The migrations of this species are not at all easy to follow, as it appears to be 

 somewhat irregular in its movements, as detailed above by Mr. Hudson. Where tlie large 

 numbers which pass through Buenos Aires spend the winter has yet to be ascertained for 

 a certainty. The records quoted above show that the species breeds in the neighbourhood 

 of Buenos Aires in October, and further south, in the Chuput valley, it was breeding in 

 December. In the latter locality it had left by the 1st of March, but it remains at 

 Buenos Aires till April ; and Mr. Barrows states that it returns early in July, though 

 Mr. Durnford considers August the month of arrival. It is therefore absent, as a rule, 

 from May till August, and Mr. White met with it in Corrientes in May, and batterer 

 at Matogrosso in June ; bvit that some birds remain in their northern home seems to be 

 proved by its occurrence in San Paulo and in Peru in October. 



The description and measurements are taken from the series in the British Museum. 

 The specimen figni'ed is from Pelotas, and is evidently a young bird, not having the pure 

 white rump of the adult. (See our remarks above.) 



