THE ANALYSIS OF CHARACTERS IN CORN 
 AND THEIR BEHAVIOR IN TRANSMISSION 
 
 W B. GERNERT 
 
THE ANALYSIS OF CHARACTERS IN CORN 
 AND THEIR BEHAVIOR IN TRANSMISSION 
 
 BY 
 
 WALTER BYRON GERNERT 
 
 'I 
 
 Champaign. Illinois 
 1912 
 
 Submitted (May IS, 1911) in partial fulfillment of the requirements for the degree of 
 Doctor of Philosophy in Agronomy, in the Graduate School of the University 
 of Illinois. 
 
 Published by the Author 
 
IN EXCHANGE 
 
 Univ of Illinois 
 
 JUL 22 I 
 
 h\. 
 
INDEX 
 
 Page 
 
 INTRODUCTION 5 
 
 Variation and Selection — 
 
 Environmental Influences. 6 
 
 Abnormalities — 
 
 Vigor of Growth 7 
 
 Inbreeding 8 
 
 1 [ybridization and Mendel's Law 9 
 
 Useful Correlations 11 
 
 Maize Groups — 
 
 TECHNIC 12 
 
 Recording r — 
 
 Pollenating — 
 
 Terminology 14 
 
 PLANT CHARACTERS 16 
 
 Germination — 
 
 Plantlets 18 
 
 Albinism — 
 
 Suckers 21 
 
 Leaves 23 
 
 Leaf Arrangem ent — 
 
 Leaf Shape 24 
 
 Leaf Texture 25 
 
 Disease Resistance — 
 
 Leaf Color 2'j 
 
 Green — 
 
 Red Tinge 
 
 Dark Red 2; 
 
 Variegated Foliage — 
 
 Pubescence 2S 
 
 Culm — 
 
 Height of Plants • 20 
 
 Internodes 32 
 
 Aerial Roots 34 
 
 Poriod of Growth — 
 
 Inflorescence 37 
 
 Tassels 3S 
 
 Structure — 
 
 Color 39 
 
 Male Flowers — 
 
 Anthers 4 ) 
 
 Size ■ • — 
 
 Color — 
 
 Sterility • • 41 
 
 Pollen 42 
 
 Viability 44 
 
 3 
 
Page 
 
 Shoots • 45 
 
 Barrenness 46 
 
 Husks : 47 
 
 Laminae 49 
 
 Silks 50 
 
 Position of Upper Shoot 51 
 
 Number of Shoots 52 
 
 REFERENCES 56 
 
 4 
 
INTRODUCTION. 
 
 I he work, of which this paper is a partial report, has in reality been a 
 search for characters in the corn plant ; how main and what are they ; how do 
 they behave in transmission? In the general plan of the investigation, it was 
 decided to begin with the germinating seed and continue through the life history 
 of the plant, going as far as time and facilities would permit.* The plan followed 
 was to study characters wherever they might be found, rather than to make a 
 group study of varieties or strains as a whole. The topics used as headings in 
 the report of the investigation are not necessarily characters of last analysis, but 
 may represent more or less complex bundles of genetic units. 
 
 To a very large extent, investigators in hybridization, before Mendel's law 
 was applied, saw their material in generalities. Mendel's far-sightedness in com- 
 paring individual characters only and not plants as a whole, together with his 
 careful watch upon the behavior of these individual characters through several 
 generations of the progeny, enabled him to secure the facts which led to the 
 discovery of this great law, overwhelming in its importance, which bears his 
 name. Following its rediscovery in 1900, methods in the study of heredity have 
 been revolutionized. Thus the era of mass selection in corn breeding is also 
 rapidly passing and the practice is to be superseded by more rational methods. 
 
 Some general principles of heredity, which have a direct bearing on the 
 subject in hand, will be considered briefly before entering into the discussion of 
 the characters themselves. 
 
 VARIATION AND SELECTION. 
 
 In a wind-pollenated plant like corn, we must expect great diversity unless the 
 parents have been inbred for a number of generations. In ordinary seed corn, 
 then, we have the opposite to what Johannsen has called a "pure line". "A pure line 
 may be defined as the descendants from one single homozygotic organism, exclu- 
 sively propogating by self-fertilization. 'Pure line' is merely a genealogical term, 
 indicating nothing as to the qualities of the individuals in question. A 'line' 
 ceases to be pure when hybridization (or even intercrossing) disturbs the continuity 
 of self-fertilization. "46) 
 
 We cannot conceive of variation being a haphazard phenomenon, but rather 
 that it is governed by definite laws. In such a study as this, the question naturally 
 arises as to the methods and causes of variations observed and the possibility of 
 controlling them. Anything gained in this direction, adds to our knowledge of 
 what we are dealing with in the selection of variations already existing. 
 
 Since we have a mixed ancestry in corn, there is much significance in the 
 suggestion, — that, in general, gains secured in selection in corn are not fluctuations 
 inherited, but rather the products from a mixture of types long existing and which 
 selection partially segregates. 
 
 Considerable confusion exists in literature in regard to the word "selection". 
 It is improbable that any thoughtful writer claims that the act of selection itself — of 
 either mutations or fluctuations — produces any genetic change in an organism, since 
 
 •The tabulation of the /•/an! characters consumed so much time that the writer was unable, 
 in the time allotted, to compile and report the data secured on the car and kernel characters. 
 
 In order to save expense in publishing, the original thesis has also been considerably abbre- 
 viated by the removal of the voluminous tabular matter and the discussion relating to it. This 
 paper, therefore, contains only a genera] discussion of some of the plant or "vegetative." char 
 acters studied in the investigation, of which this is only a partial report. 
 
tlie differences, which we see expressed in the soma or body, existed before the 
 selection was made, otherwise there would be nothing to select. The question to 
 be decided finally is not ; whether "selection" produces anything new ; but, are 
 small variations (so-called "continuous fluctuations") inherited? Are siYiall varia- 
 tions due to differences in environment and food only? May they not be of 
 greater significance in total differentiation than a limited number of mutations 
 ("discontinuous fluctuations") in the evolution of a race? Selection is the tool, 
 not the cause. 
 
 One writer has the following to say about selection : "Within pure lines, 
 if no mutations or other disturbances have been at work, or within a population in 
 which there is no genotypical difference, selection will have no hereditary influence." 
 This must be granted at once, since the clause "mutation or other disturbances" 
 excludes everything else. Neither is there anything left to work on if we allow 
 the term "small mutations", which is used by some writers to explain the results 
 gained by the use of fluctuations, granting that fluctuations are cumulative. 
 
 ENVIRONMENTAL INFLUENCES. 
 
 It has long been known that corn is very responsive to environment and in 
 order to get normal development, which the farmer calls "best results", the seed 
 should come from the region in which the crop is grown. Tests at the various 
 experiment stations have substantiated this principle. In investigations, such as 
 the present one. it is evident that environmental influences must be carefully 
 considered in the interpretation of results and especially upon corn which has not 
 come from "home grown" seed. 
 
 A number of varieties used in my study were imported from other states, 
 Mexico, and Central America. The behavior of these varieties including widely 
 different habits of growth, enabled me to compare adaptations to the environment 
 here in Illinois. Some of the varieties produced few or no viable seeds, because 
 of the short season to which they were poorly adapted. In some cases, pollen 
 only was secured and this was applied to silks of earlier maturing varieties. It 
 may be that this is the most successful way (hybridization) in which desirable 
 characters of varieties requiring a long growing season may be introduced into a 
 region having a short season. Collins 18) reported that first generation hybrids 
 are relatively free from the new place effects. 
 
 ABNORMALITIES. 
 
 Many abnormalities were found in this work, some of which are of particular 
 interest and which will be mentioned in connection with the characters to which 
 they are related. Blaringhem 9) described several new races which came from 
 mutilated plants of an eight-rowed yellow flint variety. He reports obtaining a 
 constant hermaphrodite form, another with white grains, and other differences in 
 plant structure. He states that abnormalities were much more constant in the 
 progeny of the mutilated plants than when found in plants not mutilated. Why 
 this should be, he does not explain, except in the cases that were close — or self — 
 fertilized. Sturtevant 80) mentioned a number of similar obnormalities, some being 
 secured by mutilation, but of which none were reproduced in one trial with open 
 pollenated seed. We have found all the abnormalities mentioned by Blaringhem 
 and Sturtevant, and many others during the two seasons in which our work was 
 in progress, and none of them, so far as known, were due to mechanical mutilations. 
 It is possible that the white strain which Blaringhem secured, may be the result 
 of segregation from the yellow variety as Shull has secured a yellow strain from 
 a mixed white variety by self-breeding. The corn plant undoubtedly furnishes a 
 rich field for investigation in the above mentioned class of study. 
 
 6 
 
VIGOR OF GROWTH. 
 
 In varieties that are thoroughly acclimated, the question of vigor is an important 
 factor, and one which may obscure the development of certain characters in both 
 pure and hybrid plants. One of the objects of the study was to get additional 
 information upon this much discussed topic. Vigor in corn has been investigated 
 from two standpoints ; wide-bred, and inbred populations. 
 
 The question was perhaps first studied from the wide-bred standpoint by 
 W. J. Beal 6), who concluded as follows: "To improve or infuse new vigor 
 into varieties — plant near each other and mix them — good results of such crossing 
 will last for several years, though most apparent the first year." Other experiment 
 stations were asked to take part in this investigation, but the reports were generally 
 indefinite. Beal's own results with yield tests were very satisfactory, as is shown 
 by the following proportionate yields which he obtained : 
 
 "From Crossed Seed." "Seed Xot Crossed." 
 
 1878 109.67 100 
 
 1881 121 100 
 
 So far as known. Beal's method of securing hybrid seed is the first reported 
 instance in which detasseling was put to any extensive use. This method was 
 used later by Burrill. Sturtevant, and others, and was finally adopted as a practical 
 method to prevent inbreeding. 
 
 Results by McCluer and Morrow and Gardner a decade later confirmed the 
 work of Professor Beal. McCluer 57) found that "The corn grown from the 
 crossed -ced was iii nearly till cases clearly increased in size [ears larger than those 
 of either parent] as a result of crossing — nearly all the corn grown a second year 
 from the crosses is smaller than that grown the first year, though most of it is 
 yet larger than the average size of the parent varieties." Morrow and Gardner 62) 
 noted that "In every instance the yield from the cross is greater than the average 
 from the parent varieties ; the average increase per acre from the five crosses being 
 9.5 bushels."* But this conclusion does not give the most important part of the 
 results as may be seen by a rearrangement of their data. 
 
 Yields in bushels of air dry corn per acre: 
 
 Increase over Increase over 
 Parent Average of 
 
 No. yields. parents. 
 
 1. A-64.2 B-61.6 62.9 
 
 2. A-64.2 C-79.2 71.7 
 
 3. D-73.6 E-65.1 69.3 
 
 4. F-60.6 D-73.6 67.1 
 
 5. A-64.2 G-58.4 61.3 
 
 Averages 66.5 
 
 Evidently nothing would be gained by making such hybrids as No's 1 and 2, 
 while such hybrids as No's 3 and 5 would be very profitable because the yields 
 were, in each case, much higher than that of either parent. What might have 
 been secured from other combinations of these same parents could be told only 
 by trial. Some of the other possible combinations might have given greater yields 
 than either of those tried, since there is evidence that some of the hybrids were 
 more favorable combinations than others. Reports of a later, similiar test by the 
 same authors 63) show somewhat less gains from hybridization than in the first 
 report. The system of checking field results, used by these early workers, was 
 imperfect and consequently the yields reported are not absolutely reliable. 
 
 Hartley 421 states that "in some instances, cross bred seed produces less than 
 the average of the strains crossed." Collins id) found similiar cases and recom- 
 mends that "we should lose no time in securing information regarding the amount 
 
 '(More exactly 9.2 bushels). 
 
 Yield of 
 
 average of 
 
 higher-yielc 
 
 Hybrid. 
 
 of 
 
 parents. 
 
 ing parent. 
 
 64.1 
 
 
 1.2 
 
 — .1 
 
 73- 1 
 
 
 1-4 
 
 —6.1 
 
 86.2 
 
 
 16.9 
 
 12.6 
 
 76.2 
 
 
 9-1 
 
 2.6 
 
 78.5 
 
 
 17.2 
 
 14 3 
 
 75-6 
 
 
 9.2 
 
 5-9 
 
of difference that should exist between the strains crossed to secure the maximum 
 increase of vigor — self-fertilization of corn inevitably leads to sterility." 
 
 Shull has investigated the question of vigor in connection with self-bred 
 strains of corn, originating from the same field mixture, somewhat upon the plan 
 of "Mr. Q. I. Simpson in breeding hogs by the combination of two strains which 
 are only at the highest quality in their first generation, thus making it necessary 
 to go back each year to the original combination, instead of selecting from among 
 the hybrid offspring the stock for continued breeding." Shull's results with 
 hybrids of self-bred strains were similiar to those of Beal, McCluer, and Morrow 
 and Gardner upon hybrids of open pollenated strains, and with which it seems lie 
 was not familiar at the time of writing his first two papers on this subject, although 
 in his first paper 72) Shull refers the reader to an article on inbreeding by Shamel 
 in which McCluer's work is mentioned and Morrow and Gardner's experiments 
 are discussed at some length. 
 
 Discussing the cause of increased vigor in hybrids, Shull 75) suggests in a 
 recent paper; "the degree of vigor is correlated with the number of characters in 
 respect to which the hybrids are heterozygous — heterozygosis in some elements may 
 be without effect upon vigor, or even depressing. The presence of unpaired genes 
 or the presence of unlike or unequal paired genes, was assumed [A. B. A. 1908] to 
 produce the greater functional activity upon which larger size and greater efficiency 
 depend — East suggests that this stimulation due to hybridity may be analogous to 
 that of ionization — A. P.. Bruce proposed that the degree of vigor depends upon 
 the number of dominant elements present rather than upon the number of heterozy- 
 gous elements." 
 
 Since the above propositions are nothing more than theories as yet, other 
 explanations might be considered. Thus, vigor (increased or decreased) might 
 be due to : 
 
 (a) — Xew size-producing factors introduced by hybridazation. 
 
 (b) — A new combination of latent size-controlling factors. 
 
 (c) — The readjustment necessary in changing from a more or less pure to a 
 complex heterozygous condition which expands or contracts size-pro- 
 ducing parts, being therefore a purely mechanical phenomenon similar to 
 the change in form and bulk in many chemical reactions. 
 In any of these explanations, segregation would have the usual scattering effect 
 upon size in the second hybrid generation. 
 
 Vgain, the theories as held at present in regard to increased vigor may be 
 the reverse of the real truth, and instead of a plant developing above the normal, 
 it may he that by means of favorable combinations the plant is only enabled to 
 develop more or less to its full capacity. A special stimulus may be required 
 to bring out what is perhaps only a small part of the possibilities within its makeup. 
 
 INBREEDING. 
 
 I lure seems to be a question lingering in the minds of many as to whether 
 very much self-pollenizatiofi takes place in corn. As evidence upon this point, the 
 following observation may be mentioned: An ear having a Targe percent of sweet 
 (wrinkled) kernels was found growing in a field of dent corn. The seed planted 
 was purchased, for pure Reid's Yellow Dent seed. The most probable explanation 
 is : that a grain of pollen from a white sweet corn was blown or carried to the 
 parent "Reid" ear, fertilizing one of its ovules, and because of the two recessive 
 characters — white and wrinkled — possessed by the pollen, the stray kernel was not 
 noticed in the seed since its external appearance was probably much the same as 
 that of the other kernels planted. 
 
 It is evident that the ear developing on this hybrid plant was very largely 
 self-pollenateil, since the lower three-fourths of the ear exhibited very fair di-hybrid 
 Mendelian proportions in regard to color and composition of the endosperm. That 
 
there may have been a difference in the blooming period of this plant and that of 
 the others surrounding it is recognized, but the upper part of the ear had nearly 
 all smooth kernels due to the effect of dent pollen only, which very probably had 
 come from the surrounding plants after its own tassel no longer furnished pollen. 
 Ears borne on isolated corn plants are generally reported to have very few 
 kernels 71). In the season of 1910 three isolated corn plants were found growing 
 far apart in a cemetery and each more or less surrounded by trees which served 
 as a wind-break. These plants probably grew from scattered grains in the manure 
 applied to the flower beds. A goodly number of kernels developed on each of 
 these ears, many more than had been expected from earlier observation on 
 isolated plants growing in gardens. 
 
 Although these plants were sheltered somewhat, the same conditions frequently 
 exist in an ordinary field of corn, and on a day when there is not much wind 
 a very large amount of self-fertilization must take place. We have frequently 
 obtained perfectly filled, selfed ears by one application of pollen in hand-pollerjating 
 work. Such observations prove conclusively the desirability of detasseling corn 
 plants from which seed is to be obtained for general field planting. 
 
 Collins 17) has suggested that — "The production of more than one ear on each 
 stalk — would to some extent, correct the tendency to self-pollination, for in prac- 
 tically all cases, the second ear must he cross-pollinated. In regions where high 
 winds prevail at the time of flowering, the percentage of self-fertilized grains 
 would be further reduced." This conclusion was made in connection with a 
 variety in which the anthers made their appearance considerably in advance of 
 the silks. Lazenby 5-'). however, in what were apparently quite conclusive observa- 
 tions, stated, that "almost without exception the silk appears earlier in comparison 
 with the maturity of the anthers, when there is more than one ear on a stalk." 
 
 In our own notes taken daily in 1909, we learned that in some of the varieties 
 producing more than one ear on a plant, the silks on the second and lower shoots 
 appeared very soon after, and in some cases, preceded those on the upper shoot 
 (bagging the upper shoot and leaving the second one exposed frequently favored 
 this apparently abnormal behavior). In general, however, the tendency to mix- 
 pollenate is greatly increased on the lower shoots. This may lie the principal 
 reason why seed from nubbins, which often represent ears delayed in development, 
 may produce as good or a better crop the next season than seed produced on large, 
 well developed ears which are more likely to be partially self-pollenated. 
 
 In 1905 Shamel 71 ) reported some observations upon the effect of self-fertiliza- 
 tion in corn; "it w-as found that in four generations of continuous self-fertilization, 
 the vitality had become so weakened that the seed failed to germinate." That 
 Shamel used a strain with an inherent weakness, or that there was some fault in 
 his test, seems to be shown by the results which Shull has been securing on corn 
 selfbred for a number of years. Shull 74 I concludes that "the decrease in size and 
 vigor, which accompanies self-fertilization is greatest in the first generation, and 
 becomes less and less in each succeeding generation until a condition is reached 
 in which there is (presumably) no more loss of vigor." 
 
 Experiments by East 26) corroborate the conclusions of Shull. East asserts 
 that in a number of families inbred two to four generations; "In all characters 
 of stalks, leaves, roots, male infloresence, and mature seed, the plants were normal. 
 It is merely in the matter of size of plant and ear. and thereby yield, that the plant 
 compares unfavorably with cross-bred plants. Further, there is n 1 continuous 
 decline in yield." 
 
 HYBRIDIZATION AND MENDEL'S I.AW. 
 
 It is essential that the student of heredity have an accurate knowledge of the 
 method of reproduction of the organism in which he is interested. There is no 
 
 9 
 
better way to gain this information than through hybridization, and the final test of 
 a character is to hybridize it. In our study, the work was planned so that certain 
 characters might meet many others in the various hybrid combinations, and that 
 their behavior upon close — and self — breeding might also be determined. 
 
 Because of the mass of Mendelian literature, in which the general principles 
 of behavior in combination and segregation have been reviewed again and again, 
 there is no need of their discussion at this place. Since Mendel's Law was redis- 
 covered in connection with a corn hybrid and because of other earlier work upon' 
 corn, which has a close relation to this topic, it is worth while to mention some 
 incidents which are of especial significance. 
 
 The history of events leading to the rediscovery of Mendel's Law is of 
 absorbing interest and will always stand as a classic example of what patient, 
 thorough search and careful consideration of a vast array of facts and theories 
 accumulated from the work of others may accomplish in throwing light upon a 
 problem in research. 
 
 In a bibliography attached to a paper on "Cross-Breeding and Hybridizing" 
 written in 1892. L. H. Bailey included the reference, — (1865 Mendel, G. Versuche 
 uber Pflanzen-Hybriden. Brunn Verhandl iv.-3.47.) 
 
 Professor Bailey had taken this reference from Focke's book "Die Pflanzenmisch- 
 linge" published in 1881, but had not, himself, seen the paper. 
 
 The following is the account in DeVries' own words in a letter to Bailey : — 
 "Many years ago you had the kindness to send me your article on Cross-Breeding 
 and Hybridization of 189.2; and I hope it will interest you to know that it was 
 by means of your bibliography therein that I learnt some years afterwards of the 
 existence of Mendel's papers, which now are coming to so high credit. Without 
 your aid, I fear I should not have found them at all." 
 
 DeVries had already produced hybrids of sweet and starchy corn in studying 
 the phenomenon of double fertilization in corn. He carried these hybrids through 
 the second generation allowing them to pollenate naturally and reported what we 
 now call "Mendelian segregation" without at first appreciating the significance of 
 it; "Tons ces epis [twenty-five] etaient de nature mixte. Environ tin quart des 
 graines etaient sucress, les trois autres quarts etaient amylacess" 81). This was 
 at the time that DeVries was making a thorough search of literature on evolution 
 in preparation for his "Mutations-theorie". a work containing a wealth of references 
 which the author used to such good advantage in establishing this theory. The 
 discovery of the significance of Mendel's paper and its announcement occurred 
 in the year following (1900). 
 
 In looking through the available literature upon corn hybridization, other 
 similiar cases are to be found in which Mendelian proportions were approximated,. 
 but not being analyzed with due care were consequently disregarded. Ten years, 
 previous to DeVries' publication. Kellerman and Swingle 48) noted some results 
 as follows: "The ear consists of about one-fourth sweet corn and the remaining 
 kernels are more or less dented at the summit" — "in the large proportion (prob- 
 ably -]<4 ) of yellow" — "probably seven-eighths of the kernels were flint" and 
 in a later publication 49) — "By actual count, there were 370 kernels on the ear. 
 Of these jo6 were blue, 71 pink, 71 orange and yellow, and 22 pure white." 
 Although they did not realize the meaning of these behaviors, they had recorded 
 very fair Mendelian results. It will be noticed that the last quotation contains 
 a very good di-hybrid ratio in which the theoretical expectation would have been 
 approximately 213:71 :ji .24. 
 
 Quoting from McCluer 57) ; "The self-fertilized ears [of a hybrid] shown! 
 the same modification of kernels as those naturally fertilized, proving that each 
 kernel of the crossed corn had in itself the power to produce both sweet and 
 dent corn— the progeny had tended strongly to run back to the parent forms, while 
 at the same time taking on other forms different from either." 
 
 10 
 
In this way the ear of corn had provided most desirable material for the 
 deduction of a general law of inheritance, but was awaiting the insight of a second 
 Gregor Mendel to be made use of. Corn did finally serve for the first corroboration 
 of Mendel's Law in connection with the first announcement of its rediscovery in 
 1900 by Hugo DeVries. 
 
 USEFUL CORRELATIONS. 
 
 The Ohio. Nebraska, and Maine Experiment Stations have been making yield 
 tests in selections of certain ear characteristics and have reported but slight 
 differences in results. Davenport, Rietz. and Smith have made use of the statistical 
 method in the study of correlation of characteristics of the ear in a mixed popula- 
 tion, and find indications of considerable correlation in some instances. Ewing 31 1 
 has recently investigated the correlation of weight of grain with characteristics 
 before or at the beginning of the blooming period, and concluded that aside from 
 the genetic coupling of unit characters, the correlation in the fluctuating variability 
 of two different characteristics is not likely to prove of much assistance to tin- 
 breeder. 
 
 The perfection which Xilsson and his co-workers in Sweden have attained 
 in utilizing character of plant detail in agricultural species has proven highly 
 profitable for economic purposes and has also served as a fruitful source of 
 knowledge in the theory of heredity. Aside from the general use of certain 
 features, such as wrinkled endosperm for table use. large size of plant with relative 
 large amount of ear corn for ensilage purposes, and other similar cases, not a 
 great many distinguishing characters have been made use of in connection with 
 selection for desirable features in corn. The mechanical method used by Hopkins 
 and others in selecting for differences in protein and oil content in the corn 
 kernel is an example of this kind. 
 
 It would be desirable to know in a similiar way the possible relations between 
 certain characters which we are beginning to recognize in corn and to have a rather 
 comprehensive list of such characters or genetic units with complete information 
 upon the behavior of each, in order that we may realize more fully the possibilities 
 in the most important American crop. It was with no small degree of satisfaction 
 that this study of the corn plant was undertaken with the knowledge that much 
 that might be learned would be of immediate economic value. 
 
 MAIZE GROUPS. 
 
 We will perhaps never have positive evidence as to the exact origin of the six 
 generally accepted groups of corn, five of which are distinguished by their 
 endosperm characters. We know that the Indians were cultivating all of them 
 when America was discovered and we can only speculate as to their sourer 
 Quoting from Sturtevant 80) ; "It seems almost certain that in the order of evolu- 
 tion (excluding from consideration the puzzling sweet-corn group) progress has 
 been from the pops, through the flints and the dents to the softs." East 25) has 
 suggested that sweet corn "may have come about through mutation in each of 
 these groups [dent and flint], but from what we know of the early sweet corns. 
 it is more likely that the change took place among the flint types and was extended 
 by hybridization." 
 
 It scents, however, that there is a much more probable source of sweet corn 
 than the one suggested by East. It is well known that there is perhaps no variety 
 of sweet corn grown that is entirely free from soft starch, aside from the very 
 considerable amount of corneous starch which is always found in all mature sweet 
 corn kernels. There are also varieties of sweet corn which have a semi-starchy 
 endosperm, and Halsted 37) has found that the proportion of starchy endosperm 
 
 11 
 
may be increased by selection in hybrid sweet corn where one of the parents 
 possessed a semi-starchy endosperm. Stnrtevant classified samples of such semi- 
 starchy corn received from the Zuni Indians under the name of "amyleasaccharata". 
 As East has also mentioned ; many of our varieties of sweet corn have been pro- 
 duced by the selection of segregates from hybrids of sweet with flint, dent, and pop- 
 corn and this has resulted in wide differences in kernel shapes, ear and stalk 
 characters. 
 
 East's conclusions were based largely on results secured in sweet corn hybrids 
 with flint and dent corns, the new combinations suggesting flint or dent origin, 
 depending on the variety of sweet corn used in the hybrid. In our own observa- 
 tions in 1910, we find that it is possible to get results similiar to those secured by 
 East when we use soft corns to hybridize with the various sweet varieties. Soft 
 corns are quite sweet at the roasting ear stage, and were used by the early settlers 
 for eating from the cob and are still so used to a large extent. The first sweet 
 corn varieties secured from the Indians bore short, eight to twelve rowed ears 
 with broad shallow kernels ; these are common characters in soft or "flour" corns, 
 to which considerable evidence points as the source of our sweet or "sugar" corn. 
 
 TECHNIC. 
 
 In this study, the arrangement and groups recognized by Stnrtevant in his 
 bulletin on "Varieties of Corn" 801 were taken arbitrarily for the purpose of 
 classifying and numbering the varieties used. 
 
 RECORDING. 
 
 A method of numbering somewhat on the plan of the Dewey library system 
 was employed. The six groups were given numbers 100 to 600 and varieties within 
 the groups designated by units. We illustrate the system as follows : — White Pearl 
 Pop was the tenth variety in the list of pop corns (200) and it was given the 
 number "210". It happened that this variety was planted on row 13 in 10x39, so in 
 that season the stake and labels on this row bore the field number 210.13. A plant 
 in this row was pollenated by a dent variety grown on row 33, the name and number 
 of which was "Fairview Speckled Dent-404". The hybrid note would then read 
 (210.13-2X404.33-14) and it would also include a description of the work done. 
 The figures following the dash ( — 2 and — 14) indicate the particular plants used 
 on the two rows in making this hybrid. 
 
 When the above hybrid was planted in 1910, all of the figures except those 
 indicating the groups 200 and 400 were dropped and as this particular hybrid was 
 .planted in row No. 162 in 1910, its field label for that season bore the number 
 24.162. This was as simple for daily use as the index of the first season and gave 
 all the information necessary for field work. Its previous history could readily 
 be found in the record book of the year before. A reciprocal (404.33-14X210.39-20) 
 of this hybrid was grown on row No. 238 in 1910. consequently its label bore the 
 number 42.238, indicating dentXpop and grown on row 238. 
 
 In igog the blooming notes were taken daily and for each individual plant, but 
 in 1910 these notes and also those pertaining to the husks were taken every four 
 days, and only for the row as a whole. The observations on size characters and 
 number of parts on the plant were made a short while before maturity, and the 
 observations on the ears were all taken in the field at harvesting time, the study 
 of kernel characters being left for the laboratory. In general, such characters 
 were chosen which could be recorded readily. Many interesting and desirable 
 features were neglected entirely for lack of time. 
 
 POLLENATING. 
 
 Some methods of technic were developed during the experience of the two 
 seasons, a knowledge of which would have been of much value in the beginning. 
 
 12 
 
As some of these may he of use to others contemplating similiar work, brief 
 mention will be made of a few of them. 
 
 The record books should always provide room for special or general notes 
 upon an individual or row in addition to the regular scheme. The labels used 
 for marking plants should be such as will not be destroyed by wind and rain. 
 Cloth, wood, or metal labels with a wire tie are indispensable. 
 
 In a locality where storms are apt to occur when the corn is in bloom, tough 
 parchment paper bags are necessary to protect the shoots, but ordinary manila 
 bags will serve to collect pollen. Parchment paper is desirable because of its 
 semi-transparency, which allows observations on the color and especially on the 
 development of the silk without removing the bag. This is very convenient at 
 pollenating time. The most satisfactory and expeditious method of bagging shoots 
 was that of bending down or preferably breaking off the leaf blade in the sheath 
 of which the shoot to be pollenated is developing. The inverted bag with one 
 side between the shoot and the culm is then pulled down over the shoot with a 
 sawing motion till there is room to fasten the bag to the leaf sheath, or to the 
 shoot about which it is folded, with an ordinary brass pin. A flexible wire may 
 be used to split the leaf sheath in case the bag is not strong enough to accomplish 
 this with ease. The bagging of shoots must be done, of course, before there is 
 any indication of a silk and the bag should be left on ten to fifteen days after 
 the silk has been pollenated. In a moist climate these shoot bags should be 
 removed when all danger of contamination is past, since the developing ears tend to 
 become moulded when the bags are left on. 
 
 The bags for collecting pollen were put on early in the morning of the clay 
 in which the pollen was to be used, preferably before the pollen begins to fall, 
 because there is here a possibility of collecting stray pollen grains which have lodged 
 on the tassel from surrounding plants. In a dry period, this may be quite effectively 
 guarded against by bagging the tassel on the day preceding, as, in the majority of 
 cases, corn pollen is no longer viable in from 24 to 36 hours after it has been 
 shed. For this reason best results are obtained in pollenating work when pollen 
 is used within a few hours or during the same day on which it has been gathered. 
 
 The inverted pollen bag is readily fastened about the base of the tassel by 
 means of a spring clothes-pin, which is a very handy article in general plant- 
 breeding work. The bag is fastened in such a manner that one side is lower 
 than the opening, thus providing a pocket in which the pollen may lodge instead 
 of sifting out through the crevices ;.t the mouth of the bag. In this way sufficient 
 pollen may lie secured at one time from one tassel to pollenate a number 
 of shoots. The pollen may lie used at any time of the day and in all kinds of 
 weather, so long as the pollen and silks are kept reasonably dry. Studies made 
 on the behavior of pollen and silks will be discussed in connection with the in- 
 florescence. 
 
 In case the silk came in contact with the hands or apparatus in the process of 
 pollenating it was carefully clipped with scissors. In general and except for 
 special study, all the silks were clipped at a short distance above the shoot as the 
 bag was being removed from it. When a shoot is young it may lie clipped below 
 the tip and opened, thus exposing a larger number of silks. Clipping of silks was 
 done not only for ease in operation, but also to insure thorough application of the 
 pollen; clipping also prevents heating of the mass after the pollen has been applied. 
 When long and tangled sii*ks are pollenated barren patches on the cob are fre- 
 quent because of the pollen failing to reach the silks on the inner or lower side 
 and sometimes due to the heating mentioned. The pollen is poured directly from 
 the hag and thus the silk and the pollen are never touched by the hands, except 
 by accident. There is therefore no need of washing the hands in alcohol after each 
 
 13 
 
pollenation, as some workers recommend. The scissors used in clipping, and the 
 hands of the operator as well, may lie cleaned with the silks which are cut away. 
 Some workers expose the shoots while pollenating them; others use an umbrella 
 with or without a side curtain for protection from falling pollen, the umbrella 
 having an extra covering to make it pollen proof; for extensive work, this re- 
 quires an extra man to hold the apparatus and is, at best, not very satisfactory. 
 During the first season, a pollenating apparatus was used which was cylindrical in 
 shape, mounted on a pedestal, covered with oilcloth, and large enough to admit 
 the operator. A slit was made in the curtain to admit the lower part of the 
 plant including the bagged shoot. On warm days, continuous work in this cloth 
 cage was very trying and a new apparatus was constructed for use in the 
 second season. 
 
 The new apparatus consisted of an inverted box of very light wood, which 
 had been carefully seasoned to prevent checking, all the joins being sealed with 
 glue. A short curtain was attached about the base to exclude falling pollen. The 
 box was large enough to provide room for the manipulation of the pollen and silk, 
 all but the arms of the operator being on the outside. The side of the curtain, 
 inserted between the shoot and the culm, was provided with a metal shield having 
 a narrow "A" shaped opening and sewed into the heavy duck curtain. In the box 
 and just above this opening was inserted a pane of glass to provide light. Glass 
 panes were also placed in the side opposite and in the top of the box. In this 
 way the operator is enabled to view his work from a number of directions, accord- 
 ing to the height of the shoot. The box was readily adjustable to any height 
 on an upright rod mounted on a "T" base with recurved sharp points to anchor it 
 in the soil and to allow it to be readily removed. These points were long enough 
 to permit straddling the ridge usually thrown up along the corn row in cultivation. 
 
 In order to secure pollen from an early blooming variety to use on a late 
 blooming variety, it was necessary to make several plantings. It was found, how- 
 ever, that the late plantings matured much more quickly, relatively, than the earlier 
 planted rows. It was frequently necessary to use the last tassels to mature on the 
 early varieties and these were found on poorly developed plants or on suckers. 
 This practice introduces difficulties in the interpretation of the results. We may 
 have thus selected genetic differences in sucker devlopment. but otherwise the 
 use of pollen from suckers would not be expected (of itself) to seriously affect 
 results. Hartley 41 ) states that progeny from ears fertilized with pollen from 
 suckers were as productive as plants from ears fertilized with pollen from the 
 main culm. In general, only one application of pollen to each shoot was made. 
 The practice of pollenating the same shoot several times is not only time consuming, 
 but also allows more opportunities for mixture by stray pollen grains. 
 
 East 27) has experienced difficulty in getting certain varieties to hybridize and 
 has concluded that there are some varieties which will not combine with others. 
 Similar experiences have been reported by early investigators. Collins 18) and 
 Emmerson ^9) have later secured successful hybrids with strains similar to those 
 used by East. In my own work, wherever I was sure of the viability of both 
 pollen and silk, I have found no difficulty in securing hybrids between varieties 
 differing widely in regard to size, period of growth, and many other characteristics. 
 
 TERMINOLOGY. 
 
 Confusion exists in literature regarding the meaning of the words "hybrid", 
 "cross", and "inbred". At present the terms "hybrid" and "cross" are used inter- 
 changeably, some writers using both words in the same sentence in referring to 
 the same individuals. While these superfluities of the language are permitted; 
 they are confusing, especially to the uninitiated. In this work some distinction was 
 necessary and the term-, will be used throughout the report as defined below; 
 
 M 
 
a 
 
 cross 
 
 to cross 
 
 a 
 
 cli ise 
 
 to close 
 
 a 
 
 self 
 
 to self 
 
 a 
 
 mix 
 
 to mix 
 
 Substantive. Infinitive. Definition. 
 
 a hybrid to hybrid to interpollenate individuals of distinct varieties or 
 species (so called ). 
 
 to interpollenate individuals of strains or "lines" of the 
 same variety, 
 to close to interpollenate individuals springing from the same 
 flower or mother plant. 
 
 to pollenate with pollen from the same flower — or other 
 flower — on the same plant. 
 
 to pollenate with pollen from either an unknown or hetro- 
 genous source (this applies to individuals pollenated 
 naturally in mixed populations). 
 
 The above terms are in common use and the distinction here made is no 
 radical departure. Although the ardent supporters of the "genotype" ideals might 
 maintain that there is no difference between hybrids and crosses, some distinction 
 should be made between a sweet cornXrice pop hybrid, as an instance, and a cross 
 between strains of slight differentiation coming from a well established variety- 
 like Reid's Yellow Dent or Boon County White corn. The general term "wide 
 bred" will be applied to hybrids and crosses; the general term "inbred", which 
 has been confused in a number of papers with "selfbred", will be applied to closes 
 and selfs. 
 
 The sign "X" is to be interpreted as it is commonly used; meaning, "pollenated 
 with", when inserted between two variety names or numbers. The sign "t" will 
 be used in the tables to indicate hybrids which have reciprocals. The expressions 
 F (or F,i), Fi, F.., F 3 , etc., are being used in papers dealing with Mendelian segrega- 
 te ui, the letter "F" being an abbreviation of the word "filial". In discussions con- 
 cerning the progeny, the words "hybrid generation" are used and not "filial genera- 
 tion". In reports on segregation, it is not unusual to find the expletives "Fi hybrid 
 generation", "Fi generation of the cross", etc.. in dealing with hybrids only. 
 
 By those who begin the use of the "F" expression and by those who are at 
 present employing it, a new term may not be desired, and at this late date a change 
 may be somewhat confusing; but in themselves the expressions X. X»> Xs 
 Xs. etc., would be intelligible almost without explanation, since the sign "X" is in 
 common use in discussions concerning reproduction. Because we have need of 
 the letter "F" in a number of cases to designate characters in discussing segreates 
 and parents in corn, the above suggested expressions will be used in this report. 
 The term "Xgeneration" will always be used in referring to the seed produced 
 b) the parent plant. When that seed has germinated, the plant developing from it 
 will be designated ">.i generation", its seed and the plants resulting therefrom as 
 "X: generation", the succeeding generations as "Xs", etc. By the word "seed", 
 in this instance, we refer to the germ and endosperm only and not to the seed 
 coat which is tissue identical with that of the generation preceding the new seed 
 We will apply the expression "X", etc.* to the progeny of hybrids, crosses, closes, 
 and selfs; and whenever it becomes necessary to discuss two or more of these 
 classes of reproduction simultaneously or in review, it will also be necessary to 
 include the name of the class to which reference is being made. 
 
 *Since this paper was submitted for publication the writer has learned that Lotsy (Biol. 
 Centralblatt 25:97-117) has employed the expressions "x-generation" and "2x-generation" as 
 Substitutes for "F," and "Fa". It seems, however, that the expressions "x", "\,". "x^" are 
 mire appropriate since they suggest only relative or subsequent generations rather than giving 
 the idea of increasing complexity or hybridization only, as Lotsy's "2x — " and "ox — " terms 
 would imply. 
 
 15 
 
PLANT CHARACTERS 
 
 Those who are familiar with variation in corn will appreciate the handicap 
 under which this work was started, with no inbred strains, and many of the 
 varieties imported. These difficulties are especially perplexing in the plant char- 
 acters to which very little attention has been paid in the developing of new varieties 
 and strains. From an economic standpoint, the significance of plant or "stalk" 
 characters was not recognized until quite recently. This fact accounts to a large 
 extent, for the individual variation in plant characters. Differences are often as 
 great within a variety as between distinct varieties whose classification is based 
 largely upon kernel and car character-. 
 
 GERMINATION 
 
 A number of influences within the corn kernel affect germination and the 
 early growth of the seedling. Some individual ears and also varieties exhibit a 
 much more vigorous germination than others. It has been observed that some of 
 the high yielding varieties frequently germinate poorly and this has been asso- 
 ciated with the old mistaken idea, prevalent among animal breeders at one time, 
 that improved strains showed the often observed loss of "vitality" as a direct con- 
 sequence of the "improvement". We know now. that vitality was lost in some 
 of these cases by selecting the wrong segregates. It is also evident that many 
 economically desirable characteristics in plants and animals are in a sense degenera- 
 tion, in that they are not favorable to the perpetuation of the species except by the 
 intervention of man. It is well known that single-eared and high-yielding varieties 
 tend to have deep, closely packed kernels and mature late. Such ears do not dry 
 out well, are apt to ferment and mould, and therefore germinate poorly or not at all. 
 
 It has been noticed that starchy kernels are apt to germinate more quickly than 
 hard, flinty kernels of the same variety. In germination, the scutellum or coty- 
 ledon (the organ in which enzyme secretion takes place) uses the soft-starch portion 
 of the endosperm first and the horny portions, if there are any, are used more 
 slowly. A soft-starchy endosperm thus furnishes the more available food and 
 shi uld favor a more rapid development of the germ. In order to test this theory 
 a number of ■ • kernels from hybrids of flint, soft, and sweet corn were germinated 
 in flats of soil in the laboratory. In this way two and. in some cases, three of the 
 above types of kernels could be obtained from the same hybrid ear. All the 
 kernels were planted at exactly the same depth by means of a wood pin having a 
 collar to permit uniform planting through a heavy wire screen placed on the surface 
 of the soil. This method insured accuracy in spacing as well as in depth of planting. 
 
 In the majority of cases, grown in the laboratory, the starchy kernels slightly 
 preceded the flinty kernels in germination, but the difference was not especially 
 noticeable and the resulting seedlings were quite similar in height and vigor. The 
 most noticeable difference was between either starchy or flint as compared to 
 sweet kernels which contain very little soft starch. Plumules from the wrinkled 
 kernels first appeared above the soil at from to to 24 hours later than those from 
 either the starchy or the flint kernels, and the germination of the entire lot of sweet 
 kernels was extended over a much longer period of time. 
 
 The delayed and often lower percentage of germination of the wrinkled kernels 
 may be associated with the fact that wrinkled kernels mature later than starchy 
 and flinty kernels on the same heterozygate ear. This is very noticeable in the 
 field just before the corn ripens; at this time the kernels which become wrinkled 
 later are larger and much softer than the others on the same ear. This has been 
 noticed also by Halsted of the Xew Jersey Experiment Station 37). Still another 
 explanation must be considered. Segregation in rapidity of growth factors may 
 
 16 
 
have taken place in these X= kernels borne on Xi cobs and there may or may not be 
 an association of this behavior with a character in the endosperm. 
 
 Walls 84) selected kernel- with large and small germs, finding that "the 
 heaviest grains do not necessarily have the best germinating qualities — The germ- 
 inating properties of the kernels containing different sizes of germs may be equal." 
 
 Correns 19) stated that he observed no change in rapidity of germination of 
 the X hybrid seeds. In our own work, however, we noticed that in general, the 
 X hybrid seeds gave a more rapid germination than the parents. In a number of 
 cases, these hybrid plantlets were more vigorous and maintained a larger size 
 throughout the season. For lack of time we were unable to get exact field data on 
 germination in 1910, but during the last winter several sets of observations were 
 made in the laboratory in connection with other tests on the influence of the 
 endosperm on early growth and on which the above conclusions were based. 
 
 Indications were found that popcorn pollenated with other groups germinated 
 better than other groups pollenated with popcorn, while popcorn alone gave in 
 some cases a better germination than the other groups. In corroboration we may 
 cite Sturtevant and others who have reported that popcorn has a better viability 
 than the other groups. This may be due largely to the adaptation in kernel shape, 
 composition of seed coat and endosperm of the pop group. 
 
 The specific density is not always a safe guide to the viability of a corn 
 kernel, as was shown in a case of a partially constricted ear. 
 
 Less than a fifth of this ear was normally developed at its base, the remainder 
 being abruptly smaller as if the whole ear had been developed and pollenated 
 normally and then the nourishment cut oft very suddenly from the upper four- 
 fifths. It was at first thought useless to plant any of the small shrunken kernels 
 in the field, but it was decided to give them a trial. Thirty hills were planted with 
 the large kernels and twenty hills with the small shrunken kernels. To our sur- 
 prise, every kernel grew and the viability of other kernels from the same ear 
 remained nearly as good a year later as shown by a germination test. The small 
 kernels germinated more slowly and there was a marked difference in the size 
 of the young plants, this difference diminishing somewhat as the plants grew 
 older. The average comparative sizes of 100 kernels from each of the two parts 
 of the ear near the division line were : 
 
 Ay. Wt. of 1 Ay. Vol. 1 if 1 Av. specific density 
 
 kernel (gms.). kernel (cc). of 1 kernel. 
 
 Large Kernels o..?_'i 0.256 
 
 Small Kernels 0.112 0.101 
 
 The following data represents the empirical modes of 29 plants grown ti- 
 the large kernels and 19 plants grown from the small kernels: 
 
 1 25 
 111 
 
 " om 
 
 ex u 
 
 ■— 
 
 r 
 
 OJ 
 
 - • 
 
 " 
 
 c 
 
 
 - 
 
 r- *~* 
 
 lfci 
 
 ■-— 
 
 r. ■" 
 
 <4H ™ 
 
 ■*- u 
 
 
 : 5 
 
 ■~ i- 
 
 = 
 
 c 
 
 & , 
 
 - " 
 
 ; ~ 
 
 s - 
 
 5 5 
 
 z — 
 
 - 
 
 
 
 I <-. 
 
 !_' :- 
 
 z ^ 
 
 r 
 
 ~z, .= 
 
 /■ 1 
 
 2 
 
 •z. 
 
 -J u 
 
 U a 
 
 55 J2 
 
 
 16 
 
 9 
 
 2 
 
 1 
 
 8.— 
 
 6.5 
 
 2 i 
 
 123 
 
 15 
 
 9— 
 
 I 
 
 1 
 
 8.— 
 
 6.0— 
 
 20 
 
 1-7 
 
 Large Kernels no 
 
 Small Kernels 100 
 
 The data just given will serve two purposes: — To illustrate that the hereditary 
 make-up of the small kernels, which was probablj thi same as that of the larger 
 ones, enabled the plants by the end of the season to catch up with the plants 
 from the large kernels in three out of nine more r less fluctuating characters On 
 the other hand the data serve to point out the effect of what are perhaps onlj 
 
 17 
 
physiological or somatic difference in the kernels. These differences apparent in 
 the kernel and in the plantlet stage are not entirely overcome hnt have their 
 effect throughout the life history of the plants in which the genetic make-up was 
 probably identical. 
 
 The permanent root systems of Germineae develop from the stem anfcl at or 
 near the surface of the soil. Thus, no matter at what depth a kernel of corn is 
 planted — if the depth is not excessive — the epicotyl lengthens until the node at the 
 ba^e of the coleoptile is near the surface in dry soil, or at the surface in wet soil, 
 ln-fore the secondary roots develop. The kernel itself remains hypogeal. After 
 these secondary roots have become established, the primary root is of little or no 
 value to the plant. 
 
 In germination, other temporary roots develop from the placenta of the germ. 
 In our observations, we have found these to vary from o to 8 in number, some of 
 them arising from between the epicotyl and scutellum, the remainder from the 
 outside. In tests with from 20 to 30 seedlings, from each of 67 individual hand- 
 pollenated ears in various varieties, we found differences in modal numbers of from 
 o to 4 of these rootlets. The counts were made on young plantlets from seeds 
 which had been in soil two weeks. 
 
 That these roots play a vital part in the young plantlet stage cannot be ques- 
 tioned, as they often take the place of the primary root which may become 
 diseased, cut off by insects, or fail to develop. What relation they bear, if any, 
 to the mature plants or how they behave on hybridization has not been determined. 
 Numerous, small, temporary roots may also develop on the epicotyl (below the 
 first node.) 
 
 PLAXTLETS. 
 
 It is always of great practical importance in plant breeding to be able to 
 recognize in the seedling stage of a plant what its adult characteristics win be 
 DeVries 82, 83) has made much of this point and cites a large number of instances 
 in which characters in the seedling are correlated with characters which become 
 apparent later in the mature plant. The extended use which Luther Burbank and 
 other plant breeders have made of seedling characters in eliminating undesirables is 
 well known. 
 
 It was hoped that a number of distinct characters might be found in seedling 
 corn plants and from the limited studies made it is apparent that quite a large 
 list of such characters exists. Among the varieties red tinge stems develop soon 
 after germination. This character is correlated with red tinge on the mature 
 plant in some varieties. These are constant when homozygous, and dominant to 
 pure green stems.* 
 
 Dark green leaves are distinct from light green leaves and the >G hybrid 
 plantlets show dark green. Very narrow and very broad leafed seedlings are con- 
 stant when homozygous. Certain dents and flints show very broad, blunt seedling 
 leaves, while rice pops illustrate best the narrow pointed grass-like form. Broad 
 leaves appeared in the Xi hybrid seedlings between the types. Plantlets from 
 closed and selfed ears in one strain showed marked differences in leaf-shape. 
 
 Large size of parts is dominant to small size in Xi plantlets as well as being 
 an indication of increased vigor. A number of plantlet abnormalities were found 
 but their genetic relation has not been learned. 
 
 ALBINISM. 
 
 Albinism, partial or complete, occurs in many species of plants and animals 
 and, although its cause has not been determined, its behavior has been quite well 
 
 *We will designate character? as dominant when they are epistatic to another in the x, 
 generation. The ratio or "type" of segregation tor the "plant" characters is, of course, not 
 determined r.t the time at which this paper is written. 
 
 tS 
 
studied in a number of cases. Miss Newbegin 64) found a diminished capacity 
 for assimilation in plant albinos and thus the phenomenon might be explained a 
 degeneration. That it is not a consequence of the artificial condition- of cultiva- 
 tion is shown by the fact that many cases have been observed in nature. DeVries 83 1 
 recites a number of examples arising as seedlings and bud variations from green 
 or colored ancestors to which they are more or less recessive, and generally 
 inconstant in themselves. 
 
 Cuenot 21) found albinism to be a recessive character in mice, l/arabee 32) 
 gave an account of negro albinism which indicates this character being recessivi 
 in man. Castle and Allen r3) report a similiar behavior in rabbits and guinea 
 p'igs and state that "complete albinism, without a recorded exception, behaves as a 
 recessive character in heredity." Others have investigated various legrees of 
 albinism and so far as known, with one possible exception in axolotl. reported ' \ 
 Haecker, the statement by Castle is true of albinism in its pure state. The behavioi 
 of wdiite pigment, which seems to be complicated, is not to be regarded as albinism. 
 
 Bateson 3) describes results of his own and other investigators in which the 
 hybrids from two sweet pea parents with white flowers have all been colored in 
 the Xi generation, the Xj splitting into the dihybrid (9:7) ratio. This behavior 
 is explained as due to the meeting of a complementary character from each parent, 
 producing no color when separated. 
 
 Corn in which albinism exists in the pure state cannot live beyond the plantlet 
 stage (3 to 4 leaves) since the young plants cannot elaborate plant fond because 
 of the absence of chlorophyll. Bauer 5) reports a similiar behavior in pure albino 
 seedlings arising from white margined ai d partial albinos in geranium. 
 
 Albinos are not uncommon in corn, but so far as known there are only three 
 published reports of actual counts in which close-bred or hand-pollenated seed 
 was used. The first case was that found by Hastead 35) in the > ; generation of 
 Black Mexican and Egyptian sweet corn parents. No white plantlets were found 
 in the germination tests of the parents. By various treatments he found 
 that the proportion of albinos could not be affected. It was also found that both 
 ears from the same stalk, with one exception, produced albinos, ami that the char- 
 acter was therefore "within the plant". 
 
 The important test made by Halsted was that in which "inbred" (probably self- 
 bred) and "wide bred'* (probably hybrid) seed was used from the strain in which 
 the number of albinos was pronounced. His counts were as follows (ratios com- 
 puted from Halsted's data ) : 
 
 Selfed ( ?) 
 
 corn. Albinos. 
 
 Ear No. 1 13 
 
 Ear No. 2 15 
 
 Ear No. 3 ~ 
 
 Hybrid ( ? ) 
 
 corn. Albinos. 
 
 Ear No. r o 
 
 Ear No. 2 o 
 
 Ear No. 3 o 
 
 
 
 Per cent. 
 
 Green. 
 
 Ratio. 
 
 albim >s. 
 
 47 
 
 1:3.6 
 
 -'i-; 
 
 38 
 
 1:2.5 
 
 28.3 
 
 48 
 
 1 7.0 
 
 1,0 
 
 Per cent. 
 
 Green. 
 
 Ratio. 
 
 albinos. 
 
 60 
 
 
 
 65 
 
 
 
 65 
 
 
 
 This was soon after the rediscovery of Mendel's Law and Dr. Halsted was 
 probably not studying his hybrids in sweet corn with regard to Mendelian segrega- 
 tion at that time. Although he has been making use of this all important law in his 
 hybridizing work since, we have not found that he has corrected his conclusions 
 upon this phenomenon, namely, — that close fertilization was probably the cause of 
 albinism. 
 
 19 
 
It is plainly evident that he secured on the first two ears, with only 60 and 53 
 kernels germinated, very nearly the simple Mendelian ratio with the percent of 
 albinos 21.7 and 28.3, rspectively, and an average for the two of 24.8 per cent. 
 The distribution of albinos on the third "inbred" ear must have been very uneven 
 from the charatcer of the results obtained and should be discarded. When it is 
 included, we yet have the very satisfactory average of 26.3 per cent, albinos in 
 133 kernels from three ears. On the opposite page in Dr. Halsted's report may be 
 found an interesting photograph of these three lots of plantlets by the side of the 
 three lots of "widebred" (hybrid?) green plantlets. This photograph not only 
 shows the dominance of green to albino but also, shows very strikingly, the increase 
 in vigor in the hybrids, a phenomenon which has since become so important in the 
 development of the theory of corn breeding. 
 
 Blaringhem 9) noted an instance of five albinos in 37 plantlets from a mutilated 
 parent of flint corn which was probably close pollenated. He also reported finding 
 the albinos from year to year in his various cultures. Collins ' 18) found nine 
 albinos in 48 plantlets of a hand-pollenated ear (close-bred) of dent corn, while 
 the hybrid of this with another variety produced green plantlets only. Evidently 
 neither of these cases were studied with any thought of Mendelian segregation, and 
 the counts reported are very limited but all point to the conclusion that they were 
 dealing with a simple Mendelian ratio. 
 
 Dr. Smith has observed during the last few seasons an increasing" proportion of 
 albino plantlets in the "low ear strain" of yellow dent corn grown at the Illinois 
 Experiment Station and in the season of 1910, two ears were secured from this 
 strain for use in this experiment upon inheritance. The two ears came from 
 parents which showed a large per cent, of albinos in 1909. A number of plants 
 grown from these ears were selfbred and closebred as well as hybridized with 
 other varieties including some with variegated or striped foliage. 
 
 As a preliminary test. 100 or more kernels from ears obtained in 1910 were 
 planted in flats of soil in the laboratory, 10 selfed, four closed, and six hybrid ears 
 were used. Only green plantlets were produced from four of the selfs, three of 
 the closes, and six, or all of the hybrids. From the remaining six selfs and one 
 close, the following proportions of green and albino plantlets were obtained: 
 
 Fig. 1. — Albinism in young corn plantlets. One of the rows in the germinating 
 box which gave exact 3:1 ratios of greens and albinos. 
 
 20 
 

 
 Per cent. 
 
 X umber of 
 
 Albinos. 
 
 Ratio 
 
 of 
 
 albinos 
 
 plantlets. 
 
 2/ 
 
 3-9:i 
 
 
 _i 
 
 131 
 
 29 
 
 2.6 :l 
 
 
 28 
 
 105 
 
 20 
 
 4.1:1 
 
 
 20 
 
 102 
 
 16 
 
 4.5:1 
 
 
 18 
 
 88 
 
 15 
 
 4.1 :l 
 
 
 20 
 
 76 
 
 23 
 
 3-3=1 
 
 
 -'3 
 
 99 
 
 29 
 
 2.6:1 
 
 
 28 
 
 104 
 
 SEGREGATION 01 GREEN AND ALBINO FLAXJLETS. 
 
 Greens. 
 
 424.60XS 104 
 
 424.59XS 76 
 
 424.59 <S 82 
 
 424-S9XS -2 
 
 424.S9XS 61 
 
 424.59XS 76 
 
 424.60XC 75 
 
 Total 546 159 av.=3.4 :i av .= 23 705 
 
 These cases, averaging 100 kernels from each ear, show conclusively that 
 albinism in corn is a simple Mendelian recessive. The character is especially inter- 
 esting from the theoretical as well as the economic side. It is brought to view in 
 the homozygous state only, and in this condition is disastrous to its perpetuation; 
 therefore the only way it can be transmitted is in the heterozygote. It is appar- 
 ently due to the absence of the factor which in the green plants produces the 
 chlorophyl. 
 
 What appears to be an intermediate stage of albinism was found on an ear 
 which was selfed and produced 20 yellowish greens and no albinos in 100 plantlets. 
 The first test of this ear with only 64 seedlings gave 16 yellowish greens or 23.9 
 per cent., which is near the theoretical proportion. The final test of 3 plantlets 
 gave only four yellowish green individuals, and this poor distribution caused the 
 total proportion of 4:1, or 20 per cent, yellowish greens to fall short of the 
 theoretical, although it is evident that "yellowish green" is also a true recessive. 
 These two characters, albinism and the yellowish green, which may be an inter- 
 mediate form of albinism may be added to our list of those characters appearing in 
 the plantlet stage. 
 
 SUCKERS. 
 
 The idea prevails generally that the ancestor of corn was a much branched 
 grass. Harshberger tj6) suggested that — "the original form in the wild state was 
 propagated probably by lateral offshoots — The habit of suckering. annual in the 
 north, was probably perrenial in a more southern latitude, so that in the semi- 
 tropics the non-sexual development of suckers was the ordinary method of pro- 
 pagation, the vigor of the stock being rejuvinated by an occasional distribution of 
 seed by birds." Under favorable conditions most of tin commonly grown varieties 
 produce suckers. 
 
 The tendency to sucker or tiller is decreased as a variety characteristic in the 
 corns developing stout stalks, and those varieties growing to extreme height pro- 
 duce few or no suckers under ordinary conditions. There seems also to be a 
 decrease in number of sucker- in those varieties which produce very large ears or 
 cobs. This variation in number of sucker- i- so noticeable that there is appar- 
 ently a direct correlation with the above mentioned characters but modified by 
 thickness of planting and other environmental influences. Dense planting, un- 
 fertile -oil. low humidity, and excessive heat, all have a direct limiting effect on 
 the development of suckers, while inbreeding and consequent decrease in vigor 
 also tend to check their development. 
 
 Occasionally an individual plant i- found, apparently constant otherwise, that 
 stands out from among its fellows in the production of suckers or other vegetative 
 growth and tin- is frequently attributed to reversion. Collins 15) ha- published an 
 
 21 
 
account of apcgamy in corn in the form of plant like branches developing from 
 pistillate flowers in the tassel and producing short aerial roots while attached to 
 their place or origin. These hranches, when removed and placed in soil, produced 
 roots over a foot in length and growth continued in an apparently normal manner 
 for nearly two months. 
 
 Suckers may begin to develop on corn when the seedlings are less than a foot 
 in heigth. Later, in certain varieties and especially under adverse conditions, some 
 or all of these young suckers may wither and die. Those suckers that start at or 
 beneath the surface of the soil and live through the season, develop strong, inde- 
 pendent root systems and frequently become separated from the main stalk due to 
 the expansion in growth and rupture or death of the connecting strand. 
 
 In a test concerning the value of suckers in dent corn, the Nebraska Experi- 
 ment Station 56) reported an average loss in two consecutive seasons of 17 bushels 
 per acre when the suckers were removed. Similar results were reported later 61) 
 by the same station. They found that "Removing tillers has had a tendency also 
 to increase the number of two-eared stalks and decrease the number of barren 
 plants, although this benefit has not apparently been enough to counteract loss of 
 ears ordinarily produced by tillers." In these results it was evident that the 
 production of suckers was an adjustment on the part of the plant to existing con- 
 ditions, the suckers developing when there was room and nutrition, disappearing 
 when there was not. The best yielding rows always had a good percentage of 
 ear-bearing suckers. Card 12). of the Rhode Island station, selected plants with 
 the largest total number of ears in a variety of sweet corn and found that this 
 was associated with the production of a large number of suckers. 
 
 It is doubtful whether suckers ever have much value to the main culm since 
 they at first receive nourishment from it and. later, after they have developed their 
 own root system, feed in the same soil mass, thus robbing the main part of the 
 plant continually. Lyon 55) and Hartley 41) report wide variation in the produc- 
 tion of suckers in dent corn, also that it is hereditary and can be controlled by 
 proper selection and breeding.* 
 
 During the two years in which they were under observation, several of the 
 strains and their hybrids produced no suckers ; But, perhaps, every variety known 
 will produce some suckers under certain conditions. All degrees from non-suckering 
 to profuse suckering were found. 
 
 In the hybrid (Xi) the minus-fluctuations (diminished sucker production) ap- 
 peared to be dominant to the plus fluctuations (profuse suckering). An absolutely 
 non-suckering behavior would be expected to be dominant to a suckering behavior. 
 Suckering depends very largely on environment and vigor. 
 
 Comparing these Xi results with published accounts of tillering and branching 
 in other plant hybrids, we find that Keyser 50I reports a wheat hybrid in which 
 the stooling was more abundant than with either parent. Bateson et al ,? ) states 
 that the much-branched, erect habit of the "Bush" Sweet Pea is evidently recessive 
 to the unbranched form. Shull 73) found a branched sunflower to be dominant 
 
 *Ninety-nine varieties and strains, representing the six species-groups, were used in the 
 experimental work in preparation for this thesis. The greater proportion of these strains of 
 corn used were found to be complex mixtures and in the second year many divisions were 
 made of the list as it was used at the beginning of the work. 
 
 The arrays of varieties in the parents and their hybrids, as well as the empirical modes 
 and summaries of the arrays, were tabulated in order to gain a definite idea of the constancy 
 of the character in question as well as the nature and limits of fluctuation if any variation was 
 found. Two generations of the parents and only one generation of the hybrids, selfs, and closes 
 ( X 1 ) were grown. 
 
 In the tabulation of data no aberrant individuals or populations were discarded at any 
 time for lack of conformity to the bulk of the array or table. Selections were made for a wide 
 range in combinations of characters. The hybrids, selfs, and closes, all came from carefully 
 controlled, hand-pollenated seed produced in 1909. 
 
 This tabular matter and discussion of it have been omitted in this paper and will be 
 substituted by brief general statements concerning the results. 
 
to the unbranched. Balls i) includes "branching base" in his list of dominant char- 
 acters in the cotton plant. Saunders ( cit. Bateson 3) gave the branched form of 
 stocks as dominant to an unbranched type. While Keyser's case in wheat is the 
 only one directly comparable to tillering in corn, there was no total absence of 
 tillers and perhaps no genetic difference in the parents with respect to this character, 
 and the increased stooling in the hybrid may have been due merely to increased 
 vigor. It might also have been due largely to the conditions under which the 
 hybrid was grown. 
 
 LEAVES. 
 
 During the growing season many noticeable differences in leaf characters are 
 to be found, such as ; the arrangement, shape, texture, color, pubescence, vienation, 
 and sheath characters. The misroscope would no doubt aid in identifying many 
 others in addition to those mentioned. 
 
 A corn plant may give a general impression of "leanness" from several causes — 
 alone or combined. A broad or long leaf may give it this characteristic appearance; 
 but it is more often due to a relatively large number of leaf nodes on a short 
 stem, which is equivalent to very short internodes. Suckers would naturally tend 
 to give it the same appearance. Since number of leaves is determined by the 
 number of nodes or internodes, except in rare cases where there are two or more 
 leaves at the sime node, its behavior will be identical with "number of internodes", 
 to be discussed later. 
 
 I 1 AF ARRANGEMENT. 
 
 Perhaps the most noticeable differences in aspect of the leaves are found soon 
 after blooming time when they are nearly developed, but still fresh and succulent. 
 7,ca mays exhibits the characteristic alternative, two-ranked leaf arrangement of 
 the Gramineae. Twisting occurs in the upper and more slender parts of the stem, 
 so that frequently the leaves above the ear are not in line with the lower ones 
 nor exactly on opposite sides of the stem near the tassel. In a number of cases the 
 leaves near the tassel are all warped to one side of the stem. 
 
 It is not uncommon to find plants having two or more leaves at the first and 
 second nodes below the tassel. These abnormal leaves are narrow and diminished 
 in size, giving the appearance of division rather than of addition of parts. Blar- 
 inghem 9) has noted a shortening or telescoping of internodes below the tassel on 
 poorly developed adventitious suckers produced by traumatism. Evidently these 
 abnormal growths represented efforts on the part of the mutliated plant to per- 
 petuate the species, which resulted in a contorted stem witli a tufted arrangement 
 of leaves. 
 
 In the plants observed here in Illinois in the two seasons, only one of til' 
 pase's of prolitication was associated with a deformity of the plant itself and in this 
 instance the two nodes immediately above the soil each bore a pair of leaves. Al- 
 though this plant had onlj one less internode than the empirical mode of the row, 
 it attained but little over one-half the modal height. This, and most of the cases 
 of pairing of leaves on the lirst or second node below the tassel, occurred upon Xi 
 hybrids. Three case- were 1 bserved on corn that bad been selfbred for a number 
 of generations, one of these plants producing three leaves at the first node below 
 the tassel. The plants observed were generally of average height and number of 
 internodes, more frequently above than below the average. These cases, therefore, 
 cannot be explained as being due to telescoping nor to fusion of two nodes, but are 
 simple cases of prolitication or merism. the cause of which is unknown. 
 
 During the season of 1910, a plant of rice popcorn was found having several 
 suckers, all normal but one which had paired opposite leaves at even node; thirteen 
 
 23 
 
pairs in all. There was a shoot in each of the axils of these paired leaves at the 
 eighth, ninth, tenth, and eleventh internodes above the soil. It is not expected that 
 such abnormalities are inherited. 
 
 Stray individuals having very erect leaves occur in a number of varieties. 
 The erect leaf accompanied by a one-sided arrangement about the tassel is notice- 
 able in corn which fails to develop fully. The one-sided position of leaves near 
 the tassel is frequently associated with very short internodes in that region. During 
 the season of 1910, a one-sided arrangement of husks was found to be frequent 
 on mature corn in general field culture and it is possible that such cases of orienta- 
 tion of leaves are due to warping. 
 
 Collins 18) in a paper concerning hybrids between a "Chinese corn" and other 
 varieties, reports that the erect monosticious blades seemed dominant in the Xi 
 generation hybrids at first, but later in the season the character was not so marked. 
 Second year plants from the original seed did not have this character so strongly 
 developed. 
 
 Mr. Collins kindly furnished some of the original seed for a test here at the 
 Illinois Station, where it was grown in 1910, and a large number of the plants 
 behaved as described in .Mr. Collins' original paper. It was noticed that the leaves 
 on the plants grown here in Illinois were inclined to be thick and rigid. It is' 
 possible that after it has become acclimated, this variety will still retain an erect 
 position of leaves somewhat above the average, as other varieties possessing rigid 
 leaves with heavy mid ribs tend to do. Blaringhem 9) in his paper on traumatism, 
 figures the tassel and upper leaves of what he considered to be the normal type of 
 a yellow flint variety, which exhibited this erect position of upper leaves at 
 blooming time. 
 
 A violent storm on August 23rd, 1910, ruined the entire field for observations 
 nn leaf arrangement. On many of the plants, which had not been broken off, only 
 the midribs of the leaves remained. The difference in position of these midribs 
 when stripped of the leaf blades was very noticeable. In general. Xj hybrids 
 have larger, broader leaves, consequently a heavier leaf load than the smaller of 
 the parents; but the rigid thickened blade and midrib is also, apparently, a dominant 
 character and tends to give the leaves an erect appearance. 
 
 Some varieties have very much smaller leaf blades near the tassel than 
 those lower on the stem, thus giving" an open, sparse effect; other varieties have 
 relatively large and broad leaves near the tassel frequently accompanied by shortened 
 internodes. This latter ci ndition gives a dense, crowded effect and although long 
 internodes are apparently dominant, large leaves are also dominant and various Xi 
 hybrids exhibit this leafy appearance about the tassel. 
 
 LEAF SHAPE. 
 
 Perhaps no two leaves on a single plant of any species are alike in outline. 
 Usually on the corn plant there is a gradual increase in length and leaf area till 
 the upper ear-bearing node is reached, ami then a rapid decrease in leaf dimension 
 occurs until the upper leaf at the tassel-bearing node is reached. In some varieties 
 the upper leaf is often but a mere bract, while in others this leaf is quite broad, 
 but much shortened in comparison to the leaves at the middle of the stem. 
 
 Very long leaves and very broad leaves are prominent characters of some 
 varieties and strains. Hartley 39) found a broad leafed plant of dent corn in 1900 
 which he selfed, and in 1901 the progeny was very noticeable because of their 
 broad lea\es, many being six inches in width. Seed from this row planted in 
 1902 produced plants which exhibited the same character. Collins 18) reports 
 a hybrid in which the leaf blades were slightly shorter but much broader than 
 those of either parent. 
 
 24 
 
General notes only were taken in leaf shape in this study. Varieties with 
 characteristics of different dimension in length and width were found. Several 
 hybrid rows, the product of parents one of which had long leaves, were noted as 
 having very long leaves. A number of hybrids with a broad leafed dent variety 
 also bore broad leaves. Largeness, including both dimensions, is dominant to the 
 mediocre or small leaf. Leaf blades with a broad base are apparently dominant 
 to those with a narrow base. 
 
 Abnormal leaf shapes were frequently found. Many of the hybrids as well as 
 the parents produced cleft ami distorted blades. Cleft leaves were found more 
 i if ten on suckers than on the main clum. Blaringhem reported tubular and cleft 
 leaves developing on mutilated plants, and stated that the tubular development was 
 reproduced to a considerable extent from year to year on these weakened strains. 
 DeVries 83) has pointed out associations of cleft leaves with other monstrosities, 
 such as fasciations and twisted stems in various species. 
 
 LEAF TEXTURE. 
 
 Certain external differences in leaf structure occur which serve to distinguish 
 varieties as well as individuals. Among these differences which are apparent with- 
 out miscrscopic study, are rough and smooth leaves — without reference to puhe- 
 sence ; thick, rigid blades in contrast to others which are thin and wavy, sometimes 
 crinkled. In a number of v - , hybrids the rough and the rigid types were dominant 
 to their opposites. 
 
 During the season of 1910 a small variety of white rice pop produced crumpled 
 leaves throughout. The leaf sheaths and bases of the blades, especially, gave the 
 appearance of having been crushed or jammed in the process of development. No 
 hybrids involving this character have yet been grown. Collins 18) noted a hybrid 
 between Tom Thumb pop and a tall Guatemala variety, all the plants of which had 
 crumpled and distorted leaves 
 
 DISEASE RESIS rANCE. 
 
 Disease resistance in plants is of great economic importance and remarkable 
 cases are on record of the finding of single plants which withstood attacks of fungi 
 when all their neighbors have been destroyed by the disease. Such cases have been 
 taken advantage of in a number of species 65). The production of antitoxin in 
 these disease resistant plants has been suggested as an explanation of the behavior. 
 
 Rust resistance was found to be a recessive character in wheat by Biffen 8), but 
 so far as known no other cases have been reported in which this character has been 
 studied with regard to segregation in the hybrids, although a number of cases have 
 been reported of selection of disease resistant plants after hybridization. 
 
 66). This last mentioned fact seems to indicate that disease resistance is gen- 
 erally a recessive character. Shull 72) reported increased susceptibility to smut 
 on self bred corn plants. 
 
 In iqio we noticed a number of plants which were seriously infested with rust 
 while others were practically immune. Very little rust was noticed in the previous 
 season which was a very dry year. \ hybrid row in 1910 showed advanced stages 
 of rust (probably Puccinia sorgh'i) on every plant. Another Xi hybrid row con- 
 taining only sixteen plants produced six normal green individuals which hue small 
 ears, while the leaf tissue of the remaining ten plants was thickly studded with 
 large pale yellow spots turning later to a darker yellow. These diseased plants 
 were smaller and more slender than the green plants and none produced ears. The 
 nature of the disease has not been determined 
 
 It may be that disease suscepitible combinations were unwittingly made in these 
 two hybrid cases and that in the latter case we secured what appears to lie segrega- 
 
tion. The diseases are not the same, consequently the causes in each instance may 
 differ. The rows, however, were not near each other and the conditions of infec- 
 tion may have varied. The green plants, as well as the spotted individuals in the 
 second mentioned row, were but very slightly infected with Puccinia sorghi, which 
 was the principal disease of the first case. 
 
 LEAF COLOR. 
 
 Among leaf colors in corn may be mentioned: — dark green, light green, red 
 tinge, dark red, and variegated (stripes of white, yellowish green, and red). In 
 varieties with green leaves occasional striped individuals occur, about the behavior 
 of which little is known. Pure albinos which perish in the plantlet stage will not 
 be considered under the head of leaf color. 
 
 Green. 
 
 Distinct shade-- of green are affected somewhat by humidity and plant food. 
 The Xi hybrid from dark and light green parents were dark green when growing 
 in the same soil and season (1910) with the parent varieties. It is possible that 
 the dark green of the hybrid may be partly due to vigor of growth, but segregation 
 of dark and light greens is to be looked for in the X 2 generation. Price and 
 Drinkard 69) also found the green leaf dominant to the yellowish green leaf in 
 tomatoes. 
 
 Red Tinge. 
 
 Many varieties show a more or less red tinge on the leaf sheaths and edges of 
 the leaf blade; this, in general, is associated with a red tinge on other parts of the 
 plant, especially the culm, husks, and tassel. > , hybrids frequently show this red 
 tinge intensified where one of the parents has exhibited the character. X= plantlets 
 of two such hybrids were grown in the laboratory and gave the following results 
 with respect to this character: — 
 
 Ear number. 
 23,i;oXS.. .. 
 36.205XS.... 
 
 Stems red. 
 
 Stems 
 
 10 
 
 3 
 
 10 
 
 8 
 
 
 Total Xo.of 
 
 rcent green. 
 
 plantlets. 
 
 23 
 
 13 
 
 33 
 
 24 
 
 These numbers are ton small to lie of much value, but since large numbers 
 of Xi plants growing in the field showed uniformly red tinge stems in hybrids of 
 green and red tinge, we place confidence in the above results, and in more complete 
 studies would expect perfect segregation of green and red tinge stems. Webber 85) 
 found this red tinge stem appearing in mature plants of the first generation hybrids 
 of Hickory King (green stem) and Cuzco ("purplish base"). 
 
 Red coloring matter in many plants has been generally ascribed to "anthocyan" ; 
 which very general name for a series of red. blue, and violet coloring substances is 
 usually the first and last resort in papers dealing with behavior of certain color- in 
 plants. A large and interesting field awaits the student in this very important 
 branch of heredity. In general, the theory prevails that anthocyans are produced 
 from glucosides by oxidation, etc. Similarly to chlorophyll, anthocyans depend 
 upon sunlight for their production and perhaps, in some cases, prevent the decompo- 
 sition of chlorophyll by very strong light 64). It was noticed that bagging shoots 
 retarded the production of red in the silks and on the husks which were covered 
 by the paper bags. 
 
 Anthocyans occur in the cell sap of many plants and are soluble in water, the 
 red tissue- frequently giving an acid reaction. Transpiration is reported to he less 
 
 26 
 
in red leaves than in green, and this character might be used to advantage in 
 developing a drought resistant variety. When a leaf or stem is broken on the can 
 plant during the growing season, especially during the latter part of the season, 
 the ruptured and surrounding tissue on the living end develops a tinge of ml 
 This was noticed on many plants after the violent storm in 1910 and the color was 
 apparently caused by oxidation of the cell sap. Blaringhem 9) secured a con- 
 siderable amount of red foilage presumably as a direct result of mutilation and 
 cutting of the main culm. 
 
 Dark Red. 
 
 Some varieties are of a deep blood red throughout the plant, others arc only 
 partly so in the husk. cob. seed coat, another, or silk color. Kornicke 51 ) noted a 
 variety from Peru with blood red leaves and stem. Sturtevant 80) mentions 
 receiving red-husked ears from the San Padro Indians of Mexico in 18S6, a red- 
 husked dent w^as advertised in 1885. a sweet variety with red stalks in 1889. and a 
 pop corn with red stalks in 1895. but the history of their origin is not given. Fre- 
 quently the new varieties introduced are from new combinations in hybrids and it 
 is probable that some of the above varieties originated in this way. Halsted 36) 
 noted the deep purple foilage of Ruby Sweet as dominant to the green foilage of a 
 flint variety. 
 
 Only one hybrid (green foilage X dark red) was secured in 1909 in which 
 the dark red foliage of one of the parents was constant, but of the kernels 
 planted, only two grew. These Xi plants were very dark red throughout with the 
 exception of kernel and cob colors. The endosperms were yellow and white, wdiile 
 the cobs of one plant were white and the seed coat colorless. These two plants 
 were the most striking in the field because of their deep red foilage and perfect 
 color segregation would be expected from their progeny. 
 
 Variegated Foilage. 
 
 Variegation occurs in some form in most families of flowering plants. In 
 many species this phenomenon does not appear to be constant. DeVries 83) gives 
 a large list of such instances and concludes that selection of variegated plants will 
 not of itself lead to constant forms. Occasional striped individuals occur in corn 
 as in all the Gramineae. So far as known all the groups of corn are subject to it. 
 Sturtevant 80) mentions finding them in flint, dent, and sweet varieties. In our 
 own work we found occasional striped plants in two dent varieties, one selfed pop. 
 and two dent hybrids. These plants were selfed, closed, and hybridized but the Xi 
 progeny have not yet been grown. The cause of the stripes is not known, but 
 in some cases, they are associated with degeneration or abnormalities. Correns JO) 
 and I'.aur 41 report several types of variegation in other species which (apparently) 
 do nut reproduce strictly true to seed, and consequently gave irregular segregation 
 in the hybrids. Baur 5) found a case of yellowish green variegation in Antirrhinum 
 uiu jits which was always heterozygous, segregating into one-fourth clear yellow- 
 incapable of forming pigment, one-fourth pure green, which were constant, and the 
 remaining two-fourths of the heterozygote form. 
 
 The first mention of constant striped varieties in corn, in so far as is known, 
 was in 1861 and in 1866 33). These were said to have been brought from Japan 
 and named "Japanese Striped" but the origin of the constant forms is not reported 
 Bailey 1 ) hybridized Zea Canina with Zea Japonica and in 1892 reported twelve 
 green plants out of fourteen which indicates dominance of the green. The occur- 
 rence of the two striped plants may have been due to accidental self-pollenation. 
 Variegated individuals of these Japanese Striped varieties, when selfed or closed 
 in 1909, reproduced constant in 1910 without an exception. That the parity of these 
 
 27 
 
varieties is to be questioned is evidenced by the fact that green plants were 
 occasionally to be found coming from a mixed lot of the parent seed. The striping 
 seems to be influenced by environment somewhat, more or less stripe appearing 
 under varying conditions, thus we find variations in the stripping from almost pure 
 green to almost pure albino. There is considerable fluctuation upon one plant and 
 suckers are frequently found having more stripe than the main culm. In plants 
 grown from hand-pollenated ears it was found that there were three types of the 
 striping, namely : — More white than either yellow or red, more yellow than either 
 white or red, more red than either white or yellow. It is not known whether 
 these three types will behave as units. 
 
 Without exception, all Xi hybrids of striped and green plants, including twenty- 
 seven cases and a large number of individuals, were green. This shows complete 
 dominance of green foilage. The striped varieties, however, tend to have red 
 tinge stems and leaf sheaths and, since this is a dominant character, when striped 
 plants — having the red tinge in addition — are hybridized with pure green plants, 
 the Xi progeny show the red tinge stem. There may be some variegated strains 
 which do not have red tinge stems but in case none such are found, the above 
 results might serve as an instance of broken correlation between two characters 
 on hybridization, one being a recessive, the other a dominant and variable character. 
 Whether there is an association of these characters in the_ X: segregates is not 
 known. 
 
 PUBESCENCE. 
 
 Individual plants in a number of varieties have a more or less heavy pubescence 
 on the leaf blades. Other plants were found which were perfectly glabrous. The 
 leaf sheaths of some of the rice pops have a dense, coarse pubescence. The data 
 secured, indicates that pubescence in hybrids is a dominant character. Ten 
 different Xi rows which were hybrids between varieties relatively smooth and 
 varieties pubescent were especially striking for their dense, coarse pubescence on 
 the leaf blades. 
 
 Dominance of hairs and spines has been reported in wheat, campion, stocks, 
 and thornapple 3). Miss Saunders 3) found a very interesting case of correlation 
 of purple sap color and hoariness in stocks. 
 
 The value of pubescence on the corn plant is not definitely known, although 
 it is probable that this character would be desirable in a dry region. Collins 18) 
 noted a Mexican drought resistant variety on which the leaf sheaths were densely 
 covered with long hairs borne on tubercles. 
 
 CULM. 
 
 The diameter of the culm in Gramineae, as in most monocotyledons, increases 
 by the enlargement of the cells of the young plant and not by the addition of 
 new tissue, as in the cambium of dicotyledons. Xo study of the internal structure 
 of culms was made but we might expect, from what is known of the method of 
 enlargement of the stem, that there would be considerable constancy in the cross 
 section of plants of the same strain. 
 
 There are two regions on the average corn stalk where most of the breaks 
 occur by strong winds. One common break occurs at from one to three nodes 
 above the soil, the other at the first above the earbearing node. Occasionally the 
 break occurs at a node below the ear. Of these breaks, the most objectionable is 
 at the base of the culm. Xearly all breaks occur on the upper border of nodes and 
 not along the internodes ; this may be explained by the fact that at the time of the 
 slimmer storms the upper parts of the nodes are yet succulent and spongy, con- 
 sequently they break quite easily. Every farmer who has cultivated tall corn early 
 
 '28 
 
in the morning knows from experience that corn plants break 'iff much more 
 readily at this time than later in the heat of the day when there is less moisture 
 in the stem. 
 
 It would appear that in .Tiler to prevent loss from broken stalks the breeder 
 should select individuals with linn, tough lower nodes, and large relative diameter 
 of stem at this point in addition to a strong root system to prevent lodging and 
 "down corn". Just such characters as have been mentioned, occur in different 
 varieties, and unless correlations interfere, they are probably available by means of 
 selection or by hybridization with individuals having other desirable characters. 
 
 When varieties or strains are planted side by side, the growing periods and 
 relative stage of development must also be taken into account in comparing the 
 effect of any particular storm or season. A stalk which is nearly mature is able 
 to withstand a storm far better than one which is yet green and brittle. 
 
 A large proportion of the plants in several rows, some of them hybrids, were 
 broken off near the soil in the season of iqio. These plants were usually slender 
 and had a brittle appearance. Other rows also, including hybrids, tended to break 
 above the upper shoot or ear. Plants and entire rows which lodged, and especially 
 those which failed to straighten up afterwards, were those with slender stems and 
 long internodes ; therefore, not all slender stemmed varieties tend to break. No 
 measurements of diameter of culms were made but, as was to be seen readily by 
 inspection, there is a wide difference in this regard. 
 
 The product of coarse culm X slender culm and the reciprocal gave Xi hybrids 
 with coarse, rigid culms. Slender culm X slender culm gave only a slight, it any, 
 increase in diameter; which may be attributed to increased vigor. It is therefore 
 apparent that the coarse culm is dominant to the slender culm in the Xi generation. 
 It was noticed that the coarse, stout culms were characterized by thickened cortex 
 and pericycle layers, but what relation the internal structure and the nature of 
 the vascular bundles may have to strength of stem was not determined. 
 
 During the season of 1909, an abnormally large stalk was found in a field of 
 pure-bred Learning corn belonging to Mr. F. I. Mann of Gilman, Illinois. While 
 the culm was yet green, it measured approximately two and one-half inches ifi 
 diameter at the third node above the soil. An external examination revealed no 
 other abnormality nor disease. Seed from the open-pollenated ear borne on this 
 stalk was planted at the Illinois Station in 1910 and of the plants produced on 
 medium soil, none measured more than an inch and a half in diameter. This shows 
 that this peculiarity cannot have been a dominant characteristic at least, and evidently 
 is not a genetic character since largeness in general is probably a Mendelian 
 dominant. 
 
 "Vigor of plant as shown by ability to stand upright is hereditary. Ear-n ws 
 [dent varieties] growing side by side have shown a variation of from no broken 
 plants to 56 per cent of broken plants — The ability to stand upright did not result 
 from a lighter load" 871 Webber 85) noted much taller and thicker stems in Xi 
 plants from Hickory King, a slender-stalked dent variety, pollenated with giant 
 Cuzco. 
 
 HEIGHT OF PLAXTS. 
 
 The response of the corn plant to food and environment is most significant in 
 height and size, and for this reason alone there is always a possibility of finding 
 considerable fluctuation in a study of these characters. Varieties are reported 80) 
 to vary in height from 18 inches for a "Tom Thumb" pop to 30 feet or more for 
 varieties grown in the tropics. Undoubtedly many of the difficulties which have 
 been experienced in interpreting the behavior of size or length of parts may be 
 attributed at once to abnormal environment and to impurity of strains, 
 
 29 
 
In his original paper on hybridization Mendel 58) reported complete dominance, 
 in fact, increase in height of sweet pea plants, followed by complete segregation. 
 "The longer of the two parental stems is usually exceeded by the hybrid, a fact 
 which is possibly only attributable to the greater luxuriance which appears in all 
 parts of plants when stems of very different lengths are crossed." He secured a 
 ratio of 2.84 tails to one dwarf in the next generation with 1064 plants. McCluer 57) 
 hybridized two varieties of pop corn and secured plants taller than an average 
 size of the two. Hartley 40) states that a hybrid of Mexican Dent corns produced 
 plants 23/J feet taller than the average height of the tallest parent and 4^4 feet 
 above the average of the shortest parent 
 
 Tschermak 3) reported apparent inconstancy with regard to height in bean 
 hybrids. DeVries 83) found that extracted R. dwarf antirrhinum did not breed 
 true, but gave progeny of various heights. In a hybrid of giant Cuzco and Hickory 
 King dent corns, Webber 85) noted that the Xi generation was much taller than 
 either parent. The Cuzco had been stunted, however, by being started in pots 
 and then transplanted. Hybrids of Turkey Red wheat with Russian Winter were 
 reported by Keyser 50) to have made a height growth fully 10 inches greater than 
 either parent. 
 
 Lock 53) working with corn concluded that the height of Xi hybrids "was 
 obviously intermediate — In a number of cases the cross was made between the F, 
 plants and the shorter of the parental types. The offspring of this cross showed 
 no such segregation into short and intermediate plants, as was to be expected if 
 Mendel's Law held good." Price and Drinkard 69) found "standard stature" 
 (tall) dominant to "dwarf stature" in tomatoes, giving complete segregation in 
 the X2 generation. As a result of his observations on measurements with hybrid 
 rabbits and guinea pigs. Castle 14) proposed that "Size variation is apparently 
 continuous and its inheritance blending— [hut] — Blended inheritance may possibly 
 he only a complex sort of Mendelian inheritance, in which many independent 
 factors are simultaneously concerned." 
 
 Emmerson 29) stated that in every case with which he is acquainted, there 
 has been segregation of size characters in X: following the blend in the Xi genera- 
 tion. Crossing a Tom Thumb pop 90 cm in height with a late dent 22? cm in height: 
 Emmerson secured approximate blending with regard to height (182 cm), number 
 of nodes, and growing period. He found X= plants ranging in size from the Tom 
 Thumb to above that of the Xi generation but in me were as tall as the large dent 
 parent. East 28) secured Xi plants from a medium sized flint and a tall dent 
 which were nearly equal in height to the tall dent parent. East attributes the 
 increase in size of the Xi plants over the average of the two parents to increased 
 vigor as a result of hybridizing and not to dominance of tallness. The distribution 
 of heights in the X= generation was more variable and no definite segregation 
 was apparent other than this difference in fluctuation between the Xi and Xi 
 generations. He concluded that a combination of many characters may be neces- 
 sary to produce size. 
 
 Thus, in general, the disposal of size-inheritance has been unsatisfactory. In 
 the first place a size somewhere between the heights of the parents, represents no 
 actual increase, although it is above the average of the two. Practically we would 
 not concede an increase unless we secured a hybrid larger or taller than anything 
 we used to produce that hybrid. This complication due to vigor may perhaps never 
 be overcome, but we need at least an approximate method of evaluating its 
 influence before we can say how size segregates. 
 
 Undoubtedly a system of height classification based only upon inches or 
 centimeters is unsound. It is hardly rational to expect corn plants to vary in 
 units of the yard stick or bushel measures, and to attempt an analysis of inheritance 
 from this viewpoint alone, or to expect "bushels per acre" to segregate, is extremely 
 
 30 
 
fallacious. Perhaps the division into dwarfs, intermediates, and giants is as far 
 as we may go at present with our limited knowledge of what constitutes height. 
 If it is not possible to find characters correlated with height, it will be necessary; 
 as Bateson, Castle, and others have recognized, to determine the units — of which 
 there are probably a number — that together control size. 
 
 In general, the Xi generation was more uniform than the parents ; which is 
 especially true of the selfs and closes. The arrays were made in five-inch classes 
 since the error in finding the height of a plant of this nature can scarcely be 
 limited to one inch. 
 
 In my work the fluctuation in the parent rows in 1910 was generally less than 
 in the poorer soil and season of 1909. This accords with the conclusions of 
 Davenport and Reitz 12 ) upon measurements and weights of ear corn. They 
 report variability to be slightly less on fertile land than on lands giving lower 
 yields. Love 64) has reported increased variation with increased food supply in 
 height the number of internodes of garden peas. Humbert 44) found indications 
 of a decrease in variation in night flowering catchfly with an increase of plant 
 food in the soil. 
 
 In general the Xi generation is more uniform than the parent-, which is 
 especially true of selfs and closes; this accords with observations by other in- 
 vestigators and is explained, partially at least, by the fact that the immediate 
 parantage is restricted to two individuals in the case of hybrids, crosses and closes 
 and to one parent in the selfs. For practical purposes the stragglers or poorly 
 developed plants in such cases a-; 15.100, 22.129, 25.170 etc. may be neglected. 
 
 While Gregor Mendel 58) stated in his paper on hybrid peas; "It is further- 
 more shown by the whole of the experiments that it is perfectly immaterial whether 
 the dominant character belongs to the seed-bearer or to the pollen-parent; the 
 form of the hybrid remains identical in both cases"; a number of instances have 
 since been observed where this principal does not seem to hold. Bateson 3) has 
 proposed two possible explanations of "the remarkable fact that the mother-plant 
 can impress varietal characters on her offspring by influences which are not 
 hereditary in the ordinary sense", namely, — the difference in food materials in the 
 seed, and the presence of a specific factor or factors. 
 
 From a large number of observations on plant and seed characters in corn, we 
 feel that either or both of these explanations may be justified in the same 
 individual. A summary of the reciprocals in regard to height given in Table 3 
 exhibits general uniformity in results: — 
 
 Reciprocal Hybrids. 
 No. Modal height in inches. No. of plants. 
 
 210X404 HO 36 
 
 404X210 120 4 
 
 304X503 45 47 
 
 304X503 ■■ 50 49 
 
 503X304 45 30 
 
 310X400 125 >6 
 
 408X310 125 - s 
 
 502X610 105 [ 3 
 
 610X502 105 "< 
 
 With the exception of the 404'- 210 hybrid which produced only four plants, the 
 results may lie considered as uniform. In numerous cases— not reciprocals— it was 
 noticed that w-here dent corn was used as the female parent, taller and coarser 
 
 31 
 
stalks resulted than when it was used as the male parent on the smaller sweet, flint, 
 and pop varieties; while this difference is noticeable in the mature plants, it is 
 especially striking in the plantlet stage as has been mentioned. Reciprocals made 
 in igio between rice pop and large dent varieties exhibited a wide difference when 
 growing side by side in the Xi plantlet stage in the laboratory tests. In some of 
 these instances, at least, the difference seemed to be due to unlike food material in 
 the kernel, — unlike it seems, in both quantity and quality. 
 
 Thus it is apparent that differences in vigor and size exhibited in seedlings 
 may be noticeable in the same plants when mature. It may be that the explanation 
 of increased vigor of hybrids depends largely upon this one fact — vigor of seedling — 
 due to the quantity and composition of food material in the germinating seed, which 
 favors an initial, vigorous cellular — activity upon which luxurious development 
 depends. 
 
 The modal summary of the behavior of height shows some very interesting 
 results with variations from what appear to be increases in height to those cases 
 which appear to be perfect blends and others that seem to be decreases below the 
 height of the shorter parent. Twenty-three cases out of forty-five possible ones 
 in the Xi hybrids gave modal heights greater than either parent individuals and 
 modes. In five cases Xi hybrid modes were below the parent, individuals and modes. 
 How these hybrids would behave in the X= generation would make a very 
 fascinating study indeed. The closes gave variable results, as is to be expected in 
 view of the unknown purity in the seed. With regard to height, there may have 
 been as wide difference in the majority of individuals used in the production of 
 hybrids. Had "pure" strains of corn been available, we would have expected a 
 decrease in height in the majority of Xi closes. Some of the increases in height 
 observed in the Xi closes may have been due to acclimatization ; again, w-e may 
 have had a wide difference in the individual parents of some of the closed cases. 
 
 From the experience of McCluer, Shull, East, and others, we would expect :.ll 
 the selfs to appear in the column, "Below the Lesser Parent", ie ; decrease in vigor 
 the first season because of self-breeding. To our surprise, we find one case 
 giving Xi progeny as tall as the parents (Equal to the Average of the Parents) 
 and another taller than the parents (Above the greater Parent). In the case of 
 increase (2S.170) there were only nine plants but confidence is placed in their 
 behavior, although the parent mode chosen may not have been exactly correct 
 because of the limited array given. The hybrid kernels from this same ear 
 (J.?. 170) gave plants taller than the selfed individuals although the 309 variety 
 averaged somewhat shorter than the 210 variety. It is suggestive that these two 
 varieties, both popcorns (201 and 210) may permit inbreeding for a time without 
 serious loss of vigor. The two closed populations (2C.163 and 2C.164) of 210 
 gave further evidence in favor of this conclusion. 
 
 IXTERNODES. 
 
 The question arises as to what hereditary factors determine height and, as a 
 matter of course, it is to be expected that the number of internodes and their 
 relative length would determine mere linear dimension. It is also surmised that 
 number of internodes, within a variety, is a more constant quantity than either 
 height or average length of internodes. 
 
 The possibility of correlation of number of internodes with yield or production 
 and with some other characteristics has been investigated, but so far as known, 
 no correlation tables have been constructed between number of internodes and 
 height of plant. Ewing 31) found the coefficient of variation for number of inter- 
 nodes within one variety of corn to be less than for a number of other plant 
 characters studied. The results in selection during six successive years as given 
 by Smith 76) give evidence of a direct correlation between height and number of 
 
internodes within each of the two strains, "high ear" and "low ear". In the data 
 given by Smith, the number of internodes is plainly more constant than their 
 average length. Certain varieties exist, however, in which the average length of 
 internodes remains quite constant. 
 
 Cases of "telescoping" or shortening of internodes have been found to repro- 
 duce. Hartley 39) describes stalks 4 to 5 feet tall bearing 18 to 20 broad leaves 
 in a white dent variety whose normal height was 10 feet. These short plants were 
 pollenated with each other and the dwarf character was reproduced by their 
 progeny. 
 
 In the season of 1910 in a plot of Learning, a tall dent corn, here at the 
 Illinois station, a stalk was found having 17 internodes, probably the average of 
 the variety. This plant was only 37 inches in height while the average of the 
 plot was approximately 120 inches, and the largest leaves of this plant were as 
 long as the culm of the plant itself. This average length of internode of slightly 
 over two inches as against approximately 7 inches in the normal plants presented 
 a very dense-leafed effect. No ear was produced on this plant but its pollen was 
 used on two other strains of Learning corn with success. As short internodes is 
 expected to be recessive, if there is an inheritance of the peculiarity, it cannot 
 appear till the X: generation has been grown. 
 
 With a given height, extreme length of internode is not desirable unless it is 
 accompanied with a large diameter of culm and a well developed root system. 
 Extremely short internodes are also undesirable, as such varieties usually produce 
 short leaves and small ears, with a corresponding decrease in other parts of the 
 plant. In general, the length of internode on a corn plant is least at the roots, 
 and increases toward the tassel and, in many varieties, there is again a decrease 
 in length of internode between the upper ear and the tassel. 
 
 In Emmerson's 29) Tom Thumb PopXLate Dent hybrid, an intermediate 
 number of nodes (S nodesXi9 nodes gave 12 nodes in the Xi> was obtained in 
 the Xi generation, but segregation was predicted for the X= generation. Assuming 
 that the number of nodes given by Emmerson corresponds to the number of inter- 
 nodes, we interpret his results as follows : — 
 
 A v. length of internodes 
 Height of plant. N'n. of internodes. ( Including the tassel). 
 
 Tom Thumb 90 cm. 8 11.3 cm. 
 
 Late Dent 225 cm. 19 11.9 cm. 
 
 Xi Hybrid 182cm. 12 15.2 cm. 
 
 While he secured a gain of 3.3 cm. in average length of internode, his Xi plants 
 showed a decrease of 7 internodes as compared to the taller parent. This behavior 
 indicates either dominance of small number of internodes, or that one or both 
 the individual parents used were heterozygote with respect to number of inter- 
 nodes; unless number of internodes can be shown to actually blend. 
 
 Because of the very short internodes at the base of the plant, some of which 
 may be covered in cultivation, it is impracticable in most cases to secure a count 
 of the total number with an experimental error less than 2 internodes. 
 
 It is practically impossible to get the exact number of internodes at the time 
 of maturity of the plant without dissecting the rooted portion of the stem. Care- 
 fully marking the sixth or seventh leaf of the young plantlet while the coleoptile 
 or plumule sheath is yet discernable and which is borne at the first true node, would 
 be, perhaps, the most satisfactory and definite method of counting internodes. 
 
 In the behavior of the hybrids it is significant that height of plant is affected 
 more frequently by difference in average length than by difference in number of 
 internodes. It was found that closing and selfing gave a decrease in average length 
 but no significant decrease in number of internodes. 
 
 33 
 
Among the hybrids there were a number of cases which fluctuated considerably 
 from tf)e modes of the parent strains, and the modal summary shows very clearly 
 that the fluctuation in results is largely due to hereditary differences in the individual 
 parents. The summaries show at a glance that both large number and greater 
 average length of internodes predominate in the Xi hybrids, the majority of cases 
 falling in the lists "Equal to and Above the greater Parent". Closing and selfing 
 seems not to have any decided effect upon the number of internodes, while there 
 was a decrease in average length in the majority of cases. 
 
 From the results obtained, it is evident that more reliable data and greater 
 accuracy in interpretation may be expected with regard to number of internodes 
 and their relative length than from observations on mere height of plant in units 
 of measure. These two characters do not behave alike in all cases in transmission, 
 and do not seem to be correlated. Decrease in average length of internode seems 
 to be the principal cause, at least in the first season, for "decrease in vigor due to 
 inbreeding" in so far as height of plant is concerned. 
 
 AERIAL ROOTS. 
 
 After the secondary roots, which develop at the first few nodes on the stem, 
 have become established, the primary root is of very little importance and later 
 ceases to function, as is characteristic of monocotyledonous plants. 
 
 The first few whorls of secondary roots are usually very close together, rnaking 
 it difficult to distinguish the internodes. Succeeding whorls of aerial roots, com- 
 monly known as "brace roots", frequently develop at a considerable height on the 
 culm from which they can no longer reach the soil, unless the plant is bent or 
 lodged: in that case the aerial roots always start from the side of the node nearest 
 
 
 
 the soil, sending out a mass of fibrous roots after penetrating the sod surface, thus 
 anchoring the plant firmly in its new position. 
 
 The aerial roots are forced through the epidermis of the nodes. They are 
 usually quite fleshy, and have a gummy surface which is said to be due to the 
 breaking down of the outer cell walls. During a rain or when immersed in water 
 the gum accummulates at the root tips as a transparent jelly-like mass. This 
 gelatinous substance is tasteless and gives a neutral reaction with litmus. The gum 
 prevents dessication and affords protection for the tender surface of the develop- 
 ing routs. 
 
 Some varieties have the aerial-root character well developed wdiile others do 
 not show it at all. So far as known no selection for these roots has been 
 reported. The character is fluctuating in its behavior, being affected by environ- 
 ment and vigor. Border plants frequently produce the roots higher on the culm 
 than do plants in the center of the plot or field. One plant of large Mexican corn 
 (408) on the end of a row in 1909 produced aerial roots at the thirteenth node 
 above the soil. 
 
 Counts of the aerial root wdiorls were made only in 1910. The fact that very 
 fcw of the Xi generation and only a limited number of the parents produced plants 
 with only "one rooted node" above the soil, favors the conclusion that when 
 varieties produce aerial roots at all. they tend to do so on several nodes. Certain 
 strains produced no aerial roots which failed to reach the soil. In the large 
 majority of the X» hybrids, a high number of rooted nodes predominates, when this 
 feature was found in one of the parents. Closing gave variable results depending 
 on the parentage. In the selfs the number of nodes with aerial roots decreased in 
 three out of four cases observed. 
 
 PERIOD OF GROWTH. 
 
 The origin of varieties with very short, and others with extended growing 
 periods has not been recorded, but varieties exist that are adapted to region^ 
 
 34 
 
differing widely in soil and climatic conditions. Considerable local adjustment 
 may take place m a few seasons in varieties whose requirements are not adapted 
 to the new conditions in which they may be placed. Sturtevant c'o i recorded 
 differences in growing periods of from one to seven months in varieties growing in 
 different regions. 
 
 After five years of selection for high and low ears. Smith 76) observed that 
 the "low-ear" plot was apparently a week in advance of the "high-ear" plot at 
 the tasseling stage. The low ear character is associated with a decrease in height 
 of stalk, shorter internodes, and a decrease in number of interhodes below the ear. 
 
 Pearl and Surface 67). after three years of experience in selecting sweet corn, 
 report; "Many of our result- seem to indicate that earliness, in a large part at 
 least, is a physiological rather than an hereditary phenomenon — all the gain which 
 has been made in earliness was accomplished in the first year's selection. No 
 further increase has followed the further selection practised in the two subse- 
 quent years." 
 
 Emmerson 78) has found apparently little correlation between height and 
 earliness in the second hybrid generation of two varieties of corn, some of the 
 earliest plants being above medium height and some of the latest being very short. 
 He reported the Xi generation as intermediate in maturity (100 days) between 
 parents wdiose growing periods were approximately 70 and 130 days respectively. 
 Mendel 58) failed to complete his report on the study of this character in peas, 
 but found that the first hybrid generation was intermediate and predicted segrega- 
 tion if hybrids were made in which the parents differed at least twenty days in 
 maturity. Bitten 8) also noted intermediate results with hybrids of early and 
 late wheat varieties. 
 
 In our study of this question, the "period of growth" will be designated as the 
 time in days between the planting of the seed and the date upon which the husks 
 on the upper ear or shoot are dead. We do nut assume that growth in the plant as a 
 whole ceases definitely at this stage, but the time at which all green has disappeared 
 from outside husks provides perhaps as definite and satisfactory a point as can be 
 secured for comparison in regard to the total growing period. After this stage in 
 ripening, there is undoubtedly considerable deposition of substance in the ear and 
 in other parts of the plant in some cases. Incidentally it was noted that some 
 varieties ripen their husks a number of days before the remainder of the plant 
 matures. The period of growth may properly be divided into two stages by the 
 date en which the silks of the upper shoots on most of the plants appear. This 
 point in the blooming period was taken because it is the time when the process of 
 reproduction begins. 
 
 Notes were taken daily on each plant in 1909 on the appearance and character 
 hi the tassel, anthers, and silks, and it was found in a number of instances that the 
 relative time of the appearance of these three was quite irregular. Since the plant- 
 ing was done at three different times in the first season, only data from 1910 were 
 used for comparison with the Xi generation. This data was taken every four 
 da\s ,m the appearance of anthers and silks only. The differences in length of 
 blooming periods are frequently due to poorly developed and abnormal plants, yet 
 s.ime varieties pass the period very quickly (10 days) while others have a very 
 extended period ( jo days or more). In general the blooming period of the Xi 
 hybrids tends to be longer than the average of the parents. Eleven hybrids in the 
 list of those observed had a period equal to or less than that of the parent with 
 the shortest period, while sixteen other hybrids had a period equal to or longer 
 than the average of the parents: nine being equal to or longer than that of the 
 parent with the longest blooming period. Among the closed cases, -ix out of 
 nine required more time to finish blooming than did the parent rows in the sami 
 
 35 
 
season. Two of the three selfed cases had blooming periods identical to those of 
 the parents and No. 201 (selfed) finished blooming in a much shorter time than 
 its parent. 
 
 While protandry is the most common behavior in floral development in corn, 
 no very extended observations on the inflorescence at the time of blooming are 
 necessary to bring to notice individuals which do not shed pollen till some time 
 after the silks appear. As examples of differences among the varieties, No. 203 
 had eighteen plants which showed silks on a given date while only five among 
 that number were as yet producing pollen. When this variety was closed and 
 hybridized the progeny gave a normal behavior. Other varieties, notably 503, 603, 
 and 610 produced no silks until approximately eight days after the first anthers 
 appeared. It is striking that a somewhat similar behavior prevails in the closes and 
 selfs of 503 and 603. 
 
 The reciprocals differed somewhat in time of appearance of silks after planting. 
 This difference varies from four to eight days, and may have some significance, 
 but the general result is the same; that is, the longer periods tend to be dominant 
 in the >'i hybrids. The hybrid 24.162 was approximately a week later in blooming 
 than its reciprocal 42.238. It may be that the difference in location in the field 
 may have had an influence, but it was mure probably due to the difference in the 
 individuality of the parents on the 210 side. This is also another example of 
 difference in behavior when dents are hybridized with smaller varieties. The large 
 dent kernels of 42.238 should give the plantlets an initial advantage over the 
 plantlets from the very small kernels of 24.162. Close and self pollenation delayed 
 the time of blooming in the majority of the Xi progeny. While the results do not 
 permit definite conclusions to be drawn, the balance is decidedly in favor of 
 "late blooming as being the dominant character." 
 
 A partial list of the total growing periods varied, in the parent strains used, 
 from approximately 100 days to nearly or perhaps fully twice that length of time. 
 This table also showed the number of days in the latter part of the growing period. 
 The uniformity of this latter period in comparison with the total period is sig- 
 nificant in all those cases in which maturity was reached. The exact average 
 fraction of the total time for both the parents and the Xi generation is 0.39 or 
 approximately 0.40. 
 
 If this fraction were constant for all the varieties and their progeny, it would 
 afford a ready means of calculating the total growing period when the time from 
 planting till silking is known; and thus provide a method for learning the total 
 growing period in those cases which did not mature because of the short growing 
 season. There were, however, extreme fluctuations of from 0.31 to 0.45 and there 
 must have been similar fluctuations in the varieties which did not ripen. In testing 
 this proposition on some cases which did not mature, by adding 40 percent of the 
 total (equal to % of the time from planting till the appearance of the silks) to 
 the number of days from planting till the silks appeared, a few of the cases should 
 have been ripe although they are not so shown in the field data. While further 
 and more conclusive tests of the relative lengths of the two periods of growth are 
 necessary, the results secured indicate that approximately 60 per cent of the total 
 growing period in corn preceeds the time of the appearance of the silks ; the remain- 
 ing 40 per cent of the total growing period falls between the time of silking and 
 the time wdien the husks on the ear are dead. 
 
 The data are too incomplete for a summary. In a number of cases, the Xi 
 hybrids ripened late or failed to ripen at all. There is sufficient evidence of *his 
 nature in the tallies to permit the conclusion that late maturing is dominant to 
 early maturity. Since the total growing period tends to be longer in the Xi 
 hybrids than that of their earliest parents, the period between silking and time of 
 ripe husks is as a consequence also longer. 
 
 36 
 
INFLORESCENCE. 
 
 There are numerous indications that the present monecious type of inflorescence 
 in corn has developed from a terminal hermaphrodite form. Hackel 34), Harsh- 
 berger 38), and others have recognized the origin of the ear of corn, as we have it 
 now, as due to a union of two rowed spikelets of the floral clusters on the tillers 
 or plant branches. Harshberger has pointed out further that clusters of these 
 fasciations or ears probably existed in the primitive form, and "one ear, in the 
 cultivation of corn for centuries, has enlarged at the expense of the others, furnish- 
 ing another illustration of the law of compensation in growth." These ear clusters 
 mentioned by Harshberger are of common occurrence in many varieties today, being 
 found in the terminal inflorescence of suckers and also upon the shanks and among 
 the husks of ears on the primary culm. Sturtevant 80) mentions the frequent 
 occurrence of such cases. 
 
 Montgomery 59 ) proposed that the ear has developed directly from the central 
 spike of the tassel without fasciation, the lateral branches of the tassel disappearing 
 as a consequence. It is a well known fact, bow-ever, that the large majority of 
 ears in many varieties are "spread", that is. — larger in one diameter than in the 
 one taken at right angles to the first. (The frequent occurrence of such ears 
 indicates that it is a dominant character.) Quite frequently, also, ears with 
 enormously spread or fasciated tips are found, and such features are very strong 
 evidence in favor of the fasciation theory. The fact that large husks 1 not enlarged 
 glumes) are frequently found in among the kernels near the base of the ear also 
 the occurrence of kernel baring brandies not only at the base of the ear but at 
 some distance above the base of the ear are further evidences in favor of the same 
 theory. The recurrence of these branches indicates that they were absorbed but 
 not lost. 
 
 A very interesting case arising in our own work offers evidence in favor of 
 the fasciation idea rather than the theorj presented by Montgomery. In harvest- 
 ing one of the Illinois Experiment Station plots, the men found a peculiar cone- 
 shaped ear with a prolificated cob. A section of the ear disclosed only a large 
 central pithy core to which were attached the slender branches. The kernels 
 (borne in pairs) and glumes were irregular in size and shape (due to the irregular 
 distribution ) but normal otherwise. Fifty plants were grown from this ear in 
 19.10 and two of that number bore exact duplicates of the mother ear; this fact 
 indicates that the prolification is a character and will reproduce. The pollen 
 on these two plants grown in 1910 was nearly all gone before the silks developed; 
 it is probable that the mother plant behaved similarly and that the seed was largely 
 cross pollenated by normal plants. 
 
 If this behavior, which appears to be a reversion, is a recessive character, it 
 would account for the small number of individuals which reproduced the prolifica- 
 tion. One of the ears was pollenated by a normal individual in the row and in 
 case this character behaves as a true recessive, the X= generation of the hand 
 pollenated ear should produce fifty per cent of proliferous cobs. The peculiar ear 
 character is associated with a distinct tassel type, which is also cone-shaped, bearing 
 numerous short branches. Advantage may lie taken of this fact in the selection of 
 such individuals before the silks appear. 
 
 Historical investigations indicate that the number of rows on the primitive 
 type of ear was not very large, probably ranging in number of rows from four to 
 fourteen, and in some varieties there has since been a decided increase in size and 
 number of parts in the ear as well as in other parts of the plant. It is evident that 
 large proportions within paired spikes or rows in the cob are responsible for large, 
 broad kernels. This enlargement of parts, including kernels, appears to have taken 
 place at the expense of other parts, such as anthers and glumes, as well as of one 
 
 37 
 
of the entire florets of the original pair (Embryonic kernels are paired in the row 
 similarly to the pairing of the rows themselves, but the lower flowers in each 
 spikelet are generally abortive) So). This increase in size of the female parts 
 seems also to have been associated with decrease in the leaf blade and telescoping of 
 the internodes on the ear bearing shank. 
 
 In regard to the origin of the ear-bearing branch Montgomery suggests that 
 as evolution progressed, the central tassel came to produce only staminate flowers, 
 the pistillate ears being produced on the tiller-like branches, which became shortened 
 until the leaf sheaths enclosed the ear. 
 
 Blaringhem g ) found a close correlation between the length of the supporting 
 stem and the type of terminal inflorescence on suckers; those having only female 
 flowers being borne on the shortest stems while those bearing a mixed inflorescence 
 were borne on longer stems. It is well known that extremely large pods in the pod 
 corns may develop not only at the expense of the cob, but of the entire floral parts 
 including the kernel. Pod corn frequently develops a large amount of grain or 
 very heavy glumes in the tassel, in which case it often happens that no ear 
 develops on the culm. 
 
 When a number of rows of kernels are combined in the central tassel spike, 
 there is a thickening of parts and in most cases a decrease in the number of 
 branches. 
 
 Development of scattered, medium sized, rounded kernels are frequent in tassels 
 which show no perceptible thickening or enlarging of tassel parts. This behavior 
 is very common among pod corns and tends to reproduce. Sturtevant 80) reported 
 slight or no indication of the transmissal of this character from wind-pollenated 
 seed. He also noted hermaphrodite flowers and staminate clusters on the cobs of 
 other species groups. Blaringhem a) secured hermaphrodite ears from mutilated 
 flint corn parents and reports the second generation as being constant in the 
 production of anthers on the ear. Montgomery 60) has recently reported finding 
 perfect flowers in well developed ears. In our own work, a number of similar 
 cases have been found, but of their inheritance we know nothing except in the 
 case of the pod corn. In this study no instance was found (in a large number 
 of cases observed) where both floral parts functioned in the same flower either in 
 the tassel or on the cob, although both anther and pistil are present ; one always 
 functions at the expense of the other which is poorly developed, thus insuring 
 pollenation between florets at least. In the season of 1910, an abnormal plant was 
 found which bore two medium sized ears, but no indication of a tassel. 
 
 Cases of barrenness or complete sterility will lie discussed in connection with 
 the paragraphs on "Pollen" and "Shoots". 
 
 Aside from notes on hermaphroditism and abnormalities, there are hut few 
 reports upon tassel characters. Various investigators have noted differences in 
 the number of tassel branches. Collins 18) noted frequent cases in a Mexican 
 drought resistant variety in which the tassels were unbranched and the normal 
 form itself produced but few branches. Shull 75) reports an apparent decrease 
 in the number of tassel branches in a self-bred strain. Webber 86) found a correla- 
 tion between the kernel color in Black Mexican sweet corn and green tassel 
 glums and anthers. This correlation was but rarely broken. 
 
 Structure. 
 
 Styles and sizes in tassel structures are various and, in our own work at least 
 a dozen distinct types or tonus have been found. Mo one tassel was found which 
 
 38 
 
possessed all the dominant units, consequently the various hybrids gave many new 
 combinations not observed in the parents. To the casual observer these tassels of 
 the Xi hybrid would apear a blends very analagous to the supposed blends in 
 
 Fig. 2. — Some distinct tassel types. 
 
 height of plant. The detection and sorting of the segregates with respect to certain 
 tassel characters may prove as difficult an undertaking as that of determining the 
 units which control height in the plant. It is evident that the twelve types chosen are 
 themselves the combinations of many units. 
 
 In the X. hybrids, we found the following tassel characters apparently dom- 
 inant : — large size, long .spike, erect spike, long branches, many branches, spreading 
 branches, coarse parts, thick glumes, large glumes, monstrous glumes. On many of 
 the Xi selfs and closes, we noticed a decrease in the number of branches on the 
 tassel as was reported by Shull 75) on self-bred strains. It might be assumed that 
 the behavior just mentioned is due to segregation, but a number of observations in 
 our own work led to the conclusion that it is a decrease in vigor of branching. 
 This is very similar to the decrease in the number of suckers observed on our 
 self-breds and although Shull has not mentioned this fact his photographs show 
 his "Strain A" which was originally suckered, as no longer producing them. 
 
 Color. 
 
 In varieties that were known to be fairly pure, and especially in self-breds, the 
 tassel colors were fairly constant but in the majority of the varieties used con- 
 siderable mixture existed in these characters. Red tinges in the tassel were shown 
 to be dominant to green in a number of Xi hybrids. As has been mentioned in 
 connection with colors of the culm and leaves, we were unable at the beginning 
 of the study to secure a variety that was constant with respect to deep red color 
 throughout. Dark red tassels are relatively numerous in some varieties and their 
 frequent appearance in the Xi generation indicates that this is a dominant character. 
 
 Male Flowers. 
 
 The male florets in clusters of two or a multiple of two, one pediceled, the 
 other sessile, or both sessile, are usually arranged alternately in the panicle. In 
 the four varieties secured from Arizona an opposite arrangement of spikelets 
 
 39 
 
occurs ;it many points in the tassel, similar to the behavior in a Mexican variety 
 reported by Collins 18). Cases of paired spikelets were found in hybrids with 
 these variations in 1910. 
 
 Differences in size and texture of glumes have been mentioned in connection 
 with tassel types. Colors in the glumes behave similarly to those on the stem of the 
 tassel, frequently, however, the glumes are colored when the tassel stems are not ; 
 quite often a distinct red ring is formed at the base of the florets in those varieties 
 having red tinge tassels. "Red tinge" or "red base" in the glumes and their deeper 
 colored form "reddish green" are dominant to pure greens. Dark red glumes from 
 mixed parentage gave results which indicate, dominance of that color over red 
 tinges and pure greens. 
 
 The anthers in the floret of a corn tassel are usually three in number (sometimes 
 four to six where doubling of parts occurs) and vary in different individuals and 
 varieties as to size, shape, color, and fertility. 
 
 Size. 
 
 During the hand pollenating work of the two seasons, a small quantity of 
 anthers was preserved from representative individuals in the parents and the Xi 
 generation. These anthers were thoroughly dried in the tassel bags and samples 
 stored in small glass vials for future reference. Among these cured anthers, dis- 
 tinct differences exist in regard to size and shape. Such differences had been 
 observed in the field but since there was no opportunity for taking measurements, 
 only general statements regarding their behavior can be made. 
 
 Extreme length is not always associated with large circumferences in anthers, 
 and long anthers are not always found on tall growing varieties. Anthers below 
 the average in length were found in some varieties, upon very tall plants. Long 
 anthers were found in other varieties producing small plants. Tn the V, hybrids, 
 a number of cases were found where long anthers were dominant to short anthers. 
 Many intermediate cases were found, which were attributed to distinct differences 
 and heterozygosis in parents with respect to this character. No definite statement 
 can be made in regard to the behavior of relative size and circumference in anthers, 
 but differences exist in this respect as well as in length. 
 
 Color. 
 
 Carful notes were taken on the color of anthers upon every plant grown in 
 1909 and upon every plant hand pollenated or from which pollen was secured in 
 1910 ; thus exact data are at hand on the behavior of anther color. The color in 
 dried specimens was preserved perfectly in most instances. A number of striking 
 colors and beautiful tints were found. None of the varieties used were absolutely 
 constant in anther color, consequently only a limited number of cases were found 
 in the Xi generation in which the anthers were of the same color on all the 
 plants of a population. 
 
 Among the many combinations made, we find cases as follows: — green 
 (anthers I X green, giving only green; green X reddish green, giving only reddish 
 green; and other cases giving colors from pure green, reddish green, to dark red; 
 either green or reddish green X deep red giving all deep red and some giving mixed 
 results from pure green to very dark red. From this it is evident that dark red — 
 when homozygous — is dominant to other colors, wdiile pure green is recessive to 
 colored anthers (red tinge, reddish green, and red). 
 
 40 
 
Increased uniformity was secured in another color in all the close-, and selfs. 
 This uniformity including that of glume colors was a striking feature in self-bred 
 strains. One closed case (502XC) in which one parent had pure green anthers 
 and the other reddish green, gave Xi progeny all having reddish green anthers. 
 This case gave the same results in regard to another color as was secured in hybrids 
 between varieties because — as is apparent — we closed a pure green with a pure and 
 dominant reddish green segregate. The anther color of the majority of plants of 
 the parent variety was green. This is another of the many instances of mixture 
 of characters in a variety which is popularly said to be "pure". 
 
 Tests were made of the behavior of the light green anther which Webber 86) 
 reported was correlated with kernel (aleurone) color in "Black Mexican" sweet 
 corn. Two Xi hybrid populations of this variety with others having reddish green 
 anthers gave Xi progeny all having reddish green. In Webber's own observations 
 upon X» segregates of a hybrid of this black sweet variety with one having reddish 
 green anthers, he reports almost perfect correlation of the pure "black" or purple 
 kernels with green glumes, anthers, and silks. If the correlation is not perma- 
 nently broken, we should be able to pick out the X? plants homozygote to the 
 dominant kernel character — purple — by selecting those plants bearing green anthers. 
 If such results can be secured in the X= generation it will serve not only as an 
 example of broken correlation in so far as its expression is concerned in the Xi 
 generation, but also as an instance in which a homozygote dominant can be 
 separated from the heterozygote dominants in the X? generation because of the 
 correlation of a dominant and a recessive character. 
 
 Among the Black Mexican plants growing in iyoo, four out of 108 individuals 
 produced red tinges on the anthers and silks. This may be due either to impurity 
 or to imperfect correlation. Webber also found stray individuals in the X< segre- 
 gates exhibiting red tinge silks and glumes which he attributed to breaking of 
 correlation by hybridization. 
 
 Sterility. 
 
 Kolreuter and Gartner among other early investigators thought that sterility 
 in hybrids was a means of distinguishing species. Darwin investigated many 
 cases by which he was able to point out the fallacy of this belief. In an explana- 
 tory statement in connection with an instance of total sterility in corn, DeVries 82) 
 concludes ; "barrenness is to be considered as a monstrosity, which like all other 
 monstrosities, is inherent in a race, but is developed only in a certain percentage 
 of its individuals." 
 
 A number of instances of sterile anthers have been reported in inbred corn. 
 From our knowledge of heredity gained since these cases were reported we know 
 that they have been the result of using parents in which sterility existed, rather 
 than sterility being the immediate result of inbreeding itself. Hartley 40) reported 
 two hybrid cases in corn which showed a strong tendency toward sterility. Suckers 
 developing adventitiously from mutilated plants were stated by Blaringhem 01 to 
 have produced tassels with infertile stamens, but such a condition frequently exists 
 on poorly developed suckers in corn. 
 
 In a hybrid of Maryland white dent and blue Hopi corn. Collins 18) found 
 six out of sixteen plants which failed to produce pollen Sterility has been 
 reported in hybrids with other species than corn, but the behavior does not seem 
 to be the same in all cases. Normal anthers are reported to be dominant to sterile 
 anthers in sweet peas ,; I. 
 
 In two of the parent strains used in our experiments the majority of the 
 tassels bore largely sterile anthers. A number of the plants produced a small 
 quantity of pollen insufficient for our needs, others gave no indication of sterility. 
 All the hybrids as well as the closes with these two strains gave, without exception. 
 
 41 
 
Xi progeny with partially or completely sterile anthers on the most of the plants. 
 It is therefore expected that a careful analysis of this abnormality would show 
 that it is a Mendelian dominant in corn. In many instances of total sterility, the 
 flowers never opened while in others the hard shrunken anthers were discharged. 
 These latter individuals were at first bagged, before we became aware of their 
 nature, with the expectation of securing pollen. Such a character if it is dominant 
 could of course not be transmitted unless it is pollenated by a homozygous recessive. 
 This is another instance of a character, fatal in itself, depending on heerozygosis 
 ( in this case, pollen of a recessive on a pure or a heterozygote dominant) for its 
 perpetuation. Thus the character, while it insures mixed-pollenation. is yet un- 
 desirable because it is a dominant one. Some individuals were found having a few 
 fertile anthers among the many sterile ones in the same tassel. 
 
 The daily field notes taken in the season of 1909 upon individual plants, recorded 
 the exact time of the appearance of pollen and silks upon approximately 5000 
 plants representing a large range in varieties and all the species groups. A number 
 of cases were noted in which the silks appeared on the upper shoot at from one 
 to three days before any pollen was shed from the tassel of the same plant 
 (protogyny). The large majority displayed a similar and frequently greater 
 variation in the opposite direction — anthers before silks (protandry), while many 
 instances were found in which the anthers and silks appeared on the same day 
 (synacmy ). 
 
 In the cases mentioned, we thus found a variation in the difference of time 
 in days of from minus three to infinity, if we include those plants which produced 
 no silks at all or were "barren" in so far as production of ears was concerned. 
 We may therefore expect in many varieties, a normal fluctuation of from one to 
 four or live days in the interval between the appearance of pollen and silk on the 
 same plant. Several varietial differences in this respect have already been men- 
 tioned under "Period of Growth" (see page 80). 
 
 Sturtevant 78) made similar observations on single plants, noting first — film lin- 
 ing in each row of the different varieties grown and from his data we have com- 
 piled these results: — 
 
 : 
 
 Group 
 
 is 
 
 be . be w bfl 
 
 2 > ■% £ > rt > o « fc 
 
 o ^ o^ o v in 
 
 "S *S -5 i: -5 ^ -5 JI p I ■ '-3 I 
 
 .0 . _ . o t -6 . -a 
 
 o o of o 12 o *£ x > 
 
 £ X ** % Z « Z i> W.S < .5 
 
 1 )ent 46 12 39 7 o to 9 2.6 
 
 Flint 23 12 21 2 o to 8 3.4 
 
 Sweet 40 27 39 8 2 — 2 to 8 2.4 
 
 Soft - 11 2 10 1 o to 5 1.1 
 
 Pop (common) 8 4 7 1 o to 7 3.5 
 
 Pop (pearl ) 8 4 7 1 o to 6 2.0 
 
 Pop (rice) 14 7 12 2 o to 3 1.7 
 
 With the exception of being the first to bloom in each row, these plants were sup- 
 posedly normal. 
 
 Lazenby 52) reported observations which extended over several years upon 
 15 or more distinct varieties, as follows: 
 
PER CENT OF INDIVIDUALS SHKUliING POLLEN WHEN THE SILK APPEARED. 
 
 Some pollen 54- pollen All pollen 
 
 shed. shed. shed. 
 
 Dent 6 varieties 95 76 32 
 
 Flint 4 varieties 89 60 38 • 
 
 Sweet 8 varieties 98 82 44 
 
 As in our own experience, Lazenby found that it is not always easy to tell just 
 when the pollen is all discharged. With different tassels, we found the pollen 
 production to extend over a period of from four to eight days. 
 
 In 1875 Sturtevant 77a ) counted pollen grains in a large number of anthers, 
 and the number of florets in the tassels. From his results Sturtevant concluded 
 that the average number of pollen grains in an anther is 2500 and the number 
 of stamens in a tassel 7200. Allowing two ears of 1000 kernels to each plant, 
 there would be approximately 9000 pollen grains for every ovule to be fertilized. 
 Using Sturtevant's estimate of 2500 pollen grains in an anther, Lazenby 52 I com- 
 puted the ratio of number of ovules to number of pollen grains by counting the 
 flowers in several average sized panicles in dent, flint and sweet corn. 
 
 LAZENBV's CALCULATIONS. 
 
 Approximate ratio 
 No. of No. of pollen bear- No. of pollen of ovules to 
 
 ovules. ing dowers. grains. pollen grains. 
 
 Dent Corn 1520 9300 69750000 1 : 45000 
 
 Fint Corn 940 6500 48750000 1 : 52000 
 
 Sweet Corn 445 1560 1 1700000 1 : 26000 
 
 These figures would allow a range of from 25000 to 50000 pollen grains for each 
 silk to be pollenated where there is only one ear on the plant. However, as 
 Leavenby has pointed out, there is a great variation in the number of tassel (lowers 
 on different individuals in the same variety. There is also a difference in number 
 of ears and from what we know now of differences in the size of anthers and the 
 tassels on which they are borne, the differences in number of pollen grains must 
 be much greater than in the cases selected by Lazenby. The fact that in some 
 instances a large proportion of the pollen grains are infertile must also be 
 recognized. 
 
 The first staminate flowers to mature are usually those in the upper part of 
 the axis or central spike of the tassel the sequence being in both directions, but 
 mostly downward and. on the branches, inward. When there is sunshine, more or 
 less pollen is shed throughout the day, but the most of it falls during the forenoon. 
 On days that are favorable, another, shorter period occurs in the evening, a short 
 time before sunset. Either cold, wet weather or excessive heat may retard the process. 
 Manipulation of the flower spike in the hand has an accelerating effect, probably 
 due to friction and warmth of the hand, and a considerable quantity of viable 
 pollen may be secured from mature flowers in this way. 
 
 The distance to which corn pollen is carried depends largely on the nature 
 of the wind. A rising or whirling wind might readily carry pollen a much greater 
 distance than a straight wind. Various limits have been placed upon the distance 
 to which corn pollen is carried by the wind.— varying from 50 feet to a half mile. 
 In our own work, we found the last blue kernel in a yellow variety at a distance 
 of _o rods (north) from our breeding plot wbi,ch furnished the pollen for this 
 kernel. 
 
 A3 
 
Viability. 
 In breeding work with horticultural species, advantage has been taken of the 
 fact that many kinds of pollen are viable for weeks and months. Booth 10) 
 germinated grape pollen in New York, three weeks after it had been gathered in 
 California. Sandsten 70) secured apple and plum pollen from Washington, Tennes- 
 see, Missouri, and Minnesota; testing it in Wisconsin he found that "a small per- 
 centage of apple pollen retained its vitality for six months, while but little plum 
 pollen retained its germinating power this long." Supposedly authentic instances 
 have been reported in which grape, tomato, and date pollen has remained viable 
 two, six, and twelve months respectively 30). 
 
 Because of the difficulty experienced in getting very early and very late 
 varieties of corn to bloom at the same time, it would be very desirable to store 
 corn pollen or to be able to ship it over long distances, for hybridizing work. 
 McCluer 5;) reported that corn pollen seemed to retain its viability for several 
 days if kept dry. Webber 85) stated that corn pollen retained its viability at 
 least two weeks. 
 
 In our own work, more than a thousand tests were made on the viability of 
 pollen and silks. The time in which they were retained ranged from o to 25 
 days in all the combinations made, — fresh pollen on old silks, old pollen on fresh 
 silks, and the intermediate combinations. Xo reliable cases were found, including 
 hybrids iii all the groups and many varieties, in which shoots produced kernels 
 when pollenated with pollen that had been stored 30 hours. Many successful cases 
 were found in which the pollen was used in the morning and during the forenoon 
 of the day after the oce on which it was gathered. For satisfactory results it is 
 best to use corn pollen on the same day on which it is gathered, discarding what is 
 not used and getting a fresh supply on the morning of the next day in which 
 pollenating work is to be done. 
 
 These conclusions are supported by the work of others, in which corn pollen 
 was tested under different treatments and in various germinating solutions. .lost 47) 
 states that corn pollen remained viable only one to two days under optimum condi- 
 tions, Pfundt 68) found that pollen of corn remained viable only one day. alike 
 in all the conditions under which it was kept. 
 
 There are many instances in plant species in which self-pollenation results in 
 few or no viable seeds. Similar cases of sterility in hybrids are known. In an 
 attempt at hybridizing two varieties of corn many years ago Gartner reported 
 securing only five kernels. Sturtevant 79) stated that "The agricultural species 
 [sweet, Hint. dent, soft, pop] have a strung tendency to resist cross fertilization 
 with each other." Blaringhem 9) stated that many groups of corn do not hybridize 
 among themselves but cited no authority in support of the statement. East 27) 
 was unable to secure hybrids of Giant Missouri Cob Pipe, a large late dent corn, 
 and Tom Thumb, an early dwarf pop variety. 
 
 On the other hand Kellerman and Swingle 48* concluded that "dent, Hint, 
 soft, sweet, and pop corn cross as freely with each other as with the different 
 varieties in their own class." Collins 18) hybridized successfully the smallest 
 Tom Thumb with large Guatemala and Mexican varieties. Emmerson 102) obtained 
 a hybrid of Tom Thumb pup and a large, late dent corn 
 
 As has previously been stated (see page 14) we are aware of no cases in our 
 work in which widely diverse varieties such as are mentioned above would not 
 hybridize when fresh silks and pollen are available. 
 
 In order to learn whether there are any differences in the external appearance 
 of the pollen grains, pollen from a number of varieties was observed under tlu 
 microscope. Fresh pollen was gathered and placed in the shade, part placed on 
 glass slides, both covered and uncovered, and part of it being left in the bag; 
 
 44 
 
-3 
 
 0/ 
 
 o 
 
 
 
 
 .--• 
 
 r- 
 
 
 
 
 bo 
 
 a. 
 
 
 d 
 
 OJ 
 
 c 
 
 J 
 
 C/2 
 
 in which it was gathered. It was found that the grains which were plump at first 
 began to shrink immediately, giving it a deeper yellow color under the lens. In 
 a number of samples, no plump grains were found after tfiirty minute- in an hour 
 had elapsed and yet this shrunken pollen was viable when applied to silks. Ten 
 pollen grains of from one to three representative plants, in various varieties taken 
 at random, were measured and the empirical modes of the length-breadth measure- 
 ments as well as the shape or outline of the pollen grains are given below : — 
 
 Size of freshly gathered pollen grains of various varieties (mm.). 
 
 Variety Number and Name S — • S 
 
 ~z u — 
 
 - " ■- 
 
 ■z § CO 
 
 104.4-A Pod Corn (medium sized kernels) 30 .094 .146 clips, 
 
 104.4-B Pod Corn (medium sized kernels) 30 .090 .098 pyriform 
 
 204.14 Yellow Rice Pop (very small kernels 30 .078 .0S6 elipse 
 
 206.16 Red Pearl Pop. ..(small kernels) 30 .086 .098 pyriform 
 
 211. — Yellow Pearl Pop (Tom Thumbs — small kernels) 10 .oqo 094 elipse 
 
 212.22 White Rice Pop. I very small kernels ) 30 .078 .086 
 
 213. — White Rice Pop. (very small kernels) 30 .078 .086 
 
 304. — White Flint .... (very small plants — early).... 20 .086 .098 
 
 402.37 White Dent (Hickory King — large kernels 130 .082 .098 
 
 442.77 White Dent .... (Cob pipe com — med. kernels)3o .086 .098 pyriform 
 
 504. — Soft Corn (medium kernels) 30 .090 .098 elipse 
 
 612.92 Sweet Corn (Egyptian — medium kernels).. 30 .090 .102 
 
 From these measurements it appears that there is no difference in size nr shape 
 of pollen grains which will interfere with their successful germination and subse- 
 quent fertilization of the ovules of any of the varieties used. The ratio between 
 the modal length and breadth as shown in the above data bring out the fact that 
 corn pollen is not exactly round as it has been reported b\ others. With regard 
 to the inheritance of shape, normal long pollen grains have been reported to be 
 dominant to round pollen grains in sweet peas 3). It was thought that difference 
 similar to this might be found in corn pollen. 
 
 Correns 19) reported differences in receptivity to pollen when that of several 
 varieties was applied to an ear at the same time. We have obtained similar results 
 b\ using pollen from varieties of different colors and composition of kernels all on 
 the same shoot and evidently there is a difference in germinating power or recep- 
 tivity aside from mere viability of the pollen itself. 
 
 Difference in the shade of yellow in the fresh pollen of various varieties was 
 found, some having light yellow and others darker yellow colors. Aside from 
 this difference in color and perhaps in viability, we found nothing that would in- 
 dicate distinguishable genetic differences in pollen grains of corn. 
 
 The immature female inflorencence of corn, composed of the undeveloped ear 
 with its leafy involucre of bracts or "husks" upon a short stem or "shank", is com- 
 monly known as the "shoot". This name applies until the ovules have become 
 fertilized and considerable enlargement of the young ear has taken place. 
 
 45 
 
The theory concerning the differentiation of the inflorescence of corn into 
 "tassel" and "ear", as has been mentioned previously, includes the idea that the 
 shank or stalk of the ear-bearing branch has become shortened. This idea is 
 substantiated by the frequent occurrence of very long shanks on individual plants 
 in general field culture, the increase in length of shank being most frequent on 
 the lower shoots. In the limited number of individuals of this nature observed in 
 our work, the extremely long shanks were plainly due more to the increase in 
 length of the internodes than to an increase in their number. Montgomery 59) 
 stated that the number of nodes in the ear bearing branches coincided with the 
 number of nodes found in the main culm above the point of attachment of the ear 
 bearing branch, in a sweet corn plant which he studied. 
 
 From a number of counts we have made on several dent varieties, it appears 
 that Montgomery's statement will not hold in some, and perhaps all, varieties. 
 Only distinct internodes on the shanks were included in the counts and in the 
 majority of cases there was a larger number of internodes on the shank than on 
 the main culm above the point of attachment of the shoot; this is very evident 
 where the ear is borne at the second or third node below the tassel. Incidentally 
 it was noted that the number of these upper internodes on the culm was more con- 
 stant in a variety than the number of shank internodes. If each husk on the 
 shoot is counted and an internode assigned to all that are not paired or multiple, 
 we would have many more internodes on the shoot than on the upper part of the 
 culm (above the ear-bearing node). It is often difficult to get satisfactory counts 
 on shank internodes because of telescoping near the ear. Certain varieties — 
 notably pop corns — tend to produce secondary shoots on branches in the axils of 
 the husks on the main ears; these secondary branches may have as many or more 
 apparent internodes than are found on the main shi.nk. 
 
 After four years of selection for erect and declining ears from a strain 
 which had been previously cross-bred* for seven generations. Smith 76) 
 reported : "that the ear branch or shank was on the average almost twice as long 
 in the declining ear strain as in the erect-ear strain, the averages for that year 
 being about twelve inches and seven inches respectively — the longer shanks had an 
 average of a slightly larger number of internodes and correspondingly a larger 
 number of husks." 
 
 In all the groups, individual plants may be found with very long shanks, but 
 as yet we have no data upon their inheritance. Certain varieties tend to produce 
 long shanks while others produce very short shanks but the character is a 
 fluctuating one. The behavior of the Xi hybrids, in which flint, dent, sweet, and 
 soft varieties with long shanks were used with other varieties having short shanks, 
 indicate that long shanks are probably dominant to short shanks. 
 
 Barrenness. 
 
 "Barrenness", or failure to produce ears, differs from sterility in that the entire 
 inflorescence may fail to develop, thus the fertility of the flower does not enter 
 into consideration. Barrenness in corn may include failure to produce a tassel, as 
 has already been stated, but the large majority of cases deal with the failure to 
 produce normal shoots. 
 
 Detasseling has been resorted to in order to prevent the pollen from barren 
 plants falling on the silks of individuals from which seed is to be saved for 
 planting. Some uncertainty exists in this practice because of the difficulty of 
 determining truly barren plants in some varieties at the time at which detasseling 
 is done. Bull 11) reported that barren plant-- "may lie discovered by the absence 
 of the enlarged leaf sheath just above the joint where the ear should appear"; 
 
 *Bre<l by individual? within the variety or strain. 
 
 4" 
 
while Soule and Vanatter --) stated that they were unable to distinguish barren 
 plants, at the time of tasseling, from those which develop-late and eventually pro- 
 duce ears. 
 
 We are aware of no case of reported barrenness which lias not been influenced 
 largely by either environment or decreased vigor and it was at one time suggested 
 that if barrenness is hereditary, the character would tend to eliminate itself 43, 45), 
 but this was before it was generally known that characters detrimental or even 
 fatal to the perpetuation of the species in the pure or homozygous state may yet be 
 transmitted indefinitely by means of heterozygosis. 
 
 Plants upon which shoots develop extremely late are undesirable and such 
 plants usually produce nubbins or poor ears and indicate weakness in general. 
 It has been suggested that the best way to overcome this evil would fie to discard 
 the parent ears entirely and to select from such ear-rows in a breeding plot that 
 produc vigorous shoots. Varieties differ very much in regard to size and develop- 
 ment of shoots at tasseling and silking time, some varieties producing silks before 
 the shoots leave the leaf sheaths while others develop large shoots before the silks 
 appear. By delaying detasseling till just before the pollen begins to fall it is 
 possible to differentiate between vigorous and weak plants in many varieties at least. 
 
 It sometimes happens that varieties producing some barren plants possess other 
 characters which are desirable and since breeding blocks are generally limited in 
 size the small amount of additional work necessary to detassel undesirable plants 
 would be well worth wdiile not only with regard to failure to produce good ears 
 but also with regard to other undesirable features which are known to be trans- 
 mitted to the progeny. 
 
 Since no ease has yet been reported where the progeny of a hand-pollenated ear 
 fertilized by the pollen of a barren plant have all been barren, it would seem that 
 this behavior — provided it is hereditary — must be either a recessive or a composite 
 character and he transmitted contrary to the behavior of sterile anthers mentioned 
 on a preceeding page. 
 
 Husks. 
 
 The arrangement of husks about the ear varies within an individual shoot. 
 They are most frequently two ranked and imbricate, but other arrangements are 
 to lie found, such as; paired, convolute, mixed arrangement, and multiple — not 
 imbricate — arrangements, at husking time the outer ends or tips of the husks may 
 be found on one side of the ear. This warping of the dead husks proceeds far 
 enough in many cases to expose the ears on the opposite side. 
 
 In harvesting the crop in 1909, distinct differences were noted in the amount of 
 husks about the ear and in 1910 a number of counts were made on different 
 varieties and their Xi generations. In securing these counts, all the husks on the 
 shoot including those on the shank were taken. This was done at harvest time 
 after full development had been attained. 
 
 A wide fluctuation in the varieties, 111 number of busks on the upper ear. was 
 found. This wide fluctuation may be partially accounted for by the frequent 
 prolifications of the inner husks. The number of parts in these prolifications is 
 variable. Although extremes of six and twenty-eight husks were found, practically 
 all the varieties overlapped in this character and variable results would therefore 
 be expected in their progeny. 
 
 If it can he shown that the number of internodes on the shank are the same as 
 the normal number of husks then it is evident that many of the husks are prolifica- 
 tions. Had it been possible to eliminate all proliferous husks and thus determine 
 the number of primary ones (one to each internode) we would have expected more 
 definite results. 
 
 47 
 
If prolitication is a recessive character in husks, as it appeared to he in the 
 case of the ear mentioned on page 37. many of. the contradictory cases in the Xi 
 generations might be explained by this one fact alone. In some instances there 
 may have been pairing of proliferous and non-proliferous character while in others, 
 both parents may have had the proliferous character. The conclusion reached 
 during the study of this character was that many normal husks are dominant to 
 few normal husks. 
 
 General notes were taken on other characters of the husks. Thick and thin 
 husks (without regard to number or width) were found to be varietal characters 
 and in a number of Xi hybrids, thick husks were plainly dominant to thin husks. 
 Thickness of the total layer or involucre, depends not only on the thickness of each 
 husk and on the number of husks, but also on the width or degree of wrapping of 
 the ear by each individual husk. In the Xt hybrids, the wide and well wrapped 
 husks were dominant to the narrow type. 
 
 "The Central West desires a corn with loose shucks so as to be easily husked, 
 while for protection against insects, the Southern States require a tightly fitting 
 husk" 40). Market gardeners desire young ears well covered with husks which 
 permit storing the roasting ears for a longer period than do the ears of those 
 varieties poorly covered with husks. East 24) reports finding a row of sweet corn 
 in an ear-row test of Stowell's Evergreen, in which the ears were especially well 
 covered with husks and remained in table condition for over a week ( presumably 
 after picking) while ears from other rows remained in prime condition only one day. 
 
 On some varieties used in this study the husks were very loose even though the 
 involucre was relatively thick, this was especially noticeable before the kernels on 
 the ear had enlarged to any extent. The husks remained loose and creased through- 
 out the season in some varieties while in others they were tightly stretched and 
 forced to expand as the ear developed. This firm, compact, type of shoot was 
 dominant to the loose type in the Xt hybrids and was very noticeable as the young 
 ears were developing in the field. 
 
 The Mexican variety No. 408 produced shoots with many, wide, thick husks, 
 the inner ones of which were folded or rugose longitudinally in the upper region 
 and laterally rugose at the base. These rugose husks were constant in the 408 
 variety and dominant in its Xi hybrids. Collins 18) reports a hybrid in which the 
 "inner husks were crumpled at the base of the ear, a not uncommon condition with 
 thick husked varieties." 
 
 The absolute length or size of husks on the hybrids behaved much like the 
 size of plants, being intermediate in some cases and in others equal to or greater 
 than those of either of the parents. Varieties producing long ears necessarily 
 develop long shoots and husks and length in these instances is dominant to shortness. 
 
 With regard to the relative length of the husks and ears or cobs, the behavior 
 discloses an interesting difference. When hybrids were produced between a variety 
 in which the mature husks extended beyond the tip of the ear and a variety in 
 which ear and husks were conterminous, the hybrid invariably exhibited the latter 
 character. But when either of these two types were hybridized with a type in which 
 the ear protruded, the protruded ear character was dominant. All the plants of 
 the pod variety No. 101 displayed the protruded ear character which was always 
 dominant in its Xi hybrid progeny. One of the Mexican varieties (No. 410) 
 produced approximately 50 per cent of individuals with protruded ears and this 
 was a distinguishing character in its Xi hybrids. Collins 18) found a behavior 
 similar to this in two cases in which a large number of the ears protruded beyond 
 the husks in the Xi hybrids. In my results it appears that while long husks are a 
 dominant character as well as the long ear, the long ear yet exceeds the long husks 
 in relative as well as absolute length. 
 
 48 
 
The husks of must varieties arc green liut in some a red tinge is frequent. 
 This red tinge, which is often associated with red tinge on the culm, appeared in 
 the Xi generation from parents having green husks hybridized with other parents 
 having red tinge husks. Red husked plants were frequent in the variety No. 4"* 
 as well as in the hybrids in which it was one of the parents, and homozygote red 
 husks would be expected to be dominant to red tinge and pure green husks. 
 
 Laminae. 
 
 Many years ago, Goethe, the great German naturalist-poet, announced the theory 
 that floral parts are nothing more than modified leaves. This idea is perhaps no 
 more beautifully illustrated than in the enveloping involucre or husks of an ear 
 of corn. When the shank in the shoot is relatively long the lower husks assume 
 the form of true leaves. With a tendency toward a shorter shank, there is a 
 decrease in the leaf-blade until the leaf sheath, which has become a husk, bears a 
 very small blade or none at all. 
 
 In some varieties these small leaf blades or "laminae" persist on the outer husks 
 and in some individuals (occurring in all the species groups) these laminae may 
 equal the length of leaves — two feet or more in length in extreme cases. Develop- 
 ment of unusually long laminae is generally associated with long shanks and small 
 ears. In those in which extremely bread laminae were developed, the size of the 
 ear was also generally diminished. 
 
 As a varietal character, long laminae are more often found in the flint, sweet, 
 and soft groups. Varieties producing very short or no laminae are more often 
 found in the dent and pop groups. 
 
 In studying the behavior of length of laminae, we are dealing again with a 
 fluctuating character and one which, similar to the suckering character, seems not 
 entirely lost in any variety. In addition to this fluctuation between plants in the 
 parent strains, there is also another and similar fluctuation upon the individual 
 plant. The laminae on the lower shoots are nearly always longer than the laminae 
 on the upper shoot, and a similar increase in length exists on a single shoot from 
 the tip downward. 
 
 Because of this variation in length of laminae, it was found impracticable to 
 secure exact measurement and the data were taken with respect to the length of 
 laminae on the upper shoot only and as compared to the relative length of the 
 shoot on which they were borne. While this method was only an approximate one, 
 the data were all secured by the same person and, aside from a possible bias in 
 judgment, the classification was uniformly made in all the parent varieties and Xi 
 progeny. 
 
 It was noticed that few of the varieties are constant to the "Type Group" in 
 which they fall and distinct genetic differences within a variety, and consequently 
 variable results in the hybrids, are to be expected. Seven of the ten hybrid eases 
 which gave an "Average" result were such in which the parents had like modes 
 and the hybrids produced the same. Eliminating these seven cases, the large num- 
 ber of the hybrid results would appear to fall at either extreme as shown in the 
 summary. This should not occur if there is a genetic difference in length of 
 laminae. 
 
 Among the remaining 29 cases on which data were taken, a small majority 
 indicate dominance of the long laminae. Certain hybrids displayed this dominance 
 of long laminae so markedly that we would expect this result in all cases where 
 homozygous parents are used. The mixture in varieties is evidenced by the results 
 secured in the closes in which there were distinct differences in behavior while all 
 the Xi results would have been expected to fall in the "average" list. The selfed 
 cases as usual are limited in number but indicate a decrease in the length of laminae. 
 
 49 
 
Occasional individuals occur having abnormally broad laminae, the inheritance 
 of which has not been determined. One variety (No. 214B) frequently having 
 short, broad laminae gave, in its hybrid with a variety (No. 210) having short nar- 
 n >\\ laminae, medium and broad laminae. 
 
 The colors of the laminae are generally the same as those of the leaves on the 
 plant and behave similarly in transmission. 
 
 Silks. 
 
 The most conspicuous phenomenon in connection with the development of the 
 shoot is the appearance of the silks. Their growth is very rapid and their period 
 of usefulness limited. When pollenation of these silks or "styles" is prevented they 
 may continue to grow for a number of days. Lazenby 52) reported that "Some of 
 the silks [of a dent variety] attained a total length of over eighteen inches." 
 
 In our study of the receptivity of silks of various ages, a number of individuals 
 were found which had attained a length of 8 to 14 inches beyond the tip of the 
 shoot. This behavior might be considered as an adaptation to provide a fresh 
 surface upon which the pollen may lodge, but the length to which the silks may 
 grow is apparently correlated with the length of the cob from which they arise. 
 
 As has been reported under "Technic", the silks are receptive at any point and 
 clipping them back does not interfere with the subsequent fertilization of the ovules 
 by pollen applied after the silks have been clipped. This test was also applied by 
 Kellerman and Swingle 48), and in addition they opened the husks of some shoots 
 before the silks appeared and succeeded in getting poorly filled ears by applying 
 pollen in this way. From our own observations in repeating this experiment, it 
 was found that only those silks were receptive which would have appeared very 
 soon after the time at which the husks were opened. When the husks were stripped 
 back on the shoots which had silks well exposed, and pollen applied to the portion 
 of the silks within the shoot, successful fertilization was effected. 
 
 The silks, which are double in structure, are more or less hairy near the 
 extremity and an erroneous idea has been given in popular descriptions of the 
 process of fertilization of the silks, by the statement that these hairs catch and hold 
 the pollen in their grasp. The hairs provide an increased surface upon which the 
 pollen may lodge, but the moist surfaces of the silks and pollen grains alone serve 
 to hold the grains which must germinate very quickly judging from the very short 
 period in which corn pollen is viable when exposed in a normal atmosphere. 
 
 Some writers state that the silks from the base of the ear appear first and the 
 ovules from which they come are consequently fertilized before the others on the 
 ear 11). This statement is true only 111 the broad sense — meaning the lower portion 
 of the ear. Among the many ears secured in our work by hand pollenating silks 
 of various ages, we found many imperfectly filled butts and bare tips on ears which 
 were pollenated soon after the first silks appeared. 
 
 The double pollenated ear — Xo. 2S3.170 — was an instance of this nature. The 
 dam of this particular ear was a white pearl pop (No. 210). On the first day on 
 which the silks appeared, it was pollenated from its own tassel (selfed). Six 
 days later, pollen was applied from a variety having a yellow- endosperm ( Xo. 309). 
 As has been stated, the selfed and hybrid kernels were determined by the differnce 
 in color and a distinct band of pure white kernels was found at some distance above 
 the butt of the ear. The remainder of the kernels below and above this band were 
 yellow. A number of other and similar cases were found. 
 
 At the end of the six days, the silks at the tip of this particular ear had not 
 yet developed for the reception of pollen as was shown by the bare cob at this 
 point. This regularity of distribution does not appear on all ears, many of which 
 have irregular patches of kernels evidently due to the uneven development of silks 
 
 SO 
 
on various parts of the spike. From the many hand-pollenated ears secured, the 
 longest interval of time between the hrst apeparance of-silks and the last date on 
 which the younger silks on the same ear were yet receptive was thirteen days, and 
 in this case only a few kernels at the tip of the cob were secured. Entire ears may 
 be pollenated fully in the natural way in a period ranging from two to ten days, 
 depending upon the variety and conditions. 
 
 Differences in the size and structure of silks were found but no attempt was 
 made to study their behavior in the hybrids. In varieties having colored silks 
 (other than pure green or yellowish green) the hairs on the silks are the first 
 and frequently the only parts to become colored. 
 
 No variety was found which was homozygote to silk color lint the populations 
 from selfed seed showed exact uniformity in color of silks in most of the cases. 
 In one of these Xi selfed cases (No. JOiXS) silks were produced having a dark 
 yellowish-green color; the behavior of this color on hybridization has not been 
 determined. Several cases were found in which dark green silks were dominant 
 to light green silks in the Xi generation. Green silks X green silks in most cases 
 gave green silks only, but several cases were found in which only parents with 
 green silks had been used, where the Xi generation showed some individuals with 
 red tinge or reddish green silks. Without a knowledge of the ancestry previous to 
 1909. we are unable to state whether such cases are due to heterozygosis which did 
 not shown in 1909 or whether they are due to new combinations in the X generation. 
 
 The notes on the silks were taken somewhat early in their development in 
 many cases and it is possible that some of the parents developed a red color after 
 the notes were taken. In general the order of dominance of silk color appears tc. bi- 
 as follows : dark red, reddish green, red tinge, to green. Numerous cases have 
 been reported by others in which the deep red silk appears to be dominant to the 
 green and red tinge silk. 
 
 The behavior of the various shades of green fall in a separate category from 
 that of the reds, and either of the greens may be obscured by the red color. It '•> 
 very doubtful whether red tinges which are absolutely constant to degree of colora- 
 tion can be secured. 
 
 Position of the Upper Shoot 
 
 Instances have been reported which permit the general conclusion that shoots 
 may be produced in the axil of any leaf on the culm or its branches. Among the 
 varieties used in this investigation we found ears at various heights on plants and 
 at all internodes below the tassel, while small ears at the base of the tassel are 
 not uncommon, but a rather definite location for the production of the main ear- 
 bearing shoot exists in the different varieties. The number of upper internodes 
 (between the tassel and the upper or main ear-bearing shoot) is more constant 
 throughout the varieties of corn, than is the number of lower internodes (below 
 the upper shoot ). 
 
 The results secured at the Illinois Experiment Station 76)- in selecting for 
 "high" and "low" ears, gives evidence in support of the above statement. Height 
 of ear in inches above the soil was made the basis of selection in this instance. 
 Data were also secured upon the total height of the plant, number of internodes 
 below the ear, and the total number of internodes on the plant. If from this 
 data we compare the difference between the number of internodes below the ear 
 and the total number on the plant, we find that this difference (7 internodes) was 
 always the same in both plots except in one year. In this aberrant year, the 
 "high-ear" difference was six internodes and the "low-ear" difference only bve inter- 
 nodes. The cause of this is nut known, but it is important to note that the excep- 
 tions occurred on both plots in the same year. If the deviation of one were added 
 
 51 
 
to the total number of internodes given for the "high-ear" plot and the deviation of 
 two to the number given for the "low-ear" plot in that year, it would at least make 
 the "total internode" columns for the six years more uniform as well as .to give 
 exact uniformity in the number of upper internodes in both plots for all the years. 
 
 From the data, it is also evident that there has been no change in the "high- 
 ear" plot in the height of ear, neither in inches nor in number of internodes. 
 Neither is there any change evident in this plot in regard to total height of plant 
 and total number of internodes. The difference (three feet), which now exists 
 between the two plots is due solely to a change in the "low-ear" plot in which the 
 height of ear, both in inches and number of internodes, is but little greater than 
 half of what it was at the beginning. There has also, apparently, been no change 
 in length above the ear in the low ear selections ; the average length in inches above 
 the ear being slightly more in the low-ear plot than in the high-ear plot. 
 
 This decrease, which has been found in the "low-ear" strain, suggests the 
 idea that homozygous "low-ears" have been selected and mated from within a 
 heterozygous population. The fact that the "High-ear" strain has remained con- 
 stant with only slight fluctuations, readily attributable to seasonal differences, sug- 
 gests in a similar way that the parents of this strain were either homozygous "High- 
 ears", or that this is the dominant character and that extracted "High-ears" have 
 been selected, thus maintaining but not materially increasing the height of the 
 strain. As in the case of the differences in oil and protein secured in other strains, 
 no one can say positively as yet what is happening in these experiments of far-reach- 
 ing importance. 
 
 Data were taken to secure exact information upon the behavior of the relative 
 height of the upper shoot. It is significant that there is a very close uniformity 
 between all the parent varieties in the number of upper internodes ( above the upper 
 shoot), irrespective of the number of lower internodes. It is also significant in 
 comparing cases, that this internodal height of the upper shoot or ear is not affected 
 by the number of shoots on the culm. 
 
 As was stated previously ; it is difficult to determine the internodes at the base 
 of the plant at the time of maturity, and it is evident that a shortening of inter- 
 nodes at this point is partially the cause of fluctuation in the number of lower inter- 
 nodes, apparent at maturity, in some varieties. 
 
 The reciprocals indicated small differences but in general these cases are fairly 
 uniform. 
 
 The summary of the data shows that a large majority of the hybrids are 
 above the average of the parents in number of lower internodes. The results in 
 the closes and selfs indicate that individual parents were used which had a greater 
 number of lower internodes than the mode of the variety from which they came; 
 a result contrary to that which would be expected from inbred progeny. 
 
 The summary of results with respect to the number of upper internodes is 
 indefinite and had the study of relative height of the upper shoot been made only 
 with respect to its distance from the tassel, no definite conclusions would have 
 been reached. The extremes in the number of upper internodes in all the parent 
 modes are five and eight, which, with the existing fluctuations, allows the parent 
 arrays to overlap and does not provide sufficient difference for marked differenta- 
 tion in the > , hybrids. The fact that the majority of the fourteen hybrids fall in 
 the average group is attributable to the use of parents which were very similar in 
 this respect. Xo definite concluiosns can be made on the behavior of the Xi closes 
 and selfs. 
 
 In order to check the behavior of the upper and lower internodes simultaneously 
 on the same population, the "Relative Height on the Plant" was computed from 
 the sum of the two modes. In the data secured in this way, we find that in the 
 large majority of cases the relative height of the upper shoot in the Xi hybrids 
 
 5-' 
 
is equal to or greater than that of either of the parent modes, therefore, relatively 
 high ears on the corn plant must he a dominant character. 
 
 Number of Shoots. 
 
 Perhaps no organ of the corn plant is more dependent upon environment and 
 vigor of growth than the shoot containing the undeveloped ear. While certain 
 varieties normally produce a large number of shouts in comparison to other 
 varieties, adverse conditions may inhibit their development entirely. 
 
 While the large majority of shoots on the lower part of the stem may fail to 
 produce ears, their partial development is an indication of an ear bearing tendency, 
 and as such is important in a study of characters in the corn plant. How the 
 difference among varieties in shooting tendency and prolificacy in ear production 
 came to exist, is not known, but it is evident that there must be a close correlation 
 between this and other characters in the plant. As an example, the production of 
 an extremely large ear in the upper shoot is not associated with a large number 
 of ears, and those varieties producing a large number of shoots generally have 
 small ears. 
 
 It is evident that the production of shoots is closely related to the production 
 of suckers, and individual plants are frequently found in which it is difficult to 
 distinguish between the lower shoots and the suckers. Card 12) made selections 
 in a variety of sweet corn for large number of ears on the plant, and found that 
 the increase which he secured came chiefly from the increase in the number of 
 ear-bearing suckers. 
 
 My data was obtained by counting all the shoots which extended beyond the 
 leaf-sheaths on the main culm of each plant. This data, therefore, includes a very 
 large proportion of shoots within which ears failed to develop. Data were also 
 secured upon the number of ears harvested and, as would be expected, those 
 varieties producing a large number of shoots also tended to produce a large 
 number of ears. The same parents and hybrids used in the study of behavior of 
 number of internodes were used in this instance, and it is evident that with a 
 given number of internodes sufficient to permit the development of numeri us 
 shoots, the number of shoots actually developing does not appear to have any rela- 
 tion to the relative height at which they are borne. 
 
 As usual, interesting differences in behavior were found when one parent 
 variety was used in different combinations with other varieties. The reciprocals in 
 this table gave quite uniform results as is shown by the arrays in each case. 
 
 The summary of the X, hybrids places the large majority above the average 
 of the parents in the production of shoots. As has been noted, the production of 
 shoots depends to a large extent upon vigor of growth and therefore hybridization 
 should favor the development of a large number of shoots. Leaving out the n 
 cases which gave a number of internodes above those of the greater parent, which 
 result might be attributed to increased vigor, the remaining hybrids in the sum- 
 mary yet show that a "large number cf shoots" on the main culm is dominant to 
 a "small number of shoots". 
 
 Close pollenation in the majority of cases evidently had no limiting effect upon 
 the development of shoots in this one season. Among the selfed cases only two 
 showed a decided decrease in the production of shoots, while one (310XS) of the 
 six cases produced a mode slightly above that of the parent mode of the previous 
 year. A comparison of all the parent modes in both years shows that the modes 
 for 1910 were generally higher than those in 1909; therefore, the selfed case, having 
 an apaprent slight increase, must have been the result of difference in seasons. 
 
 As is the case with all the other characters discussed in this paper we must 
 consider not only environmental influences but the probable genetic differences 
 
 53 
 
within the populations studied. The important fact brought out by the summaries 
 of the various tables is : that in order to get definite results we must have individuals 
 pure with respect to the character with which we are concerned, or at least observe 
 their behavior in the second controlled generation when the data carefully secured 
 may be just as reliable, regardless of the mixture in the original parents. 
 
 54 
 
APPENDIX 
 
 The writer, Walter Byron Geniert, graduated from the Kansas State Agricul- 
 tural College in June 1907 with the degree of Bachelor of Science in Agriculture. 
 In September 1907 he entered the Graduate School of the University of Illinois, 
 from which he received the degree of Master of Science in Agronomy in 1909. 
 
 During the years igo5-'o7 he was Student Assistant in the Kansas Experiment 
 Station and was later connected with the Illinois State Soil Survey through the 
 field seasons of 1907 and 1908. He held a Fellowship in Agronomy in the Uni- 
 versity of Illinois during the academic years 1908-'! 1. 
 
 The writer desires to express his gratitude to Dr. L. H. Smith, under whom 
 this investigation was made, and to acknowledge his indebtedness to others who 
 have rendered much valued assistance. 
 
 55 
 
REFERENCES 
 
 1. Bailey, L. H. Plant Breeding. New York, 1908. The Macmillan Co. 
 
 2. Balls, L. Some Cytological Aspects of Cotton Breeding. Proc. Amer. 
 Breeders' Assn., V. 10x19, p. 16. 
 
 3. Bateson, W. Mendel's Principles of Heredity. Cambridge, 1909. 
 
 4. Baur, E. Unters. iib. d. Erblichkeitsverhaltnisse einer.nur in Bastardform 
 lebensfahigen Sippe von Antirrhinum inajus. Ber. Deut. Bot. Ges., XXV. 1907, p. 442. 
 
 5. Baur, E. Das Wesen u. die Erblichkeitsverhaltnisse der "Varietates 
 albomarginatae hort" von Pelargonium zonale. Zts. f. Indukt. Abstain, u. Verer- 
 bnngslehre, I, 1909, p. 330. 
 
 6. Beal, W. J. Horticultural Experiments. An. Rep. Mich. State Board Agr., 
 1876, p. 206. 
 
 8. Biffen, R. H. Mendel's laws of Inheritance and Wheat Breeding. Journ. 
 Agr. Sci., I, 1905, p. 4. 
 
 9. Blaringhem, L. Mutation et Traumatismes. Paris, 1908. Felix Alcan. 
 
 10. Booth. N. O. A study of Grape Pollen. N.Y. Agr. Exp. Sta. Bui. 224, 
 1902, p. 291. 
 
 11. Bull, C. P. Corn Breeding in Minnesota. Minn. Agr. Exp. Sta. Bui. 107, 
 1908, p. 177. 
 
 12. Card, F. W. Corn Selection, R. I. Agr. Exp. Sta. Bui. 116, 1906, p. 19. 
 
 13. Castle, W. E. and Allen, G. M. The Heredity of Albinism. Proc. Amer. 
 Acad. Arts and Sci., XXXVIII. 1903, p. 603. 
 
 14. Heredity. Pop. Sci. Mo. IXVI. 1910. p. 417. 
 
 15. Collins, G. N. Apogamy in the Maize Plant. Cont. U. S. Nat. Herb., 
 XII. Part 10, 1909, p. 453. 
 
 16. The importance of Broad Breeding in Corn. Bureau Pit. 
 
 Ind. U.S.D.A. Bui. 161, 1909. 
 
 18. The Value of First-generation Hybrids in Corn. Bureau 
 
 Pit. Ind. U.S.D.A. Bui. 191, 1910. 
 
 19. Correns, C. Bastarde zwischen Maisrassen, mit besonderer Beriiucksich- 
 tingung der Xenien. Bibliotheca botanica. Heft 53, 1901. 
 
 20. Verebungsversuche mit blassgelbgrimen und buntblattrigen 
 
 Sippen bit- Mirabilis Jalapa, Urtica pilulifera, und Lunarict annua, Ztz. f. Indukt. 
 Abstain, u. Verbungslehre. T. 1909, p. 291. 
 
 21. Cuknot. L. La Ioi de Mendel et l'heredite de la pigmentation chez les 
 souris. Arch. Zool, exp. et gen. Notes et Revue, 1902. 
 
 22. Davenport, E., and Rietz. H. L. Type and Variability in Corn. Til. Agr. 
 Exp. Sta. Bui. 119, 1907, p. 1. 
 
 23. East, E. M. The Relation of Certain Biological Principles to Plant Breed- 
 ing. Conn. Agr. Exp. Sta. Bui. 158. 1907. 
 
 24. Report of the Agronomist. Conn. Agr. Exp. Sta. Bienn. Ret. 
 
 1907-1908, p. 397. 
 
 25. A Note Concerning Inheritance in Sweet Corn. Science, N.S., 
 
 XXIX. 1909, p. 173. 
 
 26. The Distinction between Development and Heredity in In- 
 breeding. Amer. Nat. XLTTT. 1909, p. 173. 
 
 27. A. Mendelian Interpretation of Variation that is Apparently 
 
 Continuous, Amer. Nat. XLIV. 1909, p. 165. 
 
 28. The Genotype Hypothesis and Hybridization. Amer. Nat. 
 
 XI.V. 101 1. p. 160. 
 
 29. Emerson, R. A. The Inheritance of Sizes and Shapes in Plants. Amer. 
 Nat. XLTV. 1910. p. 739. 
 
 30. Evans, W. H. Rcpt. Tnt'nat. Conf. Pit. Breed, and Hyb. Exp. Sta. Rec. 
 XVI. 1902, p. 208. 
 
 56 
 
3i. Ewing, E. C. Correlation of Characters in Corn. Cornell, N.Y. Agr. 
 Exp. Sta. Bui. 287, 1910, p. 67. 
 
 32. Farabee, W. C, and Castle, VV. E. Notes on Xegro Albinism. Science, 
 X.S., XVII. 1903, p. 75. 
 
 33. Gardners' Chron., 1866., p. 145. Adv. 
 
 34. Hackel, E. The True Grasses, trans. Scribner and Southworth, N.Y., 1890. 
 
 35. Halstead, B. D. Experiments in Crossing Sweet Corn. N.J. Agr. Exp. 
 Sta. Bnl. 170, 1904. 
 
 36. Report of the Botanist. Twenty-sixth Ann. Rpt. N. J. Agr. 
 
 Exp. Sta., 1905, p. 425. 
 
 37. Report of the Botanical Department X.J. Agr. Exp. Sta., 
 
 1909, p. 269. 
 
 38. Hakshberger, J. VV. Maize: A Botanical and Economic Study. Contrib. 
 Bot. Lab. Univ. Peima., I. No. 2, 1893. 
 
 39. Hartley, C. P. Improvement of Corn by Seed Selection. Yearbook 
 U.S.D.A., 1902, p. 539. 
 
 40. Corn-breeding Work in the United States. Proc. Amur. 
 
 Breeders' Assn., I. 1905, p. 33. 
 
 41. Value of Corn Pollen from Suckers vs. from Main Stalks. 
 
 Ibid, II. 1906, p. 141. 
 
 42. Progress in Methods of Producing Higher Yielding Strains 
 
 of Corn. Yearbook U.S.D.A., 1909, p. .517. 
 
 43. Hopkins, C. G. Methods of Corn Breeding. 111. Agr. Exp. Sta. Bui. 82, 
 1902, p. 525. 
 
 44. Humbert, E. P. A quantitative Study of Variation, Natural and Induced, 
 in pure Lines of Silene Noctiflora. Zts.f. Indukt. Abstam. u. Verebungslehre, IV. 
 1911, p. 161. 
 
 45. Hunt, T. F. The Cereals in America. New York, 1905. Orange Judd Co. 
 
 46. Johannsen, W. The Genotype Conception of Heredity. Amer. Nat, 
 XLV. 1911, p. 129. 
 
 47. Jost, L. Zur Physiologie des Pollens. Ber. d. Deut. Bot. Ges., XXIII. 
 
 1905, p. 504- 
 
 48. Kellerman. W. A., and Swingle, W. T. Experiments in Crossing Varieties 
 of Corn. Second Ann. Rpt. Kans. Agr. Exp. Sta., 1889, p. 288. 
 
 49. and Thompson, C. H. Crossed Varieties of Corn, Third 
 
 Year. Kans. Agr. Exp. Sta. Bui. 27, 1891, p. 139. 
 
 50. Keyser. A. Variation in Wheat Hybrids. Proc. Amer. Breeders' Assn., 
 
 II. 1906, p. S4. 
 
 51. Koerxicke. F.. and Werner, A. Handbuch des Getreidebaues. 2 vols., 
 1885, Bonn. 
 
 52. Lazenby, W. R. The Blossoming and Pollination of Indian Corn. Proc. 
 Soc. Prom. Agr. Sci., 1898, p. 123. 
 
 53. Lock. R. H. Experiments with Maize Annals Roy. Bot. Gard. Perydeniya, 
 
 III. Pa. II. 1906, p. 95. 
 
 54. Love, H. H. Influence of Food Supply on Variation. Proc. Amer. 
 Breeders' Assn., V. 1909, p. 357. 
 
 55. Lyon, T. L. Experiments witli Com Neb. Agr. Exp. Sta. Bui. 01, 1005. 
 
 56. Experiments with Corn. Neb. Agr. Exp. Sta. Bui. 112. 1000 
 
 57. McCluer. G. W. Corn Crossing. 111. Agr. Exp. Sta. Bui. 21. 1891. p. 82. 
 
 58. Mendel, G. Experiments in Plant-hybridization, trans, in Bateson's- Men- 
 del's Prin. of Hered. Cambridge, 1909. p. 317- 
 
 59. Montgomery, E. G. What is an Ear of Corn? Pop. Sci. Mo., LXVIII. 
 
 1906, p. 55- 
 
 60. Note regarding Maize Flowers. Sci, N.S., XXXIII, 191 '• 
 
 P. 435. 
 
 57 
 
61. ' Experiments with Corn. Neb. Agr. Exp. Sta. Bui. 91, 1905. 
 
 62. Morrow., E. G., and Gardner, F. D. Field Experiments with Corn. 111. 
 Agr. Exp. Sta. Bui. 31, 1893, p. 173. 
 
 63. Experiments with Corn. 111. Agr. Exp. Sta. Bui. 31, 1893, 
 
 P- 333- 
 
 64. Newbegin, Miss M. J. An Attempt to Classify Common Plant Pigments, 
 with some Observations on the Meaning of Color in Plants. Trans, and Proc. 
 Bot. Soc. Edinb. XX. 1896, p. 534. 
 
 65. Orton, W. A. Breeding Disease Resistant Plants. Proc. Araer. Breeders' 
 Assn., I. 1905, p. 202. 
 
 66. On the Theory and Practice of Breeding Disease-Resistant 
 
 Plants. Ibid, IIII. 1908, p. 144. 
 
 67. Pearl, R., and Surface, F. M. Experiments in Breeding Sweet Corn. 
 Maine Agr. Exp. Sta. Bui. 183, 1910, p. 249. 
 
 68. Pfundt, M. Der Einflusz der Luftfeuchtigkeit auf die Lebensdauer des 
 Bliitenstaubes. Jahrb. f. wiss. Botanik, XLVII. 1909, p. 1. 
 
 69. Price, L. H., and Drinkard, A. W. Inheritance in Tomato Hybrids. Va. 
 Agr. Exp. Sta. Bui. 177, p. 18. 
 
 70. Sandsten, E. P. Some Conditions which Influence the Germination and 
 Fertility of Pollen. Wis. Agr. Exp. Sta. Research Bui. 4, 1909, p. 149. 
 
 71. Shamel, A. D. The Effect of Inbreeding in Plants. Yearbook U.S.D.A., 
 
 1905, P- 377- 
 
 ■ji. Shull, G. H. The Composition of a Field of Maize. Proc. Amer. 
 Breeders' Assn., IV. 1908, p. 296. 
 
 yi. Some New Cases of Mendelian Inheritance. Bot. Gaz. XLV. 
 
 1908, p. 103. 
 
 74. A pure-line Method in Corn Breeding. Proc. Amer. Breeders' 
 
 Assn., V. 1909, p. 51. 
 
 75. The Genotypes of Maize. Amer. Nat. XLV. 191 1, p. 234. 
 
 76. Smith, L. H. The Effect of Selection upon Certain Physical Characters 
 in the Corn Plant. 111. Agr. Exp. Sta. Bui. 132, 1909, p. 51. 
 
 77. Soule, A. M., and Vanatter, P. O. The Improvement of Corn. Va. Agr. 
 Exp. Sta. Bui. 165, 1907, p. 91. 
 
 77a. Sturtevant, E. L. The Superabundance of Pollen in Indian Corn. Amer. 
 Nat., XXV. 1881, p. 1000. 
 
 78. - Botanical Notes. Second An. Rpt. N.Y. Agr. Exp. Sta., 
 
 1883, p. 37- 
 
 79. ■ - A Study of Maize. Third An. Rpt. N.Y. Agr. Exp. Sta., 
 
 1884, p. 124. ! j 
 
 80. ■ Varieties of Corn. Off. Exp. Sta. U.S.D.A. Bui. 57, 1899- 
 
 81. Vries, H.'de Sur la fecondation hybride de l'albumen Compt. rend. 
 Acad. Paris, 129: 1899, p. 973. 
 
 82. The Mutation Theory. Trans. Farmer and Darbyshire. 
 
 Chicago, 1909. Open Court Pub. Co. 
 
 83. Plant Breeding. Chicago, 1907. Open Court Pub. Co. 
 
 84. Walls, E. P. The Influence of the Size of the Grain and the Germ of 
 Corn upon the Plant. Md. Agr. Exp. Sta. Bui. 106, 1905, p. 15. 
 
 85. Webber, H. J. Xenia, or The Immediate Effect of Pollen in Maize. Div. 
 Veg. Phys. and Path. U.S.D.A. Bui. 22, 1910 
 
 86. Correlation of Characters in Plant Breeding. Proc. Amer. 
 
 Breeders' Assn., II. 1906, p. 73. 
 
 87. Williams, C. G. The Selection of Seed Corn. Ohio Agr. Exp. Sta. Cir. 
 71, 1907. 
 
 S8 
 
Flanigan-Pearson Co., 
 Printers and Binders