RELATIVE VIABILITY IX MAMMALS AND BIRDS. 545 



(9) The tabulation and analysis of the records in this memoir 

 has been a laborious task which has not been lightened by the 

 knowledge that the materials were imperfect. But I trust that 

 it may be a step towaixls obtaining systematic and more exact 

 knowledge on the subject. Such knowledge is the only sure basis 

 for that improvement in the condition of animals in captivity 

 which those in control of Zoological Collections desire to bring 

 about. The peculiarity by which a menagerie is distinguished 

 from a museum is that its zoological specimens are alive. Not 

 the obtaining of rare animals, nor the addition to the records of 

 " species new to the Collection '" should be the chief glory of a 

 Zoological Society, but that Collection should be judged most 

 valua.ble and successful in which the average dtu'ation of life of its 

 inhabitants approaches potential diu'ation most closely. 



ISTOTE ON THE ThEORY OF LONGEVITY. 



With three notable exceptions, those of Lankester, Weis- 

 mann, and Metchnikofi', discussions of the theory of longevity 

 may be dismissed briefly here, pai-tly because they have been 

 treated f idly in the litei'ature of the subject *, and partly 

 because they do not pretend to explain why the gift of life has 

 been measured out to dilTerent animals in unequal portions. 

 Bacon made a careful summary of the known facts, but refi'ained 

 from general conclusions. " In tame creatures " he wrote, as cited 

 by Lankester from Basil Montagu's tra,nslation, " their degenerate 

 life corrupteth them, in wild creatures their exposing to all 

 weathers often intercepteth them ; neither do those things which 

 may seem concomitants give any furtherance to this information 

 (the greatness of their bodies, their time of bearing in the womb, 

 the number of their young ones, the time of their growth, and 

 the rest), in regard that these things are intennixed, and some- 

 times they concur, sometimes they sever." Bufibn thought that 

 there was a relation between the total duration of life and the 

 period of growth, and set down the ratio between the two as 6 or 

 7 to 1. Flourens followed Bufibn's idea, but took as the limit of 

 growth, the age at which the long bones unite with their 

 epiphyses, estimating that the ratio of the longevity to the period 

 of growth was as 5 to 1. Bunge, without calculating an exact 

 i-atio, pointed out that there was a frequent relation between the 

 longevity and the time taken by a new born animal to' double its 

 weight. Such calculations at the best ai'e limited in theii- 

 application to the higher animals, and even amongst these have 

 to encounter many exceptions ; so far as they go they must be 

 taken as secondary and comparatively accidental correlations. 



Ra.y Lankester's ea,i-ly contribution [t. c. p. 71) was a serious 



* Seo specially E. Ray . Lankester, ^. e. ; E. Metclmikotl', ^ r. \^. 39, and article 

 Litncievitji. ill tlic XI. cditiiiii uf the Encyclopedia IJritannica. 



