STUDIES IN ANIMAL LIFE. 105 



been borrowed ; and, thirdly, that here again none 

 of the languages in which these verbal forms occur 

 possess the elements of which they are composed." 



All these languages resemble each other so close- 

 ly that they point to some more ancient language 

 which was to them what Latin was to the six Ko- 

 man languages ; and in the same way we are justi- 

 fied in supposing that all the classes of the verte- 

 brate animals point to the existence of some elder 

 type, now extinct, from which they were all de- 

 veloped. 



I have thus stated what are the two hypotheses 

 on this question. There is only one more prelimi- 

 nary which it is needful to notice here, and that is, 

 to caution the reader against the tendency, unhap- 

 pily too common, of supposing that an adversary 

 holds opinions which are transparently absurd. 

 When we hear a hypothesis which is either novel 

 or unacceptable to us, we are apt to draw some very 

 ridiculous conclusion from it, and to assume that 

 this conclusion is seriously held by its upholders. 

 Thus the zoologists who maintain the variability 

 of species are sometimes asked if they believe a 

 goose was developed out of an oyster, or a rhinoce- 

 ros from a mouse ? the questioner apparently having 

 no misgiving as to the candor of his ridicule. There 

 are three modes of combating a doctrine. The first 

 is to point out its strongest positions, and then 

 show them to be erroneous or incomplete ; but this 

 plan is generally difficult, and sometimes impossi- 

 E2 



