Variations in Bird Migration. 3 



days early to 1.7 days late, with an average variation of 0.7 

 days. That is to say the whole migration of different years 

 may vary as much as two and a half days, between the 

 earliest spring- and the latest, but the probable variation is less 

 than a day. Or to say the -same thing- in another way, the 

 notes of a first-class observer in any one year will give an 

 average that probably will not vary more than two-thirds of 

 a day from the date that would be obtained for that locality 

 as the average of long years of observation. 



The figure 2.5 days for the maximum variation is the sum 

 of two variables — the actual variation of the birds in their 

 times of arrival and the variation in the amount of time and 

 care that the observer can give to the work in the different 

 years. The fact that the sum of these two variables is so 

 small shows both the extreme uniformity of the birds in their 

 migrations and the very careful and accurate work done by 

 the four observers whose notes were selected for the test. 



The above may be summarized by saying that the migration 

 of birds as a whole is remarkably uniform. A season may 

 start out early or late and quite wide variations may occur 

 in the times of arrival of any one species, but in the course 

 of the whole season's migration, these extremes nearly bal- 

 ance and the result is very small net variations. 



The dates of arrival of individual species show considerable 

 variation between the extremes of early and late arrival, but 

 the average variation from the mean date of arrival is sur- 

 prisingly small. To arrive at a mathematical statement of 

 this variation thirty-six species were selected, which are 

 easily noted, and are common enoug'h so' that they would 

 ordinarily be seen as soon as they arrive. The records at 

 Lanesboro, Minn., for the six years 1885-1890 were used and 

 there should therefore have been six times 36 records or 21G 

 in all. Six of these records were lacking, while eight rec- 

 ords were rejected because too early to show the ordinary 

 movements of the species, and ten more because they were 

 so late as to indicate that the species was overlooked on its 

 arrival. Thus 18 records, or 9 per cent, were rejected. 



