Permanence and Evolution. 87 



that various domestic pigeons are descendants 

 of the rock dove, this ought to mean that from 

 a strain of the rock dove of known purity, under 

 conditions carefully excluded from the chance of 

 reversion or casual impurity, these several 

 varieties have been raised, and that the ex- 

 periment can be repeated by any one who has 

 the means and the patience to try it properly. 

 If that were done, we should have at least 

 a fact to start with, and on which to build 

 our superstructure. Now we have absolutely 

 none ; and if we grant that species as they 

 commonly stand can be produced by evolu- 

 tion, we are only a very little advanced, for 

 it would be arguing from the less to the 

 greater to think that this threw any light on 

 the more important structural differences which 

 separate genera and so on to natural orders, 

 etc., which can only in a few instances be re- 

 presented as differing only in degree from the 

 former. But now we have the course of scientific 



