GLACIATION OF THE SHETLAND ISLES. S07 
placed beyond all doubt by the distribution of the Boulder-clay, as 
well as by the dispersal of the stones in this deposit. On the 
western sea-board of Unst the Boulder-clay contains fragments 
of serpentine, gabbro, and graphitic schists, all of which occur 
am situ on the east side of the Vallafield range. Moreover the rela- 
tive distribution of the serpentine and gabbro stones in this deposit 
on the western shore is in direct proportion to the relative areas 
occupied by these rocks to the east of the watershed. It follows, 
therefore, that the agent which glaciated Unst must have crossed 
the watershed, carrying the bottom-moraine up the slope, and 
depositing it in the lee of the range. In Fetlar, blocks of gabbro 
and serpentine are likewise found in the Boulder-clay on the west 
coast ; while along the east coast of Yell, blocks of gabbro occur in 
this deposit which have been brought from Unst and Fetlar, testi- 
fying alike to the same westerly movement. 
The evidence derived from an examination of the Boulder-clay 
sections on the Mainland is equally conclusive; for it matters not 
whether we cross the northern, central, or southern portions of the 
island, we are compelled to admit that the ice-flow during the 
. primary glaciation must have been towards the Atlantic. In the 
central part of Northmavine it has been clearly proved that the 
Boulder-clay partakes of the physical character of the rock-forma- 
tion on which it rests, while a certain percentage of the stones is 
derived from localities which lay in the path of the glaciating 
agent. Abundant evidence has been adduced to show that the 
quartz-felsite area between Tanwick and Roeness Voe has been in- 
vaded by the diorite stones, while the area occupied by the bedded 
porphyrites has been invaded by the quartz-felsite and diorite 
stones. Moreover it is particularly observable that the blocks de- 
rived from the successive areas occupied by these rocks, which are 
present in the Boulder-clay, diminish in number in proportion to 
the distance from their parent source. 
Again, in the long tongue of land which stretches from Scallo- 
way southwards to Fitful Head, blocks of the Old-Red-Sandstone 
rocks occurring on the eastern sea-board are found, not only on the 
tops of the highest hills, but also in the Boulder-clay on the western 
shore. Moreover the distribution of the fragments of the Lerwick 
Sandstone, Brenista Flags, and basement-breccia in the Boulder- 
clay north of West Quarff is in perfect harmony with the relative 
areas oecupied by these subdivisions of the Old Red Sandstone 
south of Lerwick. The same relationship holds true in the district 
between Maywick and Fitful Head. It is clear, therefore, that the 
elaciating agent must haye overflowed the watershed, as we found 
to be the case in Unst. 
2, Insufficiency of Lecbergs or Coast-ice to account for the Phenomena. 
—Perhaps some may attribute the numerous striated surfaces, as well 
as the Boulder-clay, to the action of icebergs or coast-ice on a sink- 
ing area; but a little consideration will show that either of these 
causes is quite inadequate to explain the phenomena. We have 
shown that over the whole of Shetland the glaciating agent must 
have conformed to the inequalities of the surface, descending into 
