260 Sturtevant. 



it may be tested bj' making suitable crosses. Another series of phe- 

 nomena taken bj- Batesox ('13) as indicating somatic segregation is 

 that observed in certain variegated plants, where individuals homozygous 

 for a given gene may produce branches or smaller regions heterozygous 

 for this gene; or vice versa, heterozygous plants may produce homozy- 

 gous branches. A case of this sort in maize has recently been treated 

 by Emerson ('13 a) as probably due to somatic mutation, and this 

 explanation would seem to be perfectly applicable to all such phenomena 

 yet described. The evidence seems to me to indicate that somatic 

 segregation occurs, if at all, only as a rare anomaly. 



There is, on the reduplication hypothesis, no simple means of ex- 

 plaining cases where several genes form a group, all showing linkage 

 with each other, such as we have in Drosophila. Each additional case 

 of linkage occurring in the same individual involves an adherent of 

 reduplication in worse difficulties. Trow ('13 b) has analyzed this situ- 

 ation carefully, and, by making an extension of the reduplication hypo- 

 thesis, has worked out systems of "primary and secondary reduplication", 

 and has shown what wlU happen if several genes are all linked together. 

 This form of reduplication has been accepted and extended by Punxett 

 ('13). It seems to me, however, that the system adopted here is far 

 simpler; it has a firm cjtological basis and it gives more accurate 

 predictions, since Trow assumes in liis calculations that there is no 

 interference. 



Frequency of Linkage. 



It seems to be the conception of many geneticists that linkage is 

 a comparatively rare phenomenon — an anomaly. This is probably due 

 to its comparatively recent discovery, and also to the difficulties encoun- 

 tered by the reduplication hypothesis if several genes be supposed to 

 be linked in the same individual. In the early stages of Mendelism 

 many workers were inclined to believe that we were dealing only with 

 special cases, and were merely obsemng peculiarities of a few abnor- 

 malities. But nowadays it is usual to tliink of Mendelian segregation 

 as occurring for a vast number of genes which we cannot follow in in- 

 heritance, because we have nothing to contrast them with — because no 

 mutation has occurred in them. The work on Drosophila and the resulting 

 chromosome view of inheritance has led some of us to a similar con- 

 ception regarding linkage. We think of it as occurring everywhere — 

 being, in fact, as widespread as Mendelian segregation itself, because 



