president's address. 39 



casual visitants to a country should be ignored, but I do think that 

 it is a wrong to science to enumerate, for examples, an European 

 vagrant to America, never found but once, and not likely to recur 

 again for an indefinite time, if ever, or an American bird that has 

 been once found in Britain, in a line with the well-known members 

 of the respective faunas. Their i)lace would be most natural in an 

 appendix or foot note, and they should at least be without the 

 serial numeration, if such is given, of the catalogue of permanent 

 and seasonal members of the fauna. Let me also protest, as I have 

 done several times before, against the incorporation of Bassalian 

 types with the species of littoral faunas nearest to them geographi- 

 cally. With catalogues and data, such as I have indicated, at hand, 

 we could soon determine, as closely as practicable, the limits of most 

 of our faunas, and the general attention now paid to natural history 

 holds out the hope that the coming time may not long be deferred. 



In conclusion, I submit a few deductions that naturally result 

 from our observations. 



A distinction is to be made between the territory occupied by an 

 association of animals and the occupants thereof, and the limits of 

 founas cannot be exactly correlated with territory, except in rare 

 cases. 



The significance of animal types as indicators of zoogeographical 

 regions is, other things being equal, in ratio to their recent de- 

 velopment. 



The fresh-water types are the best indicators of the early relations 

 of the respective regions. 



The flying, and especially migratory, types are the most accord- 

 ant witli the actual relations of land areas. 



Temperature is a prime factor, and land a secondary, in the dis- 

 tribution of marine animals. 



The lay of the land is a primary, and temperature a .secondary, 

 factor in the distribution of inland animals. 



