250 ON THE OSTEOLOGY OF CRANES, RAILS, ETC. [Mar. 20, 



received from Pahang in October 1889 (see P. Z. S. 1890, p. 592, 

 pi. xlix.), which had died on tbe 27th June, 1892, so that the 

 acquisition of the present individual would be most desirable. 



A communication was read from Dr. E. W. Shufeldt, C.M.Z.S., 

 containing an account of the osteology of certain Cranes, Rails, 

 and their allies, with remarks upon their affinities. 



After a review of the opinions of previous writers upon this 

 subject, Dr. Shufeldt concluded with the following statement of 

 bis views on the taxonomy of the North-American Paludicoline 

 birds : — 



" So far as this suborder — the Paludicolce — of the United States is 

 concerned, it is primarily divided into two main stems. The first 

 of these is represented by the Cranes and Courlans ; while the second 

 contains all the Bails proper, or such generic groups as Rallus, 

 CreDB, Porzana, lonornis, Gallinula, and Fulica. 



" Structurally the Courlans possess a greater number of Gruine 

 than they do of Balline characters, and these characters are of 

 equal importance and weight. But their generic characters are by 

 no means always typical, and the differences seen are frequently of 

 a degree that distinguish families among birds rather than genera. 

 This being true, the fact settles the position of the Courlans in 

 the system as a family — the Aramidce, of the Crane-group. The 

 species which has been osteologically described here — Aramus 

 giganteus — is the only representative known to our avifauna, and 

 it is a most perfect link connecting the Cranes with the typical 

 Rcdlidce. 



" The Cranes must then constitute a family of themselves, and 

 the Gruidce has long been created to contain them. But the 

 osteological and other morphological characters held in common by 

 the Gruidce and the Aramidce are of a rank, when we come to 

 compare them wdth the corresponding ones in the llallidce, which 

 proclaim them to be higher than those commonly employed to 

 define family lines, and yet not of a rank entitling them to sub- 

 ordiiial distinction. To express this relationship a superfamily 

 Grvoidea may be made to contain all the true Gruidce, the 

 Aramidce, and perhaps the Psophidce, from another quarter of the 

 world. Another group to contain all the true Bails may be 

 created, and designated as the superfamily Ralhidca. A scheme 

 as follows would show these divisions as expressed for the forms 

 we have had under consideration : — 



Suborder. Superfamilies. Families. Genera. 



( q. ruoidea J Gruidse ... Grits. 



1 Aramidse ... Ar amies. 



f Rallus. 

 Crex. 

 Porzana. 

 Ionomis. 

 Gallinula. 

 { { Fulica, 



PALUDICOLCE .( 



Kalloidea Rallidse ... < 



