422 
BULLETIN of the bureau of fisheries. 
into two parts. It seems better to regard this structure as a partition which divides 
the blood space of the filament into two blood channels because it is known to serve 
this purpose in Area (i). In Pecten, where the circulation of the blood has been 
changed from the original type, it no longer serves this function but remains as a 
functionless membrane. 
The outer surface of each filament is ciliated and three cells on each side have a 
tuft of long cilia. (Fig. 5). The latter point outward and are usually regarded as 
having a straining function, preventing food particles from entering the interlamellar 
cavity with the water currents. If the usual interpretation be correct they have nothing 
to do with the production of water currents. 
The larger blood channels of the gill (fig. 3, d) show a structure similar to that of 
the filaments. There is a one-layered epithelium on the outside which is ciliated for the 
most part and contains numerous goblet cells (probably mucous secreting cells). Lining 
the epithelium inside the vessel there is more or less of lacunar tissue which has retained 
its primitive character. It contains scattered nuclei and its lacunar 
nature is easily made out. I find no evidence of an endothelium, 
which has been described by Bonnet (1) and Menegaux (9). The 
vessels frequently contain bundles of muscle fibers and nerves which 
run from the attached border to the free border of the gills. Their 
distribution has not been studied, but figure 3 shows their position. 
COURSE OF THE CIRCULATION IN THE GILLS. 
Before giving the course of the circulation in the gills it will be 
necessary to describe certain vessels which carry the blood to and 
from these organs, and since the circulation is the same for the two 
sides it will be necessary to describe it in but one. A very large 
vessel, which we will call X, arising from the kidney, passes pos- 
teriorly along the line of junction of the two gills as far as their 
extremities. (Fig. 20, X, pi. l). It gives off numerous branches, 
to the right and left, which pass through the interlamellar septa to the upper borders 
of the reflexed lamellae. These lateral vessels open into one which runs along the 
upper broder of the lamella. The latter which will be called vessel Y in further 
descriptions, is a distributing vessel made necessary, it would seem, because the main 
vessel from the kidney takes its origin near the middle of the gills, and hence no 
blood could pass from it to their anterior portions except through some such arrange- 
ment. Every other one of the larger vessels of the reflexed lamella (fig. 4, v), which 
lie at the reentrant angles of the folds, connects directly with vessel Y. Every alternate 
one (fig. 4, v') ends blindly at the upper border of the lamella. This was proved not 
only by the study of sections but by injections as well. In the same way one-half of the 
larger vessels of the direct lamella end blindly while the other half connect with an 
efferent vessel which runs along its upper border carrying the blood back to the heart 
Fig. 5. — Transverse sec- 
tion of a filament highly 
magnified, c. Cuticle 
lining the blood space. 
