I am not quite sure that I understand your difficulty, so I must give what seems to me the explanation of the glacial lake theory at some little length. You know that there is a rocky outlet at the level of all the shelves. Please look at my map. (530/1. The map accompanying Mr. Darwin's paper in the "Phil. Trans. R. Soc." 1839.) I suppose whole valley of Glen Spean filled with ice; then water would escape from an outlet at Loch Spey, and the highest shelf would be first formed. Secondly, ice began to retreat, and water will flow for short time over its surface; but as soon as it retreated from behind the hill marked Craig Dhu, where the outlet on level of second shelf was discovered by Milne (530/2. See note, Letter 521.), the water would flow from it and the second shelf would be formed. This supposes that a vast barrier of ice still remains under Ben Nevis, along all the lower part of the Spean. Lastly, I suppose the ice disappeared everywhere along L. Loggan, L. Treig, and Glen Spean, except close under Ben Nevis, where it still formed a barrier, the water flowing out at level of lowest shelf by the Pass of Mukkul at head of L. Loggan. This seems to me to account for everything. It presupposes that the shelves were formed towards the close of the Glacial period. I come up to London to read on Thursday a short paper at the Linnean Society. Shall I call on Friday morning at 9.30 and sit half an hour with you? Pray have no scruple to send a line to Queen Anne Street to say "No" if it will take anything out of you. If I do not hear, I will come.
LETTER 531. TO J. PRESTWICH. Down, January 3rd, 1880.
You are perfectly right. (531/1. Prof. Prestwich's paper on Glen Roy was published in the "Phil. Trans. R. Soc." for 1879, page 663.) As soon as I read Mr. Jamieson's article on the parallel roads, I gave up the ghost with more sighs and groans than on almost any other occasion in my life.
2.IX.IV. CORAL REEFS, FOSSIL AND RECENT, 1841-1881.
LETTER 532. TO C. LYELL. Shrewsbury, Tuesday, 6th {July, 1841}.
Your letter was forwarded me here. I was the more glad to receive it, as I never dreamed of your being able to find time to write, now that you must be so very busy; and I had nothing to tell you about myself, else I should have written. I am pleased to hear how extensive and successful a trip you appear to have made. You must have worked hard, and got your Silurian subject well in your head, to have profited by so short an excursion. How I should have enjoyed to have followed you about the coral-limestone. I once was close to Wenlock (532/1. The Wenlock limestone (Silurian) contains an abundance of corals. "The rock seems indeed to have been formed in part by massive sheets and bunches of coral" (Geikie, "Text-book of Geology," 1882, page 678.), something such as you describe, and made a rough drawing, I remember, of the masses of coral. But the degree in which the whole mass was regularly stratified, and the quantity of mud, made me think that the reefs could never have been like those in the Pacific, but that they most resembled those on the east coast of Africa, which seem (from charts and descriptions) to confine extensive flats and mangrove swamps with mud, or like some imperfect ones about the West India Islands, within the reefs of which there are large swamps. All the reefs I have myself seen could be associated only with nearly pure calcareous rocks. I have received a description of a reef lying some way off the coast near Belize (terra firma), where a thick bed of mud seems to have invaded and covered a coral reef, leaving but very few islets yet free from it. But I can give you no precise information without my notes (even if then) on these heads...
Bermuda differs much from any other island I am acquainted with. At first sight of a chart it resembles an atoll; but it differs from this structure essentially in the gently shelving bottom of the sea all round to some distance; in the absence of the defined circular reefs, and, as a consequence, of the defined central pool or lagoon; and lastly, in the height of the land. Bermuda seems to be an irregular, circular, flat bank, encrusted with knolls and reefs of coral, with land formed on one side. This land seems once to have been more extensive, as on some parts of the bank farthest removed from the island there are little pinnacles of rock of the same nature as that of the high larger islands. I cannot pretend to form any precise notion how the foundation of so anomalous an island has been produced, but its whole history must be very different from that of the atolls of the Indian and Pacific oceans—though, as I have said, at first glance of the charts there is a considerable resemblance.
LETTER 533. TO C. LYELL. {1842.}
Considering the probability of subsidence in the middle of the great oceans being very slow; considering in how many spaces, both large ones and small ones (within areas favourable to the growth of corals), reefs are absent, which shows that their presence is determined by peculiar conditions; considering the possible chance of subsidence being more rapid than the upward growth of the reefs; considering that reefs not very rarely perish (as I cannot doubt) on part, or round the whole, of some encircled islands and atolls: considering these things, I admit as very improbable that the polypifers should continue living on and above the same reef during a subsidence of very many thousand feet; and therefore that they should form masses of enormous thickness, say at most above 5,000 feet. (533/1. "...As we know that some inorganic causes are highly injurious to the growth of coral, it cannot be expected that during the round of change to which earth, air, and water are exposed, the reef-building polypifers should keep alive for perpetuity in any one place; and still less can this be expected during the progressive subsidences...to which by our theory these reefs and islands have been subjected, and are liable" ("The Structure and Distribution of Coral Reefs," page 107: London, 1842).) This admission, I believe, is in no way fatal to the theory, though it is so to certain few passages in my book.
In the areas where the large groups of atolls stand, and where likewise a few scattered atolls stand between such groups, I always imagined that there must have been great tracts of land, and that on such large tracts there must have been mountains of immense altitudes. But not, it appears to me, that one is only justified in supposing that groups of islands stood there. There are (as I believe) many considerable islands and groups of islands (Galapagos Islands, Great Britain, Falkland Islands, Marianas, and, I believe, Viti groups), and likewise the majority of single scattered islands, all of which a subsidence between 4,000 and 5,000 feet would entirely submerge or would leave only one or two summits above water, and hence they would produce either groups of nothing but atolls, or of atolls with one or two encircled islands. I am far from wishing to say that the islands of the great oceans have not subsided, or may not continue to subside, any number of feet, but if the average duration (from all causes of destruction) of reefs on the same spot is limited, then after this limit has elapsed the reefs would perish, and if the subsidence continued they would be carried down; and if the group consisted only of atolls, only open ocean would be left; if it consisted partly or wholly of encircled islands, these would be left naked and reefless, but should the area again become favourable for growth of reefs, new barrier-reefs might be formed round them. As an illustration of this notion of a certain average duration of reefs on the same spot, compared with the average rate of subsidence, we may take the case of Tahiti, an island of 7,000 feet high. Now here the present barrier-reefs would never be continued upwards into an atoll, although, should the subsidence continue at a period long after the death of the present reefs, new ones might be formed high up round its sides and ultimately over it. The case resolves itself into: what is the ordinary height of groups of islands, of the size of existing groups of atolls (excepting as many of the highest islands as there now ordinarily occur encircling barrier-reefs in the existing groups of atolls)? and likewise what is the height of the single scattered islands standing between such groups of islands? Subsidence sufficient to bury all these islands (with the exception of as many of the highest as there are encircled islands in the present groups of atolls) my theory absolutely requires, but no more. To say what amount of subsidence would be required for this end, one ought to know the height of all existing islands, both single ones and those in groups, on the face of the globe—and, indeed, of half a dozen worlds like ours. The reefs may be of much greater {thickness} than that just sufficient on an average to bury groups of islands; and the probability of the thickness being greater seems to resolve itself into the average rate of subsidence allowing upward growth, and average duration of reefs on the same spot. Who will say what this rate and what this duration is? but till both are known, we cannot, I think, tell whether we ought to look for upraised coral formations (putting on one side denudation) above the unknown limit, say between 3,000 and 5,000 feet, necessary to submerge groups of common islands. How wretchedly involved do these speculations become.