THE PURE LlxXE IIVPOTPIESIS. 565 



12. Variation occurred in the matter of the difficultv in set- 

 ting seed in the different plants, and it was shown that the seed 

 obtained from plants which exhibited the greatest sterility {i.e., 

 produced the smallest number of seeds) had the least germina- 

 ting power. 



13. That the hybridism of the two varieties was prejudicial 

 to fertility is further evidenced by the fact that the pollen of the 

 hybrids tended to be of poor quality. In those hybrids where the 

 proportion of bad pollen grains was greatest the number of seeds 

 obtained capable of germinating tended to be the smallest. 

 Thus in the case of nine hybrids where the proportion of .o"oo<l 

 grains was on the average only 31 per cent., the average number 

 O'i offspring raised was four only, while in nine hybrids with an 

 average of 75 per cent, good grains the average number of off'- 

 spring was 12. 



14. Habit. — The original seed-plant was pure-bred, or 

 •homozygous, with regard to straggling habit, and the original 

 pollen-plant was homozygous with reference to dwarf habit. The 

 21 hybrids obtained by the cross were all straggling. Thus 

 straggling habit was dominant. On self-fertilising these hybrids, 

 21 families, with altogether 180 individuals, were obtained, and 

 of these 131 were tall or straggling, and 49 dwarf or semi-dwarf. 

 These figures are close to the Mendelian ratio of 3 : i. 



Owing to the fact that a few of the plants were semi-dwarf, 

 the segregation which occurred must be regarded as im])erfecr, 

 although quite definitely present. 



15. Variegation. — The original seed-plant was homozvo-ous 

 with regard to the uniform greenness of the leaf, and the pollen- 

 plant with reference to tlie variegation. The 21 hybrids obtained 

 b> the cross all had uniform green leaves. Thus g'reen leaves 

 were dominant. The offspring of the hybrids included 13:; green 

 plants and 55 variegated, while according to Mendelian theory 

 the numbers should have been 143 green and 47 variep"ated. 

 There never was any question as to whether a plant should be 

 called variegated or not, and therefore the segregation may be 

 re,garded as practicallv perfect. 



The amount of variegation in the offspring varied much; 

 and according to the pure-line hypothesis there could be con- 

 siderable diversity in the amount of variegation in the different 

 individuals, but the averages or means of the families should not 

 differ appreciably from one another. There was, 'however, a clear 

 tendency for the averages to dift'er widely in the various families. 

 Thtis, for example, on a certain scale adopted, the average varie- 

 gation in one of the families was 51 per cent., and in another 8.S 

 per cent. This result can scarcely be understood except on the 

 supposition that there had occurred a modification of the factor 

 for variegation in the parents of these families. 



16. Width of Leaf. — Tlie original seed-plant was homo- 

 zygous with reference to large leaves, and the original pollen- 

 plant with regard to small leaves. The hybrids possessed large 

 leaves, and therefore large leaves are to be regarded as dominant. 



