VOl i906 CI11 ] Dwight, The White-winged Gulls. 27 



ble, and furthermore it is possible they may prove to be the con- 

 necting links between glaucus and leucopterus on the one hand 

 and glaucescens on the other, but at present there is no evidence 

 that they represent any stage of plumage of any of these three 

 distinct species. 



The material on which I have based my conclusions has been 

 most extensive, including not only the series in the large collections 

 of our own country, but I have also had opportunity for examin- 

 ing those in the British Museum, in the Rothschild Museum at 

 Tring, and in the museum at Berlin. In spite of this large amount 

 of- available material, some 350 specimens in all, the great lack 

 of proper sexing has proved a serious stumblingblock, and to 

 overcome possible errors resulting from this cause, I have con- 

 fined my measurements of adults almost wholly to birds taken in 

 the breeding season. The number of labels bearing no sex mark 

 or one that is obviously wrong is almost incredible, and among 

 the gulls where the plumages of the sexes are alike, and females 

 may be recognized only by their smaller size, the question of cor- 

 rect sex marks is of the greatest importance. In the large series 

 examined, I found an unusual proportion of moulting birds that 

 have been of the greatest value in tracing out the sequence of 

 moults and plumages, although less serviceable for measurements 

 of wings and tails. 



Relative measurements are shown on the accompanying table 

 which has been prepared by selecting, so far as possible, adult 

 breeding birds and young birds taken so late in the fall and winter 

 that they would be expected to have attained their full growth. 

 It will be observed that except for their bills the young birds 

 closely approximate to adult dimensions, and it is a well-known 

 fact that the tarsi and toes of young gulls very quickly attain their 

 full growth. It is of interest that leucopterus averages about 

 16 % and the bill 33 % smaller than glaucus, while barrovianus 

 is scarcely 3 % smaller in size and 4 % smaller in bill. Now, the 

 individual variation in any of the species under discussion amounts 

 to more than 7 °fo , and it is doubtful if any two students measuring 

 the same birds would come within 3 % of the same result. Fur- 

 thermore, in barrovianus the character of bill — "which has the 

 depth through the angle never less and usually decidedly greater 



