Classification of Sponges. . 409 



The characters upon which I had relied for establishing 

 the Sponges as a distinct series — ramified or irregular shape^ 

 absence of general cavity, great development of mesoderm, 

 ciliated chambers lined icith cJioanocgtes, absence of nemato- 

 cysts and of tentacles — were known already; but having taken 

 care to define strictly the starting-points of the nomenclature 

 that I adopted, and to explain clearly what I meant by the 

 terms series, branch, class, &c.*, they were amply sufficient 

 to establish a due conception of Sponges. The apparently 

 new character invoked by M. Delage would therefore only 

 serve to justify the importance of those upon which I had 

 dwelt, even though it should be found not to be open to 

 criticism and were freed from the^ at the very least, debatable 

 interpretations with which it has been surrounded. 



For anyone who reflects that the organization of animals 

 is dominated by general rules, it is difficult to admit that a 

 zoological group can be opposed to another, and especially to 

 the remaining groups as a whole ; in truth the idea would 

 appear to be but a metaphor to be added to those with which 

 the language of zoology is already obscured, if the author did 

 not take care to tell us that " in the Spongiarim, alone among 

 all creatures, the normal invagination of the layers is reversed, 

 the endoderm passing to the surface to form the epidermis, and 

 the ectoderm sinking into the interior to form the digestive 

 cavities^ Of the causes which could have produced so extra- 

 ordinary a reversal M. Delage tells us nothing, and yet zoology 

 is to-day too rich in materials for it to be still possible to 

 suppose, if we take our stand on the ground of transmutation, 

 that reversals of layers and transformations of organs occur 

 which could not be connected with causes that are at least 

 probable. 



But in order to arrive at the conclusion tliat the Sponges 

 are the reverse of the rest of the animal kingdom, M. Delage 

 has recourse to processes of argument and to generalizations 

 the value of which it is important to determine. 



If we were ignorant, he remarks, of the development of 

 Sponges, and the larvte of these animals were presented to 

 embryogenists with the request that they should name the 

 layers and predict their development, there is not one of them 

 who would not say that the flagellate cells are the ectoderm, 

 that the granular cells are the endoderm, and that the invagi- 

 nation of the latter within the former will ensue. There is 

 therefore no doubt as to the homologization of the larval 

 layers. 



The definition of the homologies is due to Geoffroy Saint- 



* ' Colonies animales,' p. 744, and ' Traits de Zoologie/ p. 403. 



