694 DR. p. CHALMERS MITCHELL ON THE 



second group also with small brains, with the squamosal spur 

 very highly developed, frequently reaching the postfrontal, and 

 with the latter simple, but in Tantalus showing the beginning 

 of cleavage. Cancroma and the other Herons form a third 

 group, characterized by larger brains, with the squamosal spur 

 reduced as in the first group but with the postfrontal split into 

 two separate processes of which the anterior tends to become the 

 more important. 



Mandible. — The fusion of the component parts is so complete 

 that I could not see any trace of sutures. The most notable 

 feature is the absence of the projecting spur of the angle, an 

 absence which Balceniceps shares with Scopus, all the Storks and 

 Canc)'07na, in all of which the end of the mandible behind the 

 ai'ticular cavities for the quadrate is as if abruptly sawn ofi', 

 while in Ardea and typical Herons it is produced backwards as 

 a long rounded spur. 



Vertebral column. — As Parker (33) and Gadow (16, p. 76) have 

 pointed out, there are 17 cervical vertebra? in Baloiniceps, 16 in 

 Scopus, 17 or 18 in Storks, and 18 to 20 in Herons. The carotid 

 canal is complete in all this group of birds. The individual 

 vertebra? are very much shoi'ter antero-posteriorly in Balceniceps 

 than in the Herons and Scopus ; the Storks are intermediate 

 between Balamiceps and Herons in this respect, which no doubt 

 is purely adaptive, but it suggests at least that Balceniceps is 

 not very closely related with Herons. Parker states that there 

 are no foramina for the vertebral arteries on the sides of 

 the atlas in Balcenice2)s and in its allies. They are certainly 

 absent in Balceniceps, but present, although small and limited to 

 the anterior lateral part of the atlas, in Ardea, Cancroma and 

 Scojnbs, and present and large in Storks. 



The thoracic vertebi-ie have no hfemapophyses in Balcenicejns, 

 Cancroma, Ardea and Xenorhynchus, but there is a very small 

 unpaired process in Tantalus. 



Sternum. — The posterior lateral processes are very long in 

 Bcdcenicejjs, projecting well behind the metasternum ; in Scopus, 

 Herons and Storks, they are short, not projecting behind the 

 metasternum. The notch separating the posterior lateral process 

 from the metasternum is rounded in Bcdcenicejys, Scojrus and 

 Storks, angular in Arcleco a-nd Cancroma. Parker figured a small 

 posterior intermediate process, and compai^ed it with that of the 

 Ibis, but it was completely absent in the skeleton I examined, as 

 well as in Herons, Storks and Scopus. 



The internal spine of the sternum is absent in Balcuniceps as in 

 Scopus, storks and Herons, but the external spine is also absent 

 in Balceniceps, small in Sco2nos, small or absent in Storks (l)est 

 developed in Tcciitcdus), large and proiiiinent in Herons including 

 Cancromct. 



Furcula. — In Balcenicep)s the clavicles are joined to form a 

 very short-stemmed " Y," the stem of the Y as well as the distal 

 ends of the diverging arms being strongly anchylosed to the 



