NO. II CLASSIFICATION, BIRDS OF THE WORLD — WETMORE 1 7 



confusing I prefer the conservative course. In the remarks that 

 follow I shall discuss only a few matters on which I have more or less 

 concrete ideas. 



In the superfamily Furnarioidea, von Ihering (1915, pp. 145-153) 

 united the Furnariidae and the Dendrocolaptidae, since he was unable 

 to separate two groups on the basis of the form of the posterior border 

 of the nasal opening. The variation that he showed seems valid, but 

 there are numbers of other points of supposed difference in the osteol- 

 ogy and other structural details, so that his suggestion is far from 

 established. Pycraft (1906, pp. 133-159), though seemingly uncer- 

 tain in the beginning, finally retained the two families. It may prove 

 that some genera are wrongfully allocated at present between the two 

 groups, so that their shift, when we have sufficient information, will 

 clear our understanding. 



In the Tyrannoidea, the family Oxyruncidae is known through ex- 

 ternal characters that seem to warrant separation. If the sharpbills 

 have other affinities it is doubtful that these are within the family 

 Tyrannidae, where some have placed them. 



In the family Cracticidae, recognized by Australian ornithologists, 

 the skull, according to Pycraft (1907, pp. 355-365), mainly from 

 examination of Gymnorhina, has the zygomatic process of the 

 squamosal bifurcate, the postorbital process large, the orbitosphenoid 

 ossified, the interorbital septum with a single opening, the prefrontals 

 unusually large, and the form of the palate peculiar. In his phylo- 

 genetic tree Pycraft places the group on a common stem with the 

 Artamidae, and not far from the Paradisaeidae. His account is diffi- 

 cult to summarize in concrete form. 



The family Grallinidae is likewise recognized officially by Austra- 

 lian ornithologists for Grallina cyanoleuca, the magpie-lark. The 

 principal study of the osteology is that of Shufeldt (1923, pp. 16-19, 

 pi. 6) but his account is mainly descriptive and without definite con- 

 clusion. Amadon (1950, pp. 123-127) has placed Corcorax and 

 Struthidea here tentatively, though this seems subject to further proof. 



Stonor (1937, pp. 475-490) has outlined excellent reasons for recog- 

 nition of the Ptilonorhynchidae, finding that they differ from Para- 

 disaeidae, with which they have been united, in having an apterium 

 in the center of the dorsal feather tract, the tip of the vomer convex, 

 larger, more developed maxillo-palatines, the margin of the palatines 

 angular, smaller ectethmoid, much larger lachrymal, and slender, 

 greatly elongated orbital ramus of the quadrate. The genera Loria 

 and Loboparadisea, usually included here, he transfers to the Para- 

 disaeidae. His conclusion is that "the Ptilonorhynchidae constitute 



