64 



THE EVOLUTION OF DEATH 



of temperature, which is known to have a great influence upon 

 growth. Nature makes this exclusion for us in the case of 

 warm-blooded animals. I selected for my own experiments on 

 warm-blooded animals guinea pigs for various practical reason 

 and I maintained a colony of these animals for many years. 

 Every animal of the colony was weighed at definite intervals of 

 age. After many thousands of determinations of the weight 



Fig. 26. — Graphic representation of the increase of weight in children of the Boston 



schools. — After H. P. Bowdilch. 



(Knaben, boys. Mddchen, girls. Jahre, years.) 



had been collected, they were worked over statistically.^^ 

 My first problem was to invent a method which permitted 

 the representation of the rate of growth. Formerly investi- 

 gators were satisfied to represent growth graphically in a very 

 simple way. Curves were constructed in which the abscissae 

 corresponded to the age, and the ordinates to the weight, 

 Fig. 26. Such a curve, however, although it represents the 



