Osteology of Aramus scolopaceus. 43 



tlie three dorsal vertebrae which is precisely like that in the 

 last-mentioued species : the vertebrae are united ventrally 

 but not dorsally. lu Balearica we again find an approxima- 

 tion towards the conditions observable in Aramus. Three 

 dorsal vertebrae, which are, as in the other types, the first, 

 second, and third of the series, are completely fused below ; 

 it is true that the fusion above is by no means so complete 

 as in Aramus, but there are considerable bridges of bony 

 matter uniting the spinous processes. The difference in these 

 points between Aramus and Balearica is indeed not nearly 

 so great as between either of these genera and Grus or 

 Tetrapteryx. Aramus is only a short stage in front of 

 Balearica ; and both genera, it will be observed, must be 

 regarded as comparatively specialized Cranes, since it cannot 

 be doubted that the primitive arrangement is for the vertebrae 

 to be separate and not fused at all. Such a point of resem- 

 blance, however, clearly need not be considered as one of 

 affinity ; no one can doubt that a more complete fusion 

 between certain vertebrae is a specialization which might 

 readily have been acquired several times independently. 



With regard to this fusion between separate vertebrae of 

 the doi'sal series, I may observe that Psophia shews an advance 

 upon Aramus and Balearica : in the former genus the fusion 

 is so complete that even the intervertebral nerve-foramina 

 are almost obliterated ; those of Aramus are not nearly so 

 much so. 



There are some further minutiae of structure of the last 

 cervical and the first dorsal vertebrae upon which it is 

 requisite to insist, inasmuch as they offer points for dis- 

 criminating Aramus from its allies among the Cranes. 



In this genus the last two cervicals have, as has been 

 stated, a single median hypapophysis. In addition to this 

 process, which is strongly developed, there are two small 

 lateral hypapophyses, one on either side of the median 

 downgrowth ; these two processes approach a little nearer 

 to the ventral median line on the last than on the penulti- 

 mate cervical vertebra. The first dorsal vertebra of Aramus 

 is marked off" from the cervicals by the fact that the two 

 lateral processes cease to arise independently from the 



