Dr. W. F. Hume — On Walthers Desert Erosion. 75 



coloration. In manganese films the streak remains grey, whether 

 the rock be moist or dry. The iron and manganese oxides are 

 mutually replaceable. Professor Walther (followiug Sickenberger) 

 also cites a case from Dakhla where the crust contained 8 per cent 

 of cobalt, this originating from the thermal springs there. 



The gradual development of dark protective films is well shown in 

 the biconcentric circular flints (Augensteine) which are well illustrated 

 in figs. 82-5, showing the progressive stages in depth of colour. The 

 causes noted as producing these results are traces of moisture in 

 the rocks and intense heating by the sun, the water contained in the 

 deeper layers being drawn in a capillary manner to the surface. 

 This moisture acts on the salts of iron and manganese, bringing them 

 into solution, these being redeposited on the surface by evaporation. 

 That the dark colouring is frequently observed on fossils when their 

 matrix is far lighter in tint is ascribed to the presence of phosphoric 

 acid, derived originally from the organisms, while in addition silicic 

 acid in the crystalline rocks and phosphoric acid in the limestones 

 are stated to have such an affinity for iron and manganese oxides 

 that at the point of contact chemical action is set up, which retains 

 the oxides, while the chlorides are removed by the wind. 



At p. 149 Professor Walther combats the view that the desert 

 films are directly due to rain or dew. Had this been the case, these 

 should become less marked in passing from the Nile Valley into the 

 arid desert, and again reappear in the region of the oases. On the 

 contrary, his experience has been that the brown crusts become less 

 marked as the more rainy districts are approached, and the films are 

 therefore essentially rainless desert structures, due to the evaporation 

 of highly concentrated solutions under the influence of the high 

 temperatures. 



While agreeing with the general proposition, I am certainly under 

 the impression that the crusts are more marked in that portion of the 

 desert near the Nile Valley. The oases occupy so small an area in 

 the Sahara that their effect in any case would be almost negligible. 



Section 16 deals with Erosion by Water, in which it is mentioned 

 that such streams as the Nile and the Colorado are in reality strangers 

 in the great desert system, the former especially being similar, in all 

 respects, to streams in the humid or pluvial regions, and only differing 

 from them by its long traverse across the desert belt. The essential 

 feature determining erosion by water in the desert is the sudden rain- 

 storm, and the point is emphasized by illustrations taken from the 

 neighbourhood of Helwan, in the Sinai Hills, and the Kharga Oasis. 

 The activity of the rain is short-lived, but intense while it lasts. 

 Years, nay centuries, may elapse, before another storm breaks in the 

 same locality, and in fig. 93, the Kas-el-Kadi (in Kharga Oasis), an 

 illustration is given of sucli a torrent-conglomerate based on soft 

 clayey strata, which have undergone sand erosion, leaving the torrent- 

 conglomerate as the head of a 'mushroom', of which the stalk is 

 composed of the claj's. In desert rain-storms an essential feature 

 is the widespread character of the torrent, to whicli ^lacGee has 

 given the term of ' sheet-flood ' and Passarge of ' riiichenspulung ' 

 (p. 163), the waters not following definite channels, but extending 



