556 Notices of Memoirs—J. Lomas— 
found they haye been covered by drifting sand after complete 
desiccation of the pool, and when the salt. is intimately mixed with 
the rock the wind-borne material was deposited in the pool while 
some water remained. 
Dr. Cullis has recently described the occurrence of dolomite crystals 
in the Keuper Marls of the West of England. So far.they haye not 
been found in our locality, although “carefully looked for.1 We 
cannot say at present how these erystals have been formed, but their 
presence in recent deserts and in the Triassic rocks is of interest. 
When dealing with the subterranean water supply of the desert it 
was pointed out that the impermeable bands which confine the water 
below show very striking differences of level. If these bands of clay 
can be regarded as. the silt covering the floors of filled-up pools our 
difficulties are most satisfactorily met. (See p. 512.) 
Thus far we have dealt with the part water plays in the economy 
of the desert. The waste of sand stretching as far as the eye can 
reach gives little sign of the activities below the surface, and it is 
chiefly there that water has its work to do. 
_ Surface Deposits—We turn now to discuss the surface deposits 
themselves. It must not be supposed that even in the so-called sandy 
deserts the surface is uniformly covered with loose fragmentary 
materials. In the Algerian Sahara probably less than one-third is 
eovered with sand. Solid rock forms the floor of vast barren plains, 
or stands as islands or cliffs washed by great seas of sand. However, 
it is mainly with the sand (and in this term we may for convenience 
include pebbles) that we have to deal. 
Origin of the Sand.—The sands of the deserts were naturally at first 
attributed to marine action, and the presence of salt deposits no doubt 
gave verisimilitude to the idea that they represented dried-up sea- 
bottoms. Zittel, Tissandier, and others have shown that the frag- 
mental material is of subaerial origin, and has been derived by the 
action of the ordinary disintegrating forces at work on the land. 
In a climate such as ours it is hard to appreciate the importance 
of one of the most active disintegrating agents in the desert. One 
needs to stand under the tropical sun and in the presence of the riven 
rocks to fully appreciate what sun-flaking means. \Unprotected by 
vegetation or covering soils the bare rocks become intensely heated 
during the day; when the sun’s heat is withdrawn they rapidly chill 
by radiation, and thus by alternate expansion and contraction even 
the most compact and resistant rocks yield. A characteristic form 
assumed by sun-flaked rocks is the spheroid, and one is reminded of 
the spheroidal weathering often seen in basalt. In a cuboidal mass 
of granite the solid angles are the first to split, and then the edges go, 
and so flake after flake is removed until the block resembles a well- 
worn boulder. The flakes are usually from half an inch to a little 
more than an inch in thickness, but their horizontal extent may be 
many yards. After flaking, the smooth curved surfaces give one the 
impression of roches moutonnées when viewed at a distance, and 
1 Tt might be noted that dolomite occurs in plenty in the sands dredged from the 
Mersey bar. 
