130 DARWINIAKA. 



appear to be parallel lines, as a general thing neither 

 approaching to nor diverging from each other. 



The first hypothesis assumes that they were parallel 

 from the unknown beginning and will be to the un- 

 known end. The second hypothesis assumes that the 

 apparent parallelism is not real and complete, at least 

 aboriginally, but approximate or temporary ; that we 

 should find the lines convergent in the past, if we could 

 trace them far enough ; that some of them, if produced 

 back, would fall into certain fragments of lines, which 

 have left traces in the past, lying not exactly in the 

 same direction, and these farther back into others to 

 which they are equally unparallel. It will also claim 

 that the present lines, whether on the whole really or 

 only approximately parallel, sometimes fork or send off 

 branches on one side or the other, producing new 

 lines (varieties), which run for a while, and for aught 

 we know indefinitely when not interfered with, near 

 and approximately parallel to the parent line. This 

 claim it can establish ; and it may also show that these 

 close subsidiary lines may branch or vary again, and 

 that those branches or varieties which are best adapted 

 to the existing conditions may be continued, while 

 others stop or die out. And so we may have the basis 

 of a real theory of the diversificatiorh of species ; and 

 here, indeed, there is a real, though a narrow, estab- 

 lished ground to build upon. But, as systems of 

 organic E^ature, both doctrines are equally hypotheses^ 

 are suj^positions of what there is no proof of from 

 experience, assumed in order to account for the ob- 

 served phenomena, and supported by such indirect 

 evidence as can be had. 



