1886.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 639 



feathers, loS"''" ; graduation of tail. 24"'" ; exposed culmen, 21"'" ; tar- 

 sus, 28™"^ ; middle toe with claw, 37™™. 



Treron permagna is the second species of the Green Pigeons known 

 to occur in Japan, the other species being the well-known T. sieboldii.* 

 The two species may be easily distinguished thus : 

 «! Lower breast and entire abdomen yellowish-white ; the outer tail-feathers above 



with a subtermiual black band T. sieholdii. 



a- The yellowish white of the body confiued to the middle portion of the lower 

 abdomen and the crissum ; the tail-feathers without auy subtermiual cross- 

 baud T. permar/na. 



A close ally of T. sieboldii is found in Formosa, viz, Swinhoe's T. 

 sororiusj, if, indeed, it really be separable. It would be interesting to 

 know how the corresponding bird from the intermediate islands is 

 colored. 



Megascops elegans (Cassin). 



This species is not new to the Japanese fauna, for the type was taken 

 on board a vessel while in Japanese waters, west of the northern islands 

 of the Linschoten group, consequently not far from the locality where 

 it has now been rediscovered by Mr. Kamiye, who obtained a female 

 specimen at Oroku, Okinawa Shirna, March 28, 1886. 



In order to determine this specimen I went to Philadelphia, where, 

 by the courtesy of the authorities of the Academy of Natural Sciences, 

 I was permitted to institute a direct comparison with Cassin's type of 

 '''■Epliialtes elegans^ I found, as I had anticipated, that the bird col- 

 lected by Mr. Namiye is identical with the type, agreeing as itdoes very 

 closely both in size and coloration. It is slightly more rufous all over, 

 but otherwise it matches it so closely, and especially in the amount of 

 feathering on the tarsi, that two specimens more similar are scarcely 

 found in any of the species of this variable group. The type specimen 

 has hardly any indication of occipital or cervical bands and certainly 

 less than several specimens of M.japonicus now before me. The Liu Kiu 

 specimen shows more of a cervical light band, though it is by no means 

 very pronounced. It seems not justifiable to refer the present species 

 to M. lettia and M. japonicus to M. scops as subspecies because of the 

 absence or presence of these bands, which apparently are quite an un- 

 stable character. 



The present species is certainly very closely allied to 31. japonicus, 

 from which it is easily separated, however, by its superior size and by 

 the greater extent of the naked portion of the tarsus. The coloration 

 js very much the same, the individual variation being almost endless, 

 but it seems as if J\L elegans has the blackish markings on the top of 

 the head larger and darker, thus making the crown conspicuously darker 

 than the rest of the upper parts ; a feature which I do not find in any 

 specimen of my series of M.japonicus. The ear-tufts seem to be larger 



*Columba sieboldii Temminck, PI. Color., iv, liv, 93, pi. 549. (18:?5.) 

 t Ibis, 1866, p. :511= Treron sonorius Gray, Hand-1. B. , ii, p. 224 ( 1870. ) = Sphenocercus 

 formoscB Swinhoe, Ibis, 1866, p. 122, nee Treron formosoe Swinhoe, 1863! 



