1899.] CEAXIAL OSTEOLOGY OF THE P.UIEOTS. 43 



well-marked notch on the lower margin of the squamosal process 

 at the place of the quadrate articulation, much as in Stringo2)s. 

 The auditory meatus is very narrow and crescentic in form ; the 

 space between it and the descending occipital I'idge is very wide ; 

 the basitemporal region is nearly on a level with the occipital 

 condyle; the paroccipital process is blunt (except in Pyrrhulopsis) ; 

 the orbital ring is incomplete and the postfrontal process almost 

 obsolete or represented only by a vertical ridge ; the nostril is 

 large, the inter orbital fenestra is moderately so, the mandibular 

 fenestra is extremely small or obsolete. While Melopsittacus 

 appears to differ most markedly from the above in its complete 

 orbital ring, with its bridge, as in the Cockatoos, across the temporal 

 fossa, yet at the same time it possesses an exti-emely well-marked 

 notch at the base of the squamosal and a ieeply-excavated surface 

 between this and the suprameatal tubercle ; it agrees in all the 

 other characters mentioned above with the Flatycercince, of which 

 I have no doubt it is a real, though a somewhat aberrant, member. 



The case of Calopsitlacus is a little more difficult. While in the 

 Cockatoos the auditory meatus reaches backward to the descending 

 occipital ridge, in Calopsittacus as in Melopsittacus there is a wide 

 interspace between. The auditory meatus is proportionately nar- 

 rower than in the Cockatoos. The teuiporal fossa, though bridged 

 by bone as in the Cockatoos, is much smaller and narrower than ia 

 them. There is a very distinct notch at the base of the squamosal 

 and a well-marked surface between it and the suprameatal process, 

 though this is not nearly so conspicuous a feature as in the Platy- 

 cercince. The nostrils are very large and near together as in 

 Melopsittacus, and are very different from the small, round, and 

 distant nostrils of the Cockatoos. On the whole I should say that, 

 so far as cranial osteology goes, the position of Calopsittacus is an 

 open question, and that it is by no means impossible that it may 

 really deserve to be grouped somewhere near Nympliicus and 

 Melopsittacus. While the facts suggest at least the possibility 

 of a closer affinity than that usually recognized between the 

 two Australian groups of Cacatuince and Plalycercince, this larger 

 question must also remain for the meantime in uncertainty. 



The true Lories form a natural group, and their place is, I 

 believe, not far from the Platycerchue. The auditory meatus is 

 constricted, its posterior border is crescentic and widely separated 

 from the occipital ridge. The orbit is incomplete and the post- 

 frontal process almost obsolete or (as in Eos) narrow and vertical. 

 The squamosal process is more or less distinctly notched at its 

 base, more in Lorius, much less in TricJioglossus, and the well-marked 

 suprameatal process overhangs a surface of bone, to which ascends, 

 as in Aprosinictus &c., the bar which separates the auditory cavity 

 from the region of the quadrate articulation. The excavated region 

 of the base of the squamosal is not nearly so complete as in the 

 Platycercince, but yet it is more like to them than to any other 

 family of Parrots. 



The three genera grouped by Salvadori as Psiiiacince, namely 



