480 Annals of the South African Museum. 



This may also be looked upon as as mark of aridity. Richthofen 

 considered vegetation necessary for holding in place the dust of which 

 the loess is formed: but this is only so when the area of depositon 

 is itself subject to constant winds. If the dust is deposited as a 

 result of the slackening of the force of the wind which carried it, 

 then vegetation is not a necessary adjunct; and the lack of vegetable 

 remains argues conditions adverse to their growth. 



The third difference is the absence of land-shells in the Cave 

 Sandstone. Most terrestrial mollusca require moisture for their de- 

 velopment and maintenance; and, although their absence from a 

 land deposit is a piece of purely negative evidence, it lends additio- 

 nal weight to the theory of aridity when so many other facts point 

 in the same direction. 



In spite of these minor discrepancies, however, the "loess" seems 

 to be the nearest among recent deposits to which we can liken the 

 Cave Sandstone. It is of interest to note that near the base are 

 highly irregular calcareous concretions similar to the "loess-dolls" 

 figured by Wright. 



The occurrence of mammalian bones in the loess is paralleled by 

 that of reptilian remains in the Cave Sandstone. All the forms 

 hitherto discovered are light-limbed probably cursorial forms 

 (i/ljiosaurtts, Thecodontosaurus and the like, together with Notochawpsa. 

 As has been shown, this latter was erroneously supposed to be a 

 Crocodile, but is in reality a light-limbed long-snouted armoured 

 Pseudosuchian. There is nothing of the water-loving, marsh-living 

 type in this assemblage; they are all very conceivably the inhabi- 

 tants of an arid or semiarid clime, adapted for comparatively rapid 

 transit from place to place. In no case has a complete skeleton 

 been found; but, on the other hand, isolated bones are rare and 

 the usual occurrence is in the form of articulated portions of a ske- 

 leton. It should be noted that, in the case of No1ochinps, the 

 skeleton is nearly complete, even some of the fragile bones of the 

 fingers and toes being preserved. Had the conditions been moist, it 

 is improbable that such would have been preserved in a sandstone- 

 water percolating through the porous sediment would have rapidly 

 dissolved the bones which have been preserved, as we believe, by 

 the rapid accumulation of a dry wind-borne coarse siliceous dust. 



That the climate was not absolutely arid is shown by 1) the fact 

 that the deposit is in no sense a "desert-sandstone" in the Cape - 

 0. F. S. area, and 2) the very rare local developments of greenish 

 and bluish shales containing a water-fauna. Of these, the best- 

 known is the shale-band in the Cave Sandstone at Siberia in the 



