418 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
section, which passes through the supporting membrane, shows these fibers to be con- 
tinuous with its musculature. I believe them to be muscles and the means employed 
to bring about the numerous contractions described above as taking place in the living 
gill. They do not penetrate the thickened ventral portion of the gill plates, nor do 
they seem to be present in their extreme dorsal portions ; their insertion seems to be 
upon one wall of the plate and at all points throughout their length. 
THE GILLS OE SOLENOMYA. 
The external appearances of this gill have already been mentioned. The morpho- 
logically ventral surfaces (here lateral in position) are slightly enlarged, but not 
nearly so much so as in Yoldia. These specialized edges extend from the median ven- 
tral groove of the gill to the pointed outer ends of the plates. 
The ventral edge of the plate is composed of entirely similar cells throughout, rest- 
ing upon the usual chitinous layer (Fig. 77, ve). In the preparation from which the 
section represented was made, these cells had shrunk somewhat and separated from 
one another so as plainly to show their structure. They are columnar, have a finely 
granular protoplasm, the outer edge of which is deeply stained, and each cell bears 
several cilia. The cilia of all the cells are of the same length, excepting those on the 
extreme ventral border, where they are slightly shorter. At the lateral edges of the 
ventral border — never in the midline of the ventral surface — open certain gland cells 
(gl c), which are constantly and generally easily seen in a corresponding position 
throughout the gills of lamellibranclis. 
Pelseneer (No. 17) figures two cells in the section of each plate (in the region just 
below the gland cells in Fig. 77), which are larger than the others and bear much 
louger cilia. These he probably means to compare with the u latero-froutal’’ cells of 
Peck, not so widely found among lamellibranclis, I believe, as seems to be so gener- 
ally supposed. I am confident that these large cells do not exist in Solenomya velum. 
Pelseneer possibly mistook a gland cell in this region for a large nucleus and sup- 
posed it to indicate a large cell. 
It is possible, though I think not probable, that the cilia on the sides of this ven- 
tral edge are used for joining the plates to one another, as described by Pelseneer. If 
such a junction exists it is a very slight one, for the plates, unlike those of Yoldia , are 
very easily separated. No direct interlocking of cilia is anywhere seen in sections, 
as figured for Yoldia in Fig. 82. The ends of the cilia of neighboring plates merely 
touch each other (Fig. 77), and, as in other cases, I believe that they function princi- 
pally in keeping food particles from entering between the plates, thus confining them 
to the ventral surface, where they may be rapidly swept to the median ventral groove 
of the gill and onward toward the mouth. 
The chitin of the gill plate is thickest at some distance from the extreme ventral 
edge ( ch ), and it is here relatively thicker than in Yoldia. It extends, in a long, thin 
layer, entirely to the ventral edge, and also in the opposite direction, toward the cen- 
ter of the plate. This latter extension rapidly thins out and disappears. No very 
marked expansion of the blood space takes place between the chitin plates. 
The epithelium of the walls of the plates is much like that of Yoldia , though the 
cells are distinctly marked off from each other ( w ). They are more elongated and 
their outer ends are rounded. These cells also contain a> fine, equally distributed pig- 
