3i0 Mr. F. E. Beddard on the ^ 



Applying tliis test to Heliodilus, it confirms the justice of 

 Prof. Milne-Edwards's views of the affinities of that bird. In 

 the illustration of the skeleton which he gives^ it is quite 

 obvious that the proportions of the two first phalanges of the 

 third digit are those of Strix. 



The illustration of PhotodUus is not sufficiently accurate 

 to admit of a statement with regard to this point. The 

 skeleton was apparently defective. 



The principal osteological characters of the genus Strix, 

 aud which apparently distinguish it from all others, are the 

 following : — 



(1) The skull is relatively long and narrow. 



(2) The palatines are straight, nearly parallel to each other ; 



they are of approximately the same width throughout ; 

 they almost conceal the underlying maxillo-palatines, 

 which are broader from above downwards than from 

 side to side. 



(3) Tiie prefrontal processes of the ethmoid are rounded 

 bones of some width. 



(4) The interorbital region of the skull does not form a 

 thin plate anteriorly, but is of considerable width 

 from side to side. 



(5) The sternum has but one notch on either side. 



(6), la the foot the second joint of the third toe is con- 

 siderably longer than the basal joint. 

 (7) There is no bony ridge upon the tarso-metatarsus. 



On the other hand, in the Bubonidse the skeleton has the 

 following characters : — 



(1) The skull is relatively broad and short. 



(2) The palatines are curved, the hinder part of the bone 

 being much wider than the anterior region j the max- 

 illo-palatines are very broad from side to side. 



(3) The prefrontal process of the ethmoid is a thin plate*. 



* In Athene nodua and Speotyto cnuicularia tliese processes are very 

 small, and are hidden by the palatines when the skull is viewed from the 

 ventral surface. The skull is broader in these t^vo genera than in any 

 others examined by me, and the maxillo-palatines are smaller. 



I 



