71 CELL-DIVISION AND GROWTH II. 2 



purpose alters in the same way as that of the whole body. 

 Besides weight and stature Boas has recorded measurements of 

 height sitting, head-length, head-width, length of fore-arm, 

 and hand- width. 



Though the evidence, it must be admitted, is scanty, it is 

 none the less a remarkable fact that in all the cases we have 

 examined the variability, whether of the whole organism or of 

 its parts, decreases with the decrease in the rate of growth. 

 We seem to be in the presence of a phenomenon of general 

 occurrence, though what the significance of the phenomenon is 

 is not at present clear. 



As is well known, Weldon has argued that the decline in the 

 variability of the older crabs is due to a selective death-rate, an 

 argument which is supported by the same author's observations 

 on the snail Clausilia, since in this form the variability of the 

 adult was found to be the same as the variability of the same 

 individuals when young, but less than the variability of the 

 general population of young. It is possible that the marked 

 decrease in the variability of the American as compared with 

 the British periwinkles may also be attributable to the same 

 cause, since this animal has only recently been introduced into 

 America, and may, therefore, be subjected to a more severe 

 struggle for existence in its new environment. 



It is doubtful, however, whether this explanation will fit 

 all cases. 



Vernon has suggested that at periods of rapid growth the 

 effect produced upon the organism by a change in its environ- 

 ment must be much greater than at other times, and, since he 

 has further shown that one of the effects of an adverse change 

 of circumstances is an increased variability, he argues that an 

 increase in variability would naturally accompany a high growth- 

 rate. 



Lastly, Boas points out that the rate of growth is itself 

 a variable magnitude, and this ( variation in period ' may, with 

 other causes, be a factor in producing the actual variation at 

 each stage. Should that be so, the variability would necessarily 

 increase and diminish with an increasing and diminishing growth- 

 rate, since those that are above the mean would tend to remove 



