No. 627] GERM PLASM OF THE OSTRICH 335 



tion, though not in a plus direction, beyond the present 

 limits of the race. It will of course be readily appre- 

 ciated that this possibility differs altogether from that 

 due to the ordinary selection which may go on in a race of 

 organisms where the germ plasm is static, but where all 

 grades of pure lines may be extracted between extreme 

 limits. Where the germ plasm for a race is static, as 

 demonstrated by Jennings in his work on Paramcccium, 

 we can readily understand that no further change is pos- 

 sible by selection within a pure line, as nothing inducing 

 factorial changes is present. If where germinal changes 

 are taking place it is not permissible to think of the 

 factors as changing autonomously we have to assume that 

 some causative agent is present, and may vary in degree 

 in different members and thereby form a basis for selec- 

 tive action. 



The same considerations can be applied to the state- 

 ment: mixing of germ plasms in fertilization alters 

 hereditary determiners mutually and hence is, in and of 

 itself, a cause of genetic variations." AVhen, for ex- 

 ample, two germ plasms, in each of which the causative 

 agent producing loss of factors is at its niaxiituiiii. l)0('ome 

 mixed in fertilization it is reasonable to expect that the 

 agent will be intensified and the hereditary (letei iniiiers 

 will be altered mutually, and some of them drop out. The 

 mixing will be, in and of itself, a cause of genetic varia- 

 tion, which will be expressed by a further loss of remiges. 



Though the idea of a causative agent inducing changes 

 in the germ plasm, and varying in degree and also trans- 

 missible, is altogether hypothetical yet it is stimulating 

 to further experimental effort. Of the hundreds of 

 ostriches examined not one has been found with less than 

 33 remiges, hence this number must be regarded as the 

 pi^esent minimum of the race. There is every reason to 

 expect that a pure line having this number only can be 

 built up. If by breeding these together a further reduc- 

 tion of plumes should take place we should then be fully 

 justified in assuming that the factors concerned with the 



