2$2 NEW YORK STATE MUSEUM 



duced in size, and the pygostyle is small, long, and pointed. But the 

 sternum agrees exactly with that bone in the California partridge 

 (L. californicus), while the pelvis and the rest of the 

 axial skeleton come very close indeed, though I dare say a minute 

 description would bring out a number of minor osteological dif- 

 ferences. 



On the other hand, such a species as T h a u m a 1 e a p i c t a , 

 approaches, osteologically, the members of the representatives 

 of the Xorth American genus Bonasa. This is best seen in 

 the skulls of the two types mentioned, and is extended to 

 the sterna, and other parts of the skeleton. In their 

 pelves, however, the resemblances are not so close, that of Bonasa 

 being broader and more grouselike, while in Rollulus it approaches 

 the true fowls in its morphology. Still the pelvis of Bonasa comes 

 nearer the pelvis of Rollulus than it does the pelvis of either Tym- 

 panuchus or Pediocaetes [pi. 6, fig. 26-29] and this is an interesting 

 fact. I very much regret that I have not at hand the skele- 

 tons of a great many other Gallinae for intercomparison, for 

 without them the true meaning of these resemblances and 

 differences can not be explained by me. I am therefore com- 

 pelled to suspend judgment in the premises, and they can 

 not be understood until the entire suborder is examined, not 

 only osteologically. but anatomically otherwise. I have seen 

 enough of the morphology of the Xorth American Tetraonidae to 

 convince me that the interrelationships among the various genera 

 vary greatly. That is, for example, the difference, structurally, be- 

 tween Tympanuchus and Pediocaetes is infinitely less, than a similar 

 comparison reveals to exist between either of these genera, and 

 Centrocercus. Again, osteologically, Bonasa approaches the Perdi- 

 cinae, and is considerably further removed, when thus compared, 

 from Centrocercus than is either Tympanuchus or Pediocaetes. 

 Then C e ntrocercus urophasianus not only resembles 

 a few of the true Phasianidae in the elongation of the feathers of 

 its tail, as its specific name indicates, but in certain characters it ap- 

 proaches them osteologically, so that in this particular, and in some 

 other respects, Centrocercus is nearer some of the Phasianidae than 

 it is to Pediocaetes, and still more so when we come to compare it 

 with Bonasa. 



