The question arises as to what is the origin of the stratified sands and gravels filling up the buried river channels. Are they of marine or of freshwater origin? Mr. Dugald Bell[285] and Mr. James Geikie[286] are inclined to believe that as far as regards those filling the western channel they are of lacustrine origin; that they were formed in lakes, produced by the damming back of the water resulting from the melting of the ice. I am, however, for the following reasons, inclined to agree with Mr. Bennie’s opinion that they are of marine origin. It will be seen, by a comparison of the journals of the borings made through the deposits in the eastern channel with those in the western, that they are of a similar character; so that, if we suppose those in the western channel to be of freshwater origin, we may from analogy infer the same in reference to the origin of those in the eastern channel. But, as we have already seen, the deposits extend to the Firth of Forth at Grangemouth, where they are met with at a depth of 260 feet below sea-level. Consequently, if we conclude them to be of freshwater origin, we are forced to the assumption, not that the water formed by the melted ice was dammed back, but that the sea itself was dammed back, and that by a wall extending to a depth of not less than two or three hundred feet, so as to allow of a lake being formed in which the deposits might accumulate; assuming, of course, that the absolute level of the land was the same then as it is now.

But as regards the stratified deposits of Grangemouth, we have direct evidence of their marine origin down to the bottom of the Red Clay that immediately overlies the till and its intercalated beds, which on an average is no less than 85 feet, and in some cases 100 feet, below the present surface. From this deposit, Foraminifera, indicating an arctic condition of sea, were determined by Mr. David Robertson. Marine shells were also found in this bed, and along with them the remains of a seal, which was determined by Professor Turner to be of an exceedingly arctic type, thus proving that these deposits were not only marine but glacial.

Direct fossil evidence as to the character of the deposits occupying the western basin, is, however, not so abundant, but this may be owing to the fact that during the sinking of pits, no special attention has been paid to the matter. At Blairdardie, in sinking a pit-shaft through these deposits, shells were found in a bed of sand between two immense masses of boulder clay. The position of this bed will be better understood from the following section of the pit-shaft:—

But as the shells were not preserved, we have, of course, no means of determining whether they were of marine or of freshwater origin.

In another pit, at a short distance from the above, Cyprina Islandica was found in a bed at the depth of 54 feet below the surface.[287]

In a paper read by Mr. James Smith, of Jordanhill, to the Geological Society, April 24th, 1850,[288] the discovery is recorded of a stratified bed containing Tellina proxima intercalated between two distinct boulder clays. The bed was discovered by Mr. James Russell in sinking a well at Chapelhall, near Airdrie. Its height above sea-level was 510 feet. The character of the shell not only proves the marine origin of the bed, but also the existence of a submergence to that extent during an inter-glacial period.

On the other hand, the difficulty besetting the theory of the marine origin of the deposits is this. The intercalated boulder clays bear no marks of stratification, and are evidently the true unstratified till formed when the country was covered by ice. But the fact that these beds are both underlaid and overlaid by stratified deposits would, on the marine theory, imply not merely the repeated appearance and disappearance of the ice, but also the repeated submergence and emergence of the land. If the opinion be correct that the submergences and emergences of the glacial epoch were due to depressions and elevations of the land, and not to oscillations of sea-level, then the difficulty in question is, indeed, a formidable one. But, on the other hand, if the theory of submergences propounded in Chapters XXIII. and XXIV. be the true one, the difficulty entirely disappears. The explanation is as follows, viz., during a cold period of the glacial epoch, when the winter solstice was in aphelion, the low grounds would be covered with ice, under which a mass of till would be formed. After the cold began to decrease, and the ice to disappear from the plains, the greatest rise of the ocean, for reasons already stated, would take place. The till covering the low grounds would be submerged to a considerable depth and would soon be covered over by mud, sand, and gravel, carried down by streams from the high ground, which, at the time, would still be covered with snow and ice. In course of time the sea would begin to sink and a warm and continental period of, perhaps, from 6,000 to 10,000 years, would follow, when the sea would be standing at a much lower level than at present. The warm period would be succeeded by a second cold period, and the ice would again cover the land and form a second mass of till, which, in some places, would rest directly on the former till, while in other places it would be laid down upon the surface of the sands and gravels overlying the first mass. Again, on the disappearance of the ice the second mass of till would be covered over in like manner by mud, sand, and gravel, and so on, while the eccentricity of the earth’s orbit continued at a high value. In this way we might have three, four, five, or more masses of till separated by beds of sand and gravel.

It will be seen from Table IV. of the eccentricity of the earth’s orbit, given in Chapter XIX., that the former half of that long succession of cold and warm periods, known as the glacial epoch, was much more severe than the latter half. That is to say, in the former half the accumulation of ice during the cold periods, and its disappearance in polar regions during the warm periods, would be greater than in the latter half. It was probable that it was during the warm periods of the earlier part of the glacial epoch that the two buried channels of the Midland valley were occupied by rivers, and that it was during the latter and less severe part of the glacial epoch that these channels became filled up with that remarkable series of deposits which we have been considering.

Other buried River Channels.—A good many examples of buried river channels have been found both in Scotland and in England, though none of them of so remarkable a character as the two occupying the valley of the Forth and Clyde Canal which have been just described. I may, however, briefly refer to one or two localities where some of these occur.

(1.) An ancient buried river channel, similar to the one extending from Kilsyth to Grangemouth, exists in the coal-fields of Durham, and is known to miners in the district as the “Wash.” Its course was traced by Mr. Nicholas Wood, F.G.S., and Mr. E. F. Boyd, from Durham to Newcastle, a distance of fourteen miles.[289] It traverses, after passing the city of Durham, a portion of the valley of the Wear, passes Chester-le-Street, and then follows the valley of the river Team, and terminates at the river Tyne. And what is remarkable, it enters the Tyne at a depth of 140 feet below the present level of the sea. This curious hollow lies buried, like the Scottish one just alluded to, under an enormous mass of drift, and it is only through means of boring and other mining operations that its character has been revealed. The bottom and sides of this channel everywhere bear evidence of long exposure to the abrading influence of water in motion; the rocky bottom being smoothed, furrowed, and water-worn. The river Wear of the present day flows to the sea over the surface of the drift at an elevation of more than 100 feet above this buried river-bed. At the time that this channel was occupied by running water the sea-level must have been at least 140 feet lower than at present. This old river evidently belongs to the same continental period as those of Scotland.

(2.) From extensive borings and excavations, made at the docks of Hull and Grimsby, it is found that the ancient bed of the Humber is buried under more than 100 feet of silt, clay, and gravel. At Hull the bottom of this buried trough was found to be 110 feet below the sea-level. And what is most interesting at both these places, the remains of a submerged forest was found at a depth of from thirty to fifty feet below the sea-level. In some places two forests were found divided by a bed of leafy clay from five to fifteen feet thick.

(3.) In the valleys of Norfolk we also find the same conditions exhibited. The ancient bed of the Yare and other rivers of this district enter the sea at a depth of more than 100 feet below the present sea-level. At Yarmouth the surface was found 170 feet thick, and the deep surface extends along the Yare to beyond Norwich. Buried forests are also found here similar to those on the Humber.

It is probable that all our British rivers flow into the sea over their old buried channels, except in cases where they may have changed their courses since the beginning of the glacial epoch.

(4.) In the Sanquhar Coal Basin, at the foot of the Kello Water, an old buried river course was found by Mr. B. N. Peach. It ran at right angles to the Kello, and was filled with boulder clay which cut off the coal; but, on driving the mine through the clay, the coal was found in position on the other side.

(5.) An old river course, under the boulder clay, is described by Mr. Milne Home in his memoir on the Mid-Lothian coal-fields. It has been traced out from Niddry away in a N.E. direction by New Craighall. At Niddry, the hollow is about 100 yards wide and between 60 and 70 feet deep. It seems to deepen and widen as it approaches towards the sea, for at New Craighall it is about 200 yards wide and 97 feet deep. This old channel will probably enter the sea about Musselburgh. Like the channels in the Midland Valley of Scotland already described, it is so completely filled up by drift that not a trace of it is to be seen on the surface. And like these, also, it must have belonged to a period when the sea-level stood much lower than at present.

(6.) At Hailes’ Quarry, near Edinburgh, there is to be seen a portion of an ancient watercourse under the boulder drift. A short account of it was given by Dr. Page in a paper read before the Edinburgh Geological Society.[290] The superincumbent sandstone, he says, has been cut to a depth of 60 feet. The width of the channel at the surface varies from 12 to 14 feet, but gradually narrows to 2 or 3 feet at the bottom. The sides and bottom are smoothed and polished, and the whole is now filled with till and boulders.

(7.) One of the most remarkable buried channels is that along the Valley of Strathmore, supposed to be the ancient bed of the Tay. It extends from Dunkeld, the south of Blairgowrie, Ruthven, and Forfar, and enters the German Ocean at Lunan Bay. Its length is about 34 miles.

“No great river,” says Sir Charles Lyell, “follows this course, but it is marked everywhere by lakes or ponds, which afford shell-marl, swamps, and peat moss, commonly surrounded by ridges of detritus from 50 to 70 feet high, consisting in the lower part of till and boulders, and in the upper of stratified gravels, sand, loam, and clay, in some instances curved or contorted.”[291]

“It evidently marks an ancient line, by which, first, a great glacier descended from the mountains to the sea, and by which, secondly, at a later period, the principal water drainage of this country was effected.”[292]

(8.) A number of examples of ancient river courses, underneath the boulder clay, are detailed by Professor Geikie in his glacial drift of Scotland. Some of the cases described by him have acquired additional interest from the fact of their bearing decided testimony to the existence of inter-glacial warm periods. I shall briefly refer to a few of the cases described by him.

In driving a trial mine in a pit at Chapelhall, near Airdrie, the workmen came upon what they believed to be an old river course. At the end of the trial mine the ironstone, with its accompanying coal and fire-clay, were cut off at an angle of about 20° by a stiff, dark-coloured earth, stuck full of angular pieces of white sandstone, coal, and shale, with rounded pebbles of greenstone, basalt, quartz, &c. Above this lay a fine series of sand and clay beds. Above these stratified beds lay a depth of 50 or 60 feet of true boulder clay. The channel ran in the direction of north-east and south-west. Mr. Russell, of Chapelhall, informs Professor Geikie that another of the same kind, a mile farther to the north-west, had been traced in some of the pit workings.

“It is clear,” says Professor Geikie, “that whatever may be the true explanation of these channels and basins, they unquestionably belong to the period of the boulder clay. The Chapelhall basin lies, indeed, in a hollow of the carboniferous rocks, but its stratified sands and clays rest on an irregular floor of true till. The old channel near the banks of the Calder is likewise scooped out of sandstones and shales; but it has a coating of boulder clay, on which its finely-laminated sands and clays repose, as if the channel itself had once been filled with boulder clay, which was re-excavated to allow of the deposition of the stratified deposits. In all cases, a thick mantle of coarse, tumultuous boulder clay buries the whole.”[293]

Professor Geikie found between the mouth of the Pease Burn and St. Abb’s Head, Berwickshire, several ancient buried channels. One at the Menzie Cleuch, near Redheugh Shore, was filled to the brim with boulder clay. Another, the Lumsden Dean, half a mile to the east of Fast Castle, on the bank of the Carmichael Burn, near the parish church of Carmichael,—an old watercourse of the boulder clay period—is to be seen. The valley of the Mouse Water he instances as a remarkable example.

One or two he found in Ayrshire, and also one on the banks of the Lyne Water, a tributary of the Tweed.

(9.) In the valley of the Clyde, above Hamilton, several buried river channels have been observed. They are thus described by Mr. James Geikie:—[294]

“In the Wishaw district, two deep, winding troughs, filled with sand and fine gravel, have been traced over a considerable area in the coal workings.[295] These troughs form no feature at the surface, but are entirely concealed below a thick covering of boulder clay. They appear to be old stream courses, and are in all probability the pre-glacial ravines of the Calder Water and the Tillon Burn. The ‘sand-dyke’ that represents the pre-glacial course of the Calder Water runs for some distance parallel to the present course of the stream down to Wishaw House, where it is intersected by the Calder, and the deposits which choke it up are well seen in the steep wooded banks below the house and in the cliff on the opposite side. It next strikes to south-east, and is again well exposed on the road-side leading down from Wishaw to the Calder Water. From this point it has been traced underground, more or less continuously, as far as Wishaw Ironworks. Beyond this place the coal-seams sink to a greater depth, and therefore cease to be intersected by the ancient ravine, the course of which, however, may still be inferred from the evidence obtained during the sinking of shafts and trial borings. In all probability it runs south, and enters the old course of the Clyde a little below Cambusnethan House. Only a portion of the old ravine of the Tillon Burn is shown upon the Map. It is first met with in the coal-workings of Cleland Townhead (Sheet 31). From this place it winds underground in a southerly direction until it is intersected by the present Tillon Burn, a little north of Glencleland (Sheet 31). It now runs to south-west, keeping parallel to the burn, and crosses the valley of the Calder just immediately above the mouth of the Tillon. From this point it can be traced in pit-shafts, open-air sections, borings, and coal-workings, by Ravenscraig, Nether Johnstone, and Robberhall Belting, on to the Calder Water below Coursington Bridge (Sheet 31). It would thus appear that in pre-glacial times the Calder and the Tillon were independent streams, and that since glacial times the Calder Water, forsaking its pre-glacial course, has cut its way across the intervening ground, ploughing out deep ravines in the solid rocks, until eventually it united with the Tillon. Similar buried stream courses occur at other places. Thus, at Fairholme, near Larkhall, as already mentioned (par. 94), the pre-glacial course of the Avon has been traced in pit-shafts and borings for some distance to the north. Another old course, filled up with boulder clay, is exposed in a burn near Plotcock, a mile south-west from Millheugh; and a similar pre-glacial ravine was met with in the cement-stone workings at Calderwood.[296] Indeed, it might be said with truth that nearly all the rocky ravines through which the waters flow, especially in the carboniferous areas, are of post-glacial age—the pre-glacial courses lying concealed under masses of drift. Most frequently, however, the present courses of the streams are partly pre-glacial and partly post-glacial. In the pre-glacial portions the streams flow through boulder clay, in the post-glacial reaches their course, as just mentioned, is usually in rocky ravines. The Avon and the Calder, with their tributaries, afford numerous illustrations of these phenomena.”

The question naturally arises, When were those channels scooped out? To what geological period must those ancient rivers be referred? It will not do to conclude that those channels must be pre-glacial simply because they contain boulder clay. Had the glacial epoch been one unbroken period of cold, and the boulder clay one continuous formation, then the fact of finding boulder clay in those channels would show that they were pre-glacial. But when we find undoubted geological evidence of a warm condition of climate of long continuance, during the severest part of the glacial epoch, when the ice, to a great extent, must have disappeared, and water began to flow as usual down our valleys, all that can reasonably be inferred from the fact of finding till in those channels, is that they must be older than the till they contain. We cannot infer that they are older than all the till lying on the face of the country. The probability, however, is, that some of them are of pre-glacial and others of inter-glacial origin. That many of these channels have been used as watercourses during the glacial epoch, or rather during warm periods of that epoch, is certain, from the fact that they have been filled with boulder clay, then re-excavated, and finally filled up again with the clay.