No. 641] 



THE DURATION OF LIFE 



491 



interval of time, say between x and .'^^ + 1, or dx = 

 L — These frequencies of death taken together 



constitute the so-called line of a life table. 



3. The death rates at each time (or age) .r; i.e., the 

 ratio of the deaths between time x and a; + 1 to the sur- 

 vivors at the time x. These observations together con- 

 stitute what is kno^ra as the q., line of a life table q.v = 



4. The curtate ex]XH'tation of life of individuals at a 

 given age x. This is the mean or avertige after life time of 

 all those individuals alive at age .r neglecting fractions of 

 the X interval. These obsei-vatious together constitute 

 the e.r line of a life table. 



These simple definitions state with ctitii'cly sufficient 

 accuracy for i)resent purpo.ses tht" -i-nilirniuM^ of llie con- 

 stants which we shnll present. Aii\ mir wi-hlim' lo go 

 more i)articularly into details of .-U'tunrial nictliods will 

 find a useful elementary iiilrodiiclioii in 1 1 ciuI.tx-ii (22) 

 or Dawson (23). 



Tt is our pui'posc lo pi'csciit \\i'Vr liiV tai)lt'S \'ov t'oiif 

 li-i-oupv of flie.. In MM-vc. (ii-t. lo MioxN \\u- unicrnl lax\- 

 of iiioimlilN ill l),n^ni>]nhi a^ .-.nnpaiTd with in.-nu the 

 oiilx oihrr oi-aiiiMii tor which we h;i\r .■Mm^lN.- and . \- 



ison in experimental M'ork on Drosn itliiln to h<> rrportrd 

 in subsequent papers. 



The groups for which complete tables .iic procntrd 

 are these : 



1. Lrnni-irn^n.'J ^. Thi^tabh. i ii.-hi. h- .-i 1 ' on r data 

 uptoJiin.'. 11*21. on iionnal ( ^< .. rioi oxpci-itnciit:! u\ mod- 

 ified) duration of WW- of ni.-ih^ D, <>^rh>h> iti.li\ IdnaU'at 25° 

 C.belonuinu to thr WAlnxxlw^ Mork^ u-l. p. 1^,; Old 

 Falmouth, Xcw l-'niniout h. S,.i)ia nnd I'la^h' roint. In 

 these stock-- ;ii; the in.liN IduaU ha\o wihl i>pc wings, 

 hence the designation " long -winged. " 



