462 
BULLETIN OP THE BUREAU OF FISHERIES. 
stomach with food. Probably much of the stuff which Evermann and Clark call 
'mud’ is organic. The fact that neither they nor other writers list sand [italics are 
Allen’s] in the stomach contents is further evidence of a selection of food material, 
and that river species are not an exception.” In view of this statement it seems 
important to note that sand was found in the alimentary tract of every mussel 
examined but one, whether they were taken from a lake or river. The mussels 
from the Little Sioux River, which is little more than a muddy creek, contained an 
especially large quantity of sand and mud. Allen further states (1921, p. 227) that 
powdered carborundum, starch, and carmine were never ingested by the mussels 
in his experiments, and he concludes that mussels exercise a quite rigid selection of 
the material ingested. Kellogg, as previously quoted, says “ there is no selection or 
separation of food organisms from other water-borne particles.” In view of this 
diversity of views relative to the matter, further experiments were performed. 
FEEDING EXPERIMENTS. 
INORGANIC MATERIAL. 
On page 454, under the observations on ingestion, an experiment is described in 
which a small juvenile mussel ingested sand grains in such quantities that the feces 
were more than half composed of them. 
In addition to the use of sand, experiments were made in feeding borax carmine. 
It had been found in the observations on ingestion that the mussels would take it 
to some extent. The first lot of mussels used consisted of specimens ranging in 
length from 12 to 25 mm. taken from the Mississippi River. A suspension of 
borax carmine in tap water was made and the different mussels kept in it for 
varying periods and the contents of the alimentary tract then examined microscopi- 
cally. While some of the borax carmine actually dissolved in the water, a good deal 
did not, and in the examinations search was made in the alimentary canal for dis- 
crete undissolved particles. It is to these that reference is made in the records 
below. 
Lampsilis gracilis — 1J to 2 hours. Intestine filled whole length with carmine 
as far as the anus. 
Obovaria ellipsis — 4 specimens; 1J to 4| hours. No carmine whatever in the 
alimentary tract. 
Lampsilis gracilis — -2 hours. Carmine abundant. 
Lampsilis Ixvissima — 2 specimens; li and 3 hours. Carmine present. 
Lampsilis fallaciosa — 3 hours. Some carmine. 
Obovaria ellipsis — 23 hours. Some carmine in intestine. 
Quadrula pustulata — 20 hours. Much carmine whole length of the intestine. 
Anodonta corpulenta — 24 hours. Much carmine whole length of the intestine. 
Lampsilis gracilis — 26 hours. Much carmine the whole length of the intestine. 
Examined the red material in the posterior part of the rectum and some few particles 
of carmine were found, which were as large as the small diatoms found intermingled. 
Most of the carmine, however, consisted of quite fine irregular particles. 
Quadrula pustulosa — 23 hours. Considerable carmine in the intestine; the 
style was pink. 
