364 



trochlea was apparently longer than the fourth, and the sur- 

 face for the liallux is identically the same. The small 

 relative size of the intercotylar tuberosity at the proximal 

 extremity is likewise a feature of the Ardcidcf. The 

 specimen differs, however, from all existing genera in that 

 the three distal trochleae were evidently less nearly equal 

 in length, in the absence of a distinct tubercle on the 

 anterior surface of the proximal portion for the insertion 

 of the tibialis anticus, and the much smaller and simpler 



Fig. 75. 



I'roherodius oweni. — Anterior (A) and posterior (B) aspects of the right tarso- 

 metatarsus ; from the London Clay. ^. a, facet for the hallux. 



talon, which has no closed channels and very shallow 

 grooves. These are, however, precisely the generalized 

 characters which might be expected in an ancestral type 

 of the Ardeido', and (assuming the bone to be rightly 

 referred) indicate that Proherodins belongs to a distinct 

 family. In its relative shortness and stoutness, and also 

 in the large size and depth of the groove on the proximal 

 part of the anterior surface, the specimen comes nearer to 

 the corresponding bone of Cancroma (see Milne-Edwards, 

 ' Oiseaux Fossiles de la France,' pi. xcii. fig. 7) than to 

 that of any existing form. Made in ihe Museum, ISfJ] , 



