368 BR. ST. GEOEGE MIVART ON THE SKELETON OF [May 7, 



In L. flavcpalliatus there is none of all this, save that a groove 

 appears on the mid-dorsum at a point coinciding with the level of 

 the antero-posterior middle of the nares, and then similarly 

 bifurcates — its two branches running forwards and similarly ter- 

 minating in foramina on each side of the beak at some distance 

 from its apex. 



The tentral aspect of the prosopium (see figs. 6 & 7, p. 378) 

 presents a palatal sm'face slightly more concave both antero-pos- 

 teriorly and transversely than in P. erithacus. In the latter species 

 this surface is crossed by a very slightly marked ridge the median 

 part of which is the most preaxial, the two lateral halves of the 

 ridge thence diverging backwards at an angle of about 140° ; this 

 median point of divergence is rather nearer to the apex of the pro- 

 sopium than to the hinder margin of its palatal surface. Trom 

 immediately behind it, an antero-posteriorly directed groove runs 

 postaxiad aud leads to a small foramen, beneath which the margins 

 of the groove medianly unite, and thence an antero-posterior 

 prominence may continue on in the same direction as the groove 

 and for about the same length, subsiding entirely before reaching 

 the hinder margin of the palate. 



In L. fiavopalliatus there is no transverse ridge but only a slight 

 smooth swelling of the palatal surface in about the position of the 

 middle of that transverse ridge in P. eritliacus. In front of this is 

 a short median antero-posteriorly directed groove, while from 

 behind it a similarly directed prominence runs backwards almost 

 to the very postaxial margin of the palate. The middle of that 

 margin in P. erithacus projects postaxiad rather more than the 

 parts of the margin right and left of it, because the postaxial con- 

 cavities for the palatines begin almost at once on either side of it. 



In L. flavopalliaius, on the contrary, the middle part of the 

 postaxial margin of the palate does not project postaxiad quite so 

 far as does the part of that margin on either side which is mesiad 

 to the concavity for the palatines. Nevertheless in that median 

 part there is a minute process bounded laterally by a very minute 

 notch, these notches together with the process they laterally bound 

 appear in the middle of the postaxial median marginal concavity of 

 the palate. 



The free margin of that palate is relatively much more extensive 

 in this species, the part of it interposed between the two surfaces 

 for the palatines being fully equal to the extent of both those 

 surfaces, while in P. erithacus it is but about equal to one of them. 

 The postaxiad extensions of the prosopium on either side of the 

 palatines are short, yet a little longer, relatively, than in P. erithacus. 

 Thus the palatines do not advance so far forwards into the palatal 

 region of the prosopium in the last-named species. 



These lateral prolongations are strongly convex antero-poste- 

 riorly in both species, but more so in P. erithacus, since in 

 L. jJavopalliatus (as before said) the jugal process so projects as 

 to produce a concavity towards the postaxial end of the tomial 

 margin. Here also these lateral prolongations may be said to be 



