GEOLOGY OF MAEOCCO AND GREAT ATLAS. 459 



not a trace ; and alluvial drift and valley gravels are very 

 limited in their distribution, being confined to the borders of a 

 few insignificant rivers that intersect the plain and the localities 

 of occasional waterflows ; but as soon as the flanks of the Atlas 

 are reached, new and distinct drift phenomena present them- 

 selves. It was on our second day's journey from Marocco to 

 the Atlas that the great boulder-beds came under our notice, 

 first in a valley leading up from Mesfioua to Tasseremout, as 

 scattered blocks of red sandstone, remarkable for their large 

 average size, many of them of from ten to twenty cubic yards ; 

 but here the method of their disposition scarcely enabled us to 

 decide that they were other than stream-borne masses from the 

 higher ground. From Tasseremout we turned west, and at the 



Boulder-mounds, skirting Atlas Plateau Escarpment. (Section,) 



mouth of a second valley, two miles from the village, suddenly 

 came upon a huge development of these Red Sandstone boulder- 

 beds as great ridge-like and very symmetrical masses with ter- 

 minal faces three or four hundred feet high, and, like the more 

 scattered blocks NW. of Tasseremout, intermixed with but aVery 

 small proportion of fine matter. From this valley we turned out 

 northwards, skirting the escarpment facing the plain ; and for 

 more than ten miles no lateral valley breaks into the cliff-like face ; 

 but below it the great boulder-beds (figs. 5, 6) stUl occur in huge 

 masses not resting directly against the escarpment, but as iso- 

 lated mounds two or three hundred feet in advance, sloping 

 down towards the escarpment in one direction, and in the other 

 rolling away in great wave-like ridges and undulating sheets, 



