Hyatt.] 252 [May 17, 



tarsals. This may afford a useful hint in any search for the ancestral 

 stock or primitive type of the Spheniscidse. As well as we can 

 gather from the isolated fossil data at our service, birds have grad- 

 ually coalesced both metacarpals and metatarsals that were free 

 in a primitive condition. The metacarpals appear to have run 

 together later than the metatarsals ; for nearly all birds to-day show 

 partial separation of the former, while the latter are confluent in all 

 but the Spheniscidse ; while the oldest known bird, Archceopteryx 

 macrurus, with confluent metatarsals, shows unanchylosed metacarpals, 

 as well as two unguiculate digits on the radial side of the manus, 

 a condition only elsewhere found in Struthio and Rhea. Reasoning 

 upon this, we may infer that, cceteris paribus, the existing species of 

 Spheniscidse with the broadest and most largely fenestrated metatar- 

 sus, comes nearest the original stock, from which the several genera 

 have been differentiated in the process of derivation. 



The sternum likewise is positively diagnostic. To general pygo- 

 podous features, it adds a special configuration not found outside the 

 family. The postero-external angles each send off a long slender 

 apophysis that runs backward beyond the termination of the sternal 

 body, and curves mesiad, approximating, at the end, to its fellow of 

 the opposite side. There is a deep emargination between each apo- 

 physis and the rather narrow but blunt median extremity of the 

 bone. Each one of the four families of Pygopodes shows a different 

 modification of the posterior border of the sternum; comparing 

 which, we may infer that in a very early condition, the sternum of 

 Spheniscidas extended solidly as. far as these apophyses now reach. 

 In Uria for example, which has a relatively much longer sternum 

 the posterior border is rounded and continuous, with only indications 

 of the apophyses in two small fenestras; in Colymbus, also with a 

 long sternum, the median portion is very long and broad, and sepa- 

 rated from the much shorter apophyses by a wide emargination; in 

 Podiceps, with a shorter sternum, the median portion is abrupt, with a 

 reentrance, and separated from the longer, broad and clavate apophy- 

 ses by a very narrow emargination — little more than a fenestration. 

 And in view of the fact that lengths of apophyses and of sternum 

 proper seem somewhat complementary, it would appear that these 

 long apophyses of Spheniscidas have remained in partial compensa- 

 tion for the abbreviation of the sternum that has taken place. This 

 would be the more probable, if the longest sternum, relatively, 

 should be found coexistent with the greatest fenestration of the 

 metatarsus. 



