596 MR. A. H. GARKOD ON THE [Nov. 3, 



the specimen presented by Mr. Murdoch to the Society, I have twice 

 had the opportunity of dissecting Stringops habroptilus. As a 

 Parrot it is not so strikingly peculiar as many seem to think. Its 

 wings are useless, and the carina sterni is correspondingly reduced, it 

 is true ; but as points of classiflcational importance, I regard these 

 as insignificant. The points of special anatomical interest which it 

 does possess, however, are particularly instructive. The proximal 

 ends of the incomplete furcula are well developed, so much so that 

 it might at first sight seem that their symphysial ends are only lost 

 in correlation with the excessive reduction of the powers of flight ; 

 though this is probably not the case, because the allied similarly 

 modified genera Euphema &c. do not keep to the ground. 



Further, in the Society's specimen above mentioned, though the 

 ambiens muscle did not cross the knee, yet its fleshy belly was well 

 differentiated on both sides, its thin tendon being lost over the cap- 

 sule of the joint. In the College of Surgeons' specimen, however, 

 this muscle was entirely absent in the only knee which was in a fit 

 state for dissection, the other being much shot. It is only in the 

 genus (Edicnemus that I have elsewhere found a similar partial loss 

 of the ambiens*. The partial development of this muscle in this 

 particular instance shows that the tendency to lose it is not of great 

 antiquity ; and it is to be noted that there is no other Parrot with 

 normal carotids in which any trace of an ambiens is to be found. 

 These considerations suggest, what may perhaps be the case, as is 

 suggested by the peculiarities of their geographical distribution, that 

 Agapornis may be the representative among the normal-carotid 

 Parrots of the Platycercine branch from the Arinae, whilst the 

 Stringopinae proper (including Geopsittacus, Melopsittacus, and 

 Euphema) are more direct continuations of the main stem, as indi- 

 cated by the dotted portions of the diagrams (Plates LXX. and 

 LXXI.), Stringops itself being the nearest living representative of 

 the common ancestor of the whole suborder. 



Further, it may be worth while taking a glance at some of the most 

 important changes which my classification would necessitate. Taking 

 Mr. Sclater's ' Revised List of Vertebrated Animals in the Society's 

 Gardens' as a good representative of current opinion, the Order 

 Psittaci is there divided into two families and seven subfamilies, 

 thus : — 



Order PSITTACI. 



Family I. StringotidjE. 

 Family II. Psittacide. 



Subfamilv 1. Cacatuine. 



2. Arine. 



3. Platycercine 



„ 2. Arine. 



» 



„ 4. PsittacinjE. 

 „ 5. Lorine. 

 „ 6. Nestorine. 



* Vide P. Z. S. 1873, p. 640. 



