:>r>6 



THE AMERICAN NATURALIST [Vol. XL VII 



of centers of cell keratinization. In some statoblasts the 

 number of these centers is small ; in others great. 



To test the hypothesis that the size of the cells in the 

 setigerous membrane covering the statoblast influences 

 the number of folds arising in it, we measured the diam- 

 eter of the facets on the disc and on the float of stato- 

 blasts with 20 hooks and those with 12 hooks. The aver- 

 age diameter of a facet on the disc in 25 measurements 

 (each based on a row of facets) was, in statoblasts with 

 20 hooks, 8.37 m; in statoblasts with 12 hooks (14 sets of 

 measurements) 8.25/*. On the float, in statoblasts with 

 20 hooks, 9.54 p, in statoblasts with 12 hooks, 9.40 /*. It 

 results first, that the facets (cells?) of the float are 

 slightly larger than those of the disc, but that the differ- 

 ence in size of the facets in statoblasts with many and 

 those with few hooks is neglible. 



Since there seems to be nothing in the interrelation of 

 parts to determine that the number of hooks shall be 

 great or small one is naturally led to suspect that in these 

 varying statoblasts we are actually dealing with distinct 

 biotypes. We turn, consequently, to that phase of the 

 question. The ideal conditions for an answer to the in- 

 quiry whether there are distinct biotypes in respect to 

 number of hooks are these : To plant several statoblasts 

 (with varying number of hooks) from each of the sev- 

 eral independently arisen colonies and count the number 

 of hooks on the statoblasts that are produced therefrom. 

 We have not abandoned the hope of meeting these condi- 

 tions, but our attempts to do so have hitherto been frus- 

 trated. Of nine statoblasts affixed (by shellac) to sub- 

 merged wood none hatched. Also, colonies observed 

 daily from hatching were eaten up by the larvre of cad- 

 dis flies {Eydropsyche). Finally after we had secured a 

 good development of colonies free from predaceous in- 

 sects all our work was brought to naught by the destruc- 

 tion of our floats. 



We have, however, sought to get the required informa- 

 tion in a more indirect way. We have studied the num- 



