2 2 EINAR LONNBERG 



and I am thus able to confirm Mai'HEWs' interpretation. The shafts of the 

 remiges of the type specimen are entirely white, and those of the rectrices 

 only partly on the concealed parts pale brownish. I have also the opportunity 

 of stating that the feathering of the forehead of this type specimen extends 

 well forward on the bill approaching the nostril within less distance than the 

 length of the latter. As this feature also is characteristic for the Gygis of the 

 Atlantic, it is evident that Sparrman's name alba must be used for the same. 



For the Gygis of the Pacific several different names have been used, and 

 it is not easy to decide which of them ought to be applied to the birds from 

 Easter Island. G. micro rhy tic ha Saunders is, of course, excluded on account of 

 its white shafts, different tail &ca. Gmelin's name Candida belongs typically 

 to the White Tern of the Hawaiian group, while the bird from the Kermadec 

 Islands, which is said to be considerably larger than the former, has been named 

 royaiia by Mathews. As the dimensions recorded by Mathews for the latter 

 come rather near to those of the present specimens from Easter Island it does 

 not appear impossible that they belong to the same race, although Mathews 

 says that he believes the Easter Island birds to » represent another race». 



Not long ago E. W. Gifford has published a paper (Birds of the Gala- 

 pagos Islands &ca. Exp. of the Calif Acad, of Sc. to the Galap. Isl. 1905 

 —06. Proc. Calif. Acad, of Sc. 4th Ser. Vol. II. Aug. 19 13) which throws 

 some light on the variation of the Pacific Gygis, which, evidently, is rather 

 considerable. It will thus appear that it is difficult to attach so great im- 

 portance to the difference of a few millimetres with regard to the length of the 

 wing of these birds as some authors have done. GiFFORD has examined and 

 measured 20 specimens of Gygis from Clipperton and Cocos Islands and found 

 that the length of the wing of the males varies from 223 to 260 mm. and that 

 of the females from 231 to 254 mm. (average in the former case 244, in the 

 latter 241). Mathews's G. a. royana as well as the Easter Island birds fall 

 thus within the limits of the variation of the Cocos Island birds in this respect. 

 The culmen of Gifford's birds varied between 36,7 and 43 mm. with an 

 average for males of 40,2, for females of 39,2 mm. 



I have not had any opportunity of examining specimens from Hawaii, 

 but professor Reichenow has kindly provided me with measurements of 10 

 specimens of Gygis from the Palau Islands, and from this communication it is 

 seen that four males have the length of wing from 245 to 250 mm., two fe- 

 males have the same measurement 248 and the remaining four unsexed ones 

 from 233 ($.^) to 250 (J*?), while the length of the culmen in the whole lot 

 varies between 38 and 43 mm. 



The difference between these dimensions and those recorded by Mathews 

 for G. a. royana are not very great. It is thus rather doubtful if not most of 

 the White Terns of the Pacific simply should be termed Gygis alba Candida 

 Gmelin, and it is only with hesitation that I have used the name royana above. 



Mr. Backstrom found White Terns on Motu Nui as well as on Rano 

 Raraku of the main island. The nesting places were similar to those of Pro- 

 eels terna. 



