J884.] 187 [Hyatt. 



this group, and we can homologize them with stomodeal, procto- 

 deal and possibly other invaginations occurring in the embryos 

 of higher animals which are also traceable to a tubular origin. 

 These homologies show that the cloaca of sponges are true oscula, 

 though they must be estimated as primitive oscula. The stomodea 

 of many higher animals are, however, not primitive oscula. That 

 is to say they do not connect with the primitive archenteron as 

 do the oscula of Porifera, but arise for example in Vertebrata, as 

 secondary invaginations connecting with the mesenteron. 



McAllister in his " Animal Morphology " first correctly desig- 

 nated the sponges as a distinct branch equivalent to Coelenterata 

 un ler the name of Poriferata. This view was supported by the 

 author (Proc. Bost. Soc. Nat. Hist., vol. 19, 1878 ; and in Johnson's 

 Encyclopedia, Appendix, p. 1668) from an independent stand- 

 point, and before having seen McAllister's statement. Polaejeff, 

 in his Calcispongiae of the Challenger, lately advanced a similar, 

 but more reserved opinion, apparently without having read either 

 of these views. 



These observations and conlusions in so far as the sponges are 

 concerned can be summarized as follows : 



The Porifera are peculiar in the plasticity of their forms, pre- 

 senting great variation within the same species, and tendency to 

 coalescence; in the possession of an inclosed skeleton formed by 

 the ectoderm ; in the universal presence and essential necessity of 

 pores, and the peculiarities of the external aspect arising from 

 the presence of these structures, and in the absence of any marks 

 of bilateral distribution of the parts. The interior of primitive 

 forms has a unique character due to the protozoonal forms of cells 

 in the endoderm of the lowest type throughout life, and to the 

 pores perforating the wall ; while the interior of higher forms has 

 a still more marked appearance due to the vast number and irreg- 

 ular radiatory and often branching arrangement of the coelomic 

 sacs or ampullae and their primitive shapes ; the presence of two 

 systems of gastro-vascular tubes ; the epithelium of the archen- 

 teron, and the protozoonal forms of the cells of the ampullae. The 

 habitat is universally sedentary, and complicated colonial associ- 

 ations of spongozoons are unknown. The sieve-like organization 

 and mode of growth of the coelomic sacs, and middle layer, and 

 the comparative uniformity of the habitat have all contributed to 



