ORNITHOLOGICAL EXPLORATIONS. 



263 



I do not think that it breeds on the islands, however, as it was not 

 met with from the end of May until the beginning of November. Be- 

 tween the 21st and the 28th of May, 1882, 1 met a few small troops, con- 

 sisting of three to five individuals, but their behavior convinced me 

 that they were not thinking of breeding ; they were probably on their 

 way westward. During the winter, from November till March, this race 

 was the most common of the three forms occurring on the island, al- 

 though never met with in flocks of more than six individuals. They 

 were always moving rapidly and distinguished themselves by a remarka- 

 ble shyness. 



Having discussed the affinities of the birds in question further on 

 under the heading of A. exilipes, as also in a previous paper (Auk, 

 1884, p. 148), I content myself here by only giving some comparative 

 measurements of birds from different parts of the circumpolar realm. 

 Among these are several specimens marked L. ordinaria, parvirostris, 

 or hetnlarum in Sundevall's own handwriting. It will be seen that only 

 such specimens have been selected for measurement, the locality, date, 

 and sex of which were stated by the collector. Five specimens collected 

 by myself, the sex of which I was unable to determine, are kept separate 

 in the list of specimens collected. It is too often the fault of authors 

 pretending to give the results of "careful comj^arisons" that they mix 

 birds of different sex and age, and from different seasons, together, 

 without stating the known particulars i^ertaining to every single speci- 

 meu, thus depriving the public of the means by which an intelligent 

 judgment respecting the value of the conclusion arrived at, can be 



formed. 



Li8t of specimens collected (sex determined), 

 a.— SUMMER SPECIMENS. 



