1927] Pilsbry-Bequaert, The Aquatic Mollusks of the Belgian Congo | 039 
“Actophilus africanus (Gmelin). In two stomachs of the lily-trotter I noted 
small seeds and small bits of stone in both cases, one held in addition a small snail 
with conical shell. This bird may be expected to eat many tiny mollusks, provided 
they live close to the surface of the water.” 
Several of the North American ducks devour considerable quanti- 
ties of shails along the sea-shore as well as in ponds, as may be seen from 
the following figures taken from a recent paper by Mabbott?: 
Chaulelasmus streperus (Linnzeus): but 2.15 per cent of the food is animal and 
three-quarters of this is mollusks, chiefly Neritina virginea. 
Mareca americana (Gmelin): 6.77 per cent animal food, 6.25 per cent being snails. 
Nettion carolinense (Gmelin): 4.57 per cent animal food, 3.59 per cent being 
snails (Physa, Neritina, and Planorbis). 
Querquedula discors (Linneeus): 29.47 per cent animal food, 16.82 per cent mol- 
lusks and also eggs of snails. 
Querquedula cyanoptera (Vieillot): 20.14 per cent animal food, 8.69 per cent 
mollusks. 
Dafila acuta (Linneus): 12.85 per cent animal food, 5.81 per cent mollusks. 
In North America the muskrat [Ondatra zibethica (Linneeus)] 
although normally herbivorous, often subsists, in a great measure, upon 
the flesh of river mussels (Unionidze). According to A. W. Butler,? this 
happens mainly in winter and early spring, when suitable vegetable food 
is scarce. Nevertheless this rodent when abundant in a locality may be- 
come the principal enemy of the mussels. Thus, Headlee believes that in 
Winona Lake, Indiana, Anodonta is confined to the deeper water at the 
edge of sandy and gravelly banks, beyond reach of the rodent.’ F. C. 
Baker also writes that in the Big Vermilion River, Illinois, ‘‘the presence 
of the muskrat is attested by the number of piles of opened mussel shells, 
the animals of which have provided this animal with many a meal. In 
these muskrat piles have been found the shells of many species that are 
rare or difficult to find alive in the streams.”’ * There has been some specu- 
lation as to just how the muskrat contrives to open the living mussel, 
considering the strength of the adductor muscles that lock the two valves. 
A. W. Butler believes that this is done in three ways. In species which 
withdraw the foot very slowly, the rodent may insert its paws or long 
teeth between the valves and tear them asunder.’ In others the teeth 
may cut a sufficient opening to secure the animal by pulling the valves 
apart. For the heavier mussels, however, he says: “The only way in 
see he) RY cae 
iMabbott, D. C. 1920. ‘Food habits of seven species of American shoal-w ater ducks.’ U.S. Dept. 
Bull. 862, Prof. Pa 1-67, 7 Pls. 
mdi iy A. W. un 1885." ' Observations 0 on the muskrat.’ Amer. Naturalist, XIX, pp. 1044-1055. 
*Headlee, T. J. 1906. ‘Ecological notes on the mussels of Winona, Pike, and Central Lakes of 
Kosciusko County, Indiana.’ Biolog. Bull., XI, pp. 305-318, Pl. x11. 
4 Biclosl, Monon. Vil 2 te: 
sw. Don (SBE. Row the muskrat opens the Unio.’ Journ. Trenton Nat. Hist. Soc., No. 1, 
p. 8) claims that he actually saw the animal insert its claws into the shell and then pull the valves apart. 
