32 NATURAL SCIENCE. January. 1897. 



On the Lamarckian hypothesis this result would be attributed to the 

 direct action of the environment ; on the Darwinian view to natural 

 selection, the "only assumption which it is necessary to make, being 

 that the small variations on which selection acts may always be 

 trusted to occur. 



All admit that the larger differences in structure which separate 

 animals are useful, that is, are adaptive in character. Most 

 systematists are beginning to admit that such modifications may have 

 been effected again and again, and in despair they are taking refuge in 

 minute peculiarities of form, pattern and arrangement, which they 

 hope are not adaptive. This seems to me to be a futile position, the 

 provisional acceptance of which is only rendered possible because, 

 except in the cases where closely allied forms follow each other in 

 immediately superposed strata, we have no means of determining the 

 exact ancestry of a species. What we really do when we determine 

 the structure of an animal is to unravel a series of superposed 

 adaptations, and we class together those animals in whose structure 

 we detect evidence of their having undergone the same series of 

 modifications in the same order. 



All similarity in structure between two animals is primarily due 

 to similar external conditions; the longer the period of action of the 

 same environment has been, the more complete is the likeness ; and, 

 by identity of ancestry we can only mean the extreme case in which 

 the action of similar environmental conditions has extended for 

 an indefinite time back into the past. 



Just, therefore, as it would be a hopeless task to attempt to trace 

 a single hypha in a mushroom stipe back to the spore from which it 

 arose, so it appears to me is the attempt to trace several modern 

 species back to a single original species. 



The important thing to know about any hypha is its relation to the 

 general anatomy of the plant, and especially iis level on the stem ; and 

 the important thing to know about the arthropod groups is not 

 whether they were all descended from the same species of annelid, but 

 whether they were derived from annelid ancestors by the same series 

 of modifications. The morphologist should aim at establishing and 

 defining definite grades or levels of structure, and correlating these 

 with the environmental changes which produced them. 



I do not iiatter myself that any of the points of view which I have 

 endeavoured to set forth in this paper is absolutely new. Each of 

 them will be found to have been either explicitly stated or implicitly 

 assumed by some zoologist or other ; but there is perhaps no zoologist 

 who has not argued in a manner inconsistent with some one of the 

 principles I have endeavoured to establish ; and it seemed to me most 

 desirable to try to give a coherent and exphcit account of the principles 

 vaguely recognised and imperfectly understood on which phylogenetic 

 speculation is based — and such an attempt has been made in this 

 paper. E. W. MacBride. 



