188 Degeneration in the Ostrich 



will be greatly reduced in number, following along lines similar to those 

 of the under-coverts. 



The production of a bald head patch in the northern ostrich and the 

 loss of the feathers from the legs of both the northern and southern birds 

 are examples of true mutations in the sense in which the term is em- 

 ployed by de Vries, Bateson and others. They are discontinuous, striking 

 variations, albeit coming late in the chick ontogeny. The head patch 

 has been proved to represent a Mendelian character, dominant in first 

 crosses and segregating in the second. Whatever the degenerative 

 influence may be, whether factorial atrophy, some inhibitory modifying 

 factor, or a separate factor for baldness, it acts feirly suddenly and com- 

 pletely for an entire area. The change from plumage covering to bald- 

 ness is effected as a single, complete, disconnected, retrogressive act, 

 altogether different from the slow, successional losses among the coverts 

 and remiges of the wings. If however, as Morgan and others would 

 have us do, we regard each factorial change, however small it may be, as 

 giving a unit-character or mutation, then it becomes impossible to draw 

 the line between continuous and discontinuous variation so far as magni- 

 tude alone is concerned. The slow, gradual reduction- in size of a plume 

 would represent a series of barely perceptible independent mutations, 

 while the sudden, complete losses of feathers from the head and legs 

 would be only mutations on a bigger scale. The nature of the changes 

 might be the same ; they would differ only in degree. The fundamental 

 difference would then be that the latter are disconnected changes while 

 the former are successive and cumulative, all in the same direction; 

 the two are isolated detached mutations, the others are part of a con- 

 tinuous scheme. 



That fortuitous, disconnected mutations, as contrasted with suc; 

 cessive mutations in the same direction, are the main foundation of 

 the Mendelian conception of the evolutionary process is clearly indicated 

 by Morgan in the section, Chance Mutation and Natural Selection, in the 

 paper so often referred to, where he says (p. 51) : " The mutation 

 process rests its argument for evolution on the view that among the 

 possible changes in the genes, some combinations may happen to produce 

 characters that are better suited to some place in the external world than 

 were the original characters." To mitigate this entire appeal to chance, 

 without introducing into the theory of evolution some sort of directive 

 agent to account for cases of successional changes, he then states a rela- 

 tion between chance and evolution which he conceives to be of funda- 

 mental importance : " Starting at any stage, the degree of development 



I 



