14 CM. JACKSON 



The foregoing data indicate clearly that the coefficient of vari- 

 ation in body weight is smaller in the younger rats (newborn 

 and seven days) and larger in the older (twenty days, six weeks, 

 ten weeks and five months). Less conclusive is the evidence that 

 the maximum variation, for the ages observed, occurs at twenty 

 days, and that the variation is in general greater in the male 

 than in the female. The changes in variation cannot be ascribed 

 to a selective death rate, as there were in the litters observed 

 very few deaths before the age of five months. 



For the variation in the weight of the human body at different 

 ages, fairly complete data are available for comparison. For the 

 English newborn, Pearson ('99) finds the coefficient of variation 

 for the male to be about 15.7, and for the female 14.2. For 

 Cambridge University students, nineteen to twenty-five years of 

 age, he finds the coefficient for 1030 males to be about 10.8, and 

 for 160 females 11.2. For Oxford students (male) of eighteen to 

 twenty-three years, Schuster ('11) similarly finds the average 

 coefficient of variation in body weight to be 10.8, varying from 

 10.2 to 11.1 for different years of age. 



For the period from birth to the age of sLx years, no data are 

 available. But from six years onward we have the extensive 

 observations of Porter ('95) on St. Louis school children. Porter 

 uses the 'probable deviation' (which equals approximately 0.6745 

 times the standard deviation) as a measure of dispersion, but 

 for convenience of comparison this has been reduced to the 

 standard deviation for calculating the coefficient of variation in 

 table D. 



Porter concludes that variation is correlated with rapidity of 

 growth, a conclusion previously reached by Thoma ('82). Boas 

 ('97, '04) reaches similar conclusions. This relation will not hold 

 good for the human newborn, however; for although the growth 

 rate is then far more rapid than at any subsequent period, the 

 coefficient of variation is slightly less than at the age of puberty. 

 In the rat, as we have seen, the coefficient of variation is smaller 

 in the younger animals although the growth rate is far more 

 rapid then than later. 



