1849..
BROWN ON ERECT SIGILLARI^.
359
stratum in the ascending order, forms the roof of the coal. In such
cases the surface of the peat bog could not have been level when the
shale was deposited upon it, some small patches having been still
above water ; and as no upright trees are found in the sandstone
roof, it may reasonably be inferred that plants would not vegetate
upon the bog itself, a layer of soft mud being necessary in the first
instance for germinating the seeds ; but when a plant had once taken
root in this mud its rootlets penetrated downwards into the peat,
and furnished an abundant supply of nutriment for the rapid growth
of the tree, from the rich mass of decaying vegetable matter beneath.
I may here observe that the quality of the upper part of the seam
appears invariably to be influenced by the nature of the roof, the coal
being highly charged with iron pyrites under a sandstone, but quite
free from it under a shale roof*. And since no upright trees are
found in the sandstone roofs, is it probable that the luxuriant forest
of Sigillariae growing in the mud above the peat bog has taken up
the sulphuret of iron, and thus produced such a beneficial eifect upon
the quality of the coal ? Both sulphuric acid and the oxide of iron
are inorganic constituents of plants, and it has been ascertained that
natural and artificial waters that have a sulphureous taste, when em-
ployed in irrigating meadows, give birth to a very luxuriant vegeta-
tion f, but whether the growth of a forest of Sigillarise upon the sur-
face of the peat bog, is an adequate explanation of the absence of iron
pyrites from the upper part of the coal seam, I must leave to more
skilful botanists to determine.
Having shown that Sigillaria alternans was provided with roots
peculiarly adapted for flourishing upon a soft muddy soil, and thus
furnished additional proof that coal seams have been formed from
beds of peat or other decaying vegetable matter accumulated on the
surface, it only remains for me now to direct attention to the fossil
represented in fig. 9, which it will be observed, in its external aspect,
Fig. 9.
Stump broken off close to the root, one-tivelfth of natural size.

• Mr. Buddie states in the Trans, of the Nat. Hist. Society of Newcastle, vol. i.
  

p. 217, that the coal seams in Northumberland are always more or less intermixed
with iron pyrites under a sandstone roof.