566 PTERYLOGBAPHY OF GOAT-SVCKERS AND OWLS— CLARE, vol.xvii. 



wider tlian the lower, although there is no true iuuer branch given off. 

 The upper cervical tract is very narrow, while tbe humerals are nar- 

 rower than in any other genus and the parapterum is weak. The 

 femorals are strong, but very diffuse, and are scattered over most of the 

 femur. The feet are not feathered quite to the toes, but the latter are 

 very hairy. The sternal tract is fused with the ventral, not only at its 

 origin near the furcula, but also at the other end of the breast, so that 

 the tracts are really one; very broad on the sternum, and contaiuing a 

 longitudinal apterium, and becoming abruptly narrow on the belly. 

 The hypopterum is very strongly marked, and the hook connecting it 

 with the sternal tract is composed of larger feathers, and they are much 

 more numerous than in the other owls. Indeed, the whole breast is 

 much more thickly feathered than in AHo. Another remarkable 

 peculiarity is the formula for the comparative lengths of the rectriees. 

 In all the other owls the middle pair of tail-feathers is the longest and 

 the external pair shortest, so that the formula is.l, 2, .3, 4, 5, G. In Strix, 

 however, this is exactly reversed, the outer pair being the longest and 

 the fonnula reading 6, 5, 4, 3, 2, 1. Tliere are fifteen secondaries and 

 the primaries rank as follows : 9, 8, 10, 7, 6, 5, 4, 3, 2, 1, 11. 



Specimens examined. 



Nitzsch says of Hyhris flammea : " Twenty-four remiges, the three first 

 about equally long, but the second is really the longest." This formula 

 is the same as what I have just given. There are, however, fifteen 

 secondaries, instead of fourteen, as Nitzsch says. 



COMPARISON OF THE TWO GROUPS. 



Before entering on a detailed comparison of the two groups which 

 we have been examining, it will set some of the facts more clearly 

 before us if we arrange them in tabular form. 



Comparison of the groups. 



