612 PROCEEDINGS OF THE NATIONAL MUSEUM vol. i04 



Discussion and conclusions 



The arrangement of the thoracic and cervical derivatives of the 

 aortic arch system — especially the branches of the subclavian and 

 pectoral stem arteries — in the ostrich, rhea, and cassowary clearly 

 substantiates the view that these forms are secondarily modified and 

 are not extant species derived from holocursorial ancestors. In 

 the kiwi, on the other hand, the arrangement and supply of the 

 arteries can hardly be interpreted as having been derived from an 

 ancestor possessing the normal flight-function, although a volant 

 ancestor cannot be excluded from theh* line of evolution. 



Carotid evolution of the Colymbiformes is such as to indicate a 

 considerable advance over the bicarotid ancestral forms, and the 

 wide distribution of species of this order tends to further substantiate 

 the view that the origin of this order of birds occurred early in avian 

 history. A similar interpretation may be made for the very early 

 origin of the idwis. 



The dextral evolution of the carotids in both the Ardeidae and 

 Phoenicopteridae tends to further indicate the affinities of the 

 flamingos wdth the herons rather than with the ducks, which are 

 altogether bicarotidinae normales. In addition, the laevocarotid 

 arrangement in Balaeniceps rex is fm*ther evidence for its singular 

 position among the Ciconiiformes, and warrants its aUocation to a 

 separate family (Balaenicipidae). 



Affinities between the tinamous and gallinaceous bhds are apparent, 

 but close and recent relationships can hardly be substantiated. 

 While these two orders may well have had a very early, common 

 origin, most of the gallinaceous birds have evolved along many more 

 diverse lines of specialization than has been the case among the 

 tinamous. 



In discussing the generic relationships of the parrots, Amadon 

 (1942) has pointed out that Latham considered Prosopeia tabuensis 

 and Alisterus scapularis to be varieties of one species. This I cannot 

 properly ascribe to, since two subspecies of Prosopeia tabuensis 

 were found to have the A-2-s carotid arrangement while Alisterus 

 scapularis has the A-1 carotid arrangement and is more closely allied 

 with Polytelis, but not with Cyanoramphus or Platycercus. Both of 

 these latter genera show closer affinities with Prosopeia in that they 

 aU present the A-2-s carotid arrangement. The platycercine lack a 

 furcula, as they have class 3 clavicles (Glenny, 1954b). Among the 

 Psittacinae, it is my opinion that the A-2-s forms are more closely 

 related by way of some common ancestral form than as a result of the 

 comingled relationship with A-1 carotid types. It is suggested that 

 the species Lathamus discolor be completely reexamined to determine 



