. ' 129

PART V.

PLANT ABNORMALITIES.

A few abnormalities have appeared in the maize varieties
under observation during the progress of these investigations.
They have been studied with two objects in view. The first
object was to see whether the manner of transmission of herit-
able monstrous characters gives any clue to the reason why
monstrosities have seldom obtained a foothold in nature when
in competition with normal types. The second object was
commercial. If teratological specimens appear in commercial
varieties of maize, it is desirable to know the easiest method to
destroy them.

Dwarf Forms.

The first dwarf form appeared in the 1908 culture of No. 6
Leaming dent. Thi^ strain had been selfed for the two previous
years without producing dwarfs. In the third generation, how-
ever, in a culture of 100 plants 5 dwarfs appeared. The plants
were normal in appearance, having all parts correlated as in the
full sized plants, as is shown in Plate XXV. They were from
two to three feet in height and contrasted strangely with the
other plants of the variety which were from nine to eleven
feet in height. The female flowers seemed to be normal. At
least cobs were formed and silks appeared. The pollen however
was completely sterile. The dwarf plants were pollinated
first with their own pollen and when no seeds formed were
pollinated with pollen from normal-sized plants. A few seeds
formed on two ears, which were planted the next season. From
one ear which had been borne on a plant eighteen inches high
only two plants resulted, one being a dwarf and the other
of normal height. From the other ear which came from a plant
three feet six inches high, seventeen individuals resulted, one of
which was a dwarf. The dwarfs, as in the former year, had a
normal correlation of parts. The leaves were opposite and the
ear appeared in the axil of the sixth leaf from the top as in the

130 INHERITANCE IN MAIZE.

normal plants. The pollen appeared to contain some normal
grains this year and both of the plants were selfed. No seed
set, however, and when pollinated with pollen from normal
plants it was found that the silks had passed the receptive
stage. This delay lost the strain. Seeds from the old ear
of No. 6 had again been planted and had given two dwarfs out
of sixty plants, but these had been lost in the same manner.

No. 69-5 a flint with a mosaic red pericarp also gave similar
dwarfs with a ratio of 48 normal to 14 dwarf plants. The ear
from which they came was a selfed ear from a commercial
strain obtained the year before. The commercial strain had
given no dwarfs but as only about 100 plants had been grown
it is uncertain whether or not they had ever appeared before.

A different kind of dwarf plant appeared in a commercial
strain of Stowell's Evergreen sugar corn in 1908. It was very
short (18 inches) and had short leaves of the normal breadth.
The joints were very close together and the whole appearance
of the plant suggested a normal plant that had been pushed
together like a telescope. An attempt to self this plant failed,
but four days afterward it was pollinated with pollen from a
normal strain of Stowell's Evergreen. A fairly good ear resulted
which was planted in 1909. One dwarf like the maternal
parent appeared out of thirty-seven plants. It was completely
sterile, but a selfed normal plant from the same lot gave two
dwarfs out of seventy-six plants in 1910. (See Plate XXIV.)

It is a matter of conjecture what occurred in these cases. In
the first instance, at least, controlled cultures that had produced
no dwarf plants, suddenly threw dwarfs. It was a much more
definite occurrence than De Vries' Oenothera mutations for
these were mutating when De Vries found them. If the normal
type were completely dominant, one must conclude that one
seed had been selfed in the case where the dwarf was pollinated
with pollen from normal Stowell's Evergreen. In the other
two cases the cross-pollination was made with pollen from
plants of the same strain, and as only a small number of individ-
uals were produced in the next generation, production of dwarfs
was probably continued through the pollen gametes.

The variation was transmitted by plants normal in character,
and whether one believes it to be a case of Mendelian dominance
of normals or not, there was nevertheless definite segregation.

REGULARITY OF ROWS OF SEEDS. 131

The fact that segregates appeared in ratios of less than one
abnormal to three normal, may have been due to any one of
several causes. Abnormal zygotes may have been formed and
not have been able to develop, for the germinating power of the
seeds formed on the dwarf plants was very low. On the other
hand, it may be that this result was due to the same fact that
probably gives rise to higher ratios in crosses that have been
studied thoroughly; namely, more than one chromosome
possesses the necessary material for normal height. There is
also the possibility that many abnormalities, particularly those
which show great latitude in their development, are due not to
regular Mendelian segregation, but to some abnormal chromo-
some reduction. If some reductions took place normally and
some abnormally through some disturbance of the plant's
normal physiology, abnormal and normal plants might be pro-
duced without definite and constant ratios.

Regularity of Rows oj Seeds on Cob.

The great majority of maize ears have rows of seeds running
in straight regular rows from butt to tip. Sometimes two rows
or even four rows may be dropped in going from the butt to
the tip but even then a sufficient amount of regularity exists
to call them straight-rowed ears. A varying percentage of
ears in each variety, however, have the rows quite irregular,
— the seeds often being squeezed together in such a hit and
miss manner that the number of rows can only be counted
by making cross sections of the cob. Experience with maize
cultures shows that there are two distinct kinds of irregularity,
one a physiological fluctuation which is not inherited, and one
a definitely inherited character or possibly a set of characters.
The non-inherited fluctuations are always present while the
inherited irregularity may be present or absent. Th@* latter
kind has been isolated in several varieties, the most conspicuous
being the Country Gentleman sugar corn.

Since the inherited irregularity can only be distinguished from
the fluctuation by breeding and then with difficulty owing to the
obscuring effect of the latter, it is difficult to come to any con-
clusion regarding the method of its transmission when dealing
with mixed strains. It could undoubtedly be determined by

132 INHERITANCE IN MAIZE.

careful work with a cross of which Country Gentleman formed
one of the parents. We have not made such a cross, but obser-
vations of large commercial cultures of Country Gentleman
lead us to believe that irregularity is a Mendelian dominant,
although it may not act as a simple mono-hybrid.

Ears with irregular rows appearing in our cultures have been
planted several times, but have proved to have been due ta
physiological fluctuation in all but one instance. An ear of
strain 29-2 produced some ears with irregular rows, one of
which happened to have been inbred. This ear gave 33 normal
progeny and 12 with irregular rows in the next generation.
One of these irregular ears gave 33 normal and 15 irregular
ears in a further generation, while one of the regular rowed
ears gave 125 normal and 5 irregular ears. One of these 5
irregular ears was selfed and will be tested next year. This
is about the percentage of irregular ears that the variety gives
in the commercial field, however, so the idea suggests itself,
that these five ears were fluctuations. If we regard this as the
true interpretation of the regular ears giving irregular ears,
and reduce the number of irregularities in the progeny of the
irregular ears in the same proportion, a ratio of 66 normal to 23
irregular ears is obtained. This looks like a case of mono-
hybridism with reversed dominance. It is suggested, however,
if this is a case of twice planting a heterozygous mono-hybrid;
that it is an example of fluctuating dominance in which some
apparently normal ears are really heterozygotes. One cannot
even say that only homozygotes show dominance, for it was an
irregular ear in each case that threw normals. There is no
a priori reason why this hypothesis should not be true, but it
seems probable that a more complex set of conditions exists.
The one fact that stands out clearly is that if the percentage
of irregular ears increases much over four percent in a com-
mercial progeny row culture, the whole culture must be dis-
carded to eliminate the undesirable "blood."

Bifurcated Ears.

Occasionally there is found among the eight rowed flint corns
ears which have only four rows. Their cobs are grooved so that
they appear to be almost splitting. One of these individuals

EARS WITH LATERAL BRANCHES. 133

appeared in a culture of No. 17 (Palmer's Red-nosed yellow) that
had been selfed for three generations. It was grown with the
special object of finding out whether the four rowed condition
is a final recessive condition as to number of rows. This proved
not to be the case. The condition is a secondary effect of a
heritable abnormality which causes the cob to show various
conditions of splitting into two rowed sections at the base.
The variations in this feature are shown in Plate XXIII, fig. a.
From this ear, 34 ears abnormal in varying degrees and 12 normal
ears were obtained. This ratio suggests the progeny of an ear
heterozygous for presence and absence of the abnormality.
It will be tested further.

A bifurcation of a different kind appeared in the progeny of
No. 7 Leaming dent that had been selfed for four years. In
its extreme form the tip of the growing ear becomes monstrously
fasciate; but it may vary toward the normal to such a degree
that the abnormality is shown only as a slight flattening of the
ear when observed in cross-section. The ear in which this
abnormality appeared was only slightly flattened ; its progeny,
however, showed 11 with divided tip and about 20 flattened
ears out of 44. (See Plate X XII.)

The normal-eared grand parent of No. 7 had been crossed with
No. 19, and from an extracted starchy ear of the F2 generation
there resulted the same abnormality. This ear, No. (19 x 7)-5-7
had a divided fasciate tip. It produced 29 ears with divided
tip, 33 ears abnormally flattened and 23 normal ears, — a ratio
of 62: 23. The illustration of this sort of fasciation shown in
Plate XXII, fig. b, gives an idea of how gradually the abnormal
ears intergrade with the normal ears. Yet this is a dominant
character alternatively inherited. It is difficult to tell the pure
normal ears by inspection but they appear to breed true when
isolated.

Ears with Lateral Branches.

An illustration of an ear with lateral branches which is prob-
ably nearer the ancestral type of maize appeared in the original
culture of No. 17. It is figured in Plate X XIII, fig. b. The ear
was not hand pollinated and of course no conclusion can be
drawn from the ratio in which the abnormality appeared in

134 INHERITANCE IN MAIZE.

the next generation. As a matter of fact 4 ears out of 25
progeny were so affected. One of these happened to have been
selfed, but it produced only a few seeds. Ten plants resulted
from this poor individual, two of which were abnormal.

The only valid conclusion from these data is that the
character does segregate. Normals and abnormals are produced ;
which fact suggests — as stated earlier in the paper — that the loss
of the lateral branching character of maize occurred as a retro-
gressive mutation.

Plants with Striped Leaves.

Zea mays japonica is a race which produces leaves with
longitudinal stripes with and without chlorophyll formation.
In other words, the leaves are green with white stripes. Several
races of this kind exist where the striping is apparently homozy-
gous and the race breeds true. What experience we have had
with striped races has been with another type of striping. The
phenomenon has appeared several times in our cultures, and is
clearly the same thing that Baur ( : 09) obtained in pelargoniums.
The full green type is dominant, the striped type is hetero-
zygous, while the homozygous recessives are sometimes formed
but cannot live because they lack assimilating organs. Crosses
between the striped plants and normal green plants always
gave all green progeny. Planting, in two cases, from plants
that were striped when very young, 274 normal and 27 striped
plants were obtained. This result might seem to indicate a
more complex condition than Baur obtained. It is not neces-
sarily so, however, for the plants were first examined for striping
when about 18 inches high. This may have been too late to
give them the proper classification, since it was found that many
of the 27 striped plants became greener as they aged. Several
plants without chlorophyll died when only a few inches high.
These were probably homozygous recessives.

Hermaphrodite Flowers.

Perhaps it should be mentioned in passing that the immature
sex organs, so called, of maize seem endowed with the power of
becoming either stamens or carpels. One often finds a normal

HERMAPHRODITE FLOWERS. 13S

ear ending in stamens, and nearly every plant produces lateral
branches which have carpels and stamens mixed together
indiscriminately .

A number of cases have also been observed where a few of the
ovules of an ear were surrounded by three stamens as in a perfect
flower. The only instance we have seen where all of the ovules
had three stamens within the glumes of the flower that is usually
aborted, was that of the dwarfs with wide leaves mentioned
under the heading "Dwarf forms." It might be supposed that
this was an atavistic type representing some of the characters,
at least, of the ancestral maize. We should prefer to believe,
however, that this development of stamens is merely an accom-
paniment of the dwarfing due to an endeavor to retain physio-
logic balance. That is, this type is really a healthy luxuriant
form producing very large ears for such a small plant. There
may have been developmental energy present which when
unable through inner limitations to produce a tall plant, mani-
fested itself in producing stamens.

Considered together, these various abnormalities present
several interesting features. It would be rash to make any
dogmatic statements in regard to their inheritance, yet it is
fair to say that if dominance shows progressive — and recessive-
ness retrogressive — ■ variations, both types are present. Some
of them are evidently simple in character — as far as inheritance
goes — while others are complex. It may be that the same
apparent type of variation will be found to be simple in some
races and complex in others. By this it is meant that both
3 : 1 and higher ratios will be found affecting characters which
to the eye are the same.

It does not seem probable that abnormal and degenerate types
are always or even commonly extracted recessives in which
absence of characters is concerned, as Davenport ( : 08) has
suggested. This statement has little basis from the data pre-
sented here, but the senior author has worked out certain
dominant abnormal types in the genus Nicotiana which adds
to our experience. The presumption is that they are more
often dominant like most of the abnormalities found in man.

Perhaps the fact of prime importance from these data is the

, variable dominance of characters and their obscuration by

physiological fluctuation. As stated once before, this shows

136 INHERITANCE IN MAIZE.

the extreme importance of pedigree cultures to the commercial
breeder, for some of these complex abnormalities cannot be
distinguished from normal plants by gross inspection. It is
possible that histological study might show points of difference
but these methods are not at the command of the commercial
grower.

It will be noticed that several monstrous variations occurred
in strains that had been selfed for several generations. The
effects of inbreeding in maize will form the subject matter of
another paper, but it might be well to suggest here a possible
cause for their production. Inbreeding in maize gives the same
effect as lack of nutrients, while cross-breeding gives the opposite
effect. There is retardation or acceleration of cell division,
respectively. Now such monstrosities as ears with divided
tips, occur more frequently either in cross-bred plants that
are over supplied with fertility, or in inbred plants. Perhaps
the first case represents fluctuation only, and is unlnherited;
as to this point we have no data. But disregarding this pos-
sibility, might not abnormal distribution of chromatin produce
these variations in both cases. The first kind could be caused
by abnormally accelerated division and the second kind b}-
abnormally retarded division.

General Conclusions.

The various points of genetic theor}^ discussed in this paper
are not sufficiently connected to make possible a short and at the
same time intelligible recapitulation. We simply desire to
mention our conclusions regarding the central problem of all
genetic investigations, that of laws of heredity.

When Mendel's Law of Heredity was rediscovered in 1900, it
was the general belief that it covered only a few isolated cases.
Many apparent exceptions were cited. One by one, however,
these exceptions have been found to yield to interpretation by
simple extensions of the Mendelian notation when fully under-
stood. In our experience as reported here, no exceptions to
Mendelian interpretation have been found. Such exceptions
may exist, yet it seems as unwise to say that Mendel's Law is
not general as to conclude at once that it can be made to cover
every possible case. One may say that Mendel's Law has

PLATE XXIII.

Jj#^i#*l«Si-.^*^*^

a. Heteroz\-gous bifurcated cob above; dominant and heterozygous pro-
geny below showing imperfect dominance.

b. Multiple ear. An imperfectly dominant character. Aboriginal maize
probabh' possessed a similar character.

Abnormalities.

PLATE XXIV.

^ t j: ^

A plant of the dwarf mutation appearing in Stowell's evergreen sugar
maize compared with a normal ear of the latter.

Dwarf Forms.

PLATE XXV.

A dwarf type which appeared in Learning dent maize compared with a
normal ear of that variety.

Dwarf Forms.

CONCLUSIONS. 137

covered so many cases that its generality is rendered highly
probable, although insufficient genetic investigation has been
accomplished to place it on equal terms with any of the great
laws of physics and chemistry. Yet some of the great laws
of chemistry were accepted when surrounded by seeming
exceptions. Some of these exceptions have been cleared up
by such recent advances as the Ionic Theory and the Phase
Rule; some still remain.

Is it not probable that other like generalities will be found
in biology, which, although they may entirely change our general
conception of the fundamental action of Mendel's Laws, will
nevertheless leave the facts upon which it was based as useful
and practicable as have been left the facts of chemical recombi-
nation in definite and multiple proportions in the light of the
Electron Theory ?

138 INHERITANCE IN MAIZE.

LITERATURE CITED.

Bateson, W. and Punnett, R. C.

: 08 Reports to the Evolution Committee 4 : 1-60.

Bateson, W.

: 09 Mendel's Principles of Heredity. Cambridge Uni-
versity Press, pp. 1-396.

Baur, E.

: 09 Das Wesen und die Erblichkeitsverhaltnisse der
"Varietates Albomarginatae Hort." von Pelargonium
Zonale. Zeitschrift fiir induktive Abstammungs-
und Vererbungslehre 1 : 330-351.

Correns, C.

'99 Untersuchungen iiber die Xenien bei Zea Mays.
Ber. d. Deutsch. Bot. Ges. 17 : 410-417.

: 00 G. Mendel's Regel iiber das Verhalten der Nach-
kommenschaft der Rassenbastarde. Ber. d. Deutsch.
Bot. Ges. 18 : 158.

: 01 Bastarde zwischen Maisrassen mit besonderer
Beriicksichtigung der Xenien. Bibliotheca Botanica.
53 : 1-161.

: 02 Scheinbare Ausnahmen von der Mendels'schen Spal-
tungsregel fur Bastarde. Ber. Deutsch. Bot. Ges. 20 :
159.

COURCHET.

'88 Recherches sur les chromoleucites. Ann. Sci. Nat.
Bot. Ser. 8. 7 : 263.

BIBLIOGRAPHY. 139

Davenport, C. B.

: 08 Degeneration, albinism and inbreeding. Science
N. S. 28 : 454, 455.

: 09 Inheritance of characteristics in domestic fowl.
Carnegie Institution of Washington Publication
121 : 1-100.

: 10 Imperfections of dominance and some of its conse-
quences. Amer. Nat. 44 : 129-135.

De Vries, H.

'99 Sur la fecondation hybride de I'albumen. Comptes
rend. I'Acad. Sci. 129 : 973-975.

: 00 Das Spaltungsgesetz der Bastarde. (Vorlaufige Mit-
theilung.) Ber. Deutsch. Bot. Ges. 18 : 83.

Dudley, P.

1724 An observation on Indian corn. Philosophical
Transactions 6 : 204-205.

East, E. M.

: 09 A note concerning inheritance in sweet corn. Science,
N. S. 29 : 465-467.

: 10 A Mendelian interpretation of variation that is
apparently continuous. Amer. Nat. 44 : 65-82.

Focke. W. O,

'81 Die Pfianzenmischlinge. Borntraeger. Berlin, pp. 569.

Gartner, C. F. •

'49 Versuche und Beobachtungen iiber die Bastarder-
zeugung im Pflanzenreich. Hering, Stuttgart, pp.
791.

GUIGNARD, L.

'99 Sur les antherozoides et la double copulation sexuelle
chez les vegetaux angiospermes. Rev. gener. d. Bot.
11 : 129-135. Also Comptes rend. I'Acad. Sci. 128 :
869.

140 INHERITANCE IN MAIZE.

Harshberger, J. W,

'93 Maize : a botanical and economic study. Con. Bot.

Lab. Uni. Pennsylvania 1 : 75-202.
'96 Fertile crosses of Teosinthe and Maize. Garden and

Forest. 9 : 522-523.

HiLDEBRAND, F.

'67 Unmittelbarer Einfiuss der Pflanzenbastardirung auf
die durch dieselbe erzeugte Frucht. Verh. des
naturh. Ver. d. Rheinl. u. Westf . 24 : 101 et seq.

'68 Einige Experimente und Beobachtungen iiber ....
(2) den directen Einfluss des f rem den Pollens auf die
Beschaffenheit der durch ihn erzeugten Frucht.
Botan. Zeitg. 26 : 325 et seq.

JOHANNSEN, W.

: 09 Elemente der exakten Erblichkeitslehre. Jena. Fischer
pp. 1-515.

Kellerman, W. a. and Swingle, W. T.

'89 Experiments in cross-fertilization of corn. Ann. Rep.
Kan. Exp. Station 1 : 316-337.

'90 Experiments in crossing varieties of corn. Ann. Rep.
Kan. Exp. Station 2 : 288-334.

'90a Crossed corn the second year. Ann. Rep. Kan.
Exp. Station 2 : 334-346.

'90b Bibliography of cross-fertilization of varieties of
com. Ann. Rep. Kan. Exp. Station 2 : 346-353.

'90c Preliminary study of the receptivity of corn silk.
Ann. Rep. Kan. Exp. Station 2 : 353-355.

'91 Crossed varieties of corn the third year. Kan. Exp.
Sta. Bull. 27 : 139-158.

Lamson-Scribner, E.

: 00 American Grasses — III. U. S. Dept. Agr., Div,
Agrostology Bull. 20 : 10-12.

BIBLIOGRAPHY. 141

Lock, R. H.

: 06 Studies in plant breeding in the tropics. Ill —
Annals of the Royal Botanic Gardens, Peradeniya.
3 : 95-184.

Mc Cluer, G. W.

'92 Corn crossing. 111. Agr. Exp. Sta. Bull. 21 : 82-101.

Montgomery, E. G.

: 06 What is an ear of corn? Pop. Sci. Mon. 68 : 55-62.

Morgan, T. H.

: 10 The chromosomes in heredity. Amer. Nat. 44 : 449-
496.

Morrow, G. E. and Gardner, F. D.

*92 Field experiments with corn. 111. Agr. Exp. Sta.
Bull. 25 : 173-203.

Naudin, Ch.

'63 Nouvelles recherches sur I'hybridite dans les veg6taux.
Nouvelles Archives du Museum. Tome 1.

Nawaschin, S.

'99 Resultate einer Revision des Befruchtungsvor gangs
bei Lilium Martagon und Fritillaria tenelle. Bull,
d. I'Acad. Imp. d. Sciences deSt. Petersbourg, 9 : Nr.4.

Nillson-Ehle, H.

: 09 Kreuzungsuntersuchungen an Hafer und Weizen,
Lunds Universitets Arsskrift, N. F. Afd. 2., Bd. 5,
Nr. 2. 1-122.

Puvis, A.

'37 De la degeneration et de I'extinction des varietes de
vegetaux propages par les greffes, boutures, tuber-
cules, etc., et de la creation des varietes nouvelles
par les croisemens et les semis. Paris.

142 INHERITANCE IN MAIZE.

Sargeret, E.

'26 Considerations sur la production des hybrides, des
variantes et des varietes en general, et sur celles de la
famille des Cucurbitacees en particulier. Ann. d.
Sci. Nat. T. 8.

Sturtevant, E. L.

'99 Varieties of corn, U. S. Dept. Agr., Office Exp. Sta.
Bull. 57 : 7-103.

DE ViLMORiN, Henri.

'67 Sur la fecondation du Mais. Bull. Soc. Botan. de
France. Tome 14 : 246.

Watson, S.

'91 Proc. Amer. Acad. Arts and Sci. 26 : 158.

Webber, H. J.

: 00 Xenia, or the immediate effect of pollen in Maize.
U. S. Dept. Agr. Division of Vegetable Physiology
and Pathology. Bull. 22 : 1-44.

Wheldale, M.

: 09 The colors and pigments of flowers, with special
reference to genetics. Proceedings Royal Society, B.
81 : 44-60.

: 09 On the nature of anthocyan. Proceedings Cam-
bridge Philosophical Society 15 : Pt. 2, 137-168.

: 10 Plant oxydases and the chemical inter-relationships
of color-varieties. Progressus Rei Botanicae 3 :
457-494.

Winkler, H.

: 09 Uber die Nachkommenschaft der Solanum-Propf-
bastarde und die Chromosomenzahlen ihrer Keim-
zellen. Zeitschr. f. Botanik 2 : 1-38.

5 38 8 7 i-

University of
Connecticut

Libraries

39153029221100