328 



THE AMERICAN XAITBALIST [Vol. LIII 



the strain ; all the members of a flock bred from the same 

 stock have the second row of upper-coverts complete, 

 while in other flocks all the members have a number of 

 plumes absent from the row. Crossing of birds in which 

 the complete loss of the claw on the small toe has taken 

 place with others in which the claw still appears gives 

 results on strictly factorial lines, as also does the cross- 

 ing of birds with and without a loss of the scales. In a 

 mixed assemblage of any species where only a small pro- 

 portion display a certain character it may be presumed 

 that the latter will be heterozygous with regard to the 

 particular character, seeing that the chances are much 

 against the mating of two individuals each having the 

 character. The heterozygous nature of the bird can be 

 demonstrated on mating with one in which the character 

 is absent, for if dominant it will appear in half the prog- 

 eny and be absent from the others. This proportion with 

 regard to the presence of the claw and the loss of the toe 

 scales has been found to hold in all the crosses. Out of a 

 total of 36 chicks hatched from breeding pairs where one 

 parent was clawed and not the other the numbers were 

 actually equal, namely, 18 chicks were clawed and 18 un- 

 clawed. Out of 11 chicks reared from a pair where one 

 parent showed no loss of scales on the big toe and the 

 other had a single break, 5 had no break and 6 showed the 

 break. 



It may be accepted then that all the degenerative phases 

 represent factorial changes which have come about in the 

 germ plasm of the ostrich. Presumably the changes in- 

 volve a loss of factors ; they are retrogressive or negative 

 mutations. Structures which would be expected to occur 

 either fail to appear or are seen very rarely, and may 

 then be regarded as survivals, the factorial losses not 

 having yet taken place in the particular individuals. 

 Thus, to take the case of the first row of under-coverts, 

 the principle of alternation demands that a complete row 

 of under-coverts should alternate witli tli<' row of remiges. 

 The full row actually occurs in a IVw iinlividuals, and 



