196 bulletin: museum of comparative zoology. 



47. Sauropatis sacra sacra (Gmelin). 



Alcedo sacra Gmelin, Syst. nat., 1788, 1, pt. 1. p. 453. (Tj-pe-localitj' hereby 

 restricted to Tongatabu Island). 



One male was taken on the Island of Tongatabu, 30 November. 

 There are two other birds, a male and a female from the island of 

 Xomuka taken 2 December, that have been referred to this form but 

 are not typical of it as they are slightly duller in color than the bird 

 from Tongatabu, and in addition are considerably smaller. Meas- 

 urements are as follows: — 



The birds from Xomuka are in worn plumage, which might account 

 in part for the shortness of wing and tail. The single bird from 

 Tongatabu, howcA'er, has a large heavy bill that is noticeably stronger 

 and broader at the base than in any other specimen in the various 

 forms of this species examined. Females of Sauropatis sacra in general 

 differ from males in having the superciliary stripe whiter, in being 

 flecidedly more greenish aboA'e, and in ha\'ing little or no blue appar- 

 ent in the blackish feathers on the anterior surface of the tibia. In 

 addition females are the larger of the two sexes. 



In treating the geographical forms of Sauropatis sacra it becomes 

 necessary to restrict the typical subspecies. Therefore, I designate 

 the island of Tongatabu, in the Tonga Islands, as the type-locality 

 of SauropcMis sacra sacra. The Alcedo sacra of Gmelin was based 

 upon Latham's sacred Kingfisher (Latham, Gen. syn. birds, 1782, 1, 

 pt. 2, p. 621). Latham states that his bird had a blue band on the 

 hind neck, in this agreeing with the forms found in the Tonga Islands. 

 An island in this group therefore is selected as the type-locality for 

 the typical form, as birds from the Fiji Islands have this band black 

 or with only a trace of blue. The latter are to be known as Sauro- 

 paiis sacra viticnsis (Peale) (U. S. explor. exped., 1848, 8, p. 156). 



This species has the tenth (outermost) primary- slightly shorter 

 than the sixth and longer than the fifth, and so does not agree wholly 

 with the figure of the wing of Sauropatis saiictus given by Mathews 



