1042 MR. W, p. PVCRATT ox THE [Dec. 19, 



families ; Dr. Gadow would regard these as of subordinal value. 

 Which of the two views will become ultimately adopted remaius to 

 be seen. Probably the first is a sufficiently wide separation. 



Finally, — and it had almost escaped mention, — the skeleton of 

 the Pygopodes is uon-pueumatic. 



xi. Key to the Osteology ok the Pygopodes. 

 A. Skull. (Plate LXXII.) 



Holorhinal and schizognathoiis ; nares pervious ; Toiuer cloft posteriorly ; 

 basipterygoid processes absent ; laclirjnial small, feebly developed, not extending 

 downwards to join the quadrato-jugal bar; quadrate with an elongate orbital 

 process; masillo-palatine proce.sses in the form of horizontal lauiinfe, never 

 extending as i'ai" backwards as the scroll-like antero-internal border of the 

 palatine ; basitemporal plate of the parasphenoid with an inflated anterior 

 border converting the Eustachian grooves into tubes, with a median aperture 

 below the parasphenoidal rostrum ; temporal fosste more or less well developed. 

 Dentary suture of mandible tending to disappear in the adult. Angulare 

 truncated. 



A. Supraorbital grooves very deep, with a well-developed ledge ; temporal 

 fossfe wide, separated one from another superiorly by a median sagittal 

 ridge ; lachrymal more or less completely fused with the nasal ; vomer 

 grooved and laterally expanded dorsally ; with a deep median, ventral 

 keel, and with a strongly marked ventral keel in front of the parasphenoidal 

 rostrum ; Eustachian grooves never completelj' closed ; large postorbital 



and paroccipital processes Colvmbid.*. 



(Onh" one genus — Colymbiis.) 



E. Supraorbital grooves feebly developed or ab.sent ; lachrymal free, not 

 projecting posteriorly from the sides of the supraorbital margin ; vomer 

 blade-shaped ; Eustachian grooves completely closed ; postorbital and 

 paroccipital processes obsolete Podicu'edid.e. 



Key to the Genera of the Family Podicipedidae \ 

 A. Without a broad bifid, overhanging postorbital process. 



Group a. (Type P. cristahis.) With a wide and distinct temporal fossa, 

 and strongly marked cerebral prominence; postorbital region of the 

 frontal marked by a deep scar, for the temporalis muscle, the superior 

 border of which has a rough edge ; upper jaw longer than cranium. 



* I do not feel justified in attempting to form " Keys" to the species, either 

 for the skull or any other part of tlie skeleton, of the forms comprising tlie two 

 sub-families dealt with in this paper. Inasmuch as of the Colymbida; I have 

 Dnly two species, C.glacialis and C. sepUntrionalis, and these are easily recogniz- 

 able by the difference in size alone. In the Podicipedida; I have only 8 out of 

 a possible 19 species of the genus Podicipes ; only one skeleton of Mchmophorus, 

 and no bones whatever of Podt/limbiis. From what I can gather from our 

 material, the dilfereuce between the three genera recognized in the British 

 Museum Catalogue vol. ssvi. is very slight, and that between the species 

 comprising these genera is even less. The genus Podicipes seems to divide 

 itself into two groups— one of the type of P.JluviatUis, and one of the type of 

 P. crutatus. The differences upon which such separation rests concern the 

 skull only, and depend mainly upon size : the smaller species having a 

 relatively shorter and wider skidl, and ill-defined temporal fossie. 



