504 



THE AMERICAN NATUBALIST [Vol. LV 



ago and that for the present time. As compared with 

 (Hover's 1910 U. S. tables, the Roman popuhition of 

 Africa at the dawn of our era was, in respect of the course 

 of its mortahty, even more DrosopJiila-like than Droso- 

 lihila itself! Now in Roman Africa there was relatively 

 little of what we now understand as sanitation, hygiene, 

 and preventive medicine. Men lived to old age, if they 

 did, by virtue mainly of the strength of their innate con- 

 stitutions, and their good luck in avoiding fatal accidents. 

 At the present time hospitalization, the science and art 

 of iiHMlifiiio and ])ul)lic health, and general sanitation keep 

 maii\- pcrsdiis ali\'(' well into middle age who would in 

 those days have died much earlier because of a lack of 

 constitutional ruggedness. Altogether the data of Fig. 5 

 seem highly significant in relation to the hypothesis sug- 

 gested above as to the reason for the difference between 

 Drosopliila and man in respect of mortality curves. 



Accidental Deaths 



The tacit assumption in all the foregoing is that each 

 of the 11,772 flics comprised in the four life tables died 

 a natural death, and that the time of death (or ihiration 

 of life) was in each case' determined fundamentally by 

 internal factors, since the environment was substantially 

 a constant for all. 



For certainly more than 98 per cent, of the flies this 

 assumption is unquestionably true. But in view of the 

 possibility that some few of the flies might be dying ac- 

 cidental deaths by drowning in the moisture, which some- 

 times collects on the surface of the food, it was thought 

 worth while to attempt to prevent the collection of mois- 

 ture in some of the bottles and compare the duration of 

 life of flies kept in such bottles with that of flies kept 

 imder tlie ordinary conditions. Accordingly, l)ottles of 

 food wei-c pi-epai-ed hv pntlini:' <li>c^ of several layers 

 of a ve,-v ab.orhent paper. Zorhlk. in the l.ottom of the 

 bottle and then poui'ing in the food and letting it solidify 



