44 PROF. d'aRCY W. THOilPSON ON THE [Jail. 17, 



Psittaciis, Coracopsis, and Dasyptilus, are very different from one 

 another in regard to their skulls. Coracopsis appears to show a 

 marked resemblance to Edeeius and Gtoffroyus. As to JJasi/ntilus, 

 it is certainl)' very different from both Coracojisis and Puittacus; 

 the characters of onr imperfect specimen suggest no close alliance 

 with other forms, but go some way to indicate a ^■ery isolated 

 position for the genus. 



The skull of Agcqjomis differs very materially from that of the 

 typical PaliPornithince. The difference is, in the first place, con- 

 spicuous in the extremely narrow auditory meatus and extremely 

 wide, almost square, surface between its straight posterior border 

 and the descending occipital ridge. The suprameatal tubercle is 

 moderately developed, the squamosal process long and curved, the 

 postfrontal process extremely small, the nostrils and the iiiterorbital 

 and mandibular fenesfrajall large. It is for one thing plain, from 

 the breadth of its post-auditory region, that Agajjumis differs 

 greatly from the ordinary South-American Parrots. I regret 

 that 1 have not been able to examine the skull of PsittacuJa, for it 

 M ould be extremelj- interesting to see whether this osteologiial 

 feature confirms (like the characters of the carotids) the separa- 

 tion of these two superficially similar but geographically distinct 

 genera. 



Extremely different from all the Parrots of the Old World 

 and very similar to one another are the two genera Chri/sotis and 

 Pionvs. The orbit is complete by junction of the prefrontal and 

 postfrontal processes, and the bridge of bone so formed descends 

 to form an angular prominence opposite the extremity of the 

 squamosal. The temporal fossa is extremely narrow. The inferior 

 border of the squamosal is curved but not notched. The supra- 

 meatal tubercle is low, the auditory aperture wide, and its posterior 

 border is very near to the occipital ridge which bends forward in 

 a sinuous curve to approach it. The paroccipital processes are 

 prominent but flattened or exca\ated below externally to the basi- 

 temporal ridges. The nostrils are of moderate size, the interorbital 

 fenestrse small, and the mandibular fenestra nearly obsolete. Of 

 the genera grouped w ith these by iSah adori under the name Pionince, 

 I find PacJii/mis, Cairn, and the African Pceocephahis to be very 

 different. Caica, in the extremely small size, circular form, and 

 wide distance apart of its nostrils, resembles Myopsittacus, which 

 latter is usually grouped with the Conures. In their other 

 characters Caica and j\Jyopsitiaci(s are verv similar to one another. 

 The postfrontal process is short and nearly vertical, especially in 

 Caica. The squamosal process is curved in its lo«er border and 

 more or less distinctly notclied at its base. The suprameatal 

 tubercle is distinct, the auditory meatus rathei- wide and its 

 posterior border somewhat further from the occipital ridge than in 

 Chrysotis and much farther than in Conurus. The paroccipital 

 process is very much as in Chrysotis ; the long prefrontal process 

 much as in Conurus. The interorbital fenestra is of moderate size; 



