ON FOSSIL REPTILES. 619 



The various species of Bee-eaters, when attacking butter- 

 flies, usually choose members of the Papilioniclae and 

 Pieridse. 



The Asilid diptera are formidable enemies to all other 

 living insects. They are, however, present in large 

 numbers only in the northern country, and then only for 

 a limited period of the year. 



In Ceylon a resemblance to the genera Danais and Euplcea 

 is doubtfully of value ; in fact, in the neighbourhood of 

 Wood-Swallows it is a distinct danger. 



The mimetic females of Papilio polytes are not obviously 

 protected by their mimicry, and as a whole probably 

 sutler about as much, or as little, from the attacks 

 of birds as any other fast-flying butterfly in Ceylon. 



40. On the South-African Pseudosuchian Euparheria and 

 Allied Genera. By K. Broom, M.D., D.Sc, C.M.Z.S. 



[Received and Read May 20, 1913.] 



(Plates LXXY.-LXXIX.*) 



Index. ^ Page 



XluparJceria capensis Broom 619 



Ornithos^ichus woodtvardi Newton 624 



Ornithosuchus taylori, sp. n 625 



Serpetosuchus granti Newton 626 



Jirowniella africana, gen. et sp. n 627 



Jfesositc/i«s ftrowwi Watson 627 



Scleromochlus taylori Suiith-Woodward 629 



Affinities of the Pseudosucliia 629 



EuPARKERiA CAPENsis Broom. (Pls. LXXY., LXXYI., 

 LXXVIII., LXXIX. figs. 1-8, 17-20, 22, 24.) 



In the collection of Mr. Alfred Brown of Aliwal North are 

 a number of imperfect skeletons of a small Rhynchocephaloid 

 reptile of very great interest, which he discovered in Upper 

 Triassic beds about four years ago. Mr. D. M. S. Watson examined 

 them a year ago, but unfortunately he mistakenly regarded them 

 as belonging to the little acrodont-toothed reptile named by 

 him Mesosuchus broicni. Though the two forms were met with 

 together and are of about the same size, they certainly belong to 

 different families, if not to different suborders. In a short note 

 commmunicated to the ' Records of the Albany Museum ' (6), I 

 have called attention to the confusion and proposed the name 

 Eitparkeria cajjensis for the Thecodont form. 



As the type, I take a specimen showing a beautiful skull 

 with most of the skeleton. Though this specimen was in the 



* For explanation of the Plates see p. 633. 



