3 
Bi £ d Censuses and How to Take Them. 
for several years. It will then be evident that any fluctuations in the 
bird life are not due to artificially changed environment. The re¬ 
sponse of birds to changes in their environment is an interesting 
and important study in itself; for this reason counts on land where 
conditions are changing will also be useful, when accompanied by 
full notes regarding changes in conditions from year to vear. 
KINDS OF CENSUSES DESIRED. 
Several kinds of censuses are needed for purposes of comparison 
and of learning the response of birds to different conditions. Those 
most desired are suggested below, but inability to find a tract that 
falls under one of these heads need not deter any one from taking 
bird censuses. All that can be obtained are desired. The repetition 
of a census on the same tract year after year is as important as the 
selection of a special kind of land for the purpose. 
FARM LAND. 
The farm-land tract selected should represent as nearly as possible 
the average farm conditions for the locality in regard to the propor¬ 
tion of land under cultivation, pasture, waste land, etc., and should 
not have an undue proportion of woods or orchard. The majority 
of the censuses to date have been taken on about 50 acres about the 
farm buildings, including lawns, gardens, and orchard. Censuses 
on those areas should be continued, but it is necessary to have also 
counts of the birds breeding on the wilder parts of the farm, and 
on parts of the farm more completely under cultivation; they are 
especially desired on the same farms where counts have been made 
on the land about the buildings. When two or more counts are 
made, it is preferable that they be on adjacent tracts, for then each 
report adds to the interest - and value of the other, since a larger 
area is thereby represented and at the same time the local distribu¬ 
tion is indicated. 
WOODLAND. 
Counts of the birds breeding in isolated pieces of woods compris¬ 
ing 10 to 20 acres are desired, and also data concerning the birds 
of the continuous forest. The tract selected for the latter—40 acres 
or more—should be part of a large tract of timber. 
In the case of woodland counts the report should include, besides 
a description of the area and its exact boundaries, a list of the prin¬ 
cipal varieties of trees and a statement as to whether there is much 
or little underbrush. 
IRRIGATION PROJECTS. 
In many parts of the West, where irrigation is changing the char¬ 
acter of the country from desert to farm land, the character and 
probably also the amount of the bird life will be greatly changed. 
The only means of learning what these changes will be, and how 
rapidly they will take place, is through series of bird censuses cover¬ 
ing the same tracts of land for several years before the water is 
turned on it, and for years thereafter. It is hoped that such series 
may be reported. 
