the margin in both large and small filaments. The vessels in 

 the other lamella are similar to those already described, the blood 

 being collected on this side into the vessels of the large filaments 

 (fig. 17, bv'.), and finally poured into the vessel at the bases of 

 the gills, which conducts the blood back to the heart (fig. 

 18, bv.). 



It is quite possible to make out all of the connections 

 described, in sections of the gills but the arrangement of vessels 

 has been further demonstrated by injections of the gills with 

 starch and with gelatine injecting masses. This can readily be 

 accomplished through the supplying and receiving vessels 

 of the gill with a hypodermic syringe. The animals are large 

 and the vessels are fairly distinct so with a little practice it is 

 quite easy to make successful injections. With a starch mass 

 the vessels may readily be injected different colors as the mass 

 is too coarse to pass out into the smaller vessels and complete 

 the circuit. By this method it was easy to determine that the 

 blood passing to the gill all passed along the borders of the 

 inter-lamellar junctions to the free lamella and that all of the 

 blood entering the vessel that carries the blood away from the 

 gill comes from the attached lamella. 



The general relation of the blood spaces of the gills to the rest 

 of the circulatory system will be discussed under the head of the 

 circulatory system. 



The movement of water for respiration and feeding is depend- 

 ent entirely upon the bands of cilia on the sides of the filaments. 

 These, acting like so many small paddles, force the water through 

 the inhalent ostia into the'water tubes, thus driving the water 

 along the water tubes into the space above the gills that corres- 

 ponds to a cloacal chamber, and so out of the shell along the 

 margin posterior to the adductor muscle and dorsal to the gills. 

 The current of water which enters to take the place of that 

 driven out is taken in anywhere along the ventral and anterior 

 borders of the animal. By changing the position of the margins 

 of the mantle so access is given at one place and denied at 

 another, it is possible for the animal to vary the places where 

 water is admitted and ejected. Whether this serves any definite 

 purpose or is more by way of accident caused by performing 

 other functions of the body, is not known, but powdered carmine 



28 



