542 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
The principal function of this large tube is clearly to discharge 
the secretions from the cephalic glands in which it terminates. 
These secretions evidently pass into the duct by being forced be¬ 
tween the cells which constitute its lining. The secretions from 
those cephalic glands which are situated behind the posterior end 
of the frontal organ must be passed forward through temporary 
ducts formed between the gland cells until the tube of the frontal 
organ is reached. 
Besides the true cephalic glands there occur scattered through the 
tissues of the head numerous other gland cells which are conspicu¬ 
ous because of their granular secretion, which commonly stains very 
differently from that of the cephalic glands. These gland cells lie 
mainly near the surface of the head upon which they discharge 
their granular secretion through the overlying muscular and base¬ 
ment layers. They are evidently merely specialized integumentary 
glands which have sunk beneath the basement membrane. 
Alimentary canal .— The esophagus (pi. 24, fig. 13) leaves the 
rhynchodaeum immediately anterior to the ventral brain commissure 
as a narrow tube which passes beneath the commissure. Directly 
posterior to the brain it merges into a large chamber, the stomach, 
with highly columnar, ciliated epithelium and convoluted walls. 
This conspicuous stomach was thought by Willemoes-Suhm (’74, p. 
410) to be the mouth. Behind the stomach the narrow pylorus 
passes backward directly beneath the proboscis sheath for a consid¬ 
erable distance before opening into the dorsal wall of the intestine. 
From the pyloric opening a broad intestinal caecum with numerous 
lateral diverticula extends forward nearly to the brain. The intes¬ 
tine exhibits conspicuous metameric pouches. The anus opens by a 
small pore on the dorsal side of the posterior extremity of the 
body. 
■Blood vessels .— The blood vascular svstem offers few deviations 
•/ 
from that of most related species, although it has no direct connec¬ 
tion with the nephridial canals such as Dendy (’92) describes for G. 
australiensis. Bohmig (’98, p. 503) describes certain large cells 
found in or beneath the basement layer of the blood vessels in 
Geonemertes chalicophora as acting in the nature of valves to 
regulate the direction of the flow of blood in the vessels. Similar 
cells occur in G. agricola , G. australiensis, and in numerous other 
Hoplonemertea of various genera, although such an interpretation 
