268 Heredity and Environment 



them (Fig. 100) seem to be of this sort as Gates especially has 

 emphasized; for example O. lamarckiana has 14 chromosomes, 

 O. lata 15, O. semi-gigas 21, O. gigas 28, and these variations in 

 the number of chromosomes are probably due to abnormalities 

 in the maturation divisions. It is significant that the mutants 

 lata and semi-gigas have occurred several times, whereas gigas 

 appeared but once; this may be explained by the fact that the 

 chances of the doubling of chromosomes in both germ cells 

 (gigas) are very few compared with the chances of their doub- 

 ling in one germ cell (semi-gigas) or of their increase by one in 

 a single germ cell (lata). 



But it is probable that mutations are not usually associated with 

 changes in the number of chromosomes. Where the number of 

 chromosomes remains constant the change may take place in the 

 number or composition of the chromomeres or units of the next 

 lower order, -"but it would be practically impossible to find such 

 changes in bodies so small and so numerous. Whatever the cel- 

 lular changes may be which accompany mutations, it is certain 

 that changes take place in the inheritance factors. Sometimes 

 factors drop out, as in the white sweet peas shown in Fig. 33, and 

 in many other cultivated races of plants and animals, thus pro- 

 ducing regressive mutations. Indeed most of our domestic ani- 

 mals and cultivated plants have arisen by the omission of some- 

 thing which their wild ancestors had. In most of the mutations 

 studied by deVries, Bateson, Morgan and others some factor has 

 dropped out and Bateson suggests that at present all new forms 

 arise only by the loss of factors or by the fractionation of factors 

 and that new factors are not added from without. This leads 

 him to inquire "whether the course of evolution can at all reason-- 

 ably be represented as an unpacking of an original complex which 

 contained within itself the whole range of diversity which living 

 things present." This is as extreme preformation in the field of 

 inheritance factors as was the old theory of "emboitement" in 

 the field of developed characters ; it is devolution rather than evo- 



