24 . NATURAL SCIENCE. January. 



as I am aware, has examined with care the cases of those Teleostean 

 fish, Hke some species of blenny, which come out of the water and 

 seek prey between tide-marks, where a fin has been to some extent 

 transformed into a supporting Hmb. I do not wish to dogmatise 

 against the possibihty of changes having taken place in past time 

 altogether different in kind to those of which we find evidence in 

 present specific differences; but I do maintain that if we assume such 

 changes in explaining the present structures of animals we are on 

 utterly unsafe ground. My suggestion then is that by carefully com- 

 paring and tabulating specific and even generic differences we may be 

 able by induction to arrive at " laws of successful variation." An 

 example or two will serve to make clear what I mean. If we review 

 the group of the Lamellibranchiata, we are struck by such forms as 

 Teredo and Pecten. A priori it is possible to argue that the idea that 

 these were derived from the ordinary type of Mollusca is hypothetical. 

 Practically all zoologists are agreed about it, no different explanation 

 having so far as I know ever been suggested. Now my point is that 

 cases like these give us definite data to go on ; they really enable us 

 to ascertain now, — what Bateson has looked forward to doing only 

 in the remote future — that certain changes are possible to certain 

 animals. Perhaps it may be retorted that all this is known and 

 acted on. My reply is that many changes have been postulated, 

 which have no analogue amongst variations that we know must 

 have occurred. Balfour postulated the formation of a new mouth ; 

 Dohrn that of a new arm. All through the Mollusca we find no 

 such fundamental change, nor the foundation of any new type of 

 skeleton such as the supporters of the Annelid theory are bound to 

 postulate. 



If we now turn to the difficulty of distinguishing primitive and 

 undifferentiated from degenerated structures, we shall find, I think, that 

 objections may be made to the current modes of reasoning on this 

 subject. It is often implicitly assumed (i) that if an animal can be 

 shown to be degenerate in one respect, it is not primitive at all, and 

 throws no light on the ancestry of a group ; (2) that any amount of 

 degeneracy may be followed in the history of the race by any amount of 

 forward evolution. I shall give reasons for questioning both supposi- 

 tions. To prove that the first is a real factor in theorising, we have only 

 to remember the war waged round Amphioxiis. Few would deny that 

 it is degenerate in some points ; the question at issue is. Does it in its 

 general organisation represent the ancestor of Vertebrata or not ? 

 Now it seems to me that on d priori grounds we have every reason 

 to expect that all animals, which possess on the whole a primitive 

 organisation will be degenerate in some features. For their modified 

 relatives which have departed from the ancestral type have ex hypothesi 

 been forced to do so by natural selection, and how have these primitive 

 animals been able to escape from that pressure of environment which 

 modified allied forms ? In some cases, perhaps, isolation on oceanic 



