28 BIOLOGICAL SOCIETY OF WASHINGTON. 



succumbed, the types continued to survive in the south, and the 

 deserted lands were repeopled when a gentler climate again pre 

 vailed. 



It is well known that the equatorial inhabitants of the old and 

 new worlds are quite dissimilar from one another, and it may be 

 thought that such dissimilarity is antagonistic to the law that tem 

 perature is a primary factor in distribution. The facts in the case, 

 however, seem at once to justify the law and to necessitate the in 

 vocation of another. Undoubtedly there are no hindrances offered 

 by the climates of the several equatorial regions to the life of any 

 of the tropical types of life in all of the tropical regions. Their 

 limitation, therefore, must be looked for in other causes ; the con 

 ditions imposed by climate on the north and south, where the con 

 tinents converge, have forbidden their extension, and the distinct- 

 iveness of the types characteristic of the several regions is a result 

 of the laws of evolution acting during the long ages of dissolution 

 of the respective regions. 



WHAT TYPE IS MOST FITTED TO EXPRESS FAUNAS ? 



As will be remembered, Mr. Wallace has especially insisted on 

 the super-eminent availability of the mammals for the determina 

 tion of faunal regions, and in this respect is cordially endorsed by 

 Mr. Allen. I fail, however, to appreciate the entire force or per 

 tinence of their reasons, although to some extent pertinent. The 

 mammals are certainly the best and the only factors to determine 

 the mammalian faunas, but for general faunas all animals are to be 

 considered. The survey of the chronological and geographical 

 history of the various classes of the animal kingdom reveals marked 

 discrepancies between the several types in both of these cases, and 

 we are led to postulate several propositions as tentative hypotheses 

 for proof or disproof. All other things being equal, (i) the longer 

 a type has been in existence, the more widely should it have been 

 disseminated ; (2) the later a type has supervened, the more re 

 stricted should it be, and the closer its accommodation to existing 



