THE STRUCTURE AND AFFINITIES OF HERDMANIA CLAVIFORMIS. 243 



va. rph.). It is situated between, and ventral to, the rapheal muscle-bands. There 

 is no doubt that this vessel has an endothelium that is almost if not wholly com- 

 plete. 



6. The Neural Gland and its Duct. — These, as seen from the interior of the branchial 

 sac, are shown in Figure 5. The gland is of liberal size, is somewhat pear-shaped, 

 with the broad end directed forward, and is situated on the ventral side of the ganglion 

 and somewhat farther back than this structure. 



The duct is short and trumpet-shaped, with a broad, simple, elliptical mouth trans- 

 versely placed. The mouth is situated slightly to the right of the sagittal plane, and 

 the duct in its course backward makes a sharp turn to the left so as to enter the gland 

 somewhat to the left of the median line. (Frequently the duct is relatively shorter, 

 and the gland is farther back with reference to the ganglion than in the specimen 

 shown in Figure 5.) 



The duct branches sparsely within the gland, but the branches do not penetrate 

 deeply into the glandular substance (PL XIX, Fig. 14, rm. dt.). Its main portion 

 within the gland, which as usual is very slender, continues on beyond the gland into 

 the dorsal raphe, along which it can be traced through the whole length of the branchial 

 sac, even into the anterior portion of the abdomen, where it disappears. Whether or 

 not the "rapheal duct" possesses a continuous lumen throughout its length I have not 

 been able to determine. A very small lumen is undoubtedly present in some places, 

 even far back in its course (PI. XIX, Fig. 16, dt. rph.). In many sections I am, how- 

 ever, not able to detect the opening and am inclined to believe it does not exist every- 

 where. It is always exceedingly small and gives the impression of being in a rudi- 

 mentary condition. The wall of the duct where a lumen is recognizable is composed 

 of a few rather large, nearly globular cells. At some points the duct or strand is 

 enlarged somewhat, and seems even to give off small blunt processes, which project 

 inward (Fig. 16, dt. rph'.). The duct turns rapidly dorsad after leaving the gland, 

 and from here is situated throughout the branchial sac in the dorsal blood-sinus close 

 in contact with its dorsal wall, and between the two rapheal muscle-bands. The 

 relation of this duct to a possible rapheal nerve I shall consider later. 



The histology of the gland itself deserves a few words. I have stated above 

 that the duct branches sparsely within the gland. As this is a matter that may have 

 some bearing on the question of the affinities of the animal, I shall consider it a little 

 more fully. The duct lies as usual on the dorsal surface of the gland; or, to be more 

 exact, the gland proper is a development of the ventral wall of the duct, its lumen 

 being here restricted to the portion belonging to the dorsal, unmodified half of the 

 wall, so that the duct appears in cross-sections of the organ as a small trough lying 



