1884.] 81 [Hyatt. 



of the ova of Spongia grarainea the area of the blastopore is 

 concave, in a third a deep cut is present, and in still other 

 larvae of the same size and in the same sponge this area is flat. 

 We have also drawings of the ciliated larva of Halichondria 

 Dickiei from Eastport, in which this area has the form of 

 a shallow cup, deeper and without the collar-like projection 

 figured by Keller in his gastrulated larva. The larva figured 

 by Keller in C. fertilis (op. cit., p. 338) as a probable gastrula 

 was somewhat older, and the cup is in the blastoporic area 

 itself after this part and the collar were fully formed in the cinc- 

 toplanula. 



We have considered the question whether these last might not 

 be accounted for by contraction. In Halichondria, however, the 

 cup occupied the whole of the blastoporic area, which had no signs 

 of a collar, and it was also of considerable depth, while in some- 

 what larger specimens from the same sponge the blastoporic area 

 was completely plugged up, and as is usual in Halichondria the 

 bases of spicules were visible. At still later stages in larger larvae 

 we also observed the protrusion which forms the collar. In a larva 

 of Spongia gramineafrom Florida the cup reached inwards nearly 

 half the depth of the body. While Keller's gastrula with its 

 shallow cup might possibly have been due to contraction, those 

 we have described cannot be accounted for in this way. Another 

 fact in favor of this conclusion is the absence of any similar cups, 

 after the projecting endoblastic collar is fully built out, in any of 

 the Chalinulae, or Keratosa or Silicea observed by the author. 

 The collar may vary greatly in form and even protrude like the 

 neck of a vase, as it did in one vase-shaped larva of Spongia 

 graminea, but no gastrulated cavity or cup formed into which 

 the collar or any part of it disappeared. The gastrula evidently 

 occupies a stage between that of the amphiblastula, or the paren- 

 chymula when that is present and the cinctoplanula or girdled 

 planula. After the gastrula stage has been completed and the 

 larva returned to its oval form, or during this process, the 

 collar cells appear on the rim of the blastopore, and begin to form 

 this organ. 



Marshall ( Zeitschr. Wissen. Zool. vol. xxxvn, p. 235) supposes 

 that the collar is produced by the breaking through of internal cells, 



PROCKKDINOS B. R. X. H. VOL. XXIII. 6 JANUARY, 1885. 



