ENTEROPNEUSTA FEOM THE ABROLHOS ISLANDS. 91 



Further variations are probably to be found, but I am unwilling to 

 sacrifice the few remaining specimens for the purpose of following this out. 

 It appears to be of little systematic importance. 



Punnett also finds variations in connection with the proboscis pores in the 

 varieties of Ptyclwdera flava, captured at the Maldive and Laccadive Islands. 

 There may be two pores^ both in counnunication by canals with dorsal 

 coelomic pouches ; there may be only one pore in open communication with 

 the coelom, although the two pores are present ; there may be only one 

 proboscis pore. The side of the animal — right or left, — marked by the 

 different conditions, varies just as much. Evidently every possible condition 

 may occur. 



Stomoclwrd. 



In Pt. flava, according to Willey, the Pericardium, Glomerulus, and 

 Stomochord are exactly co-extensive. Unfortunately, no longitudinal sections 

 are figured showing this. In Pt. pelsarti the pericardium and stomochord 

 are co-extensive, but the glomerulus extends over both, Ibrming a kind of cap 

 to the " central complex^-' of the proboscis (PI. 10. fig. 2,gloin.). The stomo- 

 chord (fig. 2, St.) may be divided into three regions : — (a) anterior inter- 

 glomerular region, (b) middle or csecal region, (o) the posterior nuchal region. 



The anterior portion of the structure (see fig. 2) is without any lumen 

 whatever. The latter becomes evident a little distance anterior to the 

 ventral septum of the proboscis, but there are isolated traces of it in front of 

 this. The stomochord is almost circular in section at its distal extremity. 

 More posteriorly it becomes elliptical in section, with the long axis directed 

 dorso-ventrally (fig. 4, St.). This leads next to the csecal region, where the 

 stomochord becomes wider transversely than dorso-ventrally (PL 10. fig. 6, 

 St.). Two very well-developed lateral pouches are given off at this point. 



In sections,. the cejitral lumen of the cord may be seen quite close to the 

 dorsal wall of the stomochord, whilst laterally two very well-marked lateral 

 diverticula are evident (fig. 6, ^S^.p.). It cannot be said that any ventral 

 " Blindsack " of Spengel is present at all in this form. Two well-marked 

 lateral projections of the cord are quite as distinct as the lateral pouches in 

 Glossohalanus ruJicoUis, S]?engelia, and Balanoglossus australiensis. No 

 ventral blind sac is to be seen. The lumen of the stomochord extends 

 distally for some distance after the lateral pouches are given off. Posteriorly 

 to the ventral diverticula, the stomochord loses its chorda-like character. 

 The lumen becomes larger and larger (PI. 11. fig. 8, St.c). In transverse 

 section the structure is almost quadrangular in shape, and numerous gland- 

 cells appear on all sides. Finally, as the opening of the stomochord into the 

 throat-cavity is approached, the cord becomes very much compressed dorso- 

 ventrally and drawn out laterally, until it becomes almost as wide as it is in 

 the region of the lateral diverticula. The cells of the ventral wall become 



