AUSTRALIAN WHITE TERN. 
subdivision were advisable, Gmelin’s name has priority over Hartert’s ; but 
from the series I have examined I would rather class Hawaiian birds under 
Gygis alba Candida (Gmelin) ; Christmas Island, Hawaiian group 
and preserve 
Gygis alba Icittlitzi Hartert ; Caroline group ; Mariannes. 
This is only a suggestion but it is based upon study of the variation 
existent in allied genera. Whether it be followed or not, birds from the 
South Pacific are certainly separable, and moreover more than one sub- 
species is recognisable. As Hartert pointed out, Kermadec birds are much 
longer in the wing and bill than Northern birds, and I therefore propose to 
name them 
Gygis alba royana, subsp. n. ; Kermadec Islands, 
Norfolk Island (breeding), ? Friendly Islands. 
Specimens from the Society group and Samoa agree in having longer biUs 
than the Kermadec birds, while they have shorter wings. I have not seen 
specimens from the Marquesas so cannot say what form occurs there, if any, 
belonging to the species. For this form I would use 
Gygis alba pad flea (Lesson) ; Society Group, Samoa, etc. 
and would quote Sterna nivea Bennett as an absolute synonym. 
I have examined specimens from Easter Island, Pitcairn Island, and 
Ducie Island which seem to represent another race, but from lack of series 
I forbear its nomination. Indian Ocean birds are referable to the Pacific 
Ocean type, having the bill of similar shape with the nostrils situated weU 
forward, and dark shafts to the primaries, etc. They have as long biUs as 
G, a. royana but the wing is noticeably smaller, rather less than Guam 
specimens of G, a. Icittlitzi. For these birds I propose the name 
Gygis alba monte, subsp. n. ; Indian Ocean, Seychelles (breeding). 
443 
