22 THE COMMON CROW. 



young, and at all times it swallows immense quantities of gravel or 

 sand to aid in grinding up hard food in its gizzard-like stomach. Ordi- 

 nary fowls which utilize hard substances in this way retain particles of 

 sand or gravel until they are worn to powder, or at least have lost all 

 sharp edges and corners, when they pass off through the intestines, 

 fresh supplies being taken from time to time. Crows, on the contrary, 

 eat large quantities of sand and gravel whenever their food demands 

 it, and so soon as the trituration of this food is complete the sand and 

 pebbles are expelled by the mouth, together with any indigestible sub- 

 stances of which the bird wishes to rid itself. All birds of prey are 

 known to eject similar masses of undigested material, such as bones, 

 hair, teeth, scales, and bits of shell, while some other birds, as kingfish- 

 ers and flycatchers, have similar habits. These birds, however, swal- 

 low no sand or other hard matter to assist in grinding the food, and 

 hence the ejecta, commonly called l pellets' or l castings,' consist entirely 

 of the less digestible parts of the food. 



Crow stomachs that do not contain some sand or gravel are rare at 

 any time of year, and during the colder months when hard food enters 

 most largely into the diet, the quantity of grinding material used every 

 day is enormous. 



On visiting a Crow roost for the first time one can not fail to be 

 impressed by the great quantity of refuse food and sand scattered 

 about. At the large roost in Arlington National Cemetery, Ya., just 

 across the Potomac from Washington, the deposit which covered the 

 ground late in the winter of 1886-87 appeared at first glance to consist 

 entirely of four substances: (1) Hulls of Indian corn in greatest abund- 

 ance; (2) seeds of various kinds, mainly poisonous and harmless species 

 of Rhus, with a considerable number of seeds of flowering dogwood, 

 sour gum, and a few other species; (3) sand and gravel in large 

 amounts; (4) the chalky excrement wmich matted together and more or 

 less obscured the other components of the deposit. After heavy rains 

 or melting snow the soluble excrement very largely washes away, the 

 sand settles down to the lowest layer, and the surface appears to con- 

 sist entirely of seeds, hulls of Indian corn, and some of the coarser 

 gravel, particularly bits of coke, charcoal, brick, and shell. Of course 

 other materials enter into the deposit, but those mentioned are the 

 most conspicuous. At this roost I was seldom able to find any entire 

 1 pellets' or 'castings' — doubtless because the Crows roosted well up 

 in the tops of young and thickly branching trees, so that the pellets 

 were broken to pieces by the fall or by striking on twigs and branches 

 in the course of their descent. 



The first perfect specimens of Crow 'pellets' seen were sent us by 

 Mr. Otto Widmann, who collected them from the shores and ice of 

 Hospital Island, in the Mississippi Eiver, near St. Louis, Mo. Subse- 

 quently many such pellets were obtained from caged Crows kept dur- 

 ing an entire winter ; and still later other pellets were sent us from 



