[BUREAU OF PLANT INDUSTRY—Circular No. 124-B] (May 3, 1913)


By G. N. COLLINS, Botanist, and J. H. KEMPTON, Assistant, Office of Crop Acclimatization and Adaptation Investigations.


There is a popular belief that the immediate effect of planting two varieties of maize in close proximity is to increase the yield of both varieties. Failure to appreciate some of the peculiar characteristics of the maize plant accounts for the delay in bringing this question to the test. With other crops the changes which follow crossing become apparent only when the hybrid embryos develop into plants in the next generation after the crossing takes place. In maize, however, foreign pollen often has an immediate effect in altering the color and texture of the seeds, a phenomenon that has received a special name, xenia. It seemed, therefore, not unreasonable to believe that the size of the seed as well as 'the color or texture might be affected by crossing.

1Issued May 3, 1913.
2Collins, G. N. A new type of Indian corn from China. U. S. Department of Agriculture, Bureau of Plant Industry, Bulletin 161, p. 18, 1909.
3Roberts, H. F. First generation hybrids of American X Chinese corn. American Breeders' Association, Annual Report, v. 7/8, p. 374, 1912.

In connection with a study of hybrids between United States varieties and a new type of maize from China it was observed that one of the effects of crossing the two types was an increase in the size of the seed in the same year the crossing was done.2 Similar results in crosses of this Chinese variety have been recorded by Roberts.3

Additional data on this subject were obtained in connection with an experiment planned to test the possibility of selective pollination in maize. Pollen taken from two plants belonging to different varieties was mixed and applied to the stigmas or silks of one of the varieties, thus producing pure and hybrid seed on the same ear. The phenomenon of xenia made it possible to so select the varieties with respect to color of the seed that it was possible to distinguish between the hybrid seeds and those resulting from the pollen of the same variety.

The hybrid and pure seeds from each of the ears, when weighed separately, exhibited such striking differences that it is thought advisable to place the results on record. In every instance the hybrid seed was larger than the pure seed borne on the same ear, the increase ranging from 3 to 21 per cent.


The experiments involved the following varieties:

Missouri Cob Pipe, or Collier.—The well-known, large-cobbed variety used in the manufacture of corn-cob pipes. The seeds are white. In our specimens the average weight per 1,000 seeds is 310 grams.

Gracillima.—A variety of pop corn isolated from a variegated variety, the seed of which was originally secured from Germany. The seeds are white. The average weight per 1,000 seeds is 67 grams.

Variegated.—Similar to the preceding. Isolated from a podded variety. The white seeds of this strain average 141 grams per 1,000 seeds.

Hickory King.—A large, thin-seeded, white dent variety. The seeds of the strain used in these experiments averaged 480 grams per 1,000.

Mexico Black.—A Mexican variety with slightly dented "shoe-peg" seeds and an intensely black aleurone. The original seed of this variety consisted of one ear secured from the Valley of Mexico, Seed and Plant Introduction No. 27074. The seeds of this ear averaged 295 grams per 1,000. As grown in our experiments the seeds are smaller, averaging only 231 grams per 1,000.

Algeria.—A variety of pop corn with beaked seeds and purple aleurone. The average weight of the seeds is 125 grams per 1,000.


It will be seen that the varieties involved comprise four with white seeds and two with colored seeds. The color in these seeds is located in the aleurone cells, the outer layer of the endosperm. Previous experiments have established the fact that when pollen from one of these black-seeded varieties is used to pollinate an ear of the white varieties the resulting seed is colored.

In the experiments to be described, the silks of the white-seeded varieties were dusted with a mixture of pollen from the same white variety and pollen of a different variety with colored seeds. The ears that result from an operation of this kind have white kernels that represent pure seed of the variety and colored seeds that are hybrids between the white and colored varieties. In some instances the pollen of the white variety contributed to the mixture was taken from the same plant that was being pollinated, in which case the resulting white seeds were not only pure for the variety, but were self-pollinated. The two kinds of seeds were, of course, mixed indiscriminately on the ear and developed under identical conditions. Where consistent differences in the size of the two kinds of seed occur it is therefore safe to assume that they are due to differences induced by the use of the various kinds of pollen. The number of seeds from a single ear are usually large enough to afford reliable averages and justify confidence in even relatively small differences.

Eleven such ears were secured, and the results with respect to the weight of the hybrid and pure seed are given in Table I. Individual plant numbers are given, since in many instances a plant which bears one of the pollinated ears also figures as the source of pollen in another cross. Where more than one ear of a plant was pollinated the second and third ears from the top of the plant are given the individual plant number followed by a figure 2 or 3. To avoid large decimals, the average weight is given in terms of 1,000 seeds instead of in terms of a single seed.

It will be seen that in every instance the size of the seed was materially increased by the foreign pollen. The increase ranges from 2.8 to 21.1 per cent. It should also be noted that the rate of increase bears no direct relation to the size of seed in the variety used as the source of pollen. The seeds of the Algeria variety average only half as heavy as those of Mexico Black, yet the increases secured when Algeria is used as the source of pollen are rather greater than when Mexico Black pollen is used.

TABLE 1.—Increase in size of seed secured by cross-pollination.

Ear-bearing plant Source of the mixed pollen Number
of seeds
Weight per
1,000 seeds
Increase with
Per cent
White Colored
Variety No. Variety Plant No. Variety Plant No. Pure Hybrid Pure
Pipe 259 Self   Mexican Black 643 511 220 294 337 14.6 ± 0.7
259-2 Pipe 267 " 643 541 248 268 293 9.3 ± 1.0
267 Self   " 643 553 20 371 420 13.2 ± .6
Gracillima 1212 "   Algeria 1301 171 235 65 79 21.1 ± .9
1220 "   Mexican Black 669 99 211 74 80 7.9 ± .8
1220-2 Gracillima 1212 " 669 350 94 66 74 12.3 ± 1.4
1220-3 " 1212 Algeria 1301 139 175 65 74 13.8 ± 2.1
Variegated 1227 Self   Mexican Black 653 402 81 137 156 13.9 ± 2.2
1227-2 Variegated 1228 " 653 111 245 140 148 5.7 ± .6
1228 " 1227 " 653 496 25 144 148 2.8 ± .9
Hickory King 219 Self   " 643 140 104 522 557 6.7 ± .2


The differences in weight observed in these seeds might come about through differences in size or through differences in composition, or a combination of both. To determine this point, volumetric and specific-gravity determinations were made of the pure and hybrid seed of all the ears. The results showed that the observed differences in weight were associated with corresponding differences in volume. No differences in the specific gravity of pure and hybrid seed from the same ear could be detected.


In order to distinguish between pure-bred and hybrid seed on the same ear only those hybrids with colored aleurone could be utilized. It might be urged that there was a tendency for colored seeds to be heavier than white and that we have been measuring differences between colored and white seed rather than differences between pure and hybrid seed. Fortunately, there was material at hand to test this point. In 1911 a cross had been made between Variegated and Mexico Black, two of the varieties used in these experiments. This hybrid was grown and a number of self-pollinated ears secured in 1912. These ears all had both white and colored seeds. A comparison of the weight of the white and colored seed from each of these ears showed only the ordinary fluctuating differences between the two classes.


Before accepting the increased size of the hybrid kernels on mixed ears as an indication of the increased yields that may be secured from cross-pollination, it will be necessary to consider the possibility that the hybrid kernels might develop at the expense of the neighboring pure kernels. In that case the average size of the seed would not be increased if all the seeds of the ear were cross-pollinated.

If the pure kernels are weak or develop more slowly, the hybrid kernels may gain by being less crowded, or the hybrid kernels may grow more rapidly at first because of more prompt pollination and may then rob their pure-seed neighbors by direct appropriation of a larger share of the available food materials.

A fortunate accident in pollination throws light on both of these questions. An ear of "Maryland White Dent" was pollinated by a plant of the same variety. Seven days later, in making a second application of pollen, a plant belonging to a red variety with yellow endosperm was accidentally used as the source of pollen, the mistake being noted at the time.

In the resulting ear the two kinds of seed were easily distinguished. The pure seeds resulting from the first pollination were pure white, while the hybrid seed resulting from the second pollination were yellow. Unlike the ears where mixed pollen was used, the two kinds of seed were not indiscriminately distributed. All the white seeds were on the lower portion of the ear; all the colored were on the upper portion. This segregation of the two kinds of seed must have deprived the hybrid seed of any advantage that might be secured by crowding weak neighbors, while the time which elapsed between the two applications of pollen precluded the possibility of the hybrid seed appropriating material in advance of the pure seed.

The ear produced 212 white, or pure, seeds and 161 that were yellow, or hybrid. The average weight of the pure seed was 283 grams per 1,000. The average weight of the hybrid seed was 292.5 grams per 1,000, a difference of 9.5±1.06 grams, or 3.4 per cent. The seeds in the lower portion of the ear are usually somewhat larger than those on the upper portion, a fact that should be considered in connection with the observed increase.


As soon as the phenomenon of xenia came to he recognized, the possibility that size might be among the characters thus affected was perceived by a number of investigators; but, without some method of mixed pollination similar to that used in our experiments, changes in the size of seed could only be measured by comparing the average size of seed from hybrid ears with the average size of seed in the parent variety. There is such a range of individual variation in maize with regard to the size of seed that hybrid ears would have to be produced in very large numbers to recognize any but very large differences. Since hybrid ears were secured only through artificial pollination, their number was naturally small and it is not surprising that the results were more or less contradictory.

1Correns, Carl. Bastarde zwischen Maisrassen, Stuttgart, p. 30, 81, 1901. (Bibliotheca Botanica, Heft. 53.)

Thus Correns1 made a series of crosses between races with different-sized seed and compared the weight of the hybrid and parent varieties. He viewed the results, however, from the standpoint of the inheritance of size as a character and paid little attention to the possibility of this being an increase due to crossing as such. With the limited number of ears which he secured it was hardly to be expected that increases of less than 25 per cent could be detected with certainty. In summarizing his results on this point, he states that while the size of the seed is not essentially changed by cross-pollination there is a slight increase in weight. By the method of mixed pollination used in our experiments differences averaging as low as 2 per cent are significant.

In cases where the ear-bearing plant produced two ears it was hoped that the effect of cross-pollination might be measured by pure seeding one of the ears and using foreign pollen on the other. In this way it might be possible to determine the effects of crossing two varieties with seed of the same color. But the normal range of variation in the size of seed between the first and second ears of the same plant when both were pure seeded was found to be so large as to obscure any differences that might be expected from the use of foreign pollen.


The Chinese variety in which the phenomenon of increased size was first noticed has very small seeds, suggesting that the increase that followed crossing should be looked upon as the immediate inheritance of larger seed of the pollen parent. The other alternative is to consider the increased size of the seed as connected with the greater vigor so frequently shown in the first generation of a cross.

1Roberts, H. F.

Roberts1 finds difficulty upon morphological grounds in admitting the possibility of this second view. He holds that the growth stimulus of the pericarp must be inherited from one of the parents and that additional growth that resulted from a mere increase of vigor would produce an endosperm too large for the pericarp.

In the experiments here reported there is no evident relation between the size of seed in the pollen-producing variety and the amount of increase resulting from cross-pollination. In fact, the greatest increase was secured by pollinating with a small-seeded variety, and there is one instance (plant No. 219) where an increase in the size of the seed of the large-seeded Hickory King variety was secured by pollinating with a variety whose seeds weighed only half as much as those of the Hickory King. It may be necessary to conclude that the growth of the pericarp is stimulated directly by the growth of the endosperm. In experiments thus far reported, however, the necessary increase in the size of the pericarp would be comparatively slight and to seek any explanation may be superfluous. An increase of 22 per cent in the volume of a seed, which is the largest reported, would require an increase in the superficial area of the pericarp of only about 2.8 per cent.

2Carrier, Lyman. Loc. cit. Preventing cross-pollination of corn by means of muslin screens. Paper read before the American Society of Agronomy, Washington, D. C., Nov. 14, 1910.

If an increase of from 2 to 20 per cent in the weight of the seed can be secured through the stimulation of foreign pollen, the fact is of more than scientific interest. Direct evidence that important increases can be secured by allowing two varieties of the same color to cross-pollinate has been given by Carrier,2 who secured yields of from 5 to 18 bushels more per acre when strains were allowed to. cross-pollinate than when cross-pollination was prevented.

Another idea suggested by the results of these experiments is that increase in the size of the seed in the xenia generation may serve as a means of determining in advance the hybrid combinations that will produce vigorous and productive plants the following generation. 'Whether this proves to be the case or not, the results afford additional reason for the use of first-generation hybrid seed; but even where hybrid seed is not to be used, the planting of two varieties in alternate rows may be found to increase the yields sufficiently to warrant the additional trouble.

As the increased size is evidently a manifestation of vigor, it may be considered as a factor of adaptation, like the vigor of the first-generation hybrid plants. It would seem especially desirable to take advantage of this method of increasing the yield in regions which do not produce their own seed corn.


The experiments reported in this paper afford definite evidence that the crossing of two varieties of maize is followed by an increase in the size of seed in the same year that the crossing is done. This increase is not to be confused with the increased yields secured in the year following, when the first-generation hybrid plants are grown.

By mixing pollen of a white-seeded and a colored-seeded variety and applying the mixture to the silks of the white-seeded variety pure and hybrid seed are produced on the same ear. By virtue of the xenia inheritance of seed color the two kinds of seed produced under identical conditions can be distinguished and compared.

This method makes it possible to measure the immediate increase from cross-pollination much more reliably and accurately than can be done in any experiments involving the comparison of seed produced on different plants. The results showed the hybrid seed to be heavier in every instance, the increase ranging from 3 to 21 per cent.