28 HUMERI OF TA8MANIAN LABYRINTHODONTS. 



series of the Trias."* They are coi-rehitcd by him approxi- 

 mately Avith the Ha^vkesbury l)eds and Xarrabeeii series of 

 New South Wales, and the Burrum coal fields or Mr. R. 

 L. Jack's Lower Trias-Jura of Queensland. We find ii 

 dilflcult}' in naming- more distant equivalents of these 

 sandstones. The few fossils found in them and named 

 above are consistent with an Upper Permian age. Acrolepis 

 IS a well-known Upper Carljoniferous and Permian fish ; 

 but, so far, we must confess the materials do not exist for 

 ])lacing the beds with any degree of confidence on any 

 distinct horizon in the Gondwana system of India or the 

 Karoo strata of South Africa, though they evidently 

 belonged to the ancient Gondwana land represented by 

 those systems. In a letter received this year from Pro- 

 fessor xA-malitzky, he refers to his recent discover^' of 

 Pareiosaurus, Glossopteris with its rhizome vertebraria, 

 Ta^niopteris, &c. in the Ui)per Permian of the North 

 Dwina, Russia : and we are not yet convinced that an 

 Upper Permian age for the Hol)art sandstones is definitely 

 excluded. Be this as it may, the Upper Permian and 

 Triassic stratified rocks all over the world — in England, 

 Germany, Russia, United States, South Africa, and India — 

 are known to include remains of labyrinthodont amphil^ia 

 as well as the higher reptiles. Investigators are still 

 engaged in working out the correlation of these Avidely- 

 separated sedimentarj^ formations, the exact horizon of 

 which is not yet altogether settled. There is hardly any 

 doubt that these sandstones, so similar in all the countries 

 just mentioned, were laid down in fresh water, possibly in 

 lakes, though we think more probably they belonged to 

 large river sj^stems. 



DcHcriptioii. 



The British Museum bone is GG mm. long, the Hobart 

 one, 62 mm.: the breadth of the distal end in both speci- 

 mens is 23 mm.: of the proximal end or head, 20 mm. 

 The deltoid crest is developed into a strong bony process, 

 which is prolonged as a ridge distad down the narrowest 

 part of the shaft, where it subsides. The anconal depres- 

 sion at the distal end is sub-deltoidal, being a well-defined, 

 shallow, trochlear groove, widening distad, and separating 

 the extremity into the two condyles, ulnar and radial. 

 The ends are broader across than they are thick, and are 

 fairly expanded, though not so much as is generally the 

 case in Anomodont and Dicynodont reptiles ; neither does 

 the bone shew the sigmoid shape of a lacertian humerus. 



* Historical Sketch of the Geological Relations of Australia and 

 Ta»*mania : Trans. Austr. Inst. Min. Engineers. 1895. 



