ON THE GENUS LATHAMUS. 67 



though the general character of the pterylosis remains the same in all*. 

 The outer pectoral tract is never so narrow and distinct here as it is in 

 Lathamv.s and its allies ; it is usually almost triangular in shape, and so 

 tolerably broad, shorter, and not so divergent, the interspace between it 

 and the main tract being much narrower, and frequently with a few 

 scattered feathers in it uniting the two tracts together. The inferior 

 tract on the breast is always much broader, and the carinal space narrower. 



Still better-marked characters between the two groups are to be seen 

 in the disposition of the dorso-lumbar fork. This in all the TrichoglossinaB 

 is extremely weak in front, the tracts not getting at all strongly feathered 

 till some way (in T. cincinnus j inch) from the ends of the scapular 

 fork. Each arm is much shorter (in all the forms I count about eight 

 rows of feathers to the junction), wider and more diffusely feathered 

 than in the Platycercinae, and becomes dilated and more strongly feathered 

 towards its junction with its fellow, which takes place further from 

 the tail than in the other group. The united tract is strongly feathered 

 and rather broad at first, but narrows rapidly again towards the tail. 

 Figs. 5 and 6, PL I. represent the pterylosis of Trichoglossus concinnus 

 (a bird a little larger than the " Swift Parrakeet "), and show the dif- 

 ferences between the two groups, which, if somewhat slight, are neverthe- 

 less easily appreciable after a little study, and are as well marked as any 

 others I have as yet observed in the pterylosis of this order. 



Several points in the external characters of Lathamus show that it has P. Z. S. 1879, 

 in fact no particular relationship to the TrichoglossinaB. The shape of P- ^ 2 - 

 the upper mandible, with a small but distinct tooth, is obviously (see 

 fig. 1, p. 68) much nearer to that of Psepnotus (fig. 3) than it is to that 

 of a Lory (fig. 5). The same story is told still more plainly by its 

 maxilla, which has none of the laterally compressed, elongate, and 

 pointed form characteristic of the Lories, and which induced Sundevall 

 to divide all Parrots into two groups " Psittaci proprii " and " Psittaci 

 orthognathi," the latter including only the Lories and Nestor, and cha- 

 racterized by having the "maxilla inferior recta, angusta, altitudine 

 longior." In Lathamus the maxilla is short and deep, with a broad and 

 rounded anterior margin. These differences will be seen by a glance at 

 figures 5 and 1, representing the heads of a Trichoglossus (concinnus) and 

 of Lathamus. 



In all the TrichoglossinaB I have examined, the cere is rather narrow 

 from before backwards, the anterior margin only sinuate, and the nostrils 

 elongated and ovate, with their long axis directed forwards and inwards, 

 and so somewhat transversely to the direction of the beak (fig. 5, p. 68). 

 This is very evident in the living birds, and is also to be made out in 



* I have as yet been unable to confirm Nitzsch's observation (Pterylogr., Eng. edit, 

 p. 100) that in Lorius garrulus and L. domicella the inferior tracts are continuous over 

 the lower surface of the neck. 



F2 



