NORTH AMERICAN PORIFER^. PART IT. 505 



First, that the development of the genus Spongia resembles that of Chalina, which has 

 been studied by the author, and that of the Halichondrida as expounded by Barrois, in all 

 its principal characteristics, except the entire absence in the early stages of any signs of a 

 skeleton, either in the form of threads or spicules. 



Second, that, as far as known, the development is very similar to that of the x\plysine 

 form, Verongia rosea, as studied by Barrois ; but the basal collar and area, which play so 

 prominent a part in the development of the genus Chalina' and the siliceous sponges, are 

 also well developed in these larvte, whereas they are less noticeable in Verongia. 



Third, that the Sponginaj are by these characteristics of the larvfe distinctly separated 

 from both the Aplysinte and Chalinina?, and therefore rightfully occupy the intermediate 

 place to which the structure of the full grown skeleton has hitherto caused them to be 

 assigned by authors. 



Fourth, that the sub-orders of the Keratosa, as far as known, are more easily distin- 

 guished in their early stages, in their larval than in their full grown forms, and that the 

 assumptions drawn from the adult skeletons that the larger groups, such as the Aplysinre, 

 Sponginte, Chalininse, etc., run together through the structure of their adult skeletons, may 

 perhaps be founded upon errors of observation, which will be corrected by more prolonged 

 and thorough researches upon the differences in the mode of development of the similar 

 structure exhibited by the fibres in these different groups. 



This last position is not only supported by Barrois' observations on Verongia, showing 

 that this genus of the Aplysinae has a larva with the basal collar and central area much flat- 

 tened ; by my own observations on the Spongia, showing that the larvfB have a prominent 

 basal collar and area, but no skeleton ; and by my own unpublished camera drawings of the 

 larvtB of two species of Chalina exhibiting the basal collar, area, and the skeletal spicules 

 in the interior of the endoderm, with their bases showing below^ as bright points in the 

 basal area ; but also by the observations of Barrois and Carter, upon the development of 

 Halisarca, which evidently does not belong to the Keratosa, since it is very distinct indeed 

 from all the jjreceding. 



In one alcoholic specimen of Spongia agaricina, sub-species corlosia, the egg at an early 

 stage was found to consist, as in all other sponges, of a large cell, with nucleus and nucle- 

 olus. Each egg was surrounded by a clear transparent membrane, and nearly filled the 

 middle of the ovarian capsule thus formed. As it increases in size the yolk becomes opaque 

 and granular, oljscuring the nucleus and nucleolus. Unfortunately the state of the spec- 

 imens did not permit the mode of segmentation to be seen. Later stages of larval growth 

 resembled those of Spongia graminea. 



There was also only one alcoholic specimen of >S/;o?«^ia graminea in the Palmer collection 

 containing larval forms. The earliest stage observed was considerably advanced. The 

 process of segmentation was fully completed, and the division of the body of the larva into 

 two regions had taken place. The interior appeared to be, as in the siliceous sponges at this 

 period, entirely filled by minute granules, and the external membrane also, as in the siliceous 

 sponges, to have become converted into long, cylindrical cells, but whether they were flagel- 



^ In this connection, I shall at present confine mvself to time past in studying the sponses of the New Engh^n(l coast, 

 mere statements, simply adding that I have ample evidence and hope to give the residts of my observations on Ch.alina 

 of the fact as stated, and have also been engaged for some and Isodictya in the ne.xt part of this Memoir. 



MEMOIRS BOST. SOC. NAT. HIST. VOL. II. 127 



