290 Report S.A.A. Advancement of Science. 



As the palate is essenlialU differeiil in ivpe from that found in 

 the typical Theriodonts. it seems advisable to refer ifie Theriodonts 

 of this l\pe to a new order, for which 1 proi)()se the name Thero- 

 ce|)halia. The forms previously known which belong to this same 

 order are ^lurosaurus and Tctidosuchus, and ])rol)ably Tilanosuchus. 

 Lvcosaurus has, according to Seeley. a palate similar to that in Cyno- 

 gnathus. In y^Llurosaurus there are teeth on the pterygoids, and 

 probabK on the palatines and '' vomers." in Sc\lacosaunis there 

 are a few teeth on the pterygoids, but there are apparenth none on 

 the i)alalines or " vomers. " 



The shoulder girdle is known in Ictidosui'hus, and is character- 

 ised bv the absence of a distinct acromion, and b\ ihe fact that the 

 })recoracoidal foramen is entireK formed b\ the precoracoid 

 (epicoracoid). 



As the structure of the ])alate is [)robabl\ Ifss subject to varia- 

 tion than the structure of the tem]>oral arch, it will probabl\ be 

 found convenient to place (iorgonojjs in this order even though its 

 temporal region ajipears to be completeh roofed. Deuterosaurus 

 and Rhopalodon ])robabl\ also belong to the same order, but owing 

 to our ignorance of man\ important points in the structure of the 

 skulls of these genera it is impossible at |>resent to be ccrlain. 



THERIODOXTIA. 



It is unforlimate that of the tvpe genus (lalesaurus oiiK the skull 

 is known. aJid even that imperfectK. 'l"he genera (\nognathus. 

 Gomphognathus. and Microgt)mphodon are. ho\\e\er, fairh well 

 known, and it is prof)al)le that Galesaurus agrees in its general 

 skeletal characters with these other genera. Cvnognathus agrees 

 \ery closely with Gomphognathus, but the differences in the structure 

 of the molar teeth entitle them to be regarded as the t\i>es of two 

 familie.s — the Cvnognathidae and Gomphognathid^ btu the difTer- 

 ences do not seem suflficienth marked to entitle them to subordinal 

 rank. Lvco.saurus, on the otlier hand, though \er\ im[)erfectlv 

 known, mav lie the t\pe of a distinct sub-order. 



A\()M()DO\TI.\. 



The structure of the Anomodonts is much more thoroughlv 

 known than that of almost anv other Triassic rej)tile. Dic^xnodon 

 and Oudenodon agree sufficiently closelv to be jilaced in one famiU. 

 but Lvstrosaurus is ])robaf)lv entitled to f)e made the t\pe oi a dis- 

 tinct family, as is also probably Cistecephalus. I,\strosaurus differs 

 from Dicvnodon in having (hstinct postfrontal and postorbital l)ones. 

 in having a distinct preparietal bone, and in the al)sence of a 

 cleithnmi. The Endothiodonts also form a distinct famib. Six 

 distinct genera of Endothiodonts are now known, and though con- 

 siderable difTerences exist between them, the occurrence of inter- 

 methate forms renders it advisable to group together all tho.se 

 Anomodonts with molar teeth on the maxillarv and deniarx bones to 

 form the familv Endothiodontidae. 



