USDA Yearbook (1936) pp. 462-473
The Story of Corn Breeding, from Mass Selection to Modern Hybridizing

IN ORDER to understand the relative merits of different methods of breeding corn, it is necessary to know how the corn plant reproduces. Corn ordinarily is monoecious, that is, the stamens and pistils are borne in separate inflorescences on the same plant. The staminate or male flowers are in the tassel, a panicle at the top of the stalk. The pistillate or female flowers are borne in spikes, the ears, placed in the axils of leaves usually near or below the central portion of the plant. Staminate and pistillate inflorescences of corn are shown in figure 3.

In 1694 Camerarius, an early botanist, reported the first experiments demonstrating the existence of sex in plants. His experiments were performed with the mulberry, castor-bean, and Indian corn. He removed the silks from corn and proved that pollen of corn was necessary to fertilize the ovules in order that the kernels might develop.

In 1716 Cotton Mather, an American whose name has been popularly associated with witch hunting, reported what are probably the first observations on natural crossing of corn varieties and the immediate effect of pollen on cross-fertilized seeds. His observations are described in a letter to James Petiver, dated September 24, 1716, which is quoted in part as follows:

First: my Friend planted a Row of Indian Corn that was Coloured Red and Blue; the rest of the Field being planted with corn of the yellow, which is the most usual colour. To the Windward side, this Red and Blue Row, so infected Three or Four whole Rows, as to communicate the same Colour unto them; and part of ye Fifth, and some of ye Sixth. But to the Leeward Side, no less than Seven or Eight Rows, had ye same Colour communicated unto them; and some small Impressions were made on those that were yet further off.

FIGURE 4.—Young ear shoot with the husks removed to show the development of the silk from the tip of each ovary. Each silk must receive pollen in order for that kernel to develop.
FIGURE 5.—Germinating pollen grain (after Miller).

The pollen grains of corn develop in the anthers of the staminate flowers in a manner typical of most flowering plants. The pollen mother cells divide twice to give rise to four immature pollen grains. One of these divisions is a reduction division, in which the number of chromosomes is reduced by half, and the other an equational division, where each chromosome divides and a representative of each one is present in the two daughter cells. These different kinds of cell division are explained in the general introduction and glossary of this Yearbook. The nucleus in each young pollen grain divides to form a tube nucleus and a generative nucleus. The generative nucleus again divides to form two sperm nuclei of identical genetic composition, which are present as crescent-shaped bodies in the mature pollen grain.

The formation of the embryo sac within the ovary at the base of each silk (fig. 4) also occurs in a manner similar to that of most other flowering plants. The egg mother cell divides twice, one division being a reductional division and the other an equational division, to produce four potential megaspores. Three of these degenerate and the one remaining gives rise to the embryo sac. The mature embryo sac contains eight nuclei of identical genetic composition, one of which functions as the egg. Two others, the two polar nuclei, are concerned in endosperm development.

When the yellow pollen, the "gold dust" of the Corn Belt, matures, it is liberated from the anthers and carried through the air at random by the wind. Some of the grains fall upon receptive silks, where they germinate and send pollen tubes down through the silks (fig. 5). The tube nucleus functions in the formation of the pollen tube. The two sperm nuclei travel down the tube and are discharged into the embryo sac. One of them unites with the egg cell and from this fertilized egg the embryo or "germ" of the kernel develops. The other unites with the two polar nuclei, and from this union the endosperm portion of the kernel develops. Thus the male contributions to the embryo and endosperm are identical. The female contributions are alike genetically but are twice as great to the endosperm as to the embryo.

From this brief description of reproduction in corn, it is clear that each kernel on an ear may be pollinated with pollen from a different male parent plant. For this reason selecting an open-pollinated ear from a good plant is selecting only the female half of the parentage. The male side of the parentage can be controlled only by controlling the source of the pollen.

Under favorable conditions corn pollen may be carried great distances and effect fertilization. Unfavorable conditions at pollinating time, however, may interfere seriously with seed setting. A hot, dry wind may wither the tassels so that pollen is not liberated or it may wither the silks so that they are not receptive, or it may so dry out the pollen floating through the air that it will not germinate when it reaches the silks.


Mass selection consists in choosing certain desired individuals from the main crop and planting en masse the seed harvested from them. A certain amount of mass selection has been practiced by corn growers since the earliest times. Corn ears are large, and, in the past at least, each ear had to be handled individually both at harvest and in the preparation of seed, so that the planting of a mixture of seed without selection can hardly be imagined. There is ample evidence that this selection has been of the utmost importance in improving corn and adapting it to the varying conditions under which it is grown.

The early corn breeders were artists in molding the form of the corn plant and the corn ear to fit their will. Each breeder had very definite opinions as to the ideal type of plant and of ear and selected rigorously with these ideals in mind. Mass selection was the only tool of the breeders, but they were experts in its use. They had no knowledge of the fundamental laws of inheritance and thus were limited in the range of their accomplishments.

The history of Reid Yellow dent may be cited as a typical example of the place of mass selection in the development of present-day varieties. Robert Reid moved from southern Ohio to central Illinois in 1846. He took with him a late semigourdseed variety known as Gordon Hopkins corn. The seed was not planted until quite late in the spring, but made a fairly good crop of immature corn. Although the best of this was selected for the next year's planting, it resulted in a poor stand in the spring of 1847. The missing hills were replanted with a small local variety called Little Yellow Corn, probably a flinty type. From the resulting crop, which presumably had a considerable amount of interpollination, the variety Reid Yellow Dent was developed after many years of careful selection by James L. Reid, a son of Robert Reid. Similarly many other varieties have been developed from previously existing varieties, from mixtures, or from varietal hybrids by the painstaking selection of interested breeders.

In most cases the mass selection has been among open-pollinated plants. The first case which has come to the writer's attention where consideration was given to the source of the pollen was reported by H. S. Bidwell in the American Agriculturist for December 1867, which states in part:

A remarkable fact has lately been brought to our attention by Mr. H. S. Bidwell, (Bidwell Bro's.,) of St. Paul, who was recently traveling in Tennessee, where he saw a field of common corn, which usually yields an average of not more than one good ear to the stalk, bearing almost uniformly two, and often three ears. The result, he informs us, had been brought about in this way. It occurred to the farmer that, as the kernel usually derived its origin, as we have described, from two different plants, saving the seed corn from stalks bearing two ears was not enough; he must see to it that the kernel germs were fertilized by similar stalks. So he planted every year a special patch for seed, and carefully cut off all the spindles on stalks where two or more ears were not set. The result was an improvement year by year in the quantity of corn, as well as in the manner in which it grew. The principle has a wide application in the improvement of the different kinds of farm and garden produce. In the breeding of animals, the qualities expected from the male, and those which usually are inherited from the dam, are to a degree understood, and the application of similar principles in breeding vegetables is certainly legitimate. The fact above stated has so good a foundation in sound reasoning, that we give it to our readers, anticipating its publication in the "American Agricultural Annual" which is now in press.

The report mentioned as about to appear in the Agricultural Annual appeared in the Annual for 1868. This report states that the tassels (spindles) were removed before the silks appeared.

The writer has been unable to identify the Tennessee farmer referred to. It may not be entirely a coincidence, however, that what is believed to be the first case involving mass selection and controlled pollination also is reported from Tennessee. About this time A. E. Blount began controlled pollination in a variety of white corn. The ear shoots of 100 to 300 desirable plants were covered with muslin each season. These selected plants were hand-pollinated with pollen of other desirable plants selected as male parents. The resulting variety, called Blount White Prolific, came to the attention of the Rural New Yorker about 1879 and was widely distributed by them. Accounts of Blount's breeding methods and the performance of the variety appeared frequently during the following 3 or 4 years. A history of the variety supplied to the Bureau of Plant Industry in 1910 at the request of C. P. Hartley is shown in figure 6. Blount Prolific still is grown in some sections of the South.

In later years Blount was located first at the Colorado and then at the New Mexico Agricultural Experiment Station, in each case continuing his work of improving corn by the same method of breeding.

The advent of the corn show about 1900 and the resulting enthusiastic interest it created in the crop led to the development of two quite different schools of thought on the selection of seed corn. One school regarded the ear as a thing of beauty and more or less assumed that the characteristics associated with its beauty were of value from the standpoint of production. Numerous experiments have indicated that many of the points emphasized by the corn show are not correlated with field performance. The other school placed first emphasis on productiveness and quality and stressed the necessity of field selection of seed ears from standing stalks, resistant to disease. Careful selection which gives due consideration to performance and quality has been amply shown to be of value in varietal improvement.


The ear-to-row method of breeding attempted to determine the relative breeding value of different ears by planting a portion of the seed from them, an ear to a row, and measuring the yields of the resulting plots. It was thought that, by selecting seed from the most productive rows or by planting the remaining seed from the highest yielding mother ears and continuing this process for a number of years, the yielding ability of the variety could be improved. The method was introduced by C. G. Hopkins in 1897 at the Illinois Agricultural Experiment Station in experiments to modify the protein and fat content of the corn kernel.

The first results of these investigations were published in 1899, and the method was very generally accepted almost immediately. It had certain difficulties, and modifications were suggested to overcome them. Some of these modifications had to do with the field plot technique of obtaining a reliable yield comparison among the different ear-row plots. Other modifications suggested by Hopkins and his associates and by C. G. Williams, of the Ohio Agricultural Experiment Station, were concerned with preventing inbreeding, and insuring high-yielding parentage on the staminate as well as the Distillate side.

In spite of the fact that the ear-to-row method of breeding seemed to offer great promise when it was first introduced, the results of its use over long periods of time have been disappointing. In general this method of breeding has improved the yields of relatively unselected varieties during the first few years of selection. There was no evidence of cumulative effect, however, and for this reason there has been little to recommend it.

With open pollination the different kernels on an ear are related only through the female side of their parentage. High-yielding mother ears probably owe their yielding ability largely to their female parent, and because of their hybrid nature they split up and lose vigor in the later generations. Any selection toward a single type also introduces a certain amount of inbreeding, which is always associated with reduced vigor in corn.


1This method has given outstanding results in the case of sweet corn as well as field corn. Golden Cross Bantam is such a hybrid corn. The discussion of sweet corn, however, must be reserved for a later series of papers dealing with the improvement of vegetables and fruits.

Modern corn breeding is concerned with the utilization of two breeding methods-mass selection and selection within inbred lines for the production of hybrid corn. Mass selection still is of value and, when based on selection for performance and quality, probably is the best method for maintaining the yielding ability of open-pollinated varieties adapted to a given region. As already indicated earlier in this article, hybrid corn resulting from the crossing of selected inbred lines is the most recent development in corn-breeding technique and offers greater promise than any of the methods previously tried.1

2This bulletin may be obtained from the Office of information, U. S. Department of Agriculture, Washington, D. C.

Hybrid corn has been developed as a result of researches in genetics, the science of heredity, and is an outstanding example, perhaps the most outstanding example, of the influence of theoretical scientific research in revolutionizing the production practices of an agricultural crop. Although it is a new development, the hybrids already produced have established their superiority in productiveness and in resistance to wind, disease, and other unfavorable conditions. Hybrid corn may be explained by quoting from Farmers' Bulletin 1744, The What and How of Hybrid Corn.2

Possibly the simplest way to give a general idea of what hybrid corn is, is to compare it to the mule. A corn hybrid, in fact, has many things in common with the mule. A mule is the first-generation hybrid between the mare and the ass, and partakes of the better qualities of both parents. It does not reproduce, but must be produced anew each generation for its value in itself, not for reproduction. A corn hybrid is time first-generation hybrid between two strains of corn. Its value is for seed in time production of a crop of commercial corn. This corn will grow, but cannot be used for seed without a loss in yield in the succeeding generations. A corn hybrid, then, like the mule, must be produced anew each generation for its value in itself, not for reproduction. During that generation good hybrids produce larger acre yields of high-quality corn than do the best commercial varieties. Finally, neither all mules nor all corn hybrids are efficient.

Every experiment station that has distributed hybrid corn has been impressed with the difficulty of convincing growers that seed should not be saved from hybrid plants for future planting. Such advice is so different from previous practice with corn that it seems necessary for each grower to try it at least once and learn from his own experience. The reduction in yield suffered when seed is saved from fields of hybrid corn is greatest for single crosses and progressively less as larger numbers of lines enter the cross. The first hybrids distributed by the Iowa Agricultural Experiment Station were single crosses. The writer still remembers one irate grower who suffered considerable financial loss because he planted some 15 acres with, seed he had harvested from his hybrid field and insisted that the warning he had received should have been vigorous enough to make him heed it whether he believed it or not.

Because hybrid corn is new and not generally understood, a number of undesirable practices have developed. Some of them are traceable to this general lack of understanding while others appear to be of a more malicious nature. In an effort to protect the purchaser of hybrid seed corn, several States recently have passed laws defining hybrid corn as the first generation of a cross between strains involving inbred lines and prohibiting the labeling of seed corn as hybrid corn which fails to meet these requirements.


The initiation of modern corn breeding may be considered as occurring about 1920, when the inbreeding programs began to get under way, although the foundations were being laid during a period many years earlier.

Six important contributors to early corn improvement are shown in figure 7.3 One of the most important early breeders was W. J. Beal, professor of botany at Michigan Agricultural College, who began working with corn about 1870. In 1876 he reemphasized the importance of pollen selection in corn improvement, the first mention having been that of Bidwell previously quoted. In 1878 he began experiments with variety crosses which were continued until 1881, when he crossed two varieties by detasseling one of them in what was probably the first detasseled crossing plot. He apparently was the first to hybridize corn for the sole purpose of utilizing the vigor of the first-generation hybrid to increase production. Beal was enthusiastically interested in corn improvement and promoted crossing experiments by other investigators. Notable among these are the Kansas experiments of Kellerman and Swingle and the Illinois experiments of McClure and of Morrow and Gardner.

The idea of using variety crosses did not prove popular, perhaps because of the necessity of producing new crossed seed each year, and although Morrow and Gardner, when their results were published in 1893 and 1894, outlined a plan for farmers to produce their own crossed seed, the method never became established.

Recommendations for the utilization of the excess vigor of first-generation hybrids for commercial corn production were renewed by G. H. Shull in 1908 and 1909, E. M. East in 1909, and G. N. Collins in 1910. Shull and East suggested the crossing of inbred lines while Collins reemphasized the possibilities of increased production from the use of variety crosses.

Previous to 1900 corn had been self-fertilized only occasionally and by very few investigators. In fact, some breeders were of the opinion that corn was self-sterile. About 1900 C. P. Hartley of the Bureau of Plant Industry did some inbreeding, and shortly after that A. D. Shamel produced some inbred lines. In 1905 Shamel reported the yields of two lines which he had selfed for three generations and of the first-generation cross between them, this being the first reported yield of a cross involving inbred lines. Hartley appreciated that increased yields could be obtained from crosses between inbred lines. His inbreeding experiments were discontinued, however, as he considered there were better ways of improving corn than by first tearing it down.

The Work of Shull and East Marks a New Era in Corn Breeding

In 1905 G. H. Shull began inbreeding corn at the Station for Experimental Evolution of the Carnegie Institution of Washington at Cold Spring Harbor, N. Y., in connection with some experiments to study the inheritance of the number of rows of kernels on the ears as influenced by self-pollination and cross-pollination. The performance of the crosses among the inbred lines developed in connection with these studies led to the suggestion of a new method of corn breeding. The publication of these suggestions in 1908 and 1909 marked the beginning of a new era in the breeding of this important crop. Their essential features were: (1) The isolation of desirable inbred lines which breed true for the characters they possess; (2) the determination of which lines are the more productive in crosses; and (3) the utilization of the superior crosses in the commercial production of hybrid corn. Shull's experiments with corn were continued until the end of 1916.

E. M. East also began inbreeding corn in 1905. His experiments were started at the Illinois Agricultural College and continued at the Connecticut Agricultural Experiment Station, where he went in the fall of the year, taking with him the material he had obtained during the season. East's experiments, like those of Shull, were begun for the purpose of studying purely theoretical principles. His investigations had as their primary object an interpretation of the facts to be obtained from a critical study of the effects of inbreeding and outbreeding. His results were reported in a series of publications beginning in 1909, the experiments being continued through 1914.

HYBRID corn has been developed as a result of researches in genetics, the science of heredity, and is an outstanding example, perhaps the most outstanding example, of the influence of theoretical scientific research in revolutionizing the production practices of an agricultural crop. Although it is a new development, the hybrids already produced have established their superiority in productiveness and in resistance to wind, disease, and other unfavorable conditions.

Other early inbreeding projects were those of G. N. Collins in the Bureau of Plant Industry, begun in 1908, and of E. G. Montgomery, of the Nebraska Agricultural Experiment Station, begun in 1909. In 1914 H. K. Hayes went from Connecticut, where he had been working with East, to the Minnesota Agricultural Experiment Station and began inbreeding corn there. In 1916 F. D. Richey and C. H. Kyle, of the Bureau of Plant Industry, and J. R. Holbert, of Bloomington, Ill., began their inbreeding experiments. H. A. Wallace, of Des Moines, Iowa, the first private breeder to inbreed corn, started his work in 1913 but did not go at it systematically until 1919.

Breeders were rather slow in taking up the new program in an extensive way for several reasons. In the first place, the original suggestion for the use of hybrids among inbred lines contemplated the use of single crosses or crosses involving only two inbred lines. There were certain inherent difficulties connected with the use of single crosses which did not make the method seem feasible. It necessitated that the crossed seed be produced on one of the parent inbred lines with a consequent low yield of small seeds which were unsuited for use in ordinary corn-planting machinery. To get around these difficulties, D. F. Jones suggested the use of the double cross, which is the product obtained by crossing two single crosses.

Of greater importance was the delay incident to a more complete investigation of the theory upon which the suggestions were based. It was not until sufficient information on the inheritance of genes affecting qualitative characters had been accumulated, until Emerson and East had advanced a suitable explanation for the performance of quantitative characters, and until the phenomenon of hybrid vigor was better understood genetically, that inbreeding programs began to expand. These developments are discussed in a later section on corn genetics.

Beginning shortly after 1920, there was a very rapid expansion in the number of inbreeding projects. In 1922 F. D. Richey was placed in charge of corn investigations in the Division of Cereal Crops and Diseases, Bureau of Plant Industry, United States Department of Agriculture. He at once began the work of reorganizing and expanding the activities of the Division and coordinating those of the different experiment stations into a more definite national program. In 1925 a committee was appointed to formulate a cooperative program of corn improvement under the Purnell Act, which recently had been passed by Congress. The work of this committee resulted in a plan for a group attack which functioned formally until 1932, and this had an important influence in coordinating the research, in developing a very free interchange of breeding material, and in promoting far more rapid progress than otherwise would have been possible.

Many experiment stations and a few private breeders now are engaged in the isolation and testing of inbred lines in hybrids. Inbreeding projects with corn are being carried on by experiment stations in 33 States and in some foreign countries. A list of the experiment stations in this country at present conducting such inbreeding projects with corn, showing the dates on which the projects were originated, present personnel and former investigators, and the varieties which are being inbred is given in table 2 (p. 499). An attempt was made to obtain similar information on the private breeding projects in this country and on foreign projects, but the data are so incomplete that they are not included.

The improvement of corn through selection in self-pollinated lines necessarily is a slow process, as many generations are required to develop the lines and then to test them in crosses. In an effort to speed up the process, numerous attempts have been made in recent years to obtain two crops a year. The possibilities of growing a winter crop in some of the more frost-free areas of southern Florida, in the Argentine, and in the greenhouse are being investigated. The Bureau of Plant Industry has had very good success raising winter crops of corn in the greenhouse at Arlington Experiment Farm after suitable methods had been developed. It has been found essential to lengthen the day by using electric lights. A night view of the greenhouse used in connection with the corn-breeding investigations of the Bureau, showing the lights for lengthening the day, is given in figure 8.