Variability of Body Size 49 



the shelled rhizopods. Whether or not this was en- 

 tirely due to the automatic elimination of the growth 

 factor through measurements of the shell only, is un- 

 certain. 



Only in one group of unicellular organisms, the try- 

 panosomes, has size been used as a taxonomic charac- 

 ter in more than a general way (Bruce, '11). The iden- 

 tification of species of trypanosomes at first depended 

 almost entirely upon the mean sizes and the size-dis- 

 tributions of a sample of the individuals found in a 

 given drop of the host's blood. But the coefficients of 

 variation (table 5) were always quite high, and it is 

 likely that they were high because the trypanosomes 

 were not of one race or possibly species; in other 

 words, multiple infection had usually occurred. More- 

 over, it is uncertain that exactly the same variability 

 would be found for the same clone in different hosts, 

 particularly hosts of different species. For, it has been 

 shown that different host-species modify greatly the 

 body sizes of trypanosomes (Wendelstadt and Fell- 

 mer, '10; Noller, '13). Some of the points concern- 

 ing variability were demonstrated by Pearson ('14), 

 utilizing the very numerous data of Bruce ('11, '12, '13). 



If the variabilities of size which occur under con- 

 trolled conditions have once been investigated, it is 

 then possible to use variability as a criterion of some 

 particular condition. Thus Jennings ('08b) and Jollos 

 ('13, '21) used it as a test of genetic homogeneity in 

 Paramecium. Taliaferro ('23, '24) used variability as 

 a criterion of the relative numbers of young and grow- 

 ing organisms in clonal infections of trypanosomes. 

 And it may be concluded, almost without discussion, 

 that constancy of environmental conditions is to be 

 judged by the smallness to which the residuum of vari- 

 ability can be reduced through various means of con- 

 trol. 



