570 Mr. G. T;. Bates on the Reversed [Ibis, 



ten on the manus and eight or nine on the cubitus ; overlap 

 on the cubitus and the adjacent part of the manus contrary, 

 on the rest of the hand conforming. The proportion of these 

 coverts on the manus with the conforming overlap varies 

 greatly, being often about half but sometimes much less, 

 and on one wing of one bird none at all; while sometimes it 

 is more than half, and in one specimen apparently all these 

 coverts were conforming. 



Tachornis parvus. Seven specimens examined, No. 5763, 

 and six others not saved. 



Reversed under coverts large, in one continuous row, of ten 

 on the manus and about eight on the cubitus. Wing eutaxic. 

 In my notes 1 have recorded the presence of a remicle or 

 11th manual remex, having no corresponding under covert, 

 and it is possible to consider this little terminal feather as 

 really an iltli under covert, as was certainly the case in 

 some birds having no remicle. Ov^ rlap of all reversed 

 under coverts conforming. 



Chatura sharpei. Two specimens examined, Nos. 5786 

 & 58i3. 



Reversed under coverts forming one continuous row, num- 

 bering ten on the manus and about seven on the cubitus, 

 with a small one besides at the carpal joint, in front of 

 the row of the others. Wing eutaxic. Here again the 

 presence of a remicle was noted, with the additional words, 

 in one case, "resembling the major under coverts," and 

 possibly this is to be considered as an llth reversed under 

 covert on the manus. Overlap of all reversed coverts 

 conforming. 



This group (the Swifts) seems to be out of its place, with 

 regard to the morphology of the reversed under coverts, 

 when put here next to the Coraciiform birds. For if we con- 

 sider the single series of these coverts to be made up of the 

 major coverts on the hand and the median ones on the 

 fore-arm, as in the preceding groups, then we have median 

 coverts with the conforming overlap and cubital major coverts 

 absent, which is improbable. Besides, the smaller extra 



