Passing by this extreme case, and considering the generality of islands as exhibited in these three districts (but more especially perhaps those of Finland) with regard to the way in which they are acted on by ice, I divide them into the following three groups:—
2. Those small islands which are always above sea-level, but yet are annually covered with ice, which is driven up their sides and over their summits by outside pressure.
3. Those islands whose summits are beyond the influence of ice. The tops of these are generally covered with trees and vegetation, whilst others, which are usually not so high, have only a black colour, probably due to a growth of lichen.Islands emerging from an ocean, as these appear to be doing, must successively pass through the stages I have enumerated. During the first two of these stages they are wholly within the influence of coast-ice, as may be clearly seen from those islands on the Finnish coast, which have not only been moulded, but are kept of a whitish colour by the scouring they continually undergo. This latter character is especially noticeable in the islands of the third class, the upper parts of which have been raised so high that they are now beyond the influence of the scouring agency; these parts are black with a growth of lichen, whilst those at a lower level, which are annually invaded by the ice, are kept of a whitish colour. New, between the dark upper parts of such islands as I have classed in this last group and their lower parts, so far as I could see, there is a continuity in contour between the undulations and curves along the margin of the water (which I know to have been produced by coast-ice) and those which lie at higher levels. From this it seems to me an inevitable conclusion that the upper mouldings were produced in the same way as those below, and have since been raised to their present position.
Should two or three, or, still better, a whole archipelago of these islands slowly rise to unite and form a continuous land surface, I do not think it would be unlike many parts of Finland.
Turning from the islands to the adjoining mainland, it seems natural to conclude that the actions which produced the features in the one were identical with those producing the features in the other.
In these arguments it must be remembered that not only does the action of coast-ice upon an oscillating area explain many phenomena of ordinary occurrence, but it also readily explains some phenomena, such as the appearance of erratics at levels higher than the parent rock from which they were derived, which would be difficult to account for by either the action of sheets of ice or of glaciers.
Icebergs I have not considered, because, both from my own observations and more especially from the observations of those whose labours lie amongst them, I believe them, as compared with all other forms of ice, incomparable as transporters of material, and, from
