32 



SCIENCE 



[N. S. Vol. XLIV. No. 1123 



about age 4.5. We might go farther and say 

 that for this same decade the males did not 

 advance as much at the ages at which they did 

 advance as they did in the previous decade. 

 On the other hand, the females maintained or 

 even excelled their record of 1890-1900 in the 

 decade 1900-1910. 



The most important feature of the table of 

 death rates is the group or period of ages at 

 which both males and females have retro- 

 gressed. This retrogression is significant in 

 value wherever indicated, and that it is not due 

 to faulty statistics or errors is clearly shown 

 by the fact that it appears in both decades in 

 both sexes. Further, the retrogression is not 

 spasmodic, but continues firmly from about age 

 60 on to the end of the table. 



EXPECTATIONS OF LIFE 



There are a few widely varying values at 

 the terminal ages, but, as mentioned above, the 

 statistics at these ages are so faulty that little 

 or no interpretation is possible. 



Summarizing the results indicated by the 

 table of death rates, mortality conditions seem 

 to have been improved at ages below sixty 

 during the two decades 1890-1900 and 1900- 

 1910 among both the males and the females, 

 steadily so among the females but not so much 

 so among the males. At ages sixty and above, 

 both males and females seem to have retro- 

 gressed, particularly the males whose period of 

 retrogression during the decade 1900-1910 be- 

 gan as far back as age 45. This period of 

 retrogression among death rates for both sexes 

 continues steadily toward the last ages of 

 human life. 



As indicated in the table of expectations of 

 life given above, the average future life time 

 of males at age twelve seems to have lengthened 

 2.19 years in the decade 1890-1900 and only 

 .59 of a year in the decade 1900-1910, or 2.78 

 years in both decades. The gain of only .59 

 is rather difiicult to explain, for even the gen- 

 eral death rate suffered a relapse in 1910, and 

 no one seems to know exactly why. It is pos- 

 sible that the fact that the period of retro- 

 gression encroached upon the earlier ages 

 might offer at least a partial explanation. The 

 period of retrogression among the expectations 

 of life of the males is seen to begin about age 

 fifty in the decade 1S90-1900 and about age 

 thirty in the decade 1900-1910. 



That the initial ages of the period of retro- 

 gression in both decades precede the corre- 

 sponding ages in the table of death rates from 

 10 to 15 years is what might be expected and 

 is really quite important in that it emphasizes 

 the fact that a retrogression in death rates at 

 any period of ages will affect the expectation 

 of life of all those living at any earlier ages. 

 The two initial ages, fifty and thirty, men- 

 tioned above, differ from earlier ages only in 

 the fact that these are the first ages at which 

 the effect of retrogression at the advanced 

 ages outweighs the effect of improvement at 

 the earlier ages. 



