\'ariations in Bird Migration. 5 



gration reports. The Biological Survey during the past 

 twenty-five years has received reports from more than two 

 thousand different persons and it must needs be that this 

 large number would include observers good, bad and indif- 

 ferent. A large part of the most exacting work of the pres- 

 ent writer for the past ten years has been the examining of 

 these thousands of reports and separating the wheat from the 

 chaff. No better single basis for a judgment has been found 

 than the above three-day variation. It received a striking 

 confirmation in a very extended set of records that were re- 

 ceived from Raleigh, N. C. In some twenty years of records 

 about half the years showed small variations, while the other 

 years gave less regular results even to a fifty per cent in- 

 crease in the amount of variation. Later the observer sent 

 his original notes and diary and they showed conclusively 

 that the variations were inversely according to the amount 

 of time spent in the field. The more days a month and the 

 more hours a day devoted to hunting for new records, the 

 more regular the records obtained. 



Conversely if an observer in a level district sends note-, 

 with wide variations it is certain that some of his records are 

 not representative of the normal movements of the birds au'l 

 extra care must be taken to ascertain which notes should not 

 be used. 



During all these years, as would be expected, there has 

 been some nature-faking in the reports. The percent of 

 spurious records has been very small and they have varied 

 from the crude impossibilities of ignorance to the carefully 

 worked out report of a person well up in birds and bird mi- 

 gration. But they can be detected with ease when the above 

 rule is applied and they are compared with genuine records 

 from neighboring districts. Either the variations will be too 

 small — the dates more regular than the actual movements 

 of the birds, or if they have taken pains to vary the dates 

 they have placed the wider variations on the wrong birds, 

 since species differ widely in their normal variations. It 

 would probably astonish some of these nature-fakers if they 



