246 THE STRUCTURE AND AFFINITIES OF HERDMANIA CLAVIFORMIS. 



oesophagus. The epithelium constituting this small fold is hardly half as thick as 

 that of the large folds. Posteriorly it disappears without reaching the end of the 

 stomach, the thinner wall continuing, however, to the end and marking the edge again 

 of the compressed duodenal portion of the intestine. This scheme of folding appears 

 to be pretty constant, — rather more so than in most other species, so far as I have 

 observed, — but considerable deviations from it are not uncommon, particularly in 

 the direction of increase in the number of large folds. 



The lacteal system is well developed. It communicates with the posterior end 

 of the stomach by a single distinct though not large duct. The duct has no ccecum. 

 The tubules ramify over the rectal portion of the intestine from the loop to near the 

 branchial sac. 



From the loop of the intestine to a region somewhat farther forward than the 

 stomach the digestive tract is usually so completely surrounded with yolk-laden 

 mesenchymatous cells, to be described more fully later, as to render difficult the trac- 

 ing of the digestive tube in an entire zooid. 



9. The Heart, Pericardium, and Epicardiac Tubes. — The heart is situated in 

 the postabdomen, behind the ovary and testis, though the latter when ripe may 

 extend as far back as the heart. 



The pericardium is very large (PL XVIII, Figs. 9, 10; PI. XIX, Fig. 11, jri'cr.). 

 It nearly fills the abdominal space in its region, and the heart occupies but a com- 

 paratively small part of its cavity. The heart (cr.) is horseshoe-shaped. It is sit- 

 uated toward the dorsal side of the body, the plane of the loop being at nearly a right 

 angle to the sagittal plane of the body. The cardiac raphe runs along the outer edge 

 of each limb of the loop for most of its length, but toward the base of the loop it shifts 

 gradually to the ventral side. It seems to be entirely closed for part of its course. 

 One limb of the loop, namely, the left, is considerably longer than the other (Figs. 

 10, 11). 



There are two epicardiac tubes extending throughout the length of the animal. These 

 are always wholly separate from each other so far as their cavities are concerned, 

 but a thin laminar bridge connects their w T alls in many places. This bridge is, how- 

 ever, of mesenchymatous origin. The true epithelial walls of the two tubes are wholly 

 distinct (PI. XIX, Fig. 13). This figure shows how the connection is formed. The 

 membranes which fine the body spaces adjacent to the tubes become closely applied 

 to the latter, reaching across from one to the other. These frequently inclose between 

 them mesenchyme cells and thus make the bridge seem in some places to be actually 

 a connection between the epithelial walls. 



