HISTORICAL PREFACE xxxiii. 



of Life, but it can hardly be admitted as a tenable one. 

 Knowing that a great many Gannets are bred at St. Kilda, 

 and that the young were formerly killed there in large 

 numbers for food — four thousand or more in a year — he 

 infers therefrom that if the Gannet had not been a long- 

 lived species, the stock must have declined. As a matter 

 of fact the number of young Gannets killed there probably 

 never greatly exceeded the number of young wliich 

 survived, so the argument hardly holds. For the same 

 reason the argument cannot be satisfactorily appUed either 

 to the Fulmar Petrel or the Puffin. 



When the eighteenth chapter of this book was written, 

 I had not had the advantage of reading the treatise by 

 Dr. Chalmers Mitchell, already alluded to, " On Longevity 

 and Relative Viabihty in Mammals and Birds " (" Proc. 

 Zool. Soc," 1911, p. 425) — a treatise which brings together 

 records of the duration of bird-Mfe, including Gannets 

 and other marine species. As regards the Gannets, the 

 particulars given (pp. 5G6, 536) are on the whole dis- 

 appointing. No case is recorded by Dr. Mitchell of Sula. 

 hassana and 8. serrator having exceeeded a year and a 

 half in captivity, but this is even a less time than *S'. 

 hassana has lived on my small pond ; however S. 

 leucogastra fBodd), the Brown Booby, was kept for three 

 years and four months by the Zoological Society. But 

 Gannets, like Pehcans, can evidently live much longer than 

 this — of that we have in the case of the former fairly con- 

 clusive proof. 



Besides the evidence which has been already adduced 

 in Chapter XVIII. we have P. J. Selby's definite statement 



