75 
thin layer of sedimentary matter. The thickening of a 
parting requires that the sinking which followed on the 
growth of the lower coal bed should have gone on faster at 
some spots than others. By such an adjustment, the lower 
seam would become bent down into the shape A B, Fig. 3, 
Plate II. : if the sediment accumulated up to the level G D, 
and a fresh growth of coal, I Gr D F, took place over the 
level top of the deposit, there would be a double seam of 
coal with a parting rapidly swelling out towards the right. 
The replacement of coal by sandstone is also often 
produced by what are known as "rock faults." In such 
cases a stream of water has flowed over the layer of dead 
plants before it became covered up, and eaten out in it a 
trough or channel, and this hollow has, during the subse- 
quent submergence, been filled in with sand or mud. 
But neither of these explanations will apply to the cases 
before us, for two facts have to be accounted for. First, the 
gradual breaking up of a seam of clean coal into numerous 
subdivisions by dirt bands, and the gradual diminution in 
thickness and eventual disappearance altogether of the coal 
out of the seam ; and, secondly, when a spot is reached where 
coal again sets in, the appearance of a bed difiering totally in 
thickness and character from the seam started with. 
The latter fact may be accounted for if it is supposed that 
the swamps in which the growth of the two distinct forms of 
the same bed went on were not continuous, but parted from 
one another by some barrier, and that the physical conditions 
of the two swamps were so different that the vegetable 
growth of the one differed from that of the other both in 
nature and amount. For instance, the Silkstone and Block- 
ing coals may both have been growing about the same time on 
two distinct swamps. Where the Silkstone seam accumulated, 
the growth was plentiful and rapid, and during the formation 
of the seam submergence occurred once and gave rise to a 
