THE 



AMERICAN NATURALIST 



Vol. November-December, 1921 No. 641 



EXPERIMENTAL STUDIES ON THE DURATION 

 OF LIFE 



I. Inteoductory Discussion of the Dueatiok of Life in 

 Drosophila ^ 



RAYMOND PEARL AND SYLVIA LOUISE PARKER 

 Such quantitative knowledge as exists of fundamental 

 principles in the general biologj^ of the duration of life 

 has, in the main, been derived from an examination hj 

 purely statistical methods of human mortality records. 

 Of course a good deal of information about the biology 

 of death and duration of life of a general and non-quanti- 

 tative character has been gained from experimental work 

 on lower organisms. This literature has recently been 

 reviewed by one of us (Pearl (1) to (7) inclusive). But 

 the outstanding fact is that most of the existing (jucnif 'da- 

 tive data about duration of life are purely statistical, and 

 derived from man as material. 



Tlie statistical method of acquiring knowledge of nat- 

 ural phenomena has a number of distinct and important 

 limitations (cf. Pearl (8)). It is the settled policy of this 

 department to check every conclusion drawn from purely 

 statistical methods by an independent experimental in- 

 vestigation of the same problems, wherever in the nature 

 of the case this is possible. Most problems of human 

 vital statistics can not, in the nature of the case, be in- 

 vestigated experimentally, in any direct way with man 

 liimself as material. Probably this is chiefly the reason 

 why all of the immense mass of data collected, and work 

 done upon vital statistics has contributed so little in the 



1 PaiK'is fioin the Dei^artment of Biometry and Vital Statistics, School of 

 Tr.vfriono aibl Public Health, Johns Hopkins University, No. 45. 



