OIROULATORY APPARATUS OF THE NEMERTEA. 67 
ing into the sheath was sometimes wide; most frequently, 
however, so closely contracted that only a high power enabled 
me to see any aperture. 
It is not my purpose to give an explanation of these sacs 
or to trace their histological structure, but they are not 
connected with the vascular system. 
The medial blood-vessel in the proboscidian sheath is 
attached with avery small base, so that it is almost totally 
bathed by the fluid of the sheath. But I did not observe that 
its upper wall had a different structure from that of the lower 
one. The epithelium of the sheath covers the vessel. The 
hyaline basal layer of the proboscidian sheath passes into that 
of the vessel, and just beneath the vessel a few longitudinal 
muscular fibres are situated. The median vessel, when beneath 
the proboscidian sheath, and the two lateral vessels, have outside 
the hyaline basal layer some circular muscular fibres. Within 
the hyaline basal layer the common epithelium is again 
present. The nephridial canals exhibit the same structure as 
in all other species. 
18. Drepanophorus serraticollis, Hubr., fig. 65. 
The mouth and the proboscidian sheath have each their own 
aperture. I could see the cephalic loop distinctly above the 
proboscidian sheath. Transverse vessels were present in a 
tailpiece at the tip of the tail, where the three longitudinal 
vessels pass into one another. The transverse vessels cut 
transversely resemble those of Cerebratulus marginatus 
(fig. 65). 
19. Tetrastemma candidum (O. F. M.), Oerst., 
figs. 6, 38. 
I did not observe that the cephalic loop went over the 
proboscidian sheath. The preparation was not sufficiently 
preserved to enable me to see it. I clearly saw the two vessels 
of the loop (fig. 38). In the usual way they protrude between 
the brain and the proboscidian sheath, descend along the sides 
of the sheath, and communicate just below the proboscidian 
