370 Prof. A. Agassiz on Palceontological 



ledge our inability to go beyond a certain point ; any thing 

 beyond the general parallelism I have attempted to trace, 

 which in no way invalidates the other proposition, we must 

 recognize as hopeless. 



But in spite of the limits which have been assigned to this 

 general parallelism, it still remains an all-essential factor in 

 elucidating the history of palasontological development ; and 

 its importance has but recently been fully appreciated. For, 

 while the fossil remains may give us a strong presumptive 

 evidence of the gradual passage of one type to another, we 

 can only imagine this modification to take place by a process 

 similar to that which brings about the modifications due to 

 different stages of growth — the former taking place in what 

 may practically be considered as infinite time when compared 

 to the short life-history which has given us, as it were, a 

 resume of the palaBontological development. We may well 

 pause to reflect that in the two modes of development we find 

 the same periods of rapid modifications occurring at certain 

 stages of growth or of historic development, repeating in a 

 different direction the same phases. Does it, then, pass the 

 limits of analogy to assume that the changes we see taking 

 place under our own eyes in a comparatively short space 

 of time — changes which extend from stages representing, 

 perhaps, the original type of the group to their most com- 

 plicated structures — may, perhaps, in the larger field of 

 palaBontological development, not have required the infinite 

 time we are in the habit of asking for them ? 



Palaeontologists have not been slow in following out the 

 suggestive track ; and those who have been anatomists and 

 embryologists besides have not only entered into most 

 interesting speculations regarding the origin of certain groups, 

 but they have carried on the process still further, and have 

 given us genealogical trees where we may, in the twigs and 

 branches and main limbs and trunk, trace the complete filia- 

 tion of a group as we know it today, and as it must theo- 

 retically have existed at various times to its very beginning. 

 While we cannot but admire the boldness and ingenuity of 

 these speculations upon genetic connection so recklessly 

 launched during the last fifteen years, we find that, with 

 but few exceptions, there is little to recommend in recon- 

 structions which shoot so wide of the facts as far as they 

 are known, and seem so readily to ignore them. The moment 

 we leave out of sight the actual succession of the fossils and 

 the ascertainable facts of post-embryonic development, to 

 reconstruct our genealogy, we are building in the air. Ordi- 

 narily the twigs of any genealogical tree have only a 



