132 PROCEEDINGS OF THE ACADEMY OF [April, 



Up to April 28 (fig. 1) the species had been observed at but three 

 stations, two of these being to the north and northeast of the cit}^ and 

 the other to the southwest. On April 29 (fig. 2) it was present at thir- 

 teen stations, and by Maj' 1 (fig. 3) had been reported by all but three 

 of our observers. 



Mr. Otto Herman's paper in Proc. Fourth Internal. Ornith. Congress, 

 p. 163, was not received until after my diagrams had been prepared. 

 In it he adopts practically the same plan in illustrating the migration 

 of the Swallow in Hungary, and as his maps are based upon 5900 

 returns, it is needless to say they are far more convincing than mine. 



Comparison of Records. 



As already stated most migration records so far obtained are the 

 work of one individual at each locality. Now when we come to com- 

 pare the time of arrival of birds at two points or their arrival at the 

 same point on successive years, it becomes very important for us to 

 consider the extent to which such records reflect the actual progress of 

 migration. The discussion on determining dates of bulk movements in 

 the vicinity of Philadelphia has already shown that while a date of 

 "first arrival" may be perfectly accurate for the limited area covered 

 by an observer, it would differ very materially from the earliest date 

 of arrival for the species in a circle of five or ten miles around that 

 observer's station. 



The work of the Delaware \^alley Ornithological Club for the past 

 seven years has shown that within the Philadelphia ten-mile circle, 

 covering an area with but little variation in altitude, we can detect 

 no constant difference in the time of arrival of a species at any two 

 points dependent upon their geographic position. 



The eariiest record is just as likely to come from the northern portion 

 of the circle as from the southern portion. At one time the records 

 seemed to show a slightly earlier date of arrival immediately along the 

 Delaware river, as compared with stations a few miles back on slightly 

 higher ground, but further data showed this difference to be purely 

 fortuitous. Therefore we can take the records of any one station 

 within this circle as representing the progress of migration at Phila- 

 delphia, just as well as those of any other station within the same 

 radius, and presumably the average dates of arrival of a species for a 

 number of years at several stations within the circle will be the same. 



For certain species which are very conspicuous and which usually 

 arrive in force on the first day of their appearance this is true, but in 

 the majority of species it is by no means so. 



