44 PROCEEDINGS OP THE 



April 6 and 7, brought Hermit Thrushes, Ruby-crowned 

 Knights and Vesper Sparrows. April was, however, for the 

 most part cloudy with much rain until the last few days, when 

 a rise in temperature, culminating on May 1st, brought hosts of 

 migrants which had been held back by unfavorable conditions. 

 May migration was concentrated on the 1st, 7th, 10th and 16th, 

 and, as is usual when the movement is congested, many Cape 

 May Warblers and other rarities were observed. 



Considering the records in the tables for Philadelphia and 

 vicinity we find that 19 records of first arrivals fell on February 

 26-28, 20 on March 11-13 and 23 on March 26, the number for 

 other days being from one to four. In April there were 33 re- 

 corded, April 7-9, while other days varied from one to seven, 

 until April 23, when the rise in temperature began, and we find 

 13 first-arrival records for April 23 ; for April 24, 4 ; April 25, 

 6 ; April 26, 21 ; April 27, 14 ; April 28, 15 ; April 29, 13, and 

 April 30, 44. This shows clearly how the movement was con- 

 centrated on a few days. 



It has been customary during the last few years to compare 

 the dates of bulk arrival, computed, as explained before,^ with 

 the average of previous years to show how many days early or 

 late each species may be. By comparing several years in this 

 manner it is interesting to see how birds which usually arrive 

 together are affected in the same way and vary from the average 

 by nearly the same number of days. This may be seen by 

 studying the following table. Notice that the earliest migrants 

 were all notably early (+) in 1906 and 1909 and notably late (— ) 

 in 1907, while in other years they came closer to normal, especi- 

 ally in 1908. Notice also the different character of the migration 

 as the spring advanced, i. e. in 1906 the earliest migrants were 

 early, those coming about the end of March were notably late, 

 those at the end of April almost exactly normal and the May 

 migrants all early. In 1907 conditions were exactly reversed, 

 earliest migrants late, March migrants early, April migrants 

 normal (as in 1906) and May migrants all late. 



' i. e. , by dropping the last quarter of the records in the Philadelphia cir- 

 cle and selecting the date on which the species had reached one-half of the 

 remaining stations. The disregarded records are considered as probably rot 

 actaal dates of arrival. 



