FOOD AND FEEDING IN FRESH-WATER MUSSELS. 
467 
material that falls directly upon the inner. These observations confirm those of 
Allen and Kellogg made upon adult mussels. 
3. At the palps the material passes between their apposed faces. Some of it 
is carried down off the palps by downward-beating cilia in the bottoms of the vertical 
grooves. The cilia in these grooves beat downward, as stated by Wallengren (1905) , 
and not upward, as thought by Siebert (1913). Some of the material is carried 
forward to the mouth by the forward-beating cilia on the crests of the ridges between 
the grooves. These observations also agree with those of Allen and Kellogg. 
4. A longitudinal groove was found on the apposing face of each palp immedi- 
ately ventral to the line of union of the two palps. No previous mention of these 
grooves has been found in the literature. Some material is carried forward to the 
mouth by cilia in this groove. This groove functions especially when the mussel is 
feeding in heavy suspensions. 
5. A vertical groove along the posterior side of each ridge, near its base, is 
described, and another was found along its anterior side near the top. The former 
is figured by Allen but not discussed; the latter is not mentioned in the literature 
reviewed by the authors. The theory is here advanced that these grooves mark the 
boundaries between the forward-beating cilia on the ridge and the downward-beating 
cilia in the groove. 
6. There is never any reversal in the direction of the beating of the cilia. The 
palps at times curl outward at the posterior ends, thus making their inner sides 
convex and longer, widening the grooves. More material would therefore fall into 
them. 
7. There is no selection whatever of the kind of material passing down off the 
palps and that going forward to the mouth. Carmine grains could be seen going 
both routes at the same time. The mechanism of the palps operates quantitatively 
'only, reducing the amount of material in such a way that it can be handled at the 
mouth. These observations confirm those of Kellogg, but are opposed to those 
of Allen. 3 
8. Mussels can and do feed when the water is so heavily loaded with material 
that the animal is invisible in the suspension. This feeding is accomplished by the 
action of the longitudinal ciliated grooves near the dorsal side of the inner faces of 
the palps. Food may pass along this groove while all the lower part of the palp is 
engaged in removing the heavy accumulation below. This observation disagrees 
with that of Kellogg, but confirms those of Grave and Nelson. The ingestion is not 
accomplished, however, at least in the fresh-water mussel, by the selection of food 
material and the rejection of the silt, fftc., but by taking a limited amount of all the 
3 Since this paper went to press there has been received by the authors a preliminary paper by Thurlow C. Nelson entitled 
“The mechanism of feeding in the oyster” (Proceedings, Society for Experimental Biology and Medicine, Vol. XXI, 1923, pp. 
166-168). Nelson studied ingestion in oyster “spat” by direct observation through the transparent shell. He found 
that the larger particles of material caused more mucus to be secreted than did the smaller, thus increasing the mass. The 
larger masses were not admitted between the palps, being too large to enter. He also found that the cilia in the more anterior 
grooves of the palps beat upward and carry small particles up, and that these particles reach the mouth. The present au- 
thors found in the fresh-water mussel, as stated in the body of this paper, that the cilia in these grooves beat downward in all 
cases. Nelson’s general conclusion is that there is a “mechanical sorting of the material ingested,” apparently on the basis of 
size, judging from his paper. This agrees with the findings of the present authors in the mussel. Nelson, however, considers 
that the inorganic material, the nonfood, is mostly larger than the food, and the separation on the basis of size results in a 
separation at the same time of the two classes of material. As will be seen from the present paper, these conclusions differ 
from those of the authors. 
