1874.] ANATOMY OF THE PARROTS. 597 



As far as the major division is concerned, the facts brought forward 

 in the present communication suggest a different arrangement, as 

 shown above, which would approximately distribute these subfamilies 

 thus : — 



Family I. Family II. 



Stringopidce. Arincc. 



Lorince. Platycercince (in part). 



Cacatuince. Psittacinee (in part). 



Platycercince (in part). Nestorince. 



Psittacinee (in part). 



The generally received families Platycercinse and Psittacinse are 

 avowedly rather incongruous mixtures. Mr. Gould, with acute per- 

 ception, was able to differentiate Aprosmictus from Platycercus, 

 chiefly by its habits of life ; and anatomical considerations show that 

 Dr. Finsch's attempt to reabsorb it in the older genus is a retrograde 

 step. If Dr. Meyer is right in stating that the sexes in Eclectus are 

 of different colours, its relations to Aprosmictus may be very intimate. 



It may at first sight seem very heretical to remove Lathamus from 

 the Lorinae, the brush-tongue being considered characteristic of that 

 subfamily. To the unbiased student, however, the brush-tongue 

 is a character not more important than several of those that have 

 been above considered. It is only an excessive development of the 

 papillae which are always present on the lingual surface, and is seen 

 in a slightly different form in the lion and other Felidae. The cha- 

 racter of the papillae is somewhat different in Lathamus from what 

 it is in Lorius, they being blunter and shorter in the former genus 

 than in the latter. 



The totally different geographical distribution of Paleeornis and the 

 true Arinae is quite opposed to Dr. Finsch's proposition that Conurus 

 and Brotogerys should be the neighbours of the Palaearctic genus. 



Brotogerys entirely agrees in structure with Chrysotis and 

 Pionus, differing greatly from Conurus ; whilst in itself Conurus, as 

 generally received, embodies the red-tailed species, with the fourth 

 primary not acuminate, and the green-tailed species, with an acumi- 

 nate fourth primary. In the former section (Pyrrhurd) the 

 ambiens muscle is quite lost, whilst in the latter {Conurus) it is 

 always well developed. 



Prof. Huxley is not the only naturalist who has been puzzled by 

 the geographical distribution of Psittacula. This genus in its wide 

 sense, however, is broken up into far separated genera — the Old- 

 World Psittinus and Agapornis differing entirely from the New- 

 World Psittacula, Agapornis being the homologue, as it may be 

 termed, in the normal-carotid Parrots of Psittacula in the other 

 group, whilst Psittinus is a less-differentiated genus of the former 

 division. 



Nestor no doubt stands rather isolated ; but possessing the ambiens 

 muscle, as well as all the other characters of Psittacus and the true 

 Arinae, it must be classed with them. 



My subfamily Pyrrhurinae seems a mixture ; and all I can say in 



