1886.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. ^37 



It may be well to remark that the individual variation as to color is 

 very great in the Hemipodes, and that the females of the present group 

 are larger than the males, and have the chin, throat, and fore-neck uni- 

 form blackish. 



Treron permagna, sp. n. 



DiAGN. — Rectrices underneath uniform slate black, the tips bordered 

 with ashy; tail slightly graduated; third primary sinuated in the inner 

 web; under tail-coverts dark-olive green broadly bordered with pale 

 creamy yellow ; secondaries and their great coverts narrowly edged with 

 the same ; primary coverts uniform blackish ; lower abdomen of a creamy 

 primrose-yellow ; rest of plumage dull olive-green above, clearer on the 

 rump, clear yellowish oil-green on forehead and under parts. Size very 

 large: wing over 200'"'". 



Type. — No. 17, Coll. Namiye ; Napa, Okinawa Shima, Liu Kiu Islands, 

 March 9, 1886. 



Habitat. — Okinawa Shima, middle group, Liu Kiu Islands, Japan. 



The identification of the present species has caused me considerable 

 difficulty. It is much larger than Treron formosce Swinhoe* from 

 Formosa, being, so far as I can see, the largest species in the whole 

 genus Treron in its widest sense, though otherwise evidently closely 

 resembling the female of that species. But the specimen before me, 

 which Mr. Namiye collected at Napa, March 9, 1886, is marked on the 

 label as being a male. Should the determination of the sex be cor- 

 rect, the present species would be unique amongst its nearest allies in 

 having the small upper wing-coverts olive-green in the male and not 

 chestnut. 



But even if the specimen be a female, it differs sufficiently from Swin- 

 hoe's description of the types of his T. formosce to warrant the separation. 



First, as to dimensions : His female has the wing 7.2 inches long and 

 the "tail -' 1 inches, while Namiye's bird measures, wing 8.3 and tail 6 

 inches ! It is consequently much larger even than the male T. formosce, 

 which, according to Swinhoe, has the wing 7.8 and the tail 5 inches, 

 with a graduation of 0.5 inch, while in permagna the latter measurement 

 is nearly 1 inch. It should also be remarked that my mode of meas- 

 uring gives the smallest possible dimensions, for the wing is not flat- 

 tened, and the tail is measured by thrusting the point of the dividers 

 between the central tail-feathers down to the base. 



In general coloration my bird agrees tolerably well with Swinhoe's 

 description (when we remember that ^'■yellow on the head and rump" is 

 a misprint for '■ yelloicer^^), except that it has the throat uniform with 

 the rest of the under parts, and not " grey, each feather margined with 

 yellowish green." In the details, however, there seem to be some dif- 

 ferences, which will be apparent from the subjoined full description of 



*Ibis, 1863, p. 396, and 1866, p. 312 ; nee Sphenocercusformosce Swinhoe, Ibis, 1866 

 p. 122, quae T. sororia SWinh. 



