264 Stiirtevant. 



certain phenomena in several species of Lepidoptera, but I doubt the 

 validity of his assumptions. Sex-linkage in Pygaera (Fedeeley '11) is 

 very probable, but the case is not yet thoroughly tested. Hurst ('11) 

 has made an interesting suggestion that linkage occurs in some families 

 of horses between racing power and certain color genes, but more evidence 

 is needed. Sex-linkage in ducks was suggested by Spillman ('08) and 

 reported by Goodale ('11), but Goldschmidt ('13a) has denied the 

 validity of the evidence, so here again we must suspend judgment. 



Leaving out of account all these doubtful cases, linkage is definit- 

 ely known in 8 dicotyledonous plants, 4 insects, 3 birds, and 2 mammals 

 — a total of 17 species. This hardly seems to indicate a rare phenomenon, 

 but points rather to the conclusion that linkage is, as stated above, as 

 widespread as Mendelian segregation itself — a conception which follows 

 naturally from the chromosome hypothesis. 



The Nature of Mendelian genes. 

 Theories such as the one adopted here, which postulate definite 

 material bodies as the bearers of Mendelian genes are sometimes criticized 

 on the grounds that characters are very complex, and that we cannot 

 suppose a given character to be actually borne by any material particle 

 in the germ plasm. The fallacy of this line of argument has often been 

 pointed out, perhaps most effectively by Wilson ('12). It arises from a 

 confusion of characters and genes, as I shall try to show. There are 

 two main aspects of the study of genetics, which may be characterized 

 as genetics proper, or the study of heredity in the strict sense — of 

 chromosome mechanics on our view; and the study of the manner of 

 action of the genes, which should fall rather under the heading of the 

 physiology of development. The first of these has formed the basis for 

 the main part of this paper, but in order to make my position clear it 

 will be necessai-y to consider briefly the question of the physiology of 

 development. As yet we know little about the subject, but we are 

 forced to make many assumptions about it which we are unable to verify 

 directly; and unfortunately we sometimes make such assumptions when 

 we are not justified by necessity. For example, the term "inhibiting 

 factor" is often found in Mendelian literature. This term, implying as 

 it does that there is a factor which keeps in check some process, is 

 unjustifiable; for we can never be certain, in any given case, whether 

 the process is checked, as it were by a poison, or whether it fails to 

 occur because some material upon which it depends is lacking. The 



