CHAPTER XXX.

TUNNELS, VIADUCTS, EMBANKMENTS, AND WEIRS. In the laying out of a canal there generally comes a juncture at which the engineer has to choose between a tunnel, a flight of locks, a lift, or series of lifts, and, finally, an embankment. There are also cases, although these are comparatively rare, in which a valley has been crossed, and the level of the water maintained, by aqueducts. Tunnelling is always a costly operation, and it seldom happens that it gives a considerable advantage, if any, over locks in the matter of speed. Nevertheless, many of the early canal engineers were partial to tunnels, and hence there are many examples of such structures on the canal system of Great Britain. Of some of these we may suitably furnish particulars before proceeding to refer to more recent works of the same character. One of the earliest canal tunnels of which we have any record was constructed by Brindley on the Bridgwater canal. This tunnel gave the Duke access from his canal into the coal works at Worsley, and after it had proceeded for some way straight into the hill, came at a great depth to be under a small brook or constant stream of water, by the side of which a large water-shaft was sunk, and a drum and a large brake-wheel erected over it, of such size that a man who stood before the lever had his two hands at liberty to pull the lines which connected the valves, and give signals to those below, while by lunging or stepping forwards, with his breast against the lever, he could in an instant stop the machinery in any part of its motion, or regulate the same at pleasure. There were two water-tubs, which were very large, and had a valve and pin to empty themselves quickly when they arrived at the bottom. They were suspended by large ropes or cables from the drum, while other large ropes descended therefrom through another or coal-shaft, by the side of the middle or principal tunnel, into and over the navigable tunnel, which is there at some 60 yards lower level. On this level, canal boats were used, similar in their dimensions to those above, and containing boxes, which being filled with coals at the several terminations of the canal, in the seams of coals, were pushed along by means of rings fixed along the roof of the tunnel at the proper height for a man, who walked on the top of the coals, to lay hold of, and shove the boat along by. The boat having arrived under the coal shaft, and one of the water-tubs being at the top of its shaft, the coal rope answering thereto was hooked on to the box of coals, and the descent of the water-tub, immediately on the ringing of a bell, drew up the same to the level of the principal canal, where, being drawn aside over an empty boat, it was lowered into the same by a slight reversion of the motion of the machine, when the interval of emptying the tub at the bottom by its valve gave time for hooking another box to the other rope which was at the bottom, when the other water-tub was filled, and the machine was suffered to move by the man who leant against the brake. This arrangement was contrived and erected by James Brindley, and it was so constructed that when coals were not drawing, the alternate descent of the water-tubs worked some very large pumps, which were sufficient to lift all the mine water of the lower level into the middle canal, and to keep the lower canal always at the proper height for navigation. The same tunnel of the Bridgwater canal was continued a considerable way farther into Worsley Hill, until, under Walkden Moor, another subterraneous canal or tunnel begins, at 35½ yards higher level, being nearly 60 yards from the surface. From the surface two shafts were sunk, one terminating in and over the upper tunnel or canal, and the other in and over the middle or principal canal. There is another canal still lower, and after passing close by the canal above. Between these shafts a large drum was erected on the surface, with a brake-wheel and a pair of strong ropes. An old account of the working of this tunnel states, that “two boats being arrived at the shafts on the upper canal, one of them loaded with boxes of limestone that was wanted at the furnaces, and another with boxes of coals intended to be transferred into an empty boat in the middle canal, the ends of the two ropes were fastened to a box of coals and a box of limestone, when the superior size and weight of the coal boxes drew the limestone to the surface, to be there landed and deposited, at the same time as the box of coals was deposited in the lower boat, ready to proceed on the canal to Manchester or other places.” This method was, in 1797, superseded by an inclined plane for letting down the boats laden with coals from the higher to the middle level, and returning the empty boats and boxes. At Brierley Hill, near Coalbrook Dale, the extremity of a branch of the Shropshire canal, great quantities of coal and iron, in crates made of iron, were let down one of two shafts, which connected with the termination of the canal above, and the ends of a railway in a tunnel below, from which limestone in similar crates was drawn up the other shaft to be placed in the boat. A barrel and brake-wheel were fixed between the tops of the shafts, and cranes with jibs, by which the crates could be raised and moved from the boat over the shaft or the reverse. These shafts, which were 120 feet deep, were not found to answer, in point of expense, so well as inclined planes, and Mr. Telford informs us (‘Plymley’s Report,’ p. 296-307) that “inclined planes have been substituted, on which crates of coal or iron pigs, or goods descend, and draw up other crates containing limestone for the use of the ironworks above, by means of ropes, a drum, and brake-wheel, with a much less portion of manual labour, and more expedition, than was done by the shafts above mentioned.”[282] Marsden tunnel, on the Huddersfield Canal, is 5280 yards in length; Sapperton, on the Thames and Severn, 4300 yards; Penfax, on the Leominster and Kington, 3850 yards; Laplat, on the Dudley Canal, 3776 yards; Blisworth, on the Grand Junction Canal, 3080 yards; Ripley, on the Cromford, 3000 yards; Dudley, on the Dudley Canal, 2926 yards; Harecastle, on the Trent and Mersey canal, 2888 yards; Norwood, on the Chesterfield Canal, 2850 yards; Westheath, on the Worcester and Birmingham Canal, 2700 yards; Morwelham, on the Tavistock Canal, 2500 yards; Oxenhall, on the Hereford and Gloucester Canal, 2192 yards; and Braunston, on the Grand Junction, 2045 yards. The longest tunnels that have been proposed, besides those stated above, were one of 5 miles on the once proposed extension of the Manchester, Bolton, and Bury Canal to the Calder river; and one of 4¼ miles on the Portsmouth and Croydon Canal, through the chalk hills to the south of the latter place. The towns of Manchester, Kidderminster, and Southampton have been partly tunnelled under by the Bridgwater, the Stafford and Worcester, and the Southampton and Salisbury Canals respectively. [Illustration: _E & F N Spon London & New York_ “INK-PHOTO.” SPRAGUE & CO. LONDON. DUDLEY TUNNEL ON THE BIRMINGHAM CANAL (SHOWING MEN IN POSITION FOR “LEGGING”)] There are several long tunnels on the Birmingham Canal system, having an aggregate length of 6¼ miles. Two of these—the Dudley Tunnel and the Netherton Tunnel—pass under the Rowley Hills, and are each about two miles in length. The former was constructed in the last century. The waterway is about nine feet in width, and there is no towing-path, the boats being propelled by two men, lying on their backs on the boat, their feet performing a sort of walking motion against the sides of the tunnel, this is called “legging.” The Netherton Tunnel was constructed in the year 1858; it has a waterway 17 feet in width on either side. Both of these tunnels have, from time to time, been seriously injured by mining operations, and in the case of the Netherton Tunnel, the injury is stated to have been caused by the mine-owner illegally working minerals that had been previously purchased by the canal company. (See illustration of Dudley tunnel.) VIADUCTS. Sir Walter Scott spoke to Southey of the viaduct on the Ellesmere Canal as the most impressive work of art he had ever seen. This viaduct is situated about 4 miles to the north of Chirk, at the crossing of the Dee, in the romantic vale of Llangollen. The north bank of the river is very abrupt; but on the south side the acclivity is more gradual. The lowest part of the valley in which the river runs is 127 feet beneath the water-level of the canal; and it became a question with the engineer, whether the valley was to be crossed, as originally intended, by locking down one side and up the other which would have involved seven or eight locks, or by carrying it directly across by means of an aqueduct. The aqueduct is approached on the south side by an embankment, 1500 feet in length, extending from the level of the waterway in the canal until its perpendicular height at the “tip” is 97 feet. Thence it is carried to the opposite side of the valley, over the river Dee, upon piers supporting nineteen arches, extending for a length of 1007 feet. The height of the piers above the low water in the river is 121 feet. The lower part of each was built solid for 70 feet, all above being hollow, for the purpose of saving masonry as well as ensuring good workmanship. The outer walls of the hollow portion are only two feet thick, with cross inner walls. Upon the top of the masonry was set the cast iron trough for the canal, with its towing-path and side rails, all accurately fitted and bolted together, forming a completely watertight canal, with a waterway of 11 feet 10 inches, of which the towing-path, standing upon iron pillars rising from the bed of the canal, occupied 4 feet, 8 inches, leaving a space of 7 feet, 2 inches for the boat. The whole cost of this part of the canal was 47,018_l._, which was considered by Telford a moderate sum compared with what it must have cost if executed after the ordinary manner. The aqueduct was formally opened for traffic in 1805. “And thus,” says Telford, “has been added a striking feature to the beautiful vale of Llangollen, where formerly was the fastness of Owen Glendower, but which, now cleared of its entangled woods, contains a useful line of intercourse between England and Ireland; and the water drawn from the once sacred Devon furnishes the means of distributing prosperity over the adjacent land of the Saxons.” The Barton Aqueduct on the Bridgwater Canal, is about 200 yards in length, and 12 yards wide, the centre part being sustained by a bridge of three semi-circular arches, the middle one being of 63 feet span. It carries the canal over the Irwell at a height of 39 feet above the river—this head room being sufficient to enable the largest barges to pass underneath without lowering their masts. The bridge is entirely of stone blocks, those on the faces being dressed on the front, beds, and joints, and with cramped iron. The canal, in passing over the arches, is confined within a puddled channel to prevent leakage, and is in as good a state now as on the day on which it was completed. The embankments formed across the low grounds on either side of the Barton viaduct were considered very formidable works at that day. A contemporary writer speaks of the embankment across Stretford meadows as “an amazing bank of earth, 900 yards long, 112 feet in breadth across the base, 24 feet at the top, and 17 feet high.” The greatest difficulty anticipated was the holding of so large a body of water within a hollow channel formed of soft materials. It was supposed at first that the water would soak through the bank, which its weight would soon burst, and wash away all before it. But Brindley, in the course of his experience, had learnt something of the powers of clay puddle to resist the passage of water, and he finished the bed of this canal, so as to make it impervious to water. Not the least difficult part of this undertaking was the formation of the canal across Trafford Moss, where the weight of the embankment pressed down and “blew up” the soft oozy stuff on either side; but the difficulty was again overcome by clay puddle. Indeed, the execution of these embankments by Brindley was regarded at that time as something quite as extraordinary in their way as the erection of the Barton Aqueduct itself. EMBANKMENTS AND WEIRS. Mr. Jebb has pointed out[283] that one of the most important duties of the canal engineer, and certainly one of the most anxious, is to take all practicable precautions for the prevention of any of the embankments giving way by the overflow of water during heavy rainfalls. In some districts at such times an enormous volume of water discharges directly into the canal; this has to be got rid of. Self-acting weirs are constructed at convenient points, and these are sufficient to keep the water within bounds at ordinary times; but in times of flood other means have to be used. The old canal “let-off,” as it was called, consisted of a wooden frame (fixed in the bed of the canal), to which was attached a hinged lid; this lid was pulled up by a chain fixed to the lid when necessity required—that is, if the chain could be found, and also sufficient power obtained for the purpose; for when the let-off had not been used for a considerable time it became covered with mud, and it was often as much as half-a-dozen men or a horse could do to pull up; this accomplished, however, the water rushed out at once with great force (as there was no means of regulating the discharge), the sudden rush often causing trouble with the owners and occupiers of the adjoining lands. Mr. Jebb has replaced some scores of these let-offs by sluice valves of similar capacity, worked by racks and pinions. The discharge of water can thus be exactly regulated, and one man only is required to work them. The valves are tested every month to see that they are in working order. For the proper and economical maintenance of the towing paths, it is necessary to have a staff of experienced men. Mr. Jebb recommends, as a material for metalling, limestone _débris_, or what is locally known in Birmingham as “raffil” or “bavin.” He finds that it sets soon, and lasts for years if properly laid down—broken furnace cinders, covered with good ashes, are largely used in the Black Country—the paths should, of course, be well drained. On the Birmingham canal, between Longford and Manchester, the sidelong ground was cut down on the upper side and embanked upon the other by means of the excavated earth. This was comparatively easy work; but a matter of greater difficulty was to accommodate the streams which flowed across the course of the canal. For instance, a stream called Cornbrook was found too high to pass under the canal at its natural level. Accordingly Brindley contrived a weir, over which the stream fell into a large basin, from whence it flowed into a small one, open at the bottom. From this point a culvert, constructed under the bed of the canal, carried the waters across to a well, situated on its further side, where the waters, rising up to their natural level, again flowed away in their proper channel. A similar expedient was adopted at the Manchester terminus of the canal, at the point at which it joins the waters of the Medlock. It was a principle of Brindley’s never to permit the waters of any river or brook to intermix with those of the canal, except for the purpose of supply; as it was clear that in a time of flood such intermingling would be a source of great danger to the navigation. In order, therefore, to provide for the free passage of the Medlock, without causing a rush into the canal, a weir was contrived, 306 yards in circumference, over which its waters flowed into a lower level, and thence to a well several yards in depth, down which the whole river fell. It was received at the bottom in a subterranean passage, by which it passed into the river Irwell, close at hand. In the earlier attempts made in the last century to deal with the cataract of Trolhätta, in Sweden, it was determined to distribute the whole fall of 113⅓ feet among three sluices only: the first to consist of 28, the second of 52, and the third of 33⅓ feet. These sluices were to be constructed alongside of the three cataracts, and were to be each 18 feet wide by 72 in length. The work advanced successfully, until the attempt to throw a weir across the river at the gulf of the last cataract, to raise and retain the water above it. The impetuosity with which the whole stream is precipitated had prevented the builders from sufficiently examining the bottom. They had conjectured, from the nature of the neighbouring mountains, that the bottom must be rock; and it was further supposed that there could not be more than 10 feet of water. Both these suppositions proved to be erroneous. The depth of the water was from 20 to 25 feet at least, and the bottom was composed of large detached stones, which were incapable of being fixed by any efforts of art. The caissons of stone, although fastened together with cramps 4 inches thick, and attached by great piles to the mountains on both flanks, were swept off and dispersed by the impetuosity of the current; and in this manner all the works were destroyed. Subsequently it was determined to avoid the pass entirely, and construct a canal 8200 feet long; and the total fall of 113⅓ feet was to be distributed, in the space of the last 3000 feet, among seven sluices or locks, each 36 feet in breadth by 200 in length. The first sluice was to be 17⅓ feet in height; the others, 16 feet. The first sluice was to stand alone; but the four following were to be close to each other, as were also the last two. Between the fifth and sixth sluice the canal was to be protected by a strong dyke against the floods of the river. There was to be a great discharger between the first sluice and the water entrance, not far from the centre; and at the entrance itself two doors or gates, to lay the canal dry when required. This plan proved more successful than the first. Forty wholly removable, regulating weirs were constructed in the Seine several years ago. When wholly closed up in summer, they maintain the required depth of water for steamboat navigation. When wholly open in floods, they cause no stoppage in the river surface. A remarkable barrage mobile has been in action for several years at a place called Port à l’Anglais, above Paris, and above the junction of the Seine and the Marne. When all is open there is not a ripple on the river flowing by. M. Gambuzat, the chief engineer of the river Seine, informed Mr. Lynam that all those wholly removable regulating weirs in the Seine were remarkably effective, and suitable for regulating that great commercial river. Mr. Lynam declared that, if in July 1861, a month previously to the great flood, the Killaloe weir-mound had been wholly removed, and a wholly removable weir, like that in the Seine at Port à l’Anglais, had been constructed, and subsequently been properly manœuvred, during the month of August, none of the crops in the level of the Shannon, above the Killaloe weir-mound, would have been materially injured.[284] The cost of high weirs on large rivers is considerable. For instance, the most recent weir on the Seine at Poses, retaining a depth of 16½ feet of water, cost 151_l._ 5_s._ per lineal foot; and the Mulatière weir, on the Saône at Lyons, retaining a depth of water of 10 feet, cost 118_l._ 11_s._ 7_d._ per lineal foot. On all navigations and canalised rivers liable to floods, the great difficulty is to be able to pass away the water without impeding the traffic, and without flooding the surrounding country. This has been accomplished on the Weaver to a very great extent by means of what are, practically, movable weirs, at Dutton, Saltersford, Hunts, and Valeroyal. They are flood-gates, or sluices, capable of being lifted clear of the water, and thus allowing an uninterrupted passage, and consist of doors 15 feet by 14 feet deep, built of rolled iron beams with timber sheathing. These are supported by masonry piers, and are lifted by means of overhead gearing, so that the attendants are entirely above water and on a permanent bridge. Friction is practically dispensed with, owing to their working on rollers. The rollers hold the doors from their seating, and would thus allow the passage of the water. To prevent this, “stopwaters” have been introduced, consisting of pieces of hard wood weighted at one end, until the specific gravity is about the same as that of water; they then float vertically, and are held in such a position that the pressure of the water forces them into the angle formed between the door and the masonry. This plan of sluice has practically reduced by one-half the flood level at Northwich, and instead of having floods of 8 to 12 feet, the highest that has occurred since their erection is one of 6 feet. On the Aire and Calder Canal a form of sluice has been invented and applied by Mr. Bartholomew which appears to have merits and originality. A large culvert is made alongside the whole length of the lock, with a very large sluice at the upper end, measuring 7 feet by 5 feet, the ordinary sluice being 2 or 3 feet square. Another sluice is provided at the other end, and when this is closed and the lock is empty, the upper sluice, which is self-balanced, like a throttle-valve, is raised. Three orifices are made into the elongated lock, which are arranged in such a way that the vessels are divided, and do not knock against each other while in the lock. In emptying the lock, the upper sluice is let down and the lower sluice is drawn, the water entering the culvert through the orifices and discharging at the lower end. In working the sluices, a man only requires to turn the handle and it raises itself, while with three turns in the other direction it is lowered. The locks on this system are 215 feet long, 22 feet wide, and have 9 feet of water on the sills. DAMS. The proposed dam on the Nicaraguan Canal is to be of concrete, faced with timber, and will be 1225 feet long on the crest, and 52 feet high. The embankment will be 6500 feet long and 51 feet high in the centre. There are, however, much larger dams than this. Of masonry dams, Verviers, a small city of Belgium, near the frontier of Prussia, with a population of about 38,000, has one—that of Gilleppe—154 feet high and 771 feet long. The water supply of the town of St. Chaumonde, in France, has a dam about 140 feet high, and the water supply of St. Etienne is held by the Furens dam, 170 feet high. The Villar dam, 162 feet high, holds the water supply of Madrid and other dams in Spain, some of them dating back to Moorish days—Puentes, Alicante, Val de Infierno, Nijar, Elche, and Almanza range from 164 feet to 68 feet in height. In England the Vyrnwy dam, at the Liverpool waterworks is 136 feet high and 1255 feet long. The San Francisco waterworks dam, 170 feet high and 700 feet long, and the Quaker Bridge dam, 278 feet high and 1300 feet long, will, when built, be still larger. Of earthen dams or embankments, some of the most notable are the Montaubry dam, on the Canal du Centre, 54 feet high; the dam, 66 feet high, by which the water supply of Dublin is impounded; the reservoir dam of the Bolton waterworks, England, over 120 feet high; the Oued Muerad dam in Algeria, 95 feet high. In India and Ceylon such examples are very numerous; the embankment of the Ashti reservoir is 58 feet high and 12,709 feet long; the Karakvasla dam is over 70 feet high; the Tansa reservoir dam (water supply of Bombay) is to be 8500 feet long and 118 feet high; the embankment of the Cummum tank in the Madras Presidency is 102 feet high, and although it ranks among the earliest works of Hindoo history, it is still in such condition as to fulfil its original intention. In Ceylon there are old tanks with embankments from 3 to 12 miles long and 50 feet to 70 feet high. The materials used for the construction of a weir or dam across a river are principally earth, timber, fascines, stone, &c. The most simple form of dam is that made of gravel protected by fascines kept in place by piles. Such dams are principally used for temporary works. Dams are often made of timber, stones, and earth combined, and covered with planking laid parallel to the current, and the bottom of the channel at the foot on the downstream side should be protected by an apron formed of a platform of planks resting on piles, or by a stone pitching. Dams of this kind built of dry stone and timber often do not become weirs except during floods; that is to say, the water does not pass over their crests except at such times, and at other seasons of the year any surplus finds its way through the interstices between the stones. Dams may be built of caissons of strong timber, filled with loose stones and covered with planking; others are filled with earth instead. A recent writer states that weirs of solid masonry, like other hydraulic works, should be founded on the natural ground on a bed of concrete, or on piles, according to circumstances. The masonry may be built in cement or hydraulic lime; the face-work is usually in dressed stone or blocks. The stones, besides being fastened together by metal cramps, are sometimes bonded by dovetailing. A good example of a masonry weir is that built across the Dora Baltea for obtaining a supply of water for the subsidiary canal of the Canal Cavour. This work consists of a mass of concrete faced with ashlar and blocks in courses roughly dressed. The crest is 1·20 metre in width, and the total length 200 metres. This dam cost 237,682 francs, or at the rate of 1188·41 francs per lineal metre. A layer of concrete alone forms a very effectual protection to a river or canal embankment. In rivers subject to excessive floods a rock-work consisting of large irregular-shaped blocks of stone—not less than one-third of a cubic metre each—is exceedingly useful for protecting the bottom of embankments or walls from scour. FOOTNOTES: [282] Paper by Mr. G. R. Jebb in the ‘Journal of the Society of Arts,’ for 1888. [283] Paper on “The Maintenance of Canals,” in the ‘Journal of the Society of Arts’ for 1888. [284] Paper read before the British Association, 1878,.