290 SLR E. NEWTON AND DE. GADOW ON THE DODO 



The single left coracoid agrees in all the features of its dorsal or scapular half with 

 A. meqacephala, but its ventral or sternal half differs considerably, first by the much 

 more strongly marked ridge of the linea intermuscularis on its ventral surface, secondly 

 by the almost straight instead of inwardly curved margin between the processus 

 lateralis and the lateral distal corner of the sternal articulation, thirdly by a very low 

 but very distinct and sharp ridge which arises from the median margin of the coracoid 

 a little above its median articulating corner. This roughness or prominent ridge is 

 entirely absent in A. megacephala and in all other Herons which we have been able 

 to examine, but at least a slight indication of it occurs in an individually varying 

 degree in Nycticorax and in Botaurits. That this coracoid bone belonged, however, 

 to an Ardeine bird is clearly indicated by its whole configuration, notably by the 

 shape and position of the precoracoid process, the various articulating facets at the 

 dorsal end, and the prominent lip on the visceral or internal surface of the median 

 portion of the sternal articulating facet. 



7. Sakcidioknis maukitianus, sp. nov. (Plate XXXIV. figs. 9, 10.) 

 The most tangible proof of the former existence of this form in the island of 

 Mauritius rests at present upon one specimen of left metacarpal bones. However, 

 this solitary specimen is sufficiently well preserved to show its aflfinities by various 

 well-marked characters. It agrees in size with the corresponding combination of 

 bones of Bernicla brenta, while it is considerably smaller than those of the common 

 domesticated Anser cinereus, and too large for the Madagascar and East-African 

 Sarcidiornis africanus and the Indian S. melanonotus. 



The generically diagnostic feature of the bones of the middle hand of Sarcidiornis 

 is the very prominent process which arises from the side of the first metacarpal, 

 proximally from the articulating facet of the pollex. The apex of this process is 

 covered in Sarcidiornis by a skin, which, although thickened and bare of feathers, is 

 not transformed into a horny callosity or spur. The same peculiar feature exists in 

 the bone before us ; the apex is rough and irregularly shaped, and since this part of 

 the process never serves for the origin or insertion of muscles or tendons, its roughness 

 plainly indicates the same purpose as that of the Madagascar species of Sarcidiornis, 

 namely its use as a fighting knuckle, although in an either arrested or incipient state. 

 Such a weapon, furnished with a sharp and long horny spur with bony core, carried 

 by the first metacarpal bone, is fully developed in Chauna. This American genus 



