242 DR JAMES GEIKIE ON 
sufficiently cooled to permit of vegetation spreading over its surface, or indeed 
before the plants themselves had time to occupy the ground. Again, it might 
occasionally happen that, owing to the nature of the volcanic surface, there could 
be little or no accumulation of vegetable matter in swamps and pools, and the 
mere “carpet of greenery” which may have covered the ground in whole or in 
part, might well be destroyed upon the advance of another lava-flow. 
The appearances which have just been described as the most likely to occur 
under the circumstances I have supposed, are precisely those which are found 
associated with the coal-beds of Suderde. The character of the clays and 
tufaceous shales with which these coals are interbedded, and the manner in 
which mineral and vegetable layers interosculate, all point to quiet deposition 
in shallow lakes, and the slow accumulation of plant-remains in swamps, 
marshes, and bogs. The thin layer of coal at Dalbofos may indicate a com- 
paratively short period of rest, while the thicker coals at the higher level seem 
to bespeak a pause of much longer duration. In like manner, many of the 
tufaceous shales which appear to be unfossiliferous, may yet have been accumu- 
lated in precisely the same manner as the shales which accompany the coal 
layers, and which in some places they closely resemble. The mere absence 
from these shales of coal or plant-remains does not necessarily prove that no 
vegetation covered the plateaux at. the time the shales in question were accumu- 
lated. The presence of the little patch of coal at Dalbofos, which is quite 
local—none ever having been seen elsewhere on the same horizon—shows us 
how easily all trace of a vegetable layer might be obliterated. But while it 
may be true that some of the darker tufaceous shales may thus be of the nature 
of ‘ surface-wash,” and owe their origin directly to the action of the weather, 
there can be little doubt that by much the greater proportion of the red and 
particoloured tuffs: are due more or less directly to igneous action. They 
consist of the finer dust and grit blown out during eruptions, and spread by 
the winds over vast areas, and partly no doubt of the same material carried down 
by rain and swept from higher to lower levels by running water—sometimes, 
perhaps, the tuffs may represent former currents and streams of volcanic mud. 
4, Position of old Volcanic Centre of Eruption —It would be interesting to 
ascertain the locality of the volcanic centre from which the old lavas of the 
Ferée Islands were ejected. I am inclined to think that it lay somewhere to 
the westward, partly for the reason that the rocks have an average easterly 
dip, and partly because the sea between the Feriée Islands and Iceland is not 
so deep as it is in the direction of Scotland. The dip, although it may have 
experienced subsequent modification, may yet indicate the original inclination 
of the ground, while the lesser depth of the sea to the west does not imply 
such extreme depression and denudation as must have taken place if the site 
of the ancient volcano lay far to the east of the present islands. At the same 
