194 
ANNALS NEW YORK ACADEMY OF SCIENCES 
or rock-covered drainage area — for it must be remembered that during the 
same period another great bed, smaller in area but attaining an extreme 
thickness of more than 60 feet, was formed in the little lakelet of les Ferrieres 
and that it is supposed to have come from the Bourrus with small contribu¬ 
tions from Chamblet and Les Boulades. The source of the material must 
be sought elsewhere. 
The form and variations of the Grande Couche, even if there were 
nothing more, would seem to leave little room for doubt that the bed repre¬ 
sents plants which grew where the coal now is. But evidence from pecu¬ 
liarities of the bed is reinforced by the apparent impossibility of deriving the 
material from any other source. The history appears to be as follows: 
The streams entering the Pegauds area had cut down their channels, so 
that for two or more miles from the water they flowed with gentle fall and 
carried comparatively little mineral matter into the basin, except in times of 
unusual flood. The filling of the lake had come to a halt. The Bourrus 
delta, composed largely of coarse material from the granites, had been 
pushed nearly to the southern border, while that of the Colombier, composed 
of less coarse material from more readily yielding rocks, had less of emerged 
area, the most of its load having been carried into the Pegauds pond. The 
plain, covered mostly by cones of dejection, was separated from the water 
by a low ill-drained beach on which a marsh originated. The principal 
outlets of the Bourrus and Colombier flowed on each side behind this curved 
beach and entered the pond near the present southward termination of the 
coal’s outcrop. Rivulets entered from the north, carrying, except at rare 
intervals, only fine silt. 
The marsh growth consisted of Cordaites trees, which form an important 
constituent of the coal, with a dense growth of ferns, lepidodendra and other 
plants. Within the little area inclosed by the curved shore, detrital deposits 
were insignificant, being for the most part only such material as eddied back 
into the stiller water from the stream-mouths at the sides; and the shore¬ 
line’s advance would be due chiefly to invasion of the shallow water by 
growth of the swamp itself. The accumulation of swamp material on the 
strand would be enormous, but the thickness would decrease abruptly 
toward the water. 
A similar condition existed in the smaller pond of les Ferrieres, where, 
according to Fayol, the formation of the great bed, 20 meters thick, was 
synchronous with that of the Grande Couche. 
The forking or branching of the Grande Couche on the western border 
was due to floods in the Bourrus, which broke across the plain and washed 
the sands into the swamp. The conflict between flood and swamp is well 
shown on that side, where the several branches of the Grande Couche have 
