The Development of Phascolosoma. 145 



hour, as ingrowths of ectoderm on each side of the ventral nerve 

 cord, in about the middle of the body. This pair of solid ingrowths 

 is covered with a layer of mesoderm, which is a part of the coelomic 

 epithelium; a cavity appears in the middle of the ectodermal 

 rudiment, and the nephrostome and nephridiopore break through. 



The prominent prostomium, flattened on its ciliated, ventral, 

 side, at the end of the first week becomes extended into two lateral 

 flat lobes, upon which in all probability the tentacles are later to 

 be developed. 



The yolk granules, which float back and forth in the coelomic 

 fluid during the constant introversion and eversion of the anterior 

 part of the body, are absorbed during the first two weeks, and the 

 larva becomes transparent. The alimentary tube in the young larva 

 is coiled in characteristic fashion, and various points of difterentiation 

 appear in it. Both pharynx and rectum are ciliated. 



Yellow^ excretory cells project from the wall of the body into 

 the coelom in Ph. gouldii, and yellow granules were observed in the 

 nephridia of Fh. vulgare. 



The characteristic band of hooks, encircling the anterior end 

 of the body of Ph. vulgare, appear at the age of six weeks. Although 

 the adult of Ph. gouldii is bookless, young individuals, 3—6 cm in 

 length, are provided with a broad circlet of hooks like those of the 

 larvae of Ph. vulgare, but none could be found upon slightly older 

 specimens. 



The larvae show no evidences of metamerism. The paired, awl- 

 shaped bristles, which Selekka found in the larva of Ph. elongatum (?) 

 at Villefranche, do not occur in Ph. elongatum at Koscoff. 



Comparisons with Sip un cuius show that the serosa of 

 that form is a highly modified prototroch, that the apical plate with 

 its rosette, dorsal cord of ectoderm in the interruption of the proto- 

 troch, and the somatic plate, correspond part by part in the two 

 allied genera. Both Ph. vulgare and >S'. nudus have a similar postoral 

 circlet of cilia in essentially the same position. 



The striking differences in development between the two 

 genera are due to the sinking beneath the yolk membrane in Sipun- 

 ctdus of the edges of the apical plate, the dorsal cord, and the 

 somatic plate, and the spreading backward and forward, over the 

 sunken areas, of the edges of the flattened prototroch cells. The 

 remnants of the prototroch in Sipunculus are shed with the yolk 

 membrane, not passed into the coelom as in Phascolosoma. 



Zool. Jahrb. XXIIl. Abt. f. Anat. 10 



