Kaymond Pearl 
69 
variation. That is to say, relatively equal degrees of aberration from the typical 
condition are, on the vvhole, produced with equal frequency in the two populations. 
Thus there is apparently nothing like a selective process in the encystment of 
this form. The last individuals to " survive " in the active condition are as 
variable as the general population. 
The third result appears to be worthy of notice. For both length and breadth 
there is a positive skewness in the variation of the individuals in the optimum 
environment. That is to say, the mean falls to the right of the mode, or the 
curve tends to " tail out " more on the side of large individuals than in the 
opposite direction. This indicates that the conditions which are favourable to the 
production of large size of body in the population as a whole, are also more 
favourable to the production of exceptionally large than of exceptionally small 
individuals. In other words, the direction of the skewness is the same as the 
direction in which the type is changing. May not this relation be generally true 
when a change of type is brought about by direct environmental action rather 
than by selection, the distribution finally becoming symmetrical when the possible 
limit of direct modification of the type is reached ? The results from both series 
of Chilomonas are in accord with such a view, but of course are altogether 
too meagre to base more than a suggestion upon. The question will be more 
definitely tested on Paramecium material collected ad hoc. 
The result that shape of body as measured by the length-breadth index is 
sensibly correlated with absolute size seems to me to have such important 
theoretical significance that it appears desirable to discuss the matter in some 
detail. At the outstart I may say that the results from Chilomonas on this 
point are by no means an isolated case. I have elsewhere shown (1906) on material 
comprising a number of fairly long series that the same thing is true for 
Paramecium, with, of course, differences of detail in the values of the constants. 
The following table gives the values of the net organic correlations between 
index and length and breadth of body in three lots of Paramecium, comprising 
altogether 544 individuals. Other data are given in the paper referred to, but 
these will be sufficient for comparison in the present instance. 
TABLE VII. 
Correlation of Index with Absolute Dimensions in Paramecium. 
Series* 
Characters 
Net Correlation (p - p^) 
A 
Length and Lidex 
•4134 ± -0386 
Breadth ,, „ 
- -2246 + -0442 
C 
Length „ „ 
•3G92 + -0410 
Breadth „ „ 
- -2497 + -0445 
E 
Length „ „ 
•3556+ -0513 
11 
Breadth „ ,, 
- -2964 + -0535 
* The letters designating the series are those used in the original paper. 
