4 BULLETIN 187, U. S. DEPARTMENT OF AGRICULTURE. 



In which case the report, in addition to the size-and exact boundaries of the wooded 

 tract, should state the principal kinds of trees and whether there is much or little 

 undergrowth. 



Still a third census desired is that of some definite area — as 40 acres, for instance — 

 forming part of a much larger tract of timber, either deciduous or evergreen. While 

 the number of birds on such an area would be far less than on an equal area of mixed 

 farm land, their correct enumeration will require considerably more care and time. 



The above are three kinds of bird censuses considered desirable, and it is hoped that 

 you will volunteer to take one or more of them this season. In this connection we 

 shall be very glad to have a statement from you concerning any changes you may 

 have noted in the bird life of your locality, especially if your observations extend 

 over a considerable number of years. 



Should you desire further information in regard to the matter, we shall be glad to 

 furnish it at any time. 



Yours, very truly, H. W. Henshaw, 



Chief Biological Survey. 



Since no funds were available for this work, the above circular was 

 sent to some 250 of the voluntary migration observers of the bureau, 

 and a call was issued through the general press for volunteer census 

 takers. There was a generous response and nearly 200 reports re- 

 sulted. Some of the censuses were not properly taken owing to a 

 misunderstanding of instructions and a few were incomplete from 

 lack of acquaintance with some of the less common birds. However, 

 there remains a sufficient number of satisfactory enumerations to 

 serve as a basis for a few interesting deductions and to indicate the 

 points which need to be more fully covered in future work. 



RESULTS IN THE NORTHEASTERN STATES. 



The larger number of censuses in 1914 were taken in the north- 

 eastern quarter of the United States, as shown by the map (fig. 1). 

 Although 44 States are represented, less than a dozen reports came 

 from the South Atlantic and Gulf States, and the number is no larger 

 from the States west of the Rocky Mountains; while the reports from 

 the Plains States — North Dakota to Oklahoma — are too few to be 

 used as representing average conditions in this region. These three 

 divisions, the South, the West, and the Plains, are so diversified in 

 climatic and agricultural conditions and vary so widely in their bird 

 life that many more bird censuses must be available before generali- 

 zations can be made for the whole country and reliable conclusions 

 drawn. 



The present report of progress concerns itself, therefore, with the 

 census of birds on farms in northeastern United States, by which is 

 meant the part north of North Carolina and east of Kansas. As the 

 principal object of the proposed bird census was to ascertain the kind 

 and number of birds on the farms of the United States, and as several 

 of the most complete censuses were made on land which was not being 



