35^^ AMERICAN MIDLAND NATURALIST 



nortli for Uirec years, but in 191 2 it was not seen until Oct. 13, 

 which was the only record for the species that autumn. (When 

 two dates are given for any species, that means the first was the 

 date of arrival from the north). It will be noted that the Hermit 

 Thrush was regular for two years in the time of its departure, 

 but 8 and 9 days earlier and 9 and 10 days later than the regular 

 dates in the other two years — 18 days being the difference between 

 the earliest and the latest date of departure. 



The Kingbird was regular in its time of leaving except in 

 191 2, when it departed on August 22. This date was 15 days 

 earlier than the late?t date, which was on Sept. 6. Barring the 

 one early date, this species left regularly in the first week of- 

 September. 



The Hummingbird is seldom recorded by the writer at any 

 season of the year, and this fact will probably explain the marked 

 difference in the dates of migration for the species. Only two 

 dates — in September — can be called regular, the third — in August — 

 being 23 days earlier than the latest date. The August date must 

 have been far from the true time of departure, yet it was obtained 

 from a fairly diligent search. 



Considerable disparity is shown in one of the dates of the 

 Vesper Sparrow — Sept. 30, 1912. This date was 26 days later 

 than the earliest date. No record was made in 19 13, which, with 

 the single late date in 191 2, would indicate that it is difficult 

 to determine the time of departure for this species. The 

 regular dates are too early for this sparrow, which — being such 

 an early spring migrant — must tarry with us as late as October. 



In the Baltimore Oriole we have a species that shows the 

 greatest regularity of any of our birds in the time of its autumn 

 migration. Only 2 days between the earliest and the latest date, 

 is the remarkable record for this species. 



The Barn Swallow shows 18 days between to be the dift'erence 

 between its earliest and latest date of migration. The time between 

 the third date — August 31 — and the latest date — Sept. 6 — the 

 writer thinks indicates the true tim.e of migration. 



Only two dates were made for the Loggerhead Shrike — one 

 in August and one in September — with the difference 24 days. 

 This species is difficult to record, especially after the nesting season; 

 and the single record for vSeptember is hardly sufficient to determine 

 with satisfaction when the species migrates. 



