84 proceedings: botanical society 



erations showed two characters where the first generation was more 

 variable than the second, six characters in which the second generation 

 was more variable than the first, and eleven characters in which there 

 was no measurable difference. In the same cross, in correlation 

 studies involving eleven contrasted characters with fifty-five possi- 

 bilities of correlation, twenty cases of correlation were observed. All 

 but five of these correlations, however, appear to be physical or physio- 

 logical in their nature rather than genetic, such as the correlation be- 

 tween branching space and number of branches. 



There are but few alternative or discontinuous characters in maize, 

 and most of these are color differences or differences in chemical com- 

 position, as in the endosperm characters. 



The discontinuous characters are for the most part Mendelian. 

 Horny or sweet endosperm is perhaps the best example of a simple 

 Mendelian character pair thus far encountered in maize. Horny and 

 waxy endosperm are completely alternative, but there are definite de- 

 partures from the expected ratios. Aleurone and endosperm color show 

 all gradation between Mendelian monohybrid ratios and continuous 

 inheritance. 



Mr. 0. F. Cook called attention to differences in behavior of dif- 

 ferent kinds of plants and animals, as indicating a possibility that the 

 expectation of finding general laws applicable to the whole organic 

 world may not be realized. Each case must be treated in a specific 

 way. Increased vigor in the first generation is one of the more general 

 phenomena attending hybridization. In cotton, for example, when 

 distinct types are crossed, there is usually evidence of increased vigor 

 and hardiness. These are expressed in larger size, better yield, and a 

 greater ability to withstand adverse conditions. By distinct types of 

 cotton are meant types which have a range comparable to the more 

 divergent races of men, or to the Indian zebu as distinguished from our 

 domestic races of cattle. The cotton types appear much more diver- 

 gent than the types of corn. • 



The discussion was largely confined to examples in hybridization of 

 cotton. As a rule the conjugate or Fi generation is intermediate be- 

 tween the parents, while splitting is pronounced in the second and later 

 generations but with no cases of complete return to either ancestral 

 type. As a rule there is a great deal of correlation or coherence in the 

 characters shown in the perjugate (F 2 and subsequent) generations. 

 It has not been possible to secure a cotton combining in stable form the 

 Upland type of vegetation and the Sea Island or Egyptian type of lint. 

 Plants which resemble the Upland type have lint with the Upland char- 

 acters and vice versa, but the plants that most nearly resemble the 

 parental types in other respects are usually very inferior with respect 

 to abundance and quality of lint. 



Thus, one of the results of hybridization may be described as a total 

 loss of the increment of selection which had been developed in the parent 

 stocks previous to crossing. Some characters are more or less dis- 

 continuous, but others appear to be continuous. The opportunities for 



