172 Defeneration in the Ostrich 



phenomenon of two wholly similar characters arising by quite different 

 methods. In the following description an " incomplete " break is under- 

 stood to be one in which the scales over the joint are so reduced as 

 to separate or become detached from one another, but where vestiges 

 still remain. A " complete " or " entire " break is one in which a clear 

 interval occurs over the joint, free of scales, except of the kind over the 

 leg generally. 



A break is, manifestly a single, complete act of degeneration, repre- 

 senting an entire loss of about seven or eight scales over the first joint 

 and four or five over the second, and the part reduction of the others in 

 each direction towards it. All stages in the process are forthcoming, 

 which show it to be gradual and sequential, the method being different 

 for the one break as compared with the other. An almost simultaneous loss 

 of a group of structural elements from a linear meristic series, at some 

 part other than the ends, offers a great contrast to the manner in which 

 the reduction of the coverts and remiges is effected. In the latter, 

 successive individual losses occur from one or both ends of a linear 

 series, in place of a disappearance from an intermediate part. Differences 

 of this kind show that any assertion as to the order or manner in which 

 past losses have taken place, without supporting evidence from the 

 intermediate steps, as well as any prognostication as to the course which 

 will be followed in future losses, is a proceeding attended with many 

 uncertainties. 



The breeding experiments serve to establish that a break is germinal 

 in its origin, and represents a dominant Mendelian character, an absence 

 dominant over a presence. Where no interval occurs in either of the 

 parents the progeny in nearly all cases reveal a like continuous scutel- 

 lation, showing that the influence responsible for the break is wholly 

 absent or non-effective. Thus in twelve different hatchings, where neither 

 the cock nor the hen showed any break, each of the 71 chicks obtained 

 had the tarsal and digital scales continuous. In these cases the factors 

 concerned in the production of the scales are manifestly in a fairly fixed 

 and stable homozygous condition in the parents, and we have uniformity 

 of somatic expression in the progeny. The stability is however by no 

 means so constant as might be expected, for in most of the series much 

 variation is to be found in the degree of narrowing of the scales over 

 the first joint, even on progeny from the same parents. 



Occasionally however an incomplete or complete break may occur in 

 the progeny of parents, both of which show no signs of any, as in the 

 two series given below. (Tables XIII and XIV.) 



a 



