9 



perhaps, is due to the lack of timber in the lowlands, and a consequent want of the 

 necessary facilities for rearing the young, rather than to any natural preference for high 

 regions. About the middle of August, while in extreme South-eastern Arizona, I noticed 

 each evening immense numbers of these birds and the Cliff-Swallows flying swiftly over- 

 head, their course leading them directly south. They only paused now and then to catch 

 an insect, immediately resuming their onward flight. All the actions of these birds 

 seem to indicate that the migration at this early date had begun, yet I have found in 

 quite a number of instances the parents feeding the just-fledged young as late as 

 August 22nd." 



A specimen from Mount Shasta, in the British Museum, seems to us to be true 

 P. purpurea, and the following note by Mr. Townsend appears to refer to this species : — 

 "Martins were not common in the localities where I collected. A few were noticed 

 about some buildings at the west base of Mount Shasta in midsummer. A colony of a 

 dozen or more was found established in a large dead pine on the edge of the forest at the 

 eastern base of Mount Lassen, on June 6. The only nest I could reach occupied a large 

 decayed cavity twenty feet from the ground. It contained four fresh eggs. There were 

 other nests higher up." 



In Western Texas Mr. Lloyd says that the Purple Martin is plentiful in summer 

 in suitable places. It breeds in colonies, arriving towards the end of February and 

 departing about the 1st of November. At San Antonio it is common in summer, 

 according to Mr. Attwater. At Fort Brown, in Southern Texas, Dr. Merrill only 

 observed the species during migration, but he noticed their arrival as early as the 

 20th of January. 



In Florida the resident species is supposed to be Progne cryptoleuca. Mr. Scott has 

 forwarded several specimens from Tarpon Springs, but after a careful examination Ave 

 have been unable to discover any specific characters for the recognition of this supposed 

 species. We add Mr. Scott's note on the Purple Martins : — 



" It seems to me probable that all Martins found breeding on the Gulf Coast of 

 Florida, at least as far north as Tarpon Springs, are referable to P. cryptoleuca, and 

 though the material that I have before me is limited, yet one of the male birds is fairly 

 intermediate between P. subis proper and what I think will ultimately have to be 

 considered as the subspecies P. subis cryptoleuca, though the latter is now given specific 

 rank. I have submitted material collected in the vicinity of Tarpon Springs to Mr. J. A. 

 Allen, who concurs in the above views and from whose letters on the subject I quote as 

 follows :— ' The Martins I should refer to Progne subis cryptoleuca, of which the single 

 female and two of the males are fairly typical. The other male I should consider an 

 intermediate between P. subis and P. subis cryptoleuca, which latter I believe at least 

 only a geographical race of P. subis.' As the birds are abundant in the breeding-season 

 in the town of Tarpon Springs, and as I am expecting additional representatives from at 

 least two points of the south on the Gulf Coast, as well as from Key West, I hope at an 

 early date with more abundant material to deal conclusively with the subject. At 



1h 2 



