464 
BULLETIN OF THE BUREAU OF FISHERIES. 
made on the process of ingestion in the smaller mussels it would seem that the 
adult mussels in the presence of carmine simply close the valves and cease to siphon 
water through the mantle cavity. Observation of them while in the suspensions 
substantiated this; they remained closed most of the time. The matter of the 
ingestion of such inorganic material as sand and the like will receive further attention 
in connection with the following experiments. 
ORGANIC MATERIAL. 
In an effort to ascertain more precisely what the food of mussels is and whether 
or not they exercise any marked degree of choice between different material or be- 
tween organic and inorganic substances a number of experiments were performed. 
Organic material alone was used in most cases, but in a few instances sand and mud 
were mixed in the cultures. In cases where the bottoms of the cultures were of sand 
or mud, search was made in the alimentary tract for particles of these substances. 
These experiments were conducted at the Iowa Lakeside Laboratory at Lake Oko- 
boji, Iowa, where the tap water consisted of filtered lake water, which contained, of 
course, some few food organisms. Mussels kept in it, however, for a day or two were 
found not to obtain much food, as was shown by an examination of the alimentary 
tract. Some of the mussels used in the experiments were previously kept in this tap 
water until the rectum was observed to be empty, and then utilized in the tests. 
These were small mussels in which examination could be made through the semi- 
transparent valves. The results of the experiments follow. 
A specimen of Anodonta grandis about 25 millimeters long, the rectum nearly 
empty, was kept for seven hours in a culture of Lyngbya, Anabsena, and Micro- 
cystis. Upon examination the stomach was found to contain considerable material, 
mainly Microcystis, a few pieces of the filamentous Lyngbya, and Anabsena being 
included, however. The rectum contained entire and disintegrating Microcystis in 
abundance and a little of the other two forms. The filaments probably could not be 
ingested in most cases because of their size and shape, while the rounded Microcystis 
could enter the mouth. A Sphserium sp., half-grown, in similar culture to that just 
mentioned for 12 hours, gave similar results. 
Anodonta grandis, 25 millimeters long, in tap water over a bottom of sand and 
mud from the Little Sioux River 18 hours, contained in its stomach Euglena, entire; 
many fine sand and dirt particles; a few fairly large sand grains, one 32 micra across, 
and some diatoms; rectum, packed full, many particles of sand and dirt, an entire 
encysted Euglena; a few round algal forms; one or two Scenedesmus. 
Anodonta grandis, 25 millimeters long, rectum empty, in culture of material 
from Spirit Lake, the bottom of which was mud from Little Sioux River (there was a 
suspension of mud in the water for at least an hour after the beginning of the experi- 
ment), after an experiment of six hours contained the following in the rectum: 
Euglenas, one dozen at least living individuals and numerous partly digested frag- 
ments; some encysted Euglenas; Scenedesmus; diatoms, elongate and rounded, some 
entire; a large entire desmid 460 micra long; bits of Microcystis; one or two frag- 
ments of filaments of diatoms; a few small living flagellates, unidentified; debris 
of plants, sand, and particles of earth; a living Vorticella head; a number of other 
