DISCOVERIES OF CONNECTING LINKS. Ill 



of these types of union, until the difficulty has not 

 only disappeared, but has been turned into a very 

 cogent positive argument. 



One very important feature in regard to these 

 connecting links is the geological position in which 

 they are found. A connecting link ought to appear 

 before the two forms which it connects, or some- 

 times between them. The former position is the 

 more natural one, and a synthetic form in this posi- 

 tion is understood to be the common ancestor of 

 the later groups which it connects. Most of our 

 fossils are of this character, being anatomically 

 intercalated between existing groups, and always 

 occurring at earlier periods. The second position 

 might sometimes occur, and would be explained by 

 supposing the fossil in question represented a tran- 

 sitional stage between the earlier and the later 

 groups which it connects. The occurrence of a 

 connecting link later than the animals connected 

 by it, would be, according to the descent theory, 

 surprising, and only explained by supposing that it 

 originally appeared earlier, but left no trace of itself, 

 while some individual survivor, in after-ages, did 

 happen to be preserved as a fossil. And this expla- 

 nation would evidently only apply to sporadic cases. 

 Now, the universal rule is, that true connecting 

 links occupy one of the first two positions, and 

 never the last. The successive fossils which connect 

 the horse family with the ancient five-toed ancestor 

 occur in successively lower and lower rocks ; while 

 the forms connecting this family with the even- 

 toed Ungulates occur still lower down. Indeed, I 



