428 DR. r, CHALMERS MITCHELL ON LONGEVITY AND 



of the young spai-row when it is hatched, on the most favourable 

 estimate cannot be more than one or two years ; the potential 

 longevity of small Passerine birds is certainly over 20 years. 



The difference between the speciiic longevity and the potential 

 longevity is a rough measure of the severity of the natural 

 conditions under which a species lives, and a similar mode of 

 comparison can be applied to assemblages of creatin'es living 

 under different kinds of unnatural conditions. The object of this 

 communication is in the first place to suggest a fashion in which 

 a system of this kind could be employed usefully in Zoological 

 Gardens, and in the second place to give the results of examina- 

 tion of a series of records regarding the duration of life of 

 mammals and birds in captivity, which, although they are 

 imperfect in many respects, are more extensive than any with 

 Avhich I am acquainted. The records in question are contained 

 in two manuscript folios, kept in the Prosectorium of this Society 

 and covering the period from 1870 to 1902. In these were 

 entered day by day the names of the animals that had died in the 

 Gardens, with the dates of their arrival and of their departure. 

 From the point of view of estimating the effects of captivity on 

 duration of life, they are defective in two important respects. In 

 the first p]ace, there is no record of the ages of the animals on 

 arrival. This of course must always be impossible in the vast 

 majority of cases, but two stages, the infantile and the senile, 

 ought to be noted and excluded, where possible. Of these, the 

 senile stage is usually easy to detect in the case of mammals, and 

 it is at least probable that senile mammals are seldom purchased 

 or accepted. On the other hand, senility in birds is extra- 

 ordinarily difficult to detect, and the presence of such individuals 

 will probably always confuse the recoixl. The infantile period, 

 corresponding to the first four years of human life, but of course 

 varying greatly in its duration in difierent animals, is easy to 

 detect ; and as the difficulties of rearing infants are entirely 

 different from the general problems of animals in captivity, infants 

 should be excluded from recoids such as those with which I am 

 dealing. In a number of cases, but not in all, I have been able 

 to discard "infantile" entries. The second important defect in 

 the statistics is that they do not include any note of the condition 

 of the animals on arrival. Owing to the mode of capture, and still 

 more to the conditions of transport, a large number of wild 

 animals arrive at Zoological Gardens in poor condition, and die 

 from causes not due to their new environment. For the kind of 

 investigation I am now discussing (as of course for other reasons) 

 new arrivals ought to be kept in quarantine. The duration of the 

 quarantine need not be for a fixed period, but should be long- 

 enough to make certain that there has been recovery from the 

 effects of capture and transpoi't, and the date of detention in the 

 Gax'dens should begin with that of liberation from quarantine. 



