1907.] xjo the osteology of birds. 375 



capture of small prey when on the wing. In the matter of the 

 deltoideus muscles the Swallows do not diiFer greatly from 

 the Muscicapine forms, but in their pterylosis they differ from 

 all other Passeres, in that the pteryla spincdis is forked. This 

 condition, however, may well have been derived from the breaking 

 up of the lower segments of an elliptical tract enclosing a space, 

 such as is seen in the Tyrannida?, or the Muscicapidje for 

 example. 



We may pass now to what I have regarded as the Laniine 

 forms. These seem to be divisible into two groups including, 

 among others, the following families of Dr. Sharpe's Hand-list: — 

 Laniidfe, Prionopidse, Artamidfe, Vangidfe, and ^I'ocharidpe. The 

 Dicruridae must also be included here, for they show, on the one 

 hand, affinities with the Malaconotidse, and on the other with the 

 Prionopidte. But many of these Families, as at present constituted, 

 contain genera that cannot be allowed to remain there. While 

 genei'a now included in other Families, outside this Group, will 

 probably have to be transferred thereto. 



We shall probably be near the truth in regarding the Laniidse, 

 Prionopidse, and Malaconotidfe as representing so many branches 

 of a common stem. Of this stem the Gymnorhine group must 

 apparently be regarded as a sister branch, similarly splitting 

 up into three branches — the Gymnoi-hinpe, Artaminfe, and 

 Vangiinje (if these two can really be separated), the Vireolanidse, 

 and probably the Paradiseidte. 



This Gymnorhine group may be distinguished by many 

 characters, the most striking of which are afforded by the skull — 

 the peculiar form of the palate, and the long postorbital processes 

 already described in the earlier pages of this paper. 



From this Gymnorhine branch it is probable the Paradiseidse 

 are derived. This latter Family, by the way, as it now stands, 

 contains many forms which must be placed elsewhere. The 

 Bower Birds (Ptilonorhynchida?), for example, do not seera to 

 possess any claim to be included here. 



The distinctness of the Gymnorhine group has been ignored by 

 many later workers, but there seems good reason to believe that 

 W. K. Parker was justified in proposing therefore the term 

 " Austro-coraces." 



As touching the Yireonida?. This Family is regarded apparently, 

 by some as allied to the Sylviidae, by others to the Laniidse. 

 The latter view is certainly the more correct. But so far as my 

 material goes, everything points to a near relationship to the 

 Muscicapidse. Of the six genera, however, usually included in 

 this Family two at least must be removed ; for Cyclorhis would 

 seem to be very nearly allied to the Laniidse, while Vireolanius is 

 almost certainly closely related to the Artamidse. This genus is 

 probably an early offshoot of the branch which eventually gave 

 rise to the Artamidje, and may therefore either be included in 

 this Family, or be allowed to rank as a separate Family ; better 



