with accuracy, though there will always be a few birds that fail to 

 sing at the expected time. 



A bird ceDsus should not be made until after the migration is 

 over and the birds are settled on their nesting grounds. Occasionally 

 belated migrants linger into the breeding season, but thorough famili- 

 arity with the local avifauna should make the elimination of these a 

 simple matter. If, however, some unusual species found is near the 

 edge of its known breeding range, its presence in the nesting season 

 should be carefully investigated and the nest located if possible, to 

 ascertain whether it is breeding there. Reports have sometimes been 

 found useless because the counts were made so early in the season 

 and certain species were so abundant that it seemed very probable 

 that some of the individuals listed were not yet nesting but were 

 migrants lingering in the southern part of their breeding ranges. 

 Birds seldom nest in large numbers on the edge of their breeding 

 range. On the other hand, if the count is made too late, the young 

 of the species that nest early may be on the wing and may have left 

 the area where they were raised. 



A previous familiarity with the area surveyed is a help, and when 

 it is necessary to make the count on land with which the enumerator 

 is not well acquainted, it is a good plan to go over it several times 

 before the actual census is taken, that the work may be done thor- 

 oughly and accurately. 



In selecting an area on which to take a census, it is better not to 

 choose a place where the birds are excessively abundant unless more 

 than one count can be made. Such places are very interesting as 

 showing how closely birds may breed, and data concerning them are 

 useful in learning the total number of birds in the country, but they 

 are not representative. If such a tract is chosen, the surrounding 

 territory over which these birds feed should also be considered, 

 preferably as a separate count. Such areas furnish excellent material 

 lor studies of local distribution and may yield interesting and valu- 

 able returns when an observer lives near enough and has the time 

 necessary to make several counts of the area of concentration and 

 of as much as possible of the surrounding territory. If these several 

 I counts are repeated year, after year, the results will form valuable 

 series of records. 



Some persons seem to have the impression that a bird census is 

 of no value unless it records a large number of birds per acre, and 

 have considered it not worth while to make a count unless such a 

 tract could be found. The actual truth concerning conditions is 

 what is wanted, and knowledge of the distribution of bird life on any 

 land contributes to this end. A count made on land that has few 

 birds is just as important, therefore, as one made in a bird paradise, 

 though probably not so interesting to the enumerator. 



The final report should include an accurate description of the tract 

 ^ surveyed, indicating its nature, whether moist bottomland or dry 

 upland; level, or broken and hilly; the number of acres in woods, 

 orchard, and garden, in lawns about buildings, in each of the farm 

 crops, and in pasture or meadow; and if there is waste land, whether 

 it is marshy, upland, brushy, or the like. The value and usefulness 

 of a report is increased if it contains information regarding the land 

 surrounding the tract surveyed, whether it is wooded, cultivated, 

 waste, or pasture: and if cultivated, whether in grain, hay. or garden 



