INSECT FOOD. 21 



owners, were visited from time to time. A brief description of the 

 latter, on which most of these subsidiary observations were made, is 

 necessary for a clearer understanding of the results here set forth. It 

 is primarily devoted to truck and fruit, though it produces also wheat, 

 corn, and tobacco. A hedgerow of large cedars cuts it into two parts, 

 each part with its house and barn. The upper section has a swamp 

 fed by a bushy brook and empt3ang into the river, while the lower 

 section is drained b}^ two ditches merging into one at their river out- 

 let. There is also a timbered dell, shallow and swamp^^, which extends 

 from the river back into the cultivated fields, and which harbored a 

 colony of breeding crow blackbirds, more than a dozen catbirds, several 

 woodcock, and at least two pairs of cardinals. Along the Hungerford 

 farm the bluff is seldom half so high as on the Bryan farm, and in 

 many places is entirely wanting. 



II.— INSECT FOOD. 



In studying data derived from the examination of stomachs collected 

 over areas widely diverse in latitude and longitude the investigator 

 seldom knows exactly what kinds of insects were available for selec- 

 tion at the time the food in the stomachs was obtained, how abundant 

 relatively the various species of insects were, and to what extent, if 

 any, they were injuring crops. He is therefore in some danger of 

 misinterpreting results, especially when he attempts to show how the 

 birds' insectivorous habits relate to agriculture in specific cases. He 

 ma}^, for instance, commend birds for having fed on a certain pest, 

 when, as a matter of fact, they had found no other food available, or he 

 may condemn them for not having eaten injurious insects when the 

 district from which they came happened to be free from such plagues. 

 For this reason, therefore, a careful study was made of the relative 

 and absolute abundance of the different kinds of insects on the farm 

 at each visit. It may be mentioned here that in recording observa- 

 tions of this kind the calendar date should be supplemented by the 

 biological date, which shows the advancement of the season and is best 

 determined by the condition of the vegetation; but this rule has not 

 always been followed in the present report. 



CRANE-FLIES. 



The most interesting visits were, naturalh^, those made when insects 

 were most numerous. Crane-flies appeared every year, but during 

 1900 were unusually abundant. The farm was visited on April 22 of 

 that year when the forests were bare and the fields brown. Peach, 

 plum, and pear were in bloom, but the apple was not yet out. Crane- 

 flies were seen everywhere, but were thickest in the grass land of lot 

 1, where they fairly swarmed on the ground and flew into one's eyes, 



