AISID OTHER EXTINCT BIRDS OF MATJEITIUS. 285 



enormous development of their jaws, a feature which makes them unlike any other 

 Parrots. Moreover, these considerations enable us to discuss with some amount of 

 certainty, or at least probability, the only other bone of a Parrot which has been found 

 in the Mare aux Songes : to wit, the sternum. 



The sternum is preserved only in its anterior part. The large spina externa agrees 

 in shape and direction exactly with that of JV. rodericanus, and excludes any possibility 

 of this sternum belonging to any other bird but a Parrot. The ventral margin of part 

 of the keel is broken off, but the line of the m. subclavius is well marked ; the whole 

 of the anterior margin and the articulating facets for several ribs are likewise uninjured. 

 This sternum appears at the first glance undoubtedly far too small for L. mauritianus, 

 but if we measure its width in level of the first pair of ribs, the height of the keel, the 

 distance from the middle of the anterior margin of the sternum (at the place where the 

 spina interna would be if it existed in these Parrots) to the highest curve of the keel, 

 or to the muscular ridge at the point (Plate XXXIII. fig. 7, S), we find that this 

 fragmentary sternum by all its dimensions indicates that it belonged to a larger bird 

 than N. rodericanus. In fact, the size of this sternum would fit one of the smaller 

 specimens of L. maicritianus ; and this is corroborated by the following calculation, 

 which gives a result which we should not have expected : — Average of total length of 

 hind limb of N. rodericanus (130) : width of its sternum (20)=Length of hind limb ot 

 smallest L. mauritianus {1%1) : width of its sternum would be 27'8, while our single 

 sternum from Mauritius actually measures 27'5 mm. across! 



There can be no doubt that the extinct IMauritian Parrot was a larger but other- 

 wise nearly allied form of the Parrot from Rodriguez ; it is, however, questionable 

 whether both might not be included in the same genus Necropsittacus, for while we 

 know from old drawings that the Mauritian form had a sort of ornamental crest, we 

 know nothing to the contrary of N. rodericanus. 



2. AsTUK ALPHONSi, sp. nov. (Plate XXXIII. figs. 9, 10.) 



Amongst numerous Asturine remains a pair of tibiae, a pair of metatarsals, and the 

 metacarpals of the left side are probably referable to one individual bird of prey. 

 The two metatarsals, with a length of 81 mm., agree perfectly with that figured by 

 M. Milne-Edwards (plate 33. fig. 2). He rightly referred them to the genus Astur, and 

 remarked that they belonged to a bird which was undescribed and unknown, unless it 

 was identical with A. melanoleucus from the Cape of Good Hope. We have been able 

 to measure the length of the tarso-metatarsus of an A. melanoleucus, and have found 

 that it agrees in this respect with the two bones in question. It would therefore seem 

 reasonable to assign these bones to A. melanoleucus, unless the absence of this South- 

 African species from Madagascar, and the numerous instances of insular forms or species 

 of Hawks, be deemed arguments sufficiently strong to distinguish the bird to which 

 these bones belonged as Astur alphonsi. 



