370 



THE AMERICAN NATURALIST [Vol.XL.VII 



constantly 15 in all colonies of the mass. So, too, in 

 Table IV with one exception the mode of the 15 colonies 

 is 16 despite the fact that the average for the whole mass 

 is near the dividing line between 15 and 16, viz., 16.03. 

 The remaining masses show a greater or less commin- 

 gling of biotypes. Thus, in Table V the empirical mode 

 varies from 15 to 17 and the range of the average num- 

 ber of statoblast hooks to a colony is .76. In Table VI 

 the mode ranges from 15 to 18 and the range of the aver- 

 age is 1.20. Tables VII and VIII show masses 5 and 6 

 to be even more variable with a range of 2.40 hooks in 

 the averages. 



Examining the standard deviations, we find no evi- 

 dence that, except for the fact that, as is usually the case, 

 the standard deviation tends to increase with the aver- 

 age, the great variability of masses 5 and 6 is due to a 

 corresponding variability inside the individual colony. 5 

 We conclude, consequently, that the difference in varia- 

 bility between masses 1 and 2, on the one hand, and 

 masses 3 to 6, on the other, is due to the fact that the 

 former are simple in origin and the latter are compound ; 

 the former represents one biotype, the latter two or more 

 biotypes. Compare the pairs of distributions in Table 

 IX for mass 1 and mass 6 — the most unalike having been 

 selected in each case. 



TABLE IX 



Comparisons of Two Unlike Distributions in 



15 16 17 ; 18 [ 19 

 i| 134 305 25S 13n 92 :•;:» 



Colony 21 39 170 332 257 



ls5 225 275 



