474 Annals of the South African Museum. 



sand-grains are not of the "millet-seed" type, but are sub-angular 

 in outline with their edges and corners slightly rounded. They are 

 uniform in size; and the ferric oxide acts as a cementing material 

 coating the grains. Mica is commonly present. 



All these features agree closely with those postulated by Barrell 

 for flood-plain deposits in a semi-arid climate; the basal portion of 

 the Red Beds thus forming a link with the Molteno Beds the 

 climate gradually changing from an intermittently rainy to a semi- 

 arid one, the aridity increasing as time went on. 



A consideration of the horizontal and vertical distribution of the 

 animal remains is of some interest. The base of the lied Beds has 

 yielded fossils chiefly in the district of Herschel. They are all large, 

 heavy-limbed forms such as Eiiskelesaiirus broivni and Plateosaurus 

 ntHingwortM while a little higher in the same area are still large 

 forms such as Melanorosaurus readi. As far as we are aware, no 

 small, lightly-huilt forms have been found at the base of the forma- 

 tion. As one ascends the forms become smaller and apparently 

 much more agile. Moxxox/iondi/lux occurs from about half-way up 

 to near the top. both in Herschel and in the north at Fouriesburg. 

 At the latter place it is associated with the somewhat larger but- 

 still light-limbed Gri/jjont/.i africanus and a small Tliecodontosaurus. 

 The southern end of the mass has yielded little from the base of 

 the Red Beds but large bones have been noted by du Toit in Elliot 

 and Maclear: the small Thecodontosaurus minor is from the upper 

 half, as are the extraordinary Cynodonts Triiheledon and Lycorhinus, 

 while the supposed Predentate Geranosaurus is from the summit of the 

 formation. Erythrochampsa is from the very top also, and is another 

 very lightly-built small form, characterised by an armoured back. 



Two features display themselves in this survey. The first is that 

 the animal type became progressively more agile; the other that the 

 majority of the fossils come from the centre and northern half of 

 the present exposures. This latter feature may be disproved by sub- 

 sequent finds: the southern portion of the area lias not been searched 

 to any appreciable extent: and the outcrops are frequently grass- 

 covered and in had condition for fossil-hunting. There can be no 

 doubt, however, that the lirst is of some importance; and it gives 

 assistance to the argument for a climate which gradually became 

 more arid. As conditions became more rigorous, there would have 

 been given an impetus towards agility and the ability to travel 

 longer distances, in search of sustenance; and it is significant that 

 deserticolous animals of to-dav are long-limbed in structure. 



It seems probable that the various animals did not live far from 



