148 Mr. W. S. Kent on the, Embryology of Sponges. 



til at different examples of these bodies, derived even from the 

 same sponge, present an extraordinary latitude of variation. 

 Among the most conspicuous and frequent of these variations, 

 is one which, indeed, in certain sponge-forms occurs almost 

 as frequently as the typical one just alluded to ; it is repre- 

 sented by Plate VI. fig. 15. The deviation in this example,, 

 as will at once be recognized, consists of the distinct character 

 of the component parts of the lower half of the organism, the 

 typical elongate flagellate units which characterize the upper 

 one being here replaced by irregularly spheroidal cells, which 

 are more or less confluent with one another. Grasping at a 

 straw, those committed to the metazoic interpretation of the 

 Porifera have selected this inconstant type for the demonstra- 

 tion of their views respecting the bilaminate or diblastic struc- 

 ture of these bodies. No distinct inner and outer lamina, as 

 first represented, being found to exist, the front flagellate 

 portion is now made to do duty for the exoderm, and the hinder 

 one for the endoderm. The constituent elements of this latter 

 region being found again occasionally retreating into the 

 central cavity of the compound body, this has been accepted 

 as a proof of the invagination of the endoderm and the forma- 

 tion of a primitive " gastrtea." The untenability of this inter- 

 pretation, however, is at once proved by the inconstant occur- 

 rence of this type, while in addition it is easy to show that the 

 basal and larger cellular elements are merely modifications or 

 more advanced stages of growth of the smaller frontal ones. 

 Two figures borrowed from Barrois (Plate VI. figs. 19, 20), re- 

 presenting two separate developmental phases of the ciliated 

 embryo of Halisarca lohularis^ assist in the demonstration of 

 this fact. In the second of these (fig. 20) we find that the 

 cellular units of the lower portion of the body, though abruptly 

 larger than those of the upper one, exhibit the same uni- 

 flagellate character, while in the preceding figure the trans- 

 ition from one to the other is perfectly gradual and uniform. 

 Another figure is given by this authority, derived from the 

 same sponge- type, corresponding with our own fig. 10, but 

 prior to the development of the flagella, and in which the 

 component cells from one end to the other present a precisely 

 similar size and character. Haeckel, again, in his ^ Kalk- 

 schwamme,' Taf. 4. fig. 6, represents the ciliate embryo of 

 Ascetta clatkrus as corresponding entirely with my delineation 

 at fig. 9 of that of Orantia compressa, the whole peripheral 

 surface consisting similarly of minute even-sized cells, exhi- 

 biting in superficial view a tessellated aspect. No distinction 

 whatever is indicated here between the cellular constituents of 

 the anterior and posterior portions of the organism, though at 



