No. 558] CLONAL VARIATION IN PECTIN ATELLA 365 



complex mass counted at the same time in October gave 

 an average of 16.0 hooks. Thus the difference between 

 two sets of counts made in the same month on two dis- 

 tinct masses is greater than between the July and Oc- 

 tober counts. The highest average number of hooks 

 found in any mass during October was in Mass B, 3,802 

 individuals, with an average of 16.6 hooks. 



Comparing with Braem's, it appears that our counts 

 run much the higher. The average of all counts made by 

 Braem is 14.34, which is decidedly lower than our July 

 average (15.3) ; and in one colony he obtained an average 

 of 12.94 hooks. A great mass found at Jackson Park. 

 Chicago, in August, 1898, gave an average of 13.78 hooks. 

 It is clear, accordingly, that however important the tem- 

 perature factor may be, 4 it is secondary in importance 

 to some other factor that determines the variation in the 

 number of hooks. 



The number of hooks is determined by the number of 

 pocket folds arising in the membrane that secretes the 

 chitinous covering of the statoblast; and the question 

 now transfers itself to the reason why in some stato 

 blasts few, in others many, such folds occur. At one time 

 we entertained the hypothesis that there was a causal re- 

 lation between thickness of membrane and the size of and 

 distance between pocket folds, such that a thin mem- 

 brane permits smaller and more numerous folds. Un- 

 fortunately, it was not feasible to measure the thickness 

 of the setigerous membrane, for by the time the number 

 of eventual hooks can be determined the membrane has 

 become relatively thin and very irregular in thieknes>. 

 Our study did serve to indicate that the number of hooks 

 can not be determined in a mechanical way by the thick- 

 ness of the membrane, but that, on the contrary, the folds 

 follow, and their number is determined by, the number 



