348 COMPARATIVE ANATOMY 



in silence in order not to avow to ourselves the contradictions of our 

 logic. And what is this intimate sense of homology which directs 

 us thus? It is nothing less than the remains of the old concept of 

 the archetype which slumbers in us and wakes occasionally. 



When, in two allied forms (the instance is found in the mollusks) , 

 the blastopore becomes in one the mouth, in the other the anus 

 (Paludina), we might conclude that the mouth of the one is the 

 homologue of the anus of the other. We do not, however, and we 

 are right. But this proves that we do not hold to the criterion which 

 we have erected in our statement of homologies. We abandon 

 ontogenesis, and take for our guide connections and structure, that 

 is, the same characteristics which we declared entirely insufficient 

 to determine homologies if they were not corroborated by develop- 

 ment. 



After phylogenesis and ontogenesis, it is to connections that we 

 attach the most importance in determining homologies, and very 

 often we content ourselves with them as a criterion. This confidence 

 in connections has its reason for existence in the fact that we know 

 that organs modify more easily their relations of vicinity than their 

 relations of continuity. But it is not wholly founded on observation; 

 it proceeds in part from the conception of an archetype with which 

 our minds are imbued, and from the fact that we conceive more 

 easily of the derivation of that archetype by the stretching and 

 displacement of organs, elastic and pliable after the manner of 

 gutta-percha, than by the transposition of masses in the way one 

 treats the first draught of a wax model. 



We must not forget, in fact, that in ontogenesis the mesodermal 

 masses change place thus in toto, without maintaining any relation 

 with their place of origin, and that even invaginations, infoldings, 

 detach themselves and become free in order to go and plant them- 

 selves elsewhere (imaginal disks of insects, enteroccelic vesicles 

 of the echinoderms, urinary bladder of the vertebrates which be- 

 comes detached from the intestine and opens with the ureters, etc.) 



Thus, the criterion of homology which, in order to be objective, 

 ought to remain wholly phylogenetic, is almost never so, and it 

 borrows, according to the case, from ontogenesis and from anatomy. 

 And according to the case, sometimes we accept the conclusions to 

 which we are led, although they may seem to clash with the most 

 reasonable comparisons, sometimes we reject them for the same 

 reason. 



It must be remarked, in fact, that when the criterion ceases to be 

 directly or indirectly phylogenetic, it ceases to be concrete. That 

 kind of material continuity between the organs of the ancestors and 

 similar organs of the descendants which exists in phylogenetic 

 homologies disappears in anatomical homologies. 



