The Distribution of Senescence 



unmoglich wird man dieser Langlebigkeit nach dem, was man 

 von anderen Tieren weiss, nicht bezeichnen diirfen' 1 (1922, 

 p. 36, footnote). 



A proper survey of the longevity of invertebrates can hardly 

 yet be undertaken — the information is mostly lacking. It seemed 

 wisest in compiling Table III, which includes a few of the 

 longest and most interesting invertebrate records, to give not 

 only the record and source, but the type of evidence upon 

 which the record is based. In invertebrates which metamor- 

 phose, length of larval life often depends entirely upon environ- 

 ment and food, while in other forms adult life can be punctuated 

 by very long spells of diapause. Figures for these forms should 

 therefore when possible indicate the circumstances of life. 

 Larval life-spans have in general been omitted from Table III. 

 The most reliable records are in all cases those of animals kept, 

 like Labitte's (1916) beetles or the Edinburgh sea anemones 

 (Ashworth and Annandale, 1904) under close observation in 

 captivity. Evidence from growth rings requires very careful 

 scrutiny. Some purely inferential evidence, as of the age of 

 termite primaries, is probably reliable. There are also some 

 surprisingly high records in the wild, especially for pelecypods, 

 where the method of ageing by rings of growth has been well 

 upheld by other evidence. The life-span of common inverte- 

 brates certainly remains a wide-open field for those with 

 facilities and an unlimited capacity for taking pains, and one 

 where any reliably-attested information is worth putting on 

 record. 



1 'No doubt these findings have little probability in themselves, but one 

 cannot dismiss such longevity records as impossible, in view of what is 

 known of other animals.' 



55 



