286 ANNALS NEW YORK ACADEMY OF SCIENCES 



But one is con fronted at once by serious difficulty in the effort to apply 

 the doctrine of transport. The Soulier-Abiracs bed rests almost directly 

 on tlie lowest deposit of the Auzits system ; its curved outcrop is from 

 one fourth of a mile to a mile and a half from the schists, so that the 

 coal must have covered much of the space inclosed by that outcrop ; it 

 may have been spread over four square miles, or even more, vpith an aver- 

 age thickness of three feet. T\niile this bed was accumulating, the three 

 streams of Longuefort, Haute Serre and Lugan must have been utterly 

 insignificant. The Campagnac coal bed had an area of at least 12 square 

 miles with an average thickness of 35 feet; the length of the streams, 

 outside of the basin, may have become as much as six miles. The Bour- 

 ran coal bed was formed when the whole basin was receiving deposits, so 

 that its material must have come wholly from outside, from an area of 

 not more than 300 square miles. One can concede that enough vegetable 

 matter, several times over, was produced on the drainage area to form 

 the coal beds and, at the same time, he would be justified in doubting the 

 possibility of its transference to the basin. Of course, every one knows 

 that small streams at all times carry twigs and leaves and that they carry 

 larger fragments during flood; but that is nothing, for one is concerned 

 here not with patches of carbonaceous matter but with coal accumula- 

 tions 70 to 100 feet thick. 



Eainfall, even when heavy, does little toward shifting the vegetation 

 growing on steep slopes, unless the rock material be loose, in which case 

 landslides may occur ; but those are of limited extent, even when greatest. 

 If there be a coating of humus, the effect of the rain is practically noth- 

 ing. It is unnecessary to go far from Decazeville in search of proof. 



At Viviez, two miles away, fumes from the zinc works have destroyed 

 almost all vegetation on the abrupt hillslope alongside ; rains have already 

 gashed the face deeply and the cover of disintegrated rock is accumulating 

 at the bottom. The limit of the devastated space is defined sharply; 

 southward from it the vegetable cover remains, and the surface is unin- 

 jured. This protective power of humus is familiar. That material accu- 

 mulates under forest cover ever\^where, even on very steep slopes; it ab- 

 sorbs water and then coheres tenaciously. Studies made near ISTew York 

 in the effort to solve the problem of water supply for some towns prove 

 that the humus coating is permanent on exceedingly steep slopes and that 

 water from such slopes is limpid, practically free even from vegetable 

 matter. The hills bounding the gorges below Eygurande and above Alias- 

 sac have slopes often reaching 30 degrees and sometimes exceeding that 

 angle, while walls in railroad cuts frequently approach the vertical. 

 During the summer of 1910, the rainfall in that portion of France was 



