142 



AGE-INCIDENCE OF CAUSES OF INFANT MORTALITY 



to do with this undertaking. Later studies of chance distribu- 

 tions included the age distribution of several fevers, and these 

 sent him back to his suspended task with a key to its completion. 

 Among the " fevers" studied were typhoid, scarlet fever, and 

 diphtheria. The typhoid curve, having the least relative altitude, 

 is the best illustrative curve, and is presented here. Typhoid 

 fever presents toward childhood such an aspect as scarlet fever 

 presents toward infancy. The average of all the frequencies for 

 typhoid falls late is the eighteenth year (17.775), but the great- 

 est frequency occurs late in the fourteenth year (13.715). Of 

 chief interest, however, is the foot of the curve toward birth, 

 bending rather sharply toward that event. This curve requires 

 that three in each thousand cases of typhoid fever shall occur 

 in antenatal time. Pearson was able to complete the mathematical 

 treatment of the chances of death when he perceived, first, that 

 "there is a mortality of infancy distinct from that of childhood," 

 and second, that the general mortality cannot be resolved into 

 its age elements without extending the figures through antenatal 

 time to a point which is "very approximately nine months before 

 birth." He also noted that while the maximum incidence of 

 diphtheria and scarlet fever fell upon the fourth and fifth year, 

 their maximum deadliness falls on the third year. It can hardly 

 be doubted, I think, that measles, scarlet fever and diphtheria 

 were important factors in determining the nature of that curve 

 which brought Pearson's investigation to a temporary standstill; 

 nor does it seem very doubtful that these diseases are in fact 

 as distinct from those of infancy on one hand as from those of 

 adult age on the other. 



What form may we expect the distinctive causes of post-natal 

 infant mortality to assume with respect to the antenatal causes, on 

 one side, and to those of childhood on the other ? Compressed into 

 two years, without divisions of a year, the statistics are absolutely 

 non-descriptive; not as intelligible as the statistics of radium in 



