4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 39 



by the possession of biconcave vertebrae. I have given a somewhat 

 more detailed discussion of this matter elsewhere (Wetmore, 1956, 

 p. 2). 



The penguins. — The question of the weight to be given the peculiari- 

 ties of uniform pterylosis, extreme specialization of the wing as a 

 flipper for submarine progression, and incomplete fusion in the meta- 

 tarsal elements, as well as such other details as erect posture in stand- 

 ing and walking and the anatomical adjustments involved, found in the 

 penguins, is one that has merited careful review. It seems reasonable 

 after this examination to retain the Impennes as a superorder, at least 

 until we have further evidence through fossils as to their line of 

 evolution. It is necessary, however, to remove the fossil family 

 Cladornithidae, since Simpson (1946, pp. 24-25) has found that the 

 two genera Cruschedula and Cladornis placed in this family have no 

 apparent relationship to the Sphenisci formes. These two, described 

 by Ameghino from the Deseado formation of Patagonia, now placed 

 in the Oligocene, are based on fragmentary, considerably flattened 

 metatarsi. The descriptions and figures that have appeared thus far are 

 not sufficiently definite to demonstrate characters of importance in 

 classification. However, from what we now know these ancient birds 

 cannot be considered as ancestral penguins of terrestrial habit, as has 

 been supposed. The only suggestion that has come to me is that pos- 

 sibly they may belong in the order Pelecani formes, in which I have 

 placed the family tentatively in a suborder Cladornithes (see p. 25). 



The Neognathae. — One important result of recent studies has been 

 the allocation to the Neognathae of the orders formerly separated as 

 the Palaeognathae. For years I have felt that recognition of the 

 Palaeognathae, as a separate group apart from other birds, on the 

 basis of a supposed peculiarity in the palate, stood on flimsy ground. 

 The studies of McDowell (1948, pp. 520-549) demonstrate that the 

 structure of the palaeognathous palate, in which the palatine and 

 pterygoid bones are articulated by a squamous suture, is variable from 

 order to order and that in fact the details of this union differ con- 

 siderably in the several groups. For example, McDowell points out 

 that in Dromiceius the palatine and pterygoid are not in contact, while 

 in a number of families placed in the Neognathae, as in the Anatidae, 

 to name only one, the two bones are in articulation. As there is no 

 clear-cut separation, the former Palaeognathae must be combined with 

 the Neognathae. 



The supposed bird Caenagnathus collinsi described by R. M. Stern- 

 berg (1940, p. 81) from the Belly River series of beds of Upper 

 Cretaceous age in Alberta has been carried tentatively in our avian 



