WILD FRUIT. 63 



illustrate this dispersion. On November 17, 1899, a large flock on the 

 wing was noticed in the distance, at a point opposite Fort Washing- 

 ton, several miles above Marshall Hall. They came on down the river 

 in a line that at times stretched almost from one bank to the other. 

 When they neared Marshall Hall the}^ circled several times and finally 

 alighted on the shore of the Bryan farm, at the mouth of Persimmon 

 Branch. The flock numbered at least a thousand, and hoarse caws 

 and croaks gave evidence that it was made up to some extent of fish 

 crows. After the birds had remained on the shore fifteen minutes 

 they were put to flight b}^ a farmer's bo}', and flew on down the river, 

 lessening to specks, and finally disappearing on the horizon. Going 

 to the place where they had alighted, I found the sandy beach cut up 

 for more than a hundred yards with their tracks. Many led out to 

 the water, and* floating black feathers here and there showed where 

 baths had been taken. The most interesting trace of their sojourn, 

 however, was several hundred pellets of fruit material, which they 

 had ejected through their mouths and dropped on the 

 ground. These pellets (fig. 22) were about an inch in 

 length and half an inch in diameter. They were of 

 a deep purplish color, due to the fruit of woodbine, 

 wild grape, and pokeberr}^, of which they were 

 mainly composed. In 50 pellets collected there were 

 only 11 seeds of other plants — namely, holly, bitter- 

 sweet, and poison ivy. Pokeberry seeds were by far 

 the most numerous. Mr. A. J. Pieters, of the 

 Botanical Division of the Department of Agriculture, 

 germinated some of them, thus demonstrating the fig. 22 —Peiiet ejected 

 fact that they were distributed uninjured. by crow. 



Examination of the pellets showed the interesting fact that they 

 were made up not only of seeds (fig. 23) and skin, but largeh^ of fruit 

 pulp in an undigested state; indeed, many pellets appeared to be com- 

 pacted masses of mashed or squeezed berries. It seems strange that 

 the birds should have rid themselves of a substance that still contained 

 a good deal of nutriment. 



Little is known of the distribution of fruit seeds by crows during 

 migration, but it is certain that they do this work effectively while 

 thev fly to and from the roosts where they congregate in winter, for 

 their feeding grounds often cover an area stretching out on all sides 

 from the roost for 50 miles or more. It appears highly probable that 

 the crows which are found in winter at Marshall Hall roost at Wood- 

 ridge, D. C, some 15 miles distant. There, in the midst of several 

 acres of woodland, a crow dormitory is established, in which prob- 

 ably 100,000 crows sleep every winter night. It was visited in Febru- 

 ary, 1901, and the ground was found to be strewn with disgorged 



