CRYPTOCHITON STELLERI MIDD. 221 



these canals and dissections show that at or near the point of 

 union of the pallial and pedal cords a connective tissue and mus- 

 cular septum (Fig. 3) completely separates the nerve sinuses from 

 the head cavity. However a short distance behind this point 

 and throughout the greater part of the succeeding portions of 

 the mantle furrow there are frequent communications between 

 the pallial and nerve sinus so that this latter space and adjacent 

 regions are probably filled in life with comparatively pure blood. 



In Cryptochiton another part of the blood passing out of the 

 head cavity enters the proboscis or snout chiefly by means of two 

 large vessels located on each side of the mid-line about level with 

 the mouth (Fig. 3). Almost immediately these break up into 

 two or three branches that pass outward to the margin of the 

 proboscis where they become continuous with an extensive sys- 

 tem of spaces forming a semicircular vessel (infi) coextensive 

 with the margin of the snout. From this marginal sinus smaller 

 vessels pass inward and furnish fresh blood to the proboscis 

 tissue, while another supply of arterial blood, much smaller in 

 amount, passes into the snout through sinuses surrounding the 

 nerves. As the figures indicate no part of the blood from any 

 of these sources passes into the tissue of the foot, at least not in 

 any appreciable quantity, but is conveyed along the region of the 

 mantle furrow by means of the relatively large pallial sinus leav- 

 ing the proboscis at its outer posterior angle, or it may be passed 

 into the visceral cavity as described presently. 



Middendorff figures (/', Fig. 2, PI. VIII.) a small curved canal 

 which surrounds the snout anteriorly and states (p. 72) that it 

 lies between the external and internal oral sphincter. Owing to 

 the fact that in one specimen he was able to force the injection 

 mass from the branchial artery into this vessel he considered it 

 to be the anterior union of the branchial arteries but as these 

 latter vessels certainly do not extend much farther forward than 

 the most anterior gill and have no direct connection with the pro- 

 boscis, I am of the opinion that the connections as described are 

 incorrect and that the marginal vessel is the same one shown in 



Fig- 3- 



Not only are the margins of the proboscis capable of consider- 

 able expansion owing to the influence of an abundant blood sup- 



