HOUSE-MARTIN. 179 



as he failed to detect any difference from our bird in some examples in the 

 museum at Omsk. 



The House-Martin has no ally so near as to make it probable that it 

 interbreeds with any other species ; but in Eastern Siberia it is replaced 

 by Pallas^s Martin, H. lagopoda, which has frequently been mistaken for 

 it. This is a well-defined species, with a shorter and squarer tail than our 

 bird : it also differs in having the longest upper tail-coverts white instead 

 of black, and the axillaries and under wing-coverts dark brown instead of 

 very light brown. I found it extremely common in the valley of the 

 Yenesay, swarming in thousands on the Arctic circle, and breeding as far 

 north as lat. 69°. It probably breeds throughout Eastern Siberia, as 

 Middendorff observed it in the Stanovoi Mountains, and Pallas records it 

 from Kamtschatka. It breeds throughout South-eastern Siberia, and is 

 probably the species found in such numbers by Prjevalsky in Mongolia. 

 Severtzow records it as passing through Turkestan on migration ; but it 

 breeds in North China, and probably winters in the Burma peninsula, 

 where it has been mistaken for the European species *. 



Between these species two intermediate species occur, doubtfully distinct 

 from each other, but both of them distinct both from the House-Martin and 

 Pallas's Martin. The larger of the two, H. dasypus, with a wing exceed- 

 ing four inches in length, breeds in Japan and winters in, Borneo; the 

 smaller one, H. cashmiriensis, with a wing less than four inches, appears 

 to be confined to the Himalayas, breeding in the higher valleys and 

 wintering in the lower ones. Both these forms have the short, slightly 

 forked tail of the eastern Martin, and the black upper tail-coverts of the 

 western species. The only other Swallow with feathered tarsi and feet 

 is the Himalayan Martin, H. nipalensis, a much smaller bird, which may 

 at once be distinguished by its black under tail-coverts. Our knowledge 

 of the geographical distribution of this section of the genus Hirundo is in 

 great confusion, owing to the carelessness of collectors, who in too many 

 instances have not thought it worth their while to shoot such a common 

 bird, taking it for granted that no mistake could possibly arise in the iden- 

 tification of so well known a species as the House-Martin. 



The House-Martin is a spring migrant to our islands, and reaches us 

 about a week after the Swallow has announced the coming of summer. In 

 the south of England it usually arrives about the middle of April ; but in 

 cold and backward seasons it sometimes does not appear until the end of 

 that month. In Greece and Asia Minor, Dr. Kruper informed me that it 

 arrived regularly during the first week in March ; whilst in the north of 



* In Colonel Tickell's manuscript Illustrations of Indian Ornithology, presented by him 

 to the Zoological Society of London, the figure and description of the example of Chelidon 

 urbica obtained in Tenasserim unquestionably refer to Pallas's House-Martin. The upper 

 tail-coverts are not only described as white, but are also figured so. 



n3 



