806 B. N. EACH AND J. HORNE ON THE 
results, as the size of the rivers would be in proportion to the area 
of drainage. We have seen also that some of the voes and inland 
valleys coincide with the outcrops of bands of limestone, the ero- 
sion of which would be aided by chemical agencies. 
There can be no doubt, however, that the sea-lochs in Shetland 
were deepened by ice-action during the primary glaciation; indeed 
numerous instances have been cited in this paper where the great 
mer de glace took advantage of the existing hollows in crossing 
the island, This produced, in certain instances, fjord-basins, of 
which we shall adduce two examples. The soundings given in 
the Admiralty chart show that Sulem Voe, which is one of the 
largest of the sea-lochs in the Mainland, measuring upwards of 
seven miles in length, varies from 10 to 15 fathoms in depth 
between Foula Ness and the mouth of Voxter Voe. Beyond 
the latter point, however, to the head of the yoe, the depth 
suddenly increases to 21 and 25 fathoms. This increase of 60 
feet in depth at the head of the sea-loch is doubtless due to the 
intense abrasion caused by the ice as it impinged on the rocky 
isthmus of Mavis Grind. We have already pointed out how dis- 
tinctly the east face of this narrow isthmus has beer polished and 
striated ; and this fjord-basin helps us to realize still better the ero- 
sive power of this agent. Still another instance occurs in Roeness 
Voe; for at the bend north of Urie Firth the depth varies from 102 
to 138 feet, while about two miles further down the loch shallows 
to 42 feet. 
There is one peculiar feature connected with these voes which may 
be dismissed in a few words. It frequently happens that spits 
of gravel are thrown up by tidal action near the head of the sea- 
loch. These banks are seen in all stages of formation in Shetland, 
sometimes extending a third, a half, or nearly the whole of the way 
across the loch. Ultimately the voe is crossed by a continuous 
bank of gravel which isolates the upper part; and this isolated 
portion is converted into a sheet of brackish water. 
VIII. Concttston. 
1. Summary of the Evidence regarding the Primary Glaciation — 
We must now, very briefly, recapitulate the evidence regarding 
the primary glaciation of Shetland, in order to show the conclusions 
which may be justly drawn from the facts, and also to determine 
the relation which the glacial phenomena of these isles bear to the 
glaciation of Norway and Scotland. 
Tt has been shown that in the islands of Unst, Fetlar, Whalsey, 
the Outskerries, Bressay, and along the eastern sea-board of the ~ 
Mainland and Yell, there is one uniform system of ice-markings 
trending W.S.W., 5.W., and in some cases 8.8.W.; while in the 
western districts of the two lattcr islands, as well as in Meikle 
2ooe, Papa Stour, and Foula, the strie swing round to the N.W. 
and N.N.W. From a careful examination of the striated surfaces 
and the Stossseite of the roches moutonnées, it is evident that the 
agent which produced them must haye crossed the islands from 
the North Sea to the Atlantic. Fortunately this conclusion is 
