NATURE AND AFFINITIES OF THE SPONGES. 179 



in the same specific type, would there seem to be much prospect of arrival 

 at a more definite result. 



The Metazoic nature of the sponges, in deference to the authoritative 

 dictum of Professor Haeckel, being accepted d priori as an article of creed, 

 it has been rendered necessary to indicate, in one and all of those diverse 

 so-called ciliated sponge-embryos, the existence of the two primary and 

 absolutely essential constituents of the Metazoic embryo, the ectoderm 

 and endoderm, as produced by the segmentation and subsequent meta- 

 morphosis of a primitive unicellular impregnated ovum. Deferrino- for a 

 while the consideration of the presumed identity of this earliest or initial 

 phase, it may be first observed that the structural type, out of the three 

 respective series just enumerated, which has been accepted as conforming 

 itself most conveniently to the Metazoic formula, is exhibited by that one 

 in which the apical pole or segment of the reproductive body is composed 

 of more minute columnar flagellate cells, and the opposite one of larger 

 but simply subspheroidal elements. Here, as typically represented in the 

 calcareous sponge, Grantia compressa (PL IX. Fig. 27), there certainly, at 

 first sight, appears to be a remarkable structural correspondence with the 

 segmented holoblastic ovum of the Mammalia, Amphibia, and various fishes, 

 \i\c\w6\\\g Amphioxus, and numerous higher Invertebrata in which one-half 

 of the primitive ovum, dividing more rapidly and abundantly, becomes con- 

 verted into numerous minute columnar blastomeres, and the opposite half, 

 dividing more slowly and less extensively, into fewer larger and sub- 

 spheroidal blastomeres. Out of these two elemental series, distinguished 

 respectively as the epiblast and hypoblast, the future ectoderm and endoderm 

 are subsequently developed, the former from the minute columnar blas- 

 tomeres or epiblast, and the latter from the larger blastomeres or hypoblast. 

 The identity of the segmentation process in the Metazoic embryo and in 

 the so-called sponge-larva being so far regarded as complete, the apparent 

 corresponding factors in either case have also been accepted as homologous 

 ectodermic and endodermic elements. Supposing, for the time, that these 

 two structural elements could be consistently correlated, what should be 

 the next step } 



In the Metazoic embryo it invariably happens that either by the 

 invagination or falling inwards, as in AmpJiioxus, upon the primitive 

 central segmentation cavity or archenteron of the hypoblast or endodermic 

 element, or by the encroachment upon or growing over the latter as in 

 the Amphibia, of the epiblast or ectodermal element, it comes to pass that 

 the endoderm is enclosed within the ectoderm, and a bilaminate structure 

 is produced roughly resembling the double-walled sac-like body or so-called 

 " gastrula " of Professor Haeckel. The outer lamina or wall of this sac- 

 like body is now the ectoderm, the inner one, closely applied to it, the 

 endoderm. The central cavity most usually enclosed within these layers 

 represents the primitive alimentary tract or archenteron, and the aperture 

 placing the latter in communication with the outer world the primitive anal 



N 2 



