456 
THE TERTIARY VOLCANOES 
BOOK VIII 
still flowing from time to time, erosion was in active progress over the sur- 
face of the volcanic plain. The records of river-action in Ganna and Sanday, 
and the buried river-channel of the Scuir of Eigg, prove that, while eruptions 
still continued, rivers descending from the mountains of the Western 
Highlands carried the detritus of these uplands for many miles across the 
lava-fields, swept away the loose material of volcanic cones, and cut channels 
for themselves out of the black rugged floor of basalt. 
The erosion thus early begun has probably been carried on continuously 
ever since. The present streams may he looked upon as practically the 
same as those which were flowing in the Tertiary period. There may have 
been slight changes of level, oscillations both upward and downward in the 
relative positions of land and sea, and shiftings of the watercourses to one 
side or other ; but there seems no reason to doubt that the existing basalt- 
plateaux, which were built up as terrestrial areas, have remained land-sur- 
laces with little intermission ever since, although their lower portions may 
have been in large measure submerged. 
In the existing valleys, fjords and sea-straits by which these plateaux 
har e been so deeply and abundantly trenched, we may recognize some of the 
drainage-lines traced out by the rivers which flowed across the volcanic plains. 
The results achieved by this prolonged denudation are of the most stupendous 
land. r J lie original lava-floor has been cut down into a fragmentary table- 
land. Hundreds of feet of solid rock have been removed from its general 
surface. Outliers of it may be seen scattered over the mountains of Morven, 
whence they look into the heart of the Highlands. Others cap the hills 
of Rum, where they face the open Atlantic. Several miles from the main 
body of the plateau in Skye, a solitary remnant, perched on the highest 
summit of Eaasay, bears eloquent witness that the basaltic tableland °once 
stretched far to the east of its present limits. 
Two lines of observation and of argument may be followed in the effort 
to demonstrate how great the denudation has been since older Tertiary time. 
In the first place, there is the evidence of the level or nearly level sheets of 
basalt that form the plateaux, and, in the second place, there is the testimony 
o the dykes, sills and bosses by which these lavas have been disrupted. 
1. The study of the denudation of the Tertiary volcanic rocks of North- 
Western Europe is most satisfactorily begun by an attempt to measure the 
minimum amount of waste which in certain places the basalt-plateaux can 
be proved to have undergone. For the purposes of this study, the stratifica- 
tion of the lavas and their nearly horizontal, or at least very slightly dis- 
turbed, position afford exceptional facilities. Amorphous rocks, such as 
granites and gabbros, or even foliated masses like the old gneisses and schists, 
may have been enormously denuded. Their mere presence at the existing 
surface may be taken as proof of such waste, yet they furnish in themselves 
no criterion by which the amount of removed material may lie estimated. 
Lut in the case of the basalt-plateaux, as in that of horizontal sedi- 
mentary formations, the successive lines of superposition of the component 
beds of the whole stratigraphical series supply admirable datum-lines which, 
