306 REY. J. F, BLAKE ON A REMARKABLE INLIER  [ May 1902, 
(See fig. 7, p. 301.) So far for the production of a breccia-bed or 
ice-foot rock in situ. 
From this comparison of the whole, we see that it is possible to 
draw up a pretty fuil description of a remarkable set of deposits, 
which might be intended either for the breccia-beds of Sutherland 
or the ice-foot rocks of Smith Sound. Such a description would 
contain the following points in common :— / 
(1) The deposits are composed largely of fragments of rock, 
(2) which are seldom striated ; 
(3) but are mixed with rounded pebbles. 
(4) They occur along the shore of a land which is bounded in places by cliffs, 
(5) 1ato the screes of which they may pass, and 
(6) of the fragments of which they are composed. 
(7) They are often, or generally, spread out in nearly flat sheets over the 
neighbouring sea-floor. 
(8) Finally, the minor phenomenon of the appearance of partial stratification 
of the breccia-beds at Gartymore reservoir (fig. 3, p. 294) may possibly 
correspond to the curious account (too long to quote at length) of how, 
when the ice-foot begins to melt, a channel is formed between it and 
the screes (the very spot required), in which the surrounding material 
is re-sorted by the water which gains access to the channel. 
In all the above cases the comparison is direct, between observa- 
tion on one side and observation on the other. But we have no 
direct evidence as to what happens to the material which is carried 
away, as on a float, by the fragments of the ice-foot, when it breaks 
up under the action of the summer-thaw. Our knowledge is 
limited to inference. We may assume that the ice-float, when 
carried up and down by the tide, will have a balance of motion 
towards the more ice-free regions to the south, and that a con- 
siderable portion will keep along shore. Any that reaches the open 
sea will scatter miscellaneous blocks at wide intervals, like the 
blocks of coal and granite in the Chalk of the North Downs,’ and 
will nowhere form beds. 
The distribution of the fragments along the path of a train of 
floats will depend on two factors: the relative proportion of the 
burden to the ice, and the size of the float. The first fragment to 
fall will be the largest, which has only just as much ice attached 
as will float it at all. When by this process the relative proportion 
of burden has lost its importance, the fragments of ice-foot with little 
burden will be distributed according to size. The smaller floats will 
travel fastest, but they will melt soonest ; the larger floats will travel 
more slowly, till they reach the last resting-place of the smaller, 
when some part of them will still be left to travel farther. The 
predicated distribution can best be compared with that observed 
in Sutherland, by uniting into one description the points that are 
common to both. Hence 
will \ 
(9) The most northerly deposits jaa contain the largest fragments. 
1 See W. P. D. Stebbing, Quart. Journ, Geol, Soe, vol. liii (1897) p. 218. 
