MIGRATION OP OUR BIRDS IN AUTUMN 357 



The four records of the Purple Martin are divided iDetween 

 August and September -the dates of each set being regular. 

 There are i6 days between the earliest and latest dates, and 

 12 days separating the other records. From these figures we can 

 see that this species may leave on dates that are i or 2 days apait 

 or within a period that is a little less or a little more than a 

 fortnight. 



The three dates of the Yellow Palm Wrabler place this species 

 among the migrants that are regular in their time of leaving, 

 two days being the greatest difference. But the same regularity 

 is not observable in the two dates of arrival — 12 days separating 

 them. 



The Bronzed Grackle presents a case of great disparity in 

 its dates of migration, no fewer than 54 days intervening between 

 the earliest and the latest dates. But such an early record as 

 September 15 may be set aside in determining the true time of 

 migration for this species; and the interval between Oct. 20 and 

 Nov. 8 — 18 days — may be taken as the probable period in which 

 this grackle usually departs. 



The Warbling Vireo showed regularity in all of its dates 

 except one, which was 10 days later than the earliest date. The 

 Redstart was more regular still — two of the dates being identical 

 and the others, 3 and 8 days later respectively. The Golden- 

 crowned Kinglet must also be ranked among the regular arrivals 

 in autumn — 5 days making the longest interval; but there is 

 less regularity in the time of departure. 



A very curious case both of irregularity in migrating and of 

 non-migration was that of the Red-headed Woodpecker. BetM^een 

 the earliest and the latest date there were 23 days; between the 

 earliest and a later date there were 8 days. In 1913 the species 

 did not migrate; but remained throughout both autumn and 

 winter. A very interesting occurrence indeed was this non- 

 migration of a species that left for three previous years in the 

 month of September. What is the explanation? The winter proved 

 a very mild one; and could the bird have known so long before- 

 hand that the season would not be severe? I think the bird must 

 have been able to determine this, but how I can not say. 



Two species that showed marked regularity in three of their 

 dates, but great disparity in one date, were the Myrtle Warbler 

 and the vSnowbird. The early arrival of the Snowbird on Sept. 4, 



