700 DR. p. CHALMERS MITCHELL OX THE 



Gcmcronia and Bcckeniceps. He selected the Heron as the type- 

 form as " it is best known, and has the characters of the family 

 moderately, but markedly developed." In his detailed examin- 

 ation, he is quite as concerned to prove that Balceniceps is like 

 jScopus or like a Stork as like a Heron. He came to the very 

 definite conclusion that Balceniceps Avas not a Pelican, but was 

 Ardeine, in his broad sense of the word. Quite certainly he did 

 not prove it to be a Heron as opposed to a Stork ; and in this 

 I entirely agree, after having gone through his own observations 

 with the skeletons before me, and having made some further 

 oompaiisons myself. 



C G. Giebel (21) in 1873 re-opened the matter, chiefly after a 

 full examination of the pterylosis and external characters. He 

 corrected Gould and Reinhardt (not quoting the latter, howevei') 

 as to the pectination, describing it carefully and showing that it 

 was present in IJalcevJceps in much the same form as in Scojyus. 

 He showed also that these birds agreed in pterylosis, where 

 they diftered from Cancroma, and came to the conclusion that 

 Balceniceps was much more closely allied with Scojras than with 

 Canc7^oincc. 



F. E. Beddard in a special memoir and in his treatise on 

 the anatomy of birds (3 and 4) assumed that Parker's osteo- 

 logical researches had proved Balceniceps to be a Heron, not 

 a Stork, overlooked the evidence supplied by Reinhardt and 

 Giebel as to the Stork-like pterylosis of Bcdceniceps and Scopus, 

 and although he referred to Giebel's paper, did not note that 

 Giebel showed Balceniceps and Scopus to be alike in the pectin- 

 ation of the claw. He placed Balceniceps definitely among the 

 Herons as opposed to the Storks, I'elying apparent!}^ chiefly 

 on the syrinx, I have already stated {sup)raj p. 651) that his 

 argument cannot support his conclusion. Possibly it may show 

 that Balceniceps is not a Stork ; it does not show that it is 

 a Heron. 



H. Gadow (16) placed both Balceniceps and Scojms imder the 

 Ardese as opposed to the Ciconife, but placed Balainiceps merely 

 as a sub-family of the Herons, distinguished from the true 

 Herons by the abseiice of pectination and various minor characters. 

 He Avas misled, I think, b}^ the literature, and does not claim to 

 have made independent observations. 



Those who have followed the old and new facts regarding 

 Balceniceps that I have been able to bring together must be 

 impressed by the number of characters in which Scopits and 

 Bcdceniceps agree, and by the much greater number of points in 

 which these two birds agree with Storks than with Herons. If 

 we adopt the method made familiar by many of those who have 

 written most copiously on the anatomy of birds, take characters 

 on their face value, and regard those birds as most nearly allied 

 which have the greatest number of characters in common, then 

 the sub-order Ardese will contain the single family Ardeidfe, and 

 the adjacent sub-order Ciconiaj will comprise the families Scopidse, 



