produce nothing for want of water. By using the water of the rivers for irrigating their lands, the inhabitants of those countries are enabled to get two and in many cases three crops annually. Even in the southern countries of Europe, where rain is very scarce in summer, and not sufficient to maintain vegetation, whilst the heat is excessive, irrigation is practised, and two crops of Indian corn are thus annually obtained, or one crop of wheat and a green crop. In those countries in which the temperature for three or four months is under the freezing-point, the rivers during that time are covered with ice, and in this state they afford to the inhabitants, in some degree, the advantages which other countries derive from railways. Travelling and the transport of goods on the smooth ice of the rivers are much less expensive, and are performed in a shorter time than in summer in the ordinary way. This is the case on some of the rivers of New Brunswick and Lower Canada. It has been observed that the outer borders of riverbasins are the most elevated parts which occur in some given places between their respective beds, though it is not always the case that the watershed is formed by mountainridges. Owing to such a disposition of the surface, the waters which are collected on or near the borders, run to one or the other of the two rivers. Up to the commencement of this century it was thought improbable, if not impossible, that two different river systems or basins could be united by a natural water communication. But it is now ascertained that a low tract of country or a deep depression of the surface may occur, by which a portion of the water of a river, after being diverted from its own channel, may join a river which otherwise is not connected with that river from which the water branches off. The instances in which this occurs are very few, and we shall therefore enumerate those whose existence is beyond all doubt. The river Arno in Tuscany, in that part where it runs between the high ridges of the Apennines and approaches the town of Arezzo, sends an arm southwards through a narrow valley, under the name of Chiana, which falls into the Chiare, an affluent of the Tiber. The Chiana had been filled up with sand, but its course has been re-established by artificial means. Another case occurs in the kingdom of Hanover, a few miles east of the town of Osnabrück, where the river Haase divides into two branches, of which one, running west to Osnabrück, preserves its name, and after a course of many miles joins the Ems; the other, running east under the name of Elz, falls after a short course into the Werre, an affluent of the Weser. In Sweden, two large rivers fall into the northern extremity of the Gulf of Bothnia, the Tornea Elf and the Calix Elf. About 100 miles from the sea, the last-mentioned river sends off an arm to the south-east, which after a course of about twelve or fifteen miles falls into the Tornea Elf: this arm is called Tarenda Elf. In these cases the rivers thus united by a natural water communication flow in the same direction or nearly so. But in South America two large rivers, the Orinoco and the Amazonas, are united in this way in a part of their extensive courses, where the Orinoco runs west and the Amazonas east. The branch of the Orinoco by which this natural water communication is is effected, is called Cassiquiare. [CASSIQUIARE.] It is a kind of established rule that the whole course of a river should bear the same name, and that this name should be continued to that branch whose sources are farthest from the mouth. But practice is frequently at variance with this rule, and it may easily be accounted for. The inhabitants of a country preserve the name of that river which does not undergo any deflection of its course. At the confluence of the Mississippi and the Missouri, the latter is the larger river, and has had a course of above 1000 miles more than the former, but it does not deflect the course of the Mississippi by its junction, and the name of the last-mentioned river is preserved. The same occurs in South America as to the Amazonas and Madeira, where we find that the last-mentioned river changes the direction of its course to meet the Amazonas, whose name is preserved. In Europe, the Rhine is joined by the Aar in Switzerland, above Laufenburg. The Aar is the larger river and brings down a greater volume of water, but the Rhine, where it is joined by it, continues its westerly course, and its name is preserved.. The extent of a few river basins is here given in round numbers, but they must only be considered as rough approximations: RIVESALTES. [PYRENBES ORIENTALES.] RIVIERA, a name given by the Italians to certain long narrow strips of land extending between mountains and the sea-coast. The most familiar instance is that of the coast of Genoa, which is divided into Riviera di Levante, or Eastern Riviera, which extends from the city of Genoa to the Gulf of La Spezia, and Riviera di Ponente, or Western Riviera, which extends from Genoa to Nizza. The Western Riviera is the more fertile and populous; but in the Eastern Riviera, which is generally more rugged and barren, there are some delightful spots, especially about Nervi, Rapallo, and Chiavari, sheltered by the mountains from the north winds, and where the orange, lemon, and other southern plants thrive in the open air. [SARDINIAN STATES.] There is also a tract along the western shore of the lake of Garda which is called Riviera di Salò, from the town of that name. [GARDA, Lake of.] RI'VÒLI, in Piedmont, a town situated about ten miles west of Turin, on the road to Mont Cenis. A wide and straight avenue, lined with fine elm-trees, leads from Rivoli to the capital, through a rich plain irrigated by canals. Rivoli has 5000 inhabitants, and a royal palace situated upon a height. There is another small town or village called Rivoli in the province of Verona, situated at the southern entrance of a defile through which the Adige coming from the Tyrol makes its way into the plain of Lombardy. Rivoli is on the right bank of the Adige, at the foot of a lofty ridge called Monte Baldo, which extends between the river and the lake of Garda. A hard-fought battle took place at Rivoli between the French under Bonaparte and the Austrians under General Alvinzi, on the 14th of January, 1797. The village was several times taken and re-taken by the two armies. At last General Massena, coming up with his division, carried the day, and Alvinzi was obliged to retire with great loss. Massena obtained afterwards, under the empire, the title of Duke of Rivoli. RIZZIO. [MARY STUART, vol. xiv., [LEUCISCUS.] p. 477.] ROACH. ROAD. Under this head it is proposed to embrace road-making, with a brief sketch of the history of roads, referring for more detailed statistical information to the geographical articles in this work, and to WAY and TURNPIKE TRUSTS for an explanation of the laws respecting the formation and maintenance of the highways in this country. The importance attached to roads by the great nations of antiquity is abundantly testified by historians, though, except in the case of the Roman roads, there are few remains existing. The Carthaginians are said to have been the inventors of paved roads, which were much used by the Romans, who were distinguished by the vast extent and solid construction of their highways, of which several thousand miles were made in Italy alone, while every country which was brought under their sway was more or less intersected by these channels of communication. Though formed mainly to facilitate military movements, the Roman roads were productive of the greatest civil benefits. Being made by a power whose resources were almost unlimited, these military roads were usually laid out in straight lines from one station to another, with little regard to natural obstacles, which were frequently passed by means of very extensive works, as excavations, bridges, and, in some instances, tunnels of considerable length. The solidity of their construction was fully equal to the boldness of their design; a fact proved by the existence of many that have borne the traffic of near two thousand years without material injury. The Roman en gineers were very particular in securing a firm bottom, which was done when necessary by ramming the ground with small stones, fragments of brick, &c. On this care fully prepared foundation a pavement of large stones was firmly set in cement, the stones being occasionally squared, but more commonly of irregular shapes, though always accurately fitted to each other. For this purpose many varieties of stone were used, but the preference seems to have been given to basalt, where it could be had, it being used in many situations in which other suitable materials might have been procured with less labour and expense. Where large blocks could not be conveniently obtained, small stones of hard quality were sometimes cemented together with lime, forming a kind of concrete, of which masses extending to a depth of several feet are still in existence. The strength of their pavements is illustrated by a fact, related by a modern traveller, who states that the substratum of one still in use has been so completely washed away by a current of water without the surface being at all disturbed, that a man may creep under the road from side to side, and carriages pass over the pavement as over a bridge. The Roman roads were generally raised above the ordinary surface of the ground, and frequently had two carriage tracks separated by a raised footpath in the centre. Though much remains to be accomplished, and the philosophy of road-making is yet very imperfectly understood by a large proportion of those to whom the care of the highways is committed, it is impossible to compare the past and present state of roads without feeling grateful for their improvement, and observing in how great a degree that improvement has benefited the agricultural, commercial, and moral interests of the community. Designing a Line of Road; Earth-works, &c.-Though formerly little attended to, the design of a line of road is a subject which requires extensive knowledge and mature deliberation. It is often advisable to survey several different lines, in order to the selection of that which, on careful comparison, appears to have the preponderance of desirable qualities. To be theoretically perfect, a road should combine the qualities of straightness and level, and its surface should be smooth and hard; and the best road, practically, will be that which makes the best compromise between unavoidable deviations from this theoretical perfection. It may be observed however that although some writers speak of the absolute perfection of each of these qualities as essential to the idea of a good road, it may be questioned whether it is desirable of any, excepting the first. Of these qualifications the two first belong to the design or laying out of the line, and the last two to the execution of the road and the materials made use of. In some parts of the continent of Europe, especially in Italy, the Roman system of road-making has been imitated, particularly in city pavements; but in Britain the attempts to follow the Roman model appear to have been very limited, and road-making has been very imperfectly practised till within a few years. Many of the existing The qualities of straightness and level, or the line of dihighways were originally mere paths or tracks from place to rection and line of draught, should be very carefully adplace, their course having been determined more by acci-justed to each other. Some remarks on this subject will dental circumstances than by a due attention to the pro- be found in the article RAILWAY, p. 250, which apply perties of a good road. Thus deviations were made from equally to the laying out of common roads, though the prothe direct course in order to cross rivers at fordable points, portionate retardation due to a given ascent is very differand the road was conducted over a hill in preference to a ent, owing to the great comparative resistance of a common more level course round its base, to take advantage of road. Among the circumstances that may authorise a natural drainage. As improvements have been introduced deviation from the straight line, are the power of obtaining in the systems of construction and repair, the direction and suitable materials for the road, avoiding valuable property levels have been frequently left unaltered, to avoid the tem- or difficult ground, and including towns or villages in the porary inconvenience and expense attending a deviation route. from the established course. The scanty information we possess as to the state of the roads in early times indicates that it was very bad; and after the introduction of turnpikes, and even down to the commencement of the present century, the greater part of the roads were, owing to injudicious modes of construction and repair, in a state very unfit for traffic. The inefficiency of the system of maintenance by parish and statute labour was proved before the passing of the first Turnpike Act in 1653; yet the necessity of improvement, and the obvious justice of maintaining roads by the produce of tolls, did not lead to the extensive adoption of the turnpike system for about a century after that time. In the latter half of the last century turnpike-roads multiplied rapidly, and superior principles of construction also made some progress. During the last forty years the attention of government has been repeatedly directed to the importance of this class of public works, and the Highland and Holyhead roads, formed by Telford and others, have done much in improving and extending the science of road-making. The Highland roads alluded to were made under the commission of 1803, and originated in the military roads formed in consequence of the rebellions of 1715 and 1745, which had been found very beneficial to the districts to which they afforded the means of access. The roads made and improved under the management of the Highland-Road Commissioners extend to more than 900 miles, the whole being in a mountainous district, but so well laid out that their inclinations are always moderate. The works executed in the formation of these roads are very extensive, and comprise upwards of 1100 bridges. The Holyhead road improvements were commenced in 1815, and in these Telford and his able assistants had the opportunity of carrying into effect, under a government commission, a plan of road-making suitable to a great traffic, on principles generally considered to be nearly perfect. The principles on which these important works have been executed are very fully detailed by Sir Henry Parnell, in his valuable Treatise on Roads, to which work the writer of this article is indebted for much of the following information. The name of Mc. Adam must not be passed over without notice in this place, as his exertions have done much towards attracting public attention to the improvement of roads, even where his peculiar principles have not been acted upon. It seems to be a prevailing opinion with modern engineers, that the line of direction has not generally been made as subordinate as it should be to the line of draught; and it will be well to remember, in laying out a new road, that while the effect of gravity must ever remain the same, the resistance occasioned by imperfections in the road and carriages will be reduced by every prospective improvement in their construction; thereby increasing the proportionate effect of gravity, and making the line of direction still more subordinate to that of draught, or, in other words, increasing the length of level road that may be traversed with the same expenditure of power as would raise the load up a given elevation. Curves increase the resistance to the motion of carriages, and add to the risk of accident; but if slight, they increase the length of the road much less than might be supposed. Edgeworth, in his Essay on the Construction of Roads and Carriages,' says, 'Á road ten miles long, and perfectly straight, can scarcely be found anywhere; but if such a road could be found, and if it were curved, so as to prevent the eye from seeing farther than a quarter of a mile of it, in any one place, the whole road would not be lengthened more than one hundred and fifty yards.' · The principle explained in p. 250 of the article RAILWAY of so arranging the inclinations on each side of the summit, or highest point unavoidably passed over, that there may be no unnecessary rise and fall, is equally deserving of attention in the design of a common road, although it has been much neglected. The following statement respecting an old road in the Isle of Anglesey, which was altered by Telford, shows how very much a road may be improved by judicious alterations; not only by shortening the line and lowering the summits, but also by diminishing the minor. undulations: raised a few feet above the surrounding land, and thereby exposed to the free action of sun and wind. Slight undulations are also considered, by most authors, to be desirable in all cases where animal power is employed; frequent changes in the amount of exertion being considered favourable to the horses. On this principle it is recommended that where an undulating road is reduced to a uniform gradient, occasional levels should be introduced to ease the draught. Any inclination exceeding the angle of repose, or that beyond which a carriage would roll down by its own gravity, occasions a loss of power; but all below it are attended with a compensating effect when the traffic in both directions is taken into account; the advantage gained by descending carriages being equal to the additional labour required in the ascent. This angle has been stated by Lardner to be about 1 in 40, with a good carriage on a broken stone road of the best quality; but the inclination allowed on the Holyhead road is 1 in 35, a slope which may be ascended at a good rate of speed, and descended at twelve miles an hour without risk. A greater slope not only occasions much additional resistance in the ascent, but, by rendering it unsafe to drive down at full speed, causes a loss of time in the descent also. Modern engineers consider it unadvisable in any case to exceed an inclination of 1 in 24, though there are hills at least twice as steep on some turnpike roads. The following table shows the effect of various inclinations in increasing the draught of a stage-coach at different velocities on the same description of road, as indicated by a dynamometer contrived by Mr. Macneill for experiments on the draught of carriages:-* Force required at It should always be borne in mind that the occurrence of one steep hill on a line of road affects the working of the whole line, as the number of horses required for ascending it must be used, although a portion of their power may be unemployed on the greater part of the road. The inconvenience of a steep inclination may be diminished by laying a stone tramway for the use of ascending vehicles; a measure which has been adopted with success on the Holyhead road, where, on a slope of about 1 in 20, the power required to draw a ton has been reduced by this means from 294 lbs. to 132 lbs. In arranging the works necessary for obtaining the required level, the preference should be given to embankments; and, wherever it is practicable, the bed of the road should be elevated two feet above the natural level, for the sake of efficient drainage. Tunnels are very rarely introduced on common roads, being very costly, and, when of considerable length, inconvenient from their darkness. When the road is in an excavation, the side-slopes should never be steeper than two horizontal to one vertical, and it is desirable to have those on the south side three to one; because, though many materials will stand at steeper inclinations, it is essential to the preservation of the road, and the comfort of horses travelling upon it, that the sun and air should have free access to its surface. Where stone can be readily procured, the erection of walls at the bottom of the slopes gives a neat and finished appearance to the road, and prevents earth, which may be loosened from the sides, from falling into the side channels or drains. The Highgate Archway road affords an example of the great difficulties that occasionally attend a deep excavation, owing to the accumulation of water; the remedy for which is described hereafter. Where embankments are required, strong fencing is especially necessary, to guard against the occurrence of accidents. Some of the roads formed by Telford are conducted across deep valleys by bridges or viaducts of great magnitude, in order to maintain the desired level without the inconvenience and expense of large earth embank ments. In old roads the bridges erected for the passage of rivers are frequently made much smaller than is advisable, so that This useful instrument, which its ingenious inventor denominates a Road Indicator, is mounted in a light phaeton, and besides marking the draught at every ten or twenty yards, points out the distance run, and the rates of acclivity or declivity on every part of the road. A full description of it is given in Parnell's Treatise on Roads.' the level of the road is made too low, and the water is in peded by the contracted arches to such a degree as to occasion much damage during floods. Modern engineers, by adopting bolder dimensions for the bridges, and forming raised approaches, avoid these inconveniences, and secure their roads from the risk of obstruction by floods. The raising of the road wherever it passes through marshy or low land is a very necessary measure. Many old roads still in use are sunk several feet below the surface of the ground, because they have originally been exposed to the destructive action of water, and the materials thus softened have been ground into mud and cleared away, until, by the repetition of these operations, the roads have been converted into deep trenches, which are frequently flooded in winter. Of the extent to which this process has been occasionally suffered to go on, an idea may be formed from the statement of Edgeworth, that 'the stag, the hounds, and the horsemen have been known to leap over a loaded waggon in a hollow way, without any obstruction from the vehicle.' In conducting a road through a mountainous district, in addition to numerous bridges for the purpose of crossing ravines (for which purpose suspension bridges have been occasionally applied, as in the passage of the Menai strait), embankments between retaining walls of stone, and walls to support the road along the face of a precipice, are frequently necessary. Some works of the latter character have excited much admiration. If the slope of a precipice be only six inches horizontal to a foot vertical, such a road may be formed by building a wall thirty feet high, based on steps cut into the rock, and cutting into the rock to the depth of ten feet on a level with the top of the wall, the space between which and the face of the precipice is filled in with earth or stone. By this means a platform twentyfive feet wide is obtained. Many works of this character have been executed by Telford and other engineers, in various parts of Scotland, in the Highland roads, and those forming the communication between Edinburgh and London; and others, the boldness of which commands universal admiration, occur in the great mountain-passes of the Simplon and Mont Cenis, which form imperishable monuments of the talent and energy of the engineers of Napoleon, by whom they were executed. When the works are completed to the proper level for receiving the hard materials that form the surface of the road, the earth should be formed into the intended width and a nearly level surface, the footpath or paths being elevated a few inches above the bed of the carriage-way. Thirty feet is the ordinary width of the carriage-way, exclusive of footpaths, of the Holyhead road; but owing to the diminution of traffic since the opening of the London and Birmingham railway, a recent Report suggested the propriety of reducing the width, in most places, to twenty-four or twenty-five feet. This width may be more or less exceeded in the vicinity of large towns, according to the amount of traffic, but should be exactly adhered to in other situations, as uniformity in this particular greatly improves the appearance of a road, and also contributes to economy, both as to the land and materials, and the cost of maintenance. Some engineers recommend that the bed should be made convex, in the same degree as the finished surface of the road; but it is quite flat in the Holyhead road, by which means a greater depth of materials is allowed in the centre than at the sides of the road. Much has been said on the subject of the best form for the transverse section of a road. Formerly it was common to make it very convex, often to a degree that was highly dangerous, with the idea of throwing off water: but this notion is very fallacious, because if a road be allowed to wear into ruts, no degree of convexity that can be given is sufficient to keep it dry; while, if the surface be good, a very moderate slope is sufficient to carry off water, and a steep inclination will cause it to run with such velocity as to wear away the road materials. Another disadvantage of too great an inclination is, that, by throwing the weight of a carriage on one side, the vehicle itself is injured, and the overloaded wheels cut up the road more than necessary. Some have gone so far in opposition to this practice as to advocate perfectly flat or even concave roads, in favour of both of which much may be said; but the general practice of modern roadmakers is to make the surface slightly convex. In Telford's roads the convexity is elliptical, the fall being half an inch at four feet from the centre, two inches at nine feet, and six inches at fifteen feet. It has been recom- | mended to form the cross section into three flat planes, that in the centre being horizontal, and the others slightly inclined from it. Very narrow roads are often sloped in one direction only, like one-half of a convex road; and roads on the face of a steep hill are occasionally treated in this man-road-making, and which appears to have been first prominer, the surface water being conducted towards the hill, and carried off by drains under the road. This plan has the advantage of checking any tendency in carriages to roll or turn over towards the least protected side of the road. Wherever the substratum of a road is wet and soft, great care is necessary to make the bed solid. If the ground be boggy or marshy, it is desirable to form an embankment of sufficient thickness to compress the elastic foundation; such embankment being sometimes supported by faggots. Telford and most engineers recommend that ramming with stone-chips should be resorted to where the bed is wet and spongy; and that where soft clay occurs, a stratum of earth should be laid between it and the road materials, a precaution which tends to diminish the injurious effect of frost on a road with a clay bottom. Though great care is usually considered necessary in order to obtain a firm foundation, Mc. Adam and some others have not only contended against its importance, but actually preferred, in certain cases, a yielding substratum to one of rock, on the supposition that the wear of the road is diminished by elasticity. Careful observations on the repairs of a road in Somersetshire, of which about seven miles are supported by a morass, and five or six by limestone rock, indicated a difference of expense in repair of about five to seven in favour of the morass, though it was so soft that the vibration caused by a carriage passing was sufficient to break the young ice in the side ditches; but extended experience seems to confirm the more general opinion in favour of a hard unyielding foundation. Deep ditches should be cut for the efficient drainage of the road, which is of paramount importance; and these should be on the field side of the fences. They should extend to a depth of from two feet six inches to four feet below the bed of the road, according to the nature of the ground. The earth thrown out from them is commonly used in forming banks for the hedges; but in wet soils, where the ditches are made larger and deeper than usual, the additional earth excavated is applied to raising the bed of the road. Where brick- or stone-covered drains are substituted for open channels, it is usual to build them with open joints, to allow the passage of water through the sides. Cross-drains of masonry are introduced at intervals to connect the side channels, and numerous minor drains filled with rubble stones or clean gravel are formed in the bed of the road. The latter are frequently of the kind called mitre-drains, which are made V shaped in plan, diverging from the centre of the bed, and extending diagonally to the sides, their angle being regulated by the longitudinal slope of the road, so that their inclination may not exceed 1 in 100. These may be placed about sixty yards apart, or closer in wet soils, and they receive the water that filters through the surface materials. In cuttings or excavations it is advisable to make drains to catch the water descending from the sides, and prevent its reaching the surface of the road. In treating of the choice and application of the hard materials which compose the surface of the road, the formation of metalled roads, or those made of broken stone and similar materials, will be first considered; and afterwards that of the principal varieties of pavement. sinking into the earth to a surprising extent. This evil is greatly diminished by good drainage, and by the use of stones of uniform size broken into angular pieces, which have a tendency to lock together into a hard and compact mass, a fact of the highest importance in the science of nently brought forward by the late Mr. Mc. Adam. A great extent of excellent road has been made on the plan advocated by Mc. Adam, who considered paving unnecessary, and laid the broken stone immediately on the surface of the earth, depending on its forming a hard crust impervious to water; so that the earth, being always kept dry, may have no tendency to work up among the metal or broken stone. Mc. Adam used no stones exceeding six ounces in weight, and gave the preference to those of about one ounce, or an inch diameter, which he spread over the road in thin layers, each being worked over by carriages till in some degree consolidated; and he objected to the use of chalk or earth mixed with the stone for the purpose of binding it together. He considered a thickness of ten inches of broken stone, well consolidated, to be sufficient for bearing any load, even where the foundation is a morass, in which case he considered no intermediate substance necessary. Near Bristol a road in which the metalling had worn down to a thickness of only four inches, was found to have kept the substratum of earth perfectly dry. But, satisfactory as this plan of road-making has proved in some cases, there are others in which it has failed, and some in which a very large quantity of stone has been applied before a firm road could be obtained. A road from Lewes to Eastbourn, made on Mr. Mc. Adam's principle, is said to have required three feet of materials in many parts before it was consolidated, though it was ultimately brought into a good state. The system of pitching or paving the bottom of a road has the advantage of preventing the subsoil from working up among the road materials, and, when well executed, of distributing the pressure of carriages over a larger base; while the size of the paving stones themselves prevents their sinking into the earth, as small stones are liable to do. The pavement also acts as a drain to the surface materials. In addition to these, the plan has, in many situations, the advantage of economy, as the cost of a pavement is considerably less than that of an equal depth of well-broken stone. In most, if not all, of the cases in which the paving system has failed, the want of success may be attributed to very imperfect execution; as, if the stones' are very irregular in size or badly set, or the thickness of metal is insufficient to protect the pavement from the shake of passing carriages, the stones become deranged, and the subsoil, working up among them, quickly spoils the road. The case of the Highgate Archway road, which has been before alluded to, is a remarkable illustration of the absolute necessity of a firm bottoming under some circumstances This road is over a subsoil of sand, clay, and gravel; and being partly in a deep cutting (originally intended for a tunnel), is much exposed to the influx of water. The road, which is rather more than a mile and a half long, was originally made of a quantity of gravel and sand laid on the natural soil, and covered with broken flints and gravel; but this plan not succeeding, the road was taken up, and pieces of waste tin were laid on the subsoil, over which were spread gravel, flints, and broken stone. This expedient did not produce the desired effect, and at length, in 1829, the road was placed under the management of the Holyhead-road Commissioners, its proprietors having failed, notwithstanding an enormous outlay, and the application of 1200 cubic yards Metalled or Broken-Stone Roads. In the formation of of gravel annually, to bring it into a satisfactory state. A metalled roads the system adopted in the great works of thick coat of broken granite was spread on a portion of the Telford and his followers is that most decidedly sanctioned road; but owing to the unsoundness of the foundation, it by experience. The distinguishing feature of this system is never consolidated, the stones wearing into smooth pebthe use of a rough pavement of hand-laid stones on the bed bles by their attrition against one another, even down to the of the road, to support the small broken stone of which the bottom of the mass. The commissioners therefore detersurface is composed. In the very imperfect mode of road-mined, as paving-stones could not be procured without great making formerly practised (which scarcely deserves the expense, to lay a coating of Roman cement and gravel as a name of a system), it was very common to cover a bad road bed for the road-metal, an experiment that has been atwith a large quantity of stones, often unbroken, and generally tended with complete success. The work was executed by of very irregular dimensions. These stones, owing to their Macneill, and consists of a composition of Roman cement rounded form and the softness of the substratum, never con- with eight times its quantity of washed gravel and sand, solidated into a hard surface, and in course of time sunk which, after being mixed in a box, was laid on the bed of into the soft earth beneath, which worked up among the the road to a thickness of six inches and a width of about stones in the form of mud. Thus enormous quantities of eighteen feet. A few minutes after being laid, the upper stone were used without producing a good road, the stone surface was indented, by means of a triangular piece of wood sheeted with iron, with numerous channels or grooves, sloping about three inches from the centre to the sides; these channels serving for the stones to lie and fasten in, and conducting any water that might percolate through them into the side drains. This measure, combined with an extraordinary extent of drainage, amounting in the whole to a length of 12,803 yards, proved so complete a remedy, that in the first winter after the cement was laid, coaches were able to go up with four horses at a trot with the heaviest loads, though before the improvement six horses had mounted with difficulty at a walking pace. The effect of the alteration on the wear of the road was equally satisfactory, four inches of quartz being worn away on the old bottom while only half an inch of the same stone was worn where laid on the cement foundation. The expense of laying the cement composition, including the formation of the bed of the road, was about ten shillings per lineal yard, part of the gravel used being old. Macneill estimated the cost at from twelve to fifteen shillings per yard if new gravel were purchased. The effect of a paved or concrete foundation in diminishing the draught appears, from the subjoined statement, founded on experiments with Mr. Macneill's road indicator, to be very great; but a more extensive series of trials is desirable for a comparison of different systems under various circumstances. The draught of a waggon weighing 21 cwt. was found to be as follows:On a well-made pavement On a road with six inches of hard broken stone on a rough pavement On a similar road, with a foundation of Roman cement and gravel in lieu of pavement 33 lbs. 46 46 On a road with a thick coating of broken stone on earth 147 On a road with a thick coating of gravel on earth For the formation of the pavement of a metalled road, almost any hard stone that may be easily dressed with the hammer may be used. The stones should be tolerably regular in size, and laid in rows with their broadest face downwards, the interstices being carefully filled up with stone-chippings, so as to pin the whole pavement together, and effectually prevent the earth from working up through the joints. In one of Telford's specifications for the Holyhead road the dimensions of the stones for a pavement thirty feet wide are given as seven inches deep in the middle of the road, five inches at nine feet from the centre, four inches at twelve feet, and three inches at fifteen feet, the stones to be laid lengthwise across the road, and the upper edge in no case to exceed four inches wide. All irregularities are to be broken off by the hammer, and the stone-chips used in packing the joints are directed to be wedged in by hand or with light hammers. No ramming is necessary, and it is desirable to prevent carts which are used in the conveyance of the road materials from being drawn upon the pavement before it is covered with broken stone. Some road-makers use a pavement even on a substratum of rock, where it is uneven, but in many cases it is unnecessary; although, if the surface be smooth, it should be picked to a degree of roughness similar to that of a pavement, in order that the road materials may not slide upon it. Where paving the whole width of a road might be too expensive, the pavement is sometimes limited to a width of sixteen or eighteen feet in the centre. In situations where coarse stone of suitable quality can be easily procured, it is found to be cheaper to make a road with six inches of broken stone and a pavement, than with ten inches depth of broken stone without paving. Mr. Wingrove, surveyor of the Bath roads, mentions the use of freestone brash, chalk, &c. for forming the foundation pavement of metalled roads. The quality and right application of the road-metal, or broken stone which forms the surface of the road, is of great importance. As a general rule the hardest stone is to be preferred, but this rule admits of some qualification, some very hard stones being found to wear much more rapidly than others of a softer but tougher quality. According to Parnell, whose experience renders him a good authority, the best descriptions of road materials 'consist of basalt, granite, quartz, syenite, and porphyry rocks. The whinstones found in different parts of the United Kingdom, Guernsey granite, Mountsorrel and Hartshill stone of Leicestershire, and the pebbles of Shropshire, Staffordshire, and Warwickshire are among the best of the stones now commonly in use. The schistus stones will make smooth roads, being of a slaty and argillaceous structure, but are rapidly destroyed by wet, by the pressure of wheels, and occasion great expense in scraping and constantly laying on new coatings. Limestone is defective in the same respect. It wears rapidly away when wet, and therefore, when the traffic is very great, it is an expensive material. Sandstone is much too weak for the surface of a road; it will never make a hard one, but it is very well adapted to the purpose of a foundation pavement. Flints vary very much in quality as a road material. The hardest of them are nearly as good as the best limestone, but the softer kinds are quickly crushed by the wheels of carriages, and make heavy and dirty roads. Gravel, when it consists of the pebbles of the hard sorts of stones, is a good material, particularly when the pebbles are so large as to admit of their being broken; but when it consists of limestone, sandstone, or flint, it is a very bad one; for it wears so rapidly that the crust of a road made with it always consists of a large portion of the earthy matter to which it is reduced. This prevents the gravel from becoming consolidated, and renders a road made with it extremely defective with respect to that perfect hardness which it ought to have.' Mr. Stevenson, in the article Road' in the Edinburgh Encyclopædia,’ states the distribution of road materials in the British islands to be partial and irregular. Throughout Scotland, and even as far south as the approaching sources of the rivers Tees and Ribble, good road-metal is generally to be met with, containing the numerous varieties of granite, greenstone, basalt, porphyry, and limestone. South of this boundary, as far as the Trent and the Dee in Cheshire, the formation is chiefly coal, sandstone, and the softer varieties of limestone. In the southern counties chalk and gravel soils chiefly occur, affording flint and gravel, both of which, under proper management, make excellent roads. In North and South Wales we have all the varieties of roadmetal which are common to Scotland. In Ireland they have excellent road materials, as granite and limestone are pretty generally distributed.' An interesting experiment has been tried on a part of the Holyhead road between London and Birmingham, as to the effect of iron amongst the road-metal in diminishing the wear of the road. The iron is cast in the form of cubes, about an inch square, and when the road was consolidated, holes large enough to receive them were picked in its surface. A single cube was then placed in each hole, so as to be level with the road, and the small stone-chips were beat down about the iron with a mallet. One of these iron cubes is placed in every four inches of surface. They very soon become firmly imbedded, so as not to be disturbed by the rolling of carriages or the feet of the horses; and to assist their consolidation, it is recommended to water the road freely, if the cubes are inserted in dry weather. The iron was applied in March, 1835, since which time the portion of road in which they are used has continued in excellent repair, and the wear is so materially diminished, that a more extended trial of the plan appears highly desirable. Mr. Macneill, the patentee of this method of road-making, considers it particularly applicable to streets, on account of its durability, and thinks that the expense would be trifling, as iron of the worst quality may be used. It may be observed that the draught on the piece of road on which the experiment is made is very easy, and that horses do not show any tendency to slip upon it. In the choice of materials, the expense of conveying them to the road must be taken into consideration, but it is often better economy to fetch good stone from a great distance than to use that which is less durable, though readily procured; as, in addition to the expense of frequent repairs to a road formed with weak materials, great additional labour is imposed upon the horses, which have to wear down repeated coats of fresh stones. This is one of the points in which the inexperience and ignorance of road surveyors have often been displayed, cases having occurred in which an inferior material has been procured from a distance at great cost, while stone of excellent quality existed in abundance on the spot. With regard to the best size of the broken stone for the surface of a road, both Telford and Mc. Adam direct that no piece should exceed six or eight ounces in weight. In some districts the surveyors have been instructed to test |