322 PROCEEDINGS OF THE ACADEMY OF [1887. 



world at lai'ge a condition which can be clearly demonstrated 

 then, manifestly, the ratio of survivals must be a definite measure 

 of this development ; and if this is the case, it must, necessarily, 

 prove the safest guide in the delimitation of chronological bounda- 

 ries. Were it possible to definitely determine the proportion of sur- 

 vivals, or what is practically equivalent to the same thing, determine 

 the amount of actual faunal peculiarities, by specific determinations 

 made with equal care for all countries, we would then have, and 

 only then, a rational basis for a systematic classification applicable 

 to the whole world. 



