obtained in 1759, was to carry a canal over the river Irwell, near Barton Bridge, to Manchester, and to. conduct a branch to Longford Bridge in Stratford. This was to be accomplished without the aid of locks, by preserving the same level through the whole course of the canal. After many difficulties had been surmounted, of sufficient magnitude to have deterred an ordinary man from the undertaking, Brindley commenced that which was by far the most gigantic part of the whole, viz., to carry the canal over the river Irwell, at a height of thirty-nine feet above the surface of the water. Brindley felt confident of the practicability of the design; but at the same time he wished the duke to take the opinion of some able engineer before the attempt was made; for though such an enterprise is now very common, at that time it was an entire novelty. A gentleman was accordingly consulted, to whom the scheme appeared to demand ridicule rather than deliberation; he is stated to have said, "that he had often heard of castles in the air, but was never before shown where any of them might be erected." But neither Brindley's confidence, nor the duke's acquiescence in his judgment, was shaken by this declaration; the work was begun in September, 1760, and in the July of the year following, a boat floated along the aqueduct. The design extended with the progress of the work; and another branch was opened from the canal, which was to be carried over the rivers Mersey and Bollan, besides many deep valleys, in its extension to the tide-way of the Mersey. Here the obstruction of locks was also avoided; high mounds of earth being raised across the valleys, the ridges of which became the bed of the canal.

The success of the Duke of Bridgewater's canal encouraged a number of gentlemen and manufacturers to entertain the idea of a navigable canal through the county of Stafford; and Brindley was engaged to survey it from the Trent to the Mersey. He reported that it was practicable to construct a canal from the one river to the other, and thereby connect the ports of Liverpool and Hull, on the two opposite coasts of the kingdom. The report being deemed favourable, and the requisite funds subscribed, this great work was commenced by Brindley in 1766, and finished by his brother-in-law in 1777. It was called the Grand Trunk canal, in reference to its importance, as a centre whence others might spring; it is ninety-three miles in length, and besides a vast number of bridges, has seventysix locks and five tunnels. The most remarkable of these tunnels is the subterraneous passage of Harecastle, nearly three thousand yards in length, and more than seventy yards below the surface of the earth. It appears that the scheme of this canal had occupied the thoughts of engineers for twenty years before, and that Harecastle hill formed the one great difficulty which baffled them all; until Brindley took up the subject, and accomplished it.

While the Grand Trunk canal was in progress, Brindley designed and executed a canal leading therefrom to the Severn; thus effecting the important object of bringing Bristol, as well as Hull and Liverpool, into mutual waterintercommunication; this canal is forty-six miles in length, and was completed in 1772. Brindley did not live to see the Metropolis brought into connexion with the northern parts by similar means; but he planned all the arrangements for the Coventry and the Oxfordshire canals, by which the Grand Trunk was connected with the Thames.

Brindley died in 1772, after having planned a vast number of canals, which were left to be executed or finished by those who succeeded him. He was beyond all comparison the greatest canal-engineer that ever lived; and so entirely were his thoughts absorbed by this subject, that it is related of him, when asked in the House of Commons for what purpose he supposed rivers to have been formed, he replied, "to feed navigable canals." In relation to Brindley's mental powers, Mr. Phillips states:-" When any extraor dinary difficulty occurred to Brindley in the execution of his works, having little or no assistance from books or the labours of other men, his resources lay within himself; in order, therefore, to be quiet and uninterrupted whilst he was in search of expedients, he generally retired to his bed, and has been known to lie there one, two, or three days, till he had attained the object in view; he would then get up, and execute his design without drawing or model-indeed it never was his custom to make either, unless to satisfy his employers. His memory was so remarkable, that he has often declared he could remember and execute all the most complex machines, provided he had sufficient time to settle in his mind the several departments, and their relation to each

other. His method of calculating the powers of any machine invented by him, was peculiar to himself; he worked the question some time in his head, and then put down the result in figures: after this, taking it up again in that stage he worked it further in his mind for a certain time, and set down the result as before; in the same way he still proceeded, making use of figures only at stated periods of the question; yet the ultimate result was generally true, though the road he travelled in search of it, was unknown to all but himself, and it was not in his power to have shown it to another. The attention which was paid by Brindley to objects of peculiar magnitude, did not permit him the common diversions of life. He never seemed in his element if not planning or executing some great work, or conversing with his friends upon subjects of importance. He was once prevailed on, when in London, to see a play, having never seen one before; it had a powerful effect upon him, and he complained for several days after that his ideas were disturbed, and rendered him unfit for business. He declared that he would not go to see another play on any account.' After the death of Brindley, other engineers took up the prosecution of enterprises which he had taught them how to surmount; and in a report on the advantages of canals, an engineer makes the following remarks, which may be taken as a representative of the tone of feeling on the subject at that time,-a tone remarkably similar to that recently manifested in respect of railroads. "The advantages of these canals to this nation would be very great, as every person must allow that the stamina of our trade is the internal produce of the kingdom; and the nearer each manufacturer is to his market, the greater price his goods will take; for instance, if a person manufactures a piece of cloth (such as is usually sent abroad to America or else where) in the county of Stafford where he lives, forty miles from the merchant's ship which is to carry it abroad, will he not be obliged to sell it, upon account of the heavy charge of land carriage, for a much less price than if he could convey it at an easy rate to a sea-port, by a canal, and there sell it at the same price that a factor would? Let us only consider that the factor buys up the goods or manufactures of fifty poor people, and conveys them to a merchant, at a large profit, which falls into the hands of the factor alone, and enriches only one; whereas, were each manufacturer to have the conveniency of conveying his own goods directly to the merchant, all the profits monopolized by the factor would be divided between the fifty poor persons, and enrich that number instead of one. The more hands any goods or merchandize go through, the dearer they are to the consumer, as every person through whose shop they go must have a profit; but in this case the factor monopolizes the whole profits of fifty poor manufacturers. In case of invasion or rebellion, by these canals Government would be able to transport their heaviest cannon to any part of the country in a short time: as also regiments with their baggage might be conveyed in a much safer way than by long, tedious harassing marches, through bad roads that are almost impassable, and oftentimes far round to the point they want to come to; besides the saving in artillery horses, and the expense of baggagewaggons to Government, which usually amounts to nearly as much as the pay of one-third part of the regiment for the time." Recent occurrences, happily rare in this country, have shown how vast an advantage railroads present for the transport of troops, beyond anything which canals could afford.

ENGLISH CANALS AT THE PRESENT DAY.

The aggregate length of the navigable canals of England is now more than twenty-two hundred miles. In addition to making canals, such of the rivers as are capable of it, have been made navigable; so that it is said there is no spot in England, south of Durham, which is more than fifteen miles distant from water communication; and in those districts which form the chief seats of manufactures, the distance is even smaller; while every considerable town of those manufacturing districts has this means of cheap and easy conveyance for receiving the raw materials which it employs, and for distributing to other parts of the country, and to the sea-ports, the finished products of its industry. There are upwards of a hundred canals in England and Wales alone, exclusive of very short cuts.

Of these several

are of considerable length, and have to surmount acclivities by means of a large number of locks. For example:-The Birmingham and Liverpool canal is forty miles in length, and has twenty-seven locks; the Chesterfield canal is forty-six

705-2

In some respects the Caledonian canal, extending across Scotland from sea to sea, is more remarkable than any canal in England, as being a government project, and much wider and deeper than any canal in Great Britain-or indeed, in the world, except the Great Ship Canal of Holland. A valley remarkable for its uniformity, straightness, and depth, and extending from sea to sea through two parallel ranges of steep mountains, divides the Highlands of Scotland into two nearly equal parts. The general direction of this chasm is from north-east to south-west; and besides being entered at each extremity by an arm of the sea, the rest of its bottom is for the most part occupied by a chain of rivers and lakes. The remarkably elongated form and contiguity of these lakes had long ago suggested the facility of forming an inland communication between the Atlantic Ocean and the German Sea; for it required only, in a distance of one hundred miles, to cut small canals to the aggregate length of twenty-three miles, as a means of joining the several lakes. In the year 1773, this line was surveyed by James Watt, who reported favourably of it, and proposed that the lakes should be connected by a canal of a very moderate size. Nothing farther was done till the early part of the present century, when Telford and Jessop were employed to make new surveys. These engineers recommended a canal of such dimensions as should admit frigates of thirty-two guns, and merchant ships of 1000 tons burden, particularly those which had been accustomed to make the voyage from Ireland to the eastern parts of Scotland round the dangerous passage of the Orkneys. The plan was agreed to, the public money voted, and the dimensions fixed on as follows: width of the canal one hundred and twenty feet at top and fifty feet at bottom, depth twenty feet, locks one hundred and eighty feet long, forty feet wide, and with a rise or fall of twenty feet. Operations were commenced in 1803, and were continued for the long period of twenty years before their completion. The canal commences on the north-east at about two miles from Inverness, where a sea-lock opens a communication with the open sea; and from thence it proceeds towards Loch Ness, to the level of which it is raised by means of four locks. This loch is a fine sheet of water twenty-four miles long by more than a mile broad, and forms part of the connected chain of water communication. Another short canal brings us to Loch Oich, one of very small dimensions, and so shallow in some parts that it had to be deepened for the purposes of the canal. A short cut connects this loch with Loch Lochy, to which succeeds Loch Eil. At the latter place there is a remarkable assemblage of eight connected locks, called "Neptune's stairs," whereby the canal descends sixty-four feet, and soon afterwards joins the Atlantic. As a commercial enterprise this canal has proved a total failure; the expenditure has been enormous, and the receipts, obtained by tolls on the vessels and boats passing through the canal, have never been equal to the annual expense of management. For many years past, the Government is scarcely decided whether to abandon the canal altogether, or to expend large additional sums in keeping it in repair.

The canals of Scotland, including the Caledonian, are less than three hundred miles in aggregate length. The Edinburgh and Glasgow is about thirty miles in length; the Forth and Clyde thirty-five miles, with about forty locks; and all the others, about seven or eight in number, are small. The Irish canals are on rather a larger scale than those of Scotland, with the exception of the Caledonian. All canals, whether foreign or British, whether of large or small dimensions, whether for the passage of sailing vessels or for small barges, are constructed on certain prin

CONSTRUCTION OF CANALS.-AQUEDUCTS.-EMBANKMENTS. The main principle whereon all canals are constructed is this, that the boat or barge shall pass along a perfect level, neither inclining in the one direction nor the other; since, if this level were not attained, the water would form a stream from the upper to the lower parts, and the boats would pass up the canal at a great disadvantage. If the ground over which the canal passes should be, as is generally the case, unequal and at different levels, the effect is attained by three modes of proceeding: 1st. by elevating the depressed portions, by means of embankments and aqueducts; 2nd. depressing the elevated portions, by means of cuttings and tunnels; and 3rd. forming a kind of series of stairs, called locks, whereby one level portion is brought into communication with another higher or lower than itself. It rests with the engineer, after a careful examination of the ground, to determine in what proportions these three methods may be most conveniently combined; for we may remark that in almost every canal, all the methods are more or less employed. The preparative survey has especial reference to this among other points; for the general level at which the canal is fixed must have a vast influence on the supply of water obtained from it, as well as the facility of junction with other canals. It is a general rule, observed as nearly as possible in the construction of a canal, to proportion the excavations and embankments so that the quantity of earth procured from the one shall be just sufficient for the formation of the other: Were this not the case, either a supply of earth would have to be obtained at great expense from some other quarter, or else a useless mound of earth would be accumulated, which would be no easy matter to get rid of. Sometimes a canal runs along the side of a declivity, in which case either one bank will require more raising than the other, or one will have to be excavated while the other is embanked; and sometimes alternate embankments and excavations succeed each other with great frequency; but it is generally contrived that the entire mass of earth accumulated in the one case shall nearly correspond with that required in the other. It sometimes happens, however, when the cuttings would have to be made through solid rock, the expense would be too great if this rule were abided by, and a deviation thus results. All these are points requiring great attention and skill on the part of the engineer.

It

Canals are not often square or perpendicular at their margins. The bottom generally slopes off upwards, the slope being regulated by the quality and difference in the tenacity of the soil. The slope sometimes becomes a good deal worn by the motion of the water, insomuch as to diminish irregularly the width of the towing-path; this is remedied in some degree by planting the sides with aquatic shrubs and also by driving stakes into the banks at short intervals along the water-line, so as to form a rude wickerwork, in places where the soil is loose and porous. frequently happens that canals have to be formed more or less by artificial embankments of loose materials, or that they are excavated in earth, sand, rock, &c., so very porous as to allow the water to escape by filtration. To remedy this, the usual way is to make the excavation much wider and deeper than the intended dimensions, and to line it to the thickness of two or three feet with a tenacious clay mixed with gravel. A trench, three or four feet in width, is also dug out in the middle of each side-bank, to at least three feet below the bottom of the canal, and this in like manner is filled up with clay; this trench is made principally with the view of preventing rats and vermin from perforating the banks and occasioning breaches. When once the water has attained a very small egress, it gradually washes it wider; so that sometimes in a few hours a breach may be formed sufficient to empty the canal, and require weeks or even months for its repair.

When a canal has to be carried over a valley, two plans are adopted according to circumstances; the one to form an artificial mound or embankment, on the surface of which the canal flows; the other to construct a kind of bridge of timber, stone, brick, or a combination of two or all of these. If the valley is very deep or extensive, the latter plan is more generally adopted. As an instance of an iron aque

duct, we may mention that constructed by Telford for the Shrewsbury canal. This canal passes over the valley of the Tern, for a distance of a hundred and eighty feet, upon an aqueduct made wholly of cast iron, excepting only the nuts and screws, which are of wrought iron. Another aqueduct, made by Mr. Jessop on the Ellesmere canal for crossing the river Dee about twenty miles from Chester, is an unusually large specimen of canal aqueducts. Nineteen massive conical pillars of stone, fifty-two feet apart, the middle one being no less than a hundred and twenty-six feet in height, support a number of elliptical cast-iron ribs, which by means of upright and horizontal bars support a castiron aqueduct about a thousand feet in length, twenty wide and six in depth, composed of massive sheets of cast-iron, cemented and rivetted together, having on its south side an iron platform and railing for the towing-path. There are not, even among the viaducts of modern railroads, many structures which will exceed in boldness and magnitude this aqueduct in the Ellesmere canal.

Embankments are formed by depositing loose earth in the line which the canal is to follow, to such a height as will bring the canal to a proper level. Between Wolverton and Cosgrove, on the Grand Junction canal, a stupendous embankment has been executed; it is half a mile in length, and in some parts thirty feet high, and by its construction a great number of locks have been avoided. A still larger amount of lofty embankment occurs over the valley of the Boyne, in Ireland; and on the same canal is one embank ment reaching to the enormous height of ninety feet.

CUTTINGS.-TUNNELS.

Where the canal is to be carried at a much lower level than the surface of the ground, the soil is either cut away down to the proper level to form a channel for the canal, or a hole or tunnel is bored through, large enough conveniently to receive the canal. One or other of these plans is adopted according to influencing circumstances, among which are the depth below the surface, the nature of the soil, and the value of the land above. In the mining districts long tunnels or adits to mines were constructed long before the introduction of canals; thus, in the neighbourhood of Matlock in Derbyshire, the Helcarr Lough has been cut through the solid rock for nearly four miles for the purpose of draining the several lead-mines in the vicinity; Worksworth Moor Lough, nearly three miles in length, and Cromford Lough, of two miles, with many others of lesser note, are also to be found at great depths in that neighbourhood. But the first tunnel that we read of expressly intended for canal navigation, was that on the the Languedoc canal, in France; and the first in this country was the entrance made by Brindley to the Duke of Bridgewater's canal at Worsley.

In the construction of a tunnel, as the workmen proceed mole-like under ground, it is necessary to take scrupulous precautions in reference to the level and bearings of the line. The first thing to be attended to is to execute a survey of the ground through which it is to pass; this is done by tracing a line in a vertical plane parallel to the direction of the tunnel; to obtain which the relative levels of the principal points along the line upon the surface are ascertained. At certain spots along this line vertical pits or shafts are dug, either for the sake of free ventilation, or for commencing and carrying on the excavation of the tunnel at several different places at the same time; through the shafts made for the latter purpose the earth is hoisted out from the tunnel, and the water pumped out if any should occur. The shafts being sunk to the requisite depth, which is ascertained from the levels which have been taken at the surface, a heading, or small tunnel, at least sufficiently high and wide to allow the men to move freely in the work of excavation, is commenced just below the crown of the intended tunnel; it is carried forward from one working shaft to another, as nearly as may be in a straight line, till a connexion is formed between all the shafts and the two ends of the tunnel. If, in the progress of the heading from one shaft to another, the air becomes so bad as to oblige the workmen to desist, an air shaft is commenced from above, and sunk vertically to the point at which the heading was left off. Thus, in excavating the tunnel of the Thames and Medway canal, the air was found to be so foul that eleven air-shafts had to be sunk in addition to twelve workingshafts. After the heading has been driven from one workingshaft to another, or a complete perforation obtained throughout the whole of the intended line, the roof of the tunnel

is commenced, and is left either in a smooth state, or is lined with brick or stone, according to the nature of the ground.

CANAL-LOCKS.

One of the most important and remarkable portions of a level, or a descent to a lower. In engineering language the canal is the lock, by which an ascent is made to a higher name of pound is applied to the perfectly level portions of a canal, between the locks, so that the canal consists of an alternating series of "locks" and "pounds." The frequency with which these alternations occur depends upon the undulations of the ground; for if the ground ascends uninterruptedly, or if it ascend and descend at short intervals, then the level "pounds" must necessarily be short, because locks must be interposed with much frequency; whereas if the ground be tolerably level, very few locks will occur. Thus it is found, in one English canal, that locks are required at intervals averaging less than half a mile each; while in another instance a canal nineteen miles in length has only one lock.

The lock may be compared to a liquid stair, down which or up which a laden boat proceeds by a change being effected in the level of the water in the lock; the boat does not cease to float, but the water on which it floats is elevated or depressed by the increase or decrease of its quantity, and the boat is influenced by this elevation or depression. If the canal sloped downwards, a boat might proceed in that direction by the force of the current only, but it could not proceed up the stream without a powerful tractive force. The aim is to equalize the draught up and down the canal, and this can only be effected by having a series of dead levels.

A rude approximation to the gates of a lock, still in use in some parts of England, is the following. A beam or sill is fixed across the bottom of the stream or canal; and directly over this, but at the water's surface, is placed a second but moveable beam. Against and behind these parallel beams a set of loose boards are placed upright and close together like a door, so as to obstruct the progress of the stream. When a boat is to pass, the boards are either removed singly, or if they can be prevented from floating away, the upper beam is lifted or moved round, so as suddenly to let go the whole system, and thus afford a temporary increase of depth, to enable the boat to pass or repass the otherwise too shallow water. Since these rude kinds of gates are supposed to be placed at the upper end of the fall or shoal over which they are temporarily to increase the depth of water, and since the boat is to pass through the gate while the accumulated water rushes out with great force, it is evident that this must render it difficult for a boat to ascend the stream, and must accelerate it too much in descending. For the navigation of boats, a preferable mode of using such gates is to place the gate in moderately deep water at the lower end of the fall or shoal, and such that a boat may pass up through the gate before the water has been raised by it; the gate is then shut till the water rises sufficiently to enable the boat to ascend the fall or shoal, which it does in almost still-water. This arrangement constitutes a kind of half-lock; and indeed it is supposed that the casual position of two of them near each other had given rise to the invention of the lock; for in that case it would soon be seen, that when the lower of these two gates was closed, and the water above it had risen to a sufficient height, such water would be nearly stationary at the upper gate, and would afford an easy passage through it; if, therefore, the boat were ascending, the upper gate being next shut, the water above it would rise, and enable the boat to proceed a stage further in the ascent.

Whether or not the invention of the lock was derived from the contrivance just noticed, the action of it may be briefly described as follows:-A single lock is generally an oblong chamber, about seventy or eighty feet long, seven or eight wide, and lined with brick. It has a gate at each end, which, when open, connects it with the upper and lower "pounds" of the canal; but when closed cuts off this communication. We will suppose that a barge is about to pass upwards through the lock. In order that the barge may enter the lock, the following conditions must be observed: 1st. that the upper gate is shut; 2nd. that the lower gate is open; and 3rd. as a necessary consequence of the second, that the water in the lock is on the same level as that in the lower pound. These being the conditions, the barge is floated into the lock, and the lower gate closed. Water is then admitted through a valve from the upper

pound, till the lock is filled to the level of the upper pound. As the water rises, so does the barge also; and when the lock is filled, the upper gate is opened, and the barge emerges from the lock, thence to pursue her journey along the upper pound of the canal. If on the contrary the barge is about to descend, the conditions must be thus: 1st. the lower gate must be closed; 2nd. the upper gate open; and 3rd. the water in the lock must be on the same level as that in the upper pound. The barge enters the lock, and the gate is closed behind it; then a valve is opened by which water flows out of the lock into the lower pound, until both are at the same level. The lower gate then opens, and the barge is floated out. This operation is called "locking-down' "the canal, whereas the reverse operation is that of "locking-up.”

By a little consideration of this routine, we shall see that an inevitable waste of water occurs; for in order that the lock, when on a level with the upper pound, may be reduced to that of the lower for the transit of a barge downwards, a quantity of water must flow from the lock into the lower pound equal to their difference of level; and when, the lock being at the low level, the level is required to be raised in order that a barge may pass upwards through the lock, water must flow from the upper pound into the lock till the latter is filled to the highest level. Hence the water always descends, at each passage of a boat; and a certain portion of water is drawn from the summit level by this means. In order to prevent or at least to lessen the serious loss, which in dry seasons sometimes threatens to exhaust the canal, engineers frequently adopt the use of double-locks, by which half the waste of water is prevented. A double-lock consists of two oblong chambers, ranged side by side, and having a sluice in the brick partition which separates them; each chamber has a gate at each end, the upper connected with the upper pound of the canal, and the lower with the lower. Supposing one chamber to be filled to the higher level, and the other to the lower, and a barge about to ascend, the ascent would be managed thus: -The barge would flow into the chamber having the lower level of the water, and the gate would be closed behind it; then all four gates being closed, the sluice in the partition wall would be opened, whereby half the water in the full chamber would flow into that which contains the barge, till both are on the same level. Then the sluice would be closed, and a valve opened, by which the chamber containing the barge would be filled with water to the level of the higher pound; after which, the upper gate of that chamber being opened, the barge would float out. In the reverse operation, a barge about to descend the canal would float into the upper-level chamber; all the gates would be shut; the sluice would be opened; the water in the two chambers would be brought to the same level; the sluice would be closed, and a valve opened between the lower pound and the chamber containing the barge; the water would sink to the lower level; the lower gate would be opened, and the barge would issue forth on her journey. It is easy to see that half the lock-full of water is thus saved, by causing half the ascent or descent to be effected by the transference of water from one chamber to another. By putting more than two chambers to the lock, still less water would be lost, but the increased cost would more than neutralize the advantage. The lift of a lock, that is, the difference of level between the upper and lower pounds, depends on circumstances; if, in a canal of a certain length, a certain amount of rise is to be effected, it may either be brought about by a small number of deep locks, or a larger number of shallow locks; the former entailing the lesser expenditure of time, and the latter the lesser expenditure of water. Locks are to be found with every amount of "lift" from one to eighteen feet; but eight feet is about an average.

SUPPLY OF WATER TO A CANAL.

In such cases as the Great Ship Canal of Holland, where both extremities of the canal communicate with the sea, the supply of water is obtained from the sea itself; but where, as in most of the canals of England, there is no immediate connexion with the sea, natural springs must be so regulated as to flow into the canal as fast as waste is occasioned by lockage. When it is necessary to establish a navigable communication between separate valleys or basins of country, and where a double lockage, that is, a descent in both directions from an intermediate eummit-level, is unavoidable, the summit is generally too high to permit a regular supply of water from being drawn

directly from the streams on either side; or these streams are not unfrequently appropriated to the service of mills and manufactories, and, therefore, cannot be applied to the feeding of the canal. Under these circumstances, it becomes necessary to collect the flood-waters of the adjacent higher grounds into proper reservoirs, to be drawn off into the canal as occasion may require.

The selection of a proper situation for a canal-reservoir requires all the skill of the engineer. Much must always depend upon local circumstances, but the principal objects attended to are the following:-1st. that the reservoir should be sufficiently low to collect the flood-waters from an ample surface of country; 2nd. that it should be so high as to enable all the water it contains to be drawn into the summit of the canal; 3rd. that the bottom and sides of the reservoir should be naturally impervious to water, otherwise much trouble and expense will be incurred in rendering them so by artificial means.

Whether the water is obtained from a reservoir, or from natural springs issuing above ground not far from the canal, a channel called a feeder" is generally provided. It sometimes happens, where reservoirs are situate some distance above a canal, and a brook-course leads from the reservoir to the canal, that the water is left to take its ancient course on being let out of the reservoir, flowing in such a channel as the nature of the ground will admit; but generally speaking, an artificial brick channel is constructed for the flow of water. There is a clause in most canal Acts of Parliament, empowering the company to search for, and divert to their use, all springs of water within certain limits on either side of their line, as feeders for the canal; thus, in the Acts for the Newcastle-under-Lyne Junction, the limit is fixed at one thousand yards; in the Aberdeen, Polbrook, Thames and Medway, Wilts and Berks, and a few others, the limit has been extended to two thousand yards. Sometimes, in very dry weather, all the sources belonging to a canal company are so exhausted, that they cannot obtain a supply adequate to the waste by lockage; and it is then often necessary to purchase water at a heavy expense from water companies.

As a means of economizing the water used in a canal, with reference to the process of "lockage," side-ponas have been occasionally employed, in the following manner:-It is most common to employ two ponds, each having the same horizontal area as the lock, and made to receive one-fourth of its fill of water. The one of these ponds has its bottom at half the height or lift of the lock, so that when a valve is opened into it from the full lock, this pond receives onefourth of its contents; this valve is now shut, and another opened between the lock and the second pond, which, in its turn, receives another fourth of the water, because its bottom is at one-fourth the height or lift of the lock; the second valve is now shut, and the remaining half-lock full of water is allowed to flow into the lower pound in the usual way. When the lock is to be filled again, the lower pound is first opened and the water allowed to flow into the lock, and then the higher one, by which half the rise is effected. This is obviously another mode of attaining the same end as results from the use of double-locks.

Mr. Field has proposed the adoption of side-ponds acting on the principle of an inverted siphon, in such a way that water would rise to the same height in one leg of the siphon as in the other. The method consists in connecting a lock with a side-pond by means of a long pipe or culvert, in such a manner, that when the water in the lock is allowed suddenly to run into the empty pond, it may rise in the latter nearly to the level in which it had fallen in the former; and the moment this occurs, the valve or sluice in the culvert is shut. The same thing should take place when the water is to be restored to the empty lock, except that whatever may have been the loss of water in the former case the total loss will now be doubled.

The whole arrangements connected with the supply of a canal with water, and the economizing of the water obtained for that purpose, are important and difficult, and are made the theme of constant investigation on the part of engineers.

MODE OF TRACTION ON CANALS.-BOATS AND Barges.

As canals are, and always have been employed principally for the conveyance of goods, the vessels which navigate them are constructed with especial reference to that kind of traffic. All coarse and heavy goods, whose value is small compared with their bulk, are conveyed in open barges similar to the coal-barges on the Thames. But those goods