COMPARATIVE ANATOMY AND MORPHOLOGY 349 



When two organs whose phylogeny is unknown resemble one 

 another in certain anatomical characteristics and differ in certain 

 others, if we affirm their homology, that is to say, if we subordinate 

 their differences to their resemblances, we do so only by comparison 

 with an entirely subjective type to which we refer them. 



In the construction of this subjective type we have no other 

 guides than certain anatomical characteristics, so that our premises 

 and our conclusions are one and the same thing; in other words, 

 we make a vicious circle. 



For example, when we homologize the endostyle of the ascidians 

 with the thyroid gland of Ammocetes and the other vertebrates, 

 we presuppose a common phylogenetic origin for these organs, 

 but since we know nothing whatever of this common ancestor, 

 in reality we conceive of an abstract form in which the endostyle 

 and the thyroid are replaced by one and the same rudiment from 

 which we suppose that these organs are derived. 



This abstract form is nothing else than the archetype of the 

 natural philosophers, more refined, more certified, fashioned more 

 by observation, but not less subjective. 



The fact is more striking in what concerns general homologies. 



No one disputes the homology of the rudimentary mammary 

 gland of man with that of woman. Nevertheless that homology 

 could not in any way be based on phylogenesis. It has been inter- 

 preted as a mark of a primitive hermaphroditism, but that opinion 

 is untenable, since the ancestors of man had ceased to be herma- 

 phrodite millions of generations before they were provided with 

 these glands. Here, then, the homology is founded solely upon the 

 unconscious conception of an archetype common to male and female 

 based upon anatomical resemblances without any possible phylo- 

 genetic significance. 



In the vertebral theory of the cranium we do exactly the reverse. 

 We conclude from anatomical and embryological observations that 

 without doubt no ancestor of the vertebrates had a cranium com- 

 posed of true vertebrae identical with those of the backbone, and 

 we go on from there to declare that no part of the skull, not even 

 the occipital, is homologous to a vertebra. 



If, however, the occipital, although developing by the localized 

 ossification of a cartilaginous cranium continued by the addition 

 of a membrane bone, were found to be identical in form with a 

 vertebra of the backbone, would we deny that homology? We 

 should have the right, but we should fall into the same exaggeration 

 as in the case of Asterias and Asterina. 



There is no use in insisting further upon having the right to con- 

 clude that the criterion of homology is nothing absolute. We 

 announce our pretension of borrowing that criterion from phylogeny; 



