THE EXTINCT MAMMALIA. 297 



dent to the specialist in each department, that like does not at 

 all times produce like. It is perfectly clear, and I will venture 

 the assertion that nearly all the biologists in this room will bear 

 me witness, that variability is practically unlimited in its range, 

 and multiplied in the number of its examples. That is to say : 

 species vary by adding or by failing to retain certain characteris- 

 tics ; and generic and other characters are found to appear or 

 disappear in accordance with some law to be discussed farther on/ 



I believe that this is the simplest mode of stating and explain- 

 ing the law of variation : that some forms acquire something 

 which their parents did not possess ; and that those which ac- 

 quire something additional have to pass through more numerous 

 stages than their ancestors ; and those which lose something pass 

 through fewer stages than their ancestors ; and these processes are 

 expressed by the terms "acceleration " and "retardation." 



Of course we are met with the opposite side of the case — the 

 law of heredity. We are told that the facts there are not ac- 

 counted for by any law of evolution ; that we can not pass from 

 one class of facts to the other class of facts ; that the law of the 

 one class is not that of the other. Here is a question of rational 

 processes, of ordinary reason. If the rules of chemistry are true 

 in America, I imagine they are true in Australia and Africa, al- 

 though I have not been there to see. If the law of gravitation is 

 effective here, I do not need to go to Australia or New Zealand 

 to ascertain whether it be true there. So, if we find in a group 

 of animals a law sufficient to account for their creation, it is not 

 necessary to know that others of their relatives have gone through 

 a similar process. I am willing to allow the ordinary practical 

 law of induction, the practical law of inference, to carry me over 

 these gaps, over these interruptions. And I state the case in this 

 way, because it is just here that some people differ from me, and 

 it is just here that I say the simple question of rationality comes 

 in. I can not believe that Nature's laws are so dissimilar, so 

 irregular, so inexact, that those which we can see and understand 

 in one place are not true in another ; I also believe that the ques- 

 tion of geological likelihood is similar to the question of geo- 

 graphical likelihood. If a given process be true in one of the 

 geological periods, it is true in another ; if it be true in one part 

 of the world, it is true in another ; because I find interruptions 

 in the series here, it does not follow that there need be interrup- 

 tions clear through from age to age. The assumption is on the 



