1906.] OP THE TKACllEOrHOJfE PASSEJIES. 153 



In Philepitta the tnhercnlam onedkis, whicli f'oi-nis a kind of 

 penthouse I'oof over the pneumatic foramen, is perfoi-ated by a 

 small hole ; and the deltoid crest is short and but feebly developed. 

 The arm and forearm ai-e pneumatic. 



The intermetacarpal plate is moderately well developed, and 

 fuses along its hinder border with the metacarpal III., which is 

 sleiider and bowed. 



In the Pittidse only the humerus is pneumatic. The deltoid 

 crest is obsolete, but the crista inferior is well developed and 

 i-ougiily triangular in shape. As in Philejnita, there is no coraco- 

 humeral gi-oove, but the ectepicondylar process is better developed 

 tha,n in Philepitta. 



In the Pipridse the humerus only is pneumatic, the deltoid 

 crest is short and feeble, and the coiuco-humeral gi-oove is wanting. 

 At the base of metacarpal I. is a deep notch continued outwards 

 along the preaxial border of metacarpal II. in the form of a channel 

 for the tendon. 



jSTone of the bones in the wing of the Pteroptochinse is pneu- 

 matic ; the deltoid crest of the humerus is obsolete and confined 

 to the extreme proximal end of the shaft, and the ectepicondylar 

 pi-ocess is wanting ; the intermetacarpal plate is well developed, 

 and the pi'eaxial border of metacarpal II. is marked by a small, 

 laterally comjDressed, mound-shaped boss of bone. 



As in Pteroptochus, so in Hylactes the wing is non-pneumatic, 

 and the deltoid crest of the humerus is feebly developed, while the 

 radial and ulnar condyles are set close together. 



Metacarpals II. and III, are unusually broad and set close 

 together, reducing the space between to a mere slit. 



Of the subfamilies Formicariinse, Dendrocolaptina^, and Synal- 

 laxinse, I can say nothing that would be of use. The wing 

 here presents a great general similarity, and it is impossible to 

 say, at present — owing to the lack of skeletons, — whether the 

 slight diffei-ences which can be made out are due to individual 

 vai'iation, or whether they obtain throughout whole genera. 



viii. The Pelvic Limb. 



The pelvic limb in some respects resembles that of the Eury- 

 la?mid0e. Although the diflerent groups herein described do not, 

 in this matter, differ very widely one from another, yet this limb 

 presents a greatei- I'ange of variation than is found in the case of 

 the wing. 



In the Pipridse only is the femur pneumatic, in this matter 

 agreeing with the Cotingidse ; the fibular crest is short, and the 

 fibula continued far down the leg in the form of a delicate style ; 

 tlie cnemial crests are modei'ately well developed. The Ph. I. of 

 1). IV. is less than half the length of Ph. I. D. III. 



Tlie most striking feature perhaps about the Philepittidje is the 

 oblong, more or less quadrangular shape of the entocnemial process, 

 which rises direct! v from the level of the articular surface of 



