390 Mr. H. J. Carter on the 



the largest have a branched tubuhir form very different from 

 the compressed one of the parent, yet their position in prox- 

 imity to the adult, together with the presence of the club- 

 shaped surface-spicule (PI. XX, fig. 20), at once identifies them 

 as belonging to it. And it seems worth remarking that while 

 the cavity of the parent is almost always occupied by a minute 

 shrimp-like crustacean devouring not only the fat young em- 

 bryos, but the whole substance of the sponge itself (especially 

 in captivity), the young individuals, which at this time have 

 nothing of the kind in them to attract these animals, are left 

 untouched. 



Having then cut off a few of the branches of the P^iVota bear- 

 ing adult specimens oiGrantia compressa, they were placed in a 

 glass vessel in fresh sea-water on the 12th of August, where 

 they remained undisturbed for seven days, when they were 

 taken out, and the residue at the bottom, after the greater part 

 of the water had been carefully drawn off by a siphon, was 

 placed in a flat glass vessel under an inch compound power for 

 examination. In this residuum several specimens of the embryo 

 in its active state were observed (PI. XX. fig. 13), together 

 with others that had become ^^assz've, on their way to become de- 

 veloped into the perfect sponge (figs. 16 & 17), and specimens 

 of the latter also (figs. 18 & 19). These were successively 

 taken out with the pipette, as before mentioned, and trans- 

 ferred to a slide for examination with a much higher compound 

 power, viz. that of 5-inch focus. 



Of the early part of the development of the ovum I need 

 say nothing beyond what has already been mentioned in the 

 " First Period," viz. that it becomes attached to the surface 

 of the excretory canal, and that in the "Second Period" it 

 undergoes segmentation, as shown by Hiickel. We shall 

 therefore go at once to the others, viz. the Third and Fourth 

 Periods. 



Third Period. 



The embryo of Grantia compressa^ commencing almost in 

 a globular form, still remains encapsuled in the parent until 

 fully prepared for an independent existence, when it breaks 

 through its capsule and leaves the parent somewhat elongated 

 (PI. XX. fig. 13). 



It is now cylindrical, a little longer than broad, obtusely 

 conical at one end (fig. 13, a), and roundly truncated at the 

 other (fig. 13, c?). The surface is covered with a layer of mi- 

 nute monociliated cells (fig. 13, h), which cells being much 

 longer than they are broad, and more or less wedge-shaped, 

 form (in juxtaposition) a crust of columnar structure, radiating 



