The almost incredibly bad state of the roads in England towards the latter part of the 17th century appears from the accounts cited by Macaulay (Hist., c. iii.). It was due chiefly to the state of the law, which compelled each parish to maintain its own roads by statute labour, but the establishment of turnpike trusts and the maintenance of roads by tolls do not appear to have effected any great improvement. At the time of Arthur Young’s six months’ tour in 1770 the roads would seem to have been almost as bad as ever, and it is doubtful if there was much improvement up to the beginning of the present century. The turnpike roads were generally managed by ignorant and incompetent men until Telford and Macadam brought scientific principles and regular system to their construction and repair. The name of Telford is associated with a pitched foundation, which he did not always use, but which closely resembled that which had been long in use in France, and the name of Macadam often characterizes roads on which all his precepts are disregarded. Both insisted on thorough drainage and on the use of carefully prepared materials, and adopted a uniform cross section of moderate curvature instead of the exaggerated roundness given before; but, while Telford paid particular attention to a foundation for the broken stone, Macadam disregarded it, contending that the subsoil, however bad, would carry any weight if made dry by drainage and kept dry by an impervious covering. Macadam was engaged more with the repair of old roads than with the construction of new ones, and, though it is not possible to agree with all his doctrines, the improvement which he effected in road management and maintenance was great and lasting.

Construction of Roads.—A road should be as short as possible between two points to be connected, but straightness must often be sacrificed to avoid difficulties and expense and to secure good gradients. The latter should be as easy as practicable, having regard to the country to be traversed, and it is desirable that there should be a ruling gradient than which none should be steeper. On a level macadamized road in ordinary repair the force which the horse has to put forth to draw a load may be taken as one-thirtieth of the load. But in going uphill the horse has also to lift the load, and the additional force to be put forth on this account is very nearly equal to the load drawn divided by the rate of gradient. Thus on a gradient of 1 in 30 the force spent in lifting is one-thirtieth of the load, and in ascending a horse has to exert twice the force required to draw the load on a level. In descending, on the other hand, on such a gradient, the vehicle, when once started, would just move of itself without pressing on the horse. Λ horse can without difficulty exert twice his usual force for a time, and can therefore ascend gradients of 1 in 30 on a macadamized surface without sensible diminution of speed, and can trot freely down them. These considerations have led to 1 in 30 being generally considered as the ruling gradient to be aimed at on first-class roads, though 1 in 40 has been advocated. Telford adopted 1 in 30 as the ruling gradient on the Holyhead road through North Wales, and there are only two gradients steeper, in places where they were unavoidable. All unnecessary rises and falls should be avoided, but a dead level is unfavourable for drainage, and on this account 1 in 100 to 1 in 150 is the flattest gradient that is desirable. Such slight rises and falls are probably rather favourable than otherwise to ease of draught by horses.

In transverse section, roads in the United Kingdom generally comprise the carriage-way, a space on each side, on one or both of which there may be a footpath, then the fences, and outside all the ditches. The width of the carriage-way may be from 15 feet, which allows of the easy passage of two vehicles, to 30 or 50 feet for roads of importance near towns. The side spaces may be from

4 or 5 to 8 or 10 feet wide; wide sides give the sun and air access to the road, and tend to keep it dry, and also afford space for the deposit of road materials and scrapings. In cuttings or on embankments the transverse section has of course to be modified. The road surface should have just enough convexity to throw the wet off freely, and a very moderate amount is sufficient when a good surface is maintained. On a too convex road the traffic keeps to the middle, and wears ruts which retain the water, so that the surface is not so dry as with a flatter section which allows the traffic to distribute itself over the whole width. Telford used a cross section differing slightly from an arc of a circle in being more convex in the middle than at the sides. Walker recommended two straight lines joined in the middle of the road by a curve, and inclined about 1 in 24 towards the sides, the objection to which is that the flat sides are liable to wear hollow. An arc of a circle is often used, and is a good form, but on the whole a curve more convex at the centre than towards the sides is the best. The rise in the curve from the sides to the centre need not exceed one-fortieth of the width, and one-sixtieth is generally enough on well-kept roads, and if seven-eighths of the total rise are given at one-fourth the distance from the centre to the sides and five-eighths at half that distance a curve of suitable form will be obtained. It is generally best to obtain the requisite convexity by rounding the formation surface or seat of the road and giving a uniform thickness to the coating of stone. When there is not a kerb there should be a “shouldering” of sods and earth on each side to keep the road materials in place, and to form with the finished surface the water tables or side channels in which the surface drainage is collected, to be conveyed by outlets at frequent intervals to the side ditches. The outlets are open cuts through the sides or drains beneath the footpaths. The side ditches should be deep enough to thoroughly drain the foundation of the road, and cross or mitre drains under the road communicating with the side ditches may be required in wet soil. A thorough drainage of the subsoil is of the greatest importance, and it is economical in the end to go to considerable expense to secure it. In a cutting, or where there are no side ditches, the surface water may be taken off by gratings and under drains beneath the side channels.

The thickness to be given to a road made altogether of broken stone will depend on the traffic it is intended for. On a good well-drained soil a thickness of 6 inches will make an excellent road for ordinary traffic, and Macadam’s opinion that 10 inches of well-consolidated material was sufficient to carry the heaviest traffic on any substratum if properly drained has proved to be generally correct. In a new road the loss of thickness during consolidation must be allowed for, and the materials should be laid about one-half thicker than the coating is intended to be. When the materials are not rolled, a thickness of 3 to 6 inches should be laid first, and when that has partly consolidated under the traffic other coats may be added to make up the full thickness. There is great wear and waste of the materials in consolidating if they are laid too thickly at once. Inferior material is sometimes used in the lower part of the road coating, especially when the surface is to be of granite or other hard expensive stone. Thus flints or gravel may be used for the lower

5 or 6 inches of a road to be coated with 3 or 4 inches of granite. Telford covered the broken stone of new roads with 1⅛ inches of gravel to act as a binding material. Macadam absolutely interdicted the use of any binding material, leaving the broken stone to work in and unite by its own angles under the traffic. An unsound road may be made by the improper use of a binding material, but there is no doubt that broken stone consolidates more quickly, and without losing its angular form, when binding is moderately used. About one half the volume of broken stone is void space, and, as the results of examinations into the composition of the coatings of roads when thoroughly consolidated prove that a very large proportion must necessarily be small stones and detritus, it is much better to give some portion of this at first rather than to obtain it by the crushing and grinding of the materials by the traffic. The binding material—fine gravel, sand, or road scrapings— should be spread over the surface after the broken stone is laid and not be mixed with it. Uniform consolidation is much aided by raking.

Whenever it is possible a new road should be finished with a roller. The materials are consolidated with less waste, and wear and tear of vehicles and horses is saved. Horse-rollers, if heavy enough to be efficient, require a number of horses to draw them and are cumbersome to use. A ton or a ton and a half weight per foot of width is desirable, and to obtain it a roller 4 feet wide must be loaded to 5 or 6 tons, and will require as many horses to draw it. In Great Britain horse-rollers have to a great extent been superseded by steam road rollers in consequence of the superiority and economy in the work done. A lδ-ton roller, 7 feet wide, giving upwards of 2 tons weight per foot, can thoroughly consolidate 1000 to 2000 square yards of newly-laid materials per day. The materials should be formed to the proper section, and not more than 4 or 5 inches in thickness; if a greater thickness is required it is better to roll two coats separately. After several passages of the roller any hollows must be filled up with small materials, and the rolling must be continued until it causes no motion among the stones. When this result has been attained the binding material may be added. It should be spread dry and uniformly in moderate quantities and should be rolled into the interstices with the aid of watering and sweeping. Provided that all the interstices in the upper stratum of stones are filled after the stones are thoroughly consolidated, the less binding that is used the better. By using binding in larger quantity, and before the stone is thoroughly consolidated, the amount of rolling required is lessened, but at the expense of durability in the road. Watering is necessary from the commencement of the rolling unless the weather is wet, but excessive watering, especially in the earlier stages, tends to soften the foundation.

A pitched foundation like that used by Telford is always desirable for a road that is subject to heavy traffic. It consists of fiat stones carefully set on edge in courses across the road with the broadest edges downwards. The upper edges should not exceed 4 inches in breadth, to hold the broken stone well. All inequalities must be knocked off, and small stones and chips must be firmly pinned into the interstices with a hammer, so as to form a regular convex surface, with every stone firmly fixed in place. The thickness of the pitching is generally 6 or 7 inches; it should not be less than 4, and it may generally be thicker without any sensible increase of cost. At least 4 inches of broken stone are required over the pitched foundation, and, when consolidated, 6 are always sufficient. A foundation of cement concrete 6 inches thick was used by Sir J. Macneill on the Highgate Archway (London) road on a bad clay bottom, and common lime concrete was subsequently used elsewhere. A bed of lias lime concrete 12 inches thick was laid as a foundation in Southwark Street and on the Thames Embankment, but it is too expensive for a macadamized road under ordinary circumstances. Burned clay, gravel, or even sand may be usefully employed as a foundation on a clay bottom, to cut off the road material from the clay.

The qualities required in a good road stone are hardness, toughness, and ability to resist the action of the weather, and these are not always found together in the same stone. Limestones possess another quality, that of furnishing a mortar-like detritus which binds the stone together, and enables it to wear better than a harder material that does not bind. For heavy traffic the best materials are traps, basalts, greenstones, and syenite; quartzose grits and cherty sandstones are also excellent materials. For moderate traffic the harder limestones are sufficiently durable and make the smoothest and pleasantest roads. Coefficients of quality for various road materials have been obtained by the engineers of the French “Administration des Ponts et Chaussées.” The quality was assumed to be in inverse proportion to the quantity consumed on a length of road with the same traffic, and measurements of traffic and wear were systematically made to arrive at correct results. These processes requiring great care and considerable time, direct experiments on resistance to crushing and to rubbing and collision have also been made on 673 samples of road materials of all kinds. The coefficients obtained by these experiments, which were found to agree fairly well with those arrived at by actual wear in the roads, are summarized in the following table. The coefficient 20 is equivalent to “excellent,” 10 to “sufficiently good,” and 5 to “bad.” [table]

Stone for a new road should be evenly broken to a size that will pass every way through a ring 2⅜ inches in diameter. For repairs, especially when the material is tough, a gauge of 2{ or 2 inches may be used with advantage, as the stone covers a larger surface, consolidates sooner, and makes a smoother surface. Stone is best broken by hand, but stone-breaking machines have been introduced which supersede hand-breaking to some extent, especially where large quantities of hard stone are to be broken. There is always a certain amount of crushing in breaking by a machine, from which softer stones suffer more, and machine-broken stone is never nearly so cubical, uniform in size, or durable as stone well broken by hand. Broken road material contains about 55 per cent, of solid stone to 45 of void space. In a well-consolidated road the void is filled up by small fragments, detritus, and mud, the result of wear, and specimens of good road surfaces weigh from 93 to 95 per cent, of the weight of the solid stone of which they are made. In the coating of a well-maintained road the proportion of stones of various sizes varies, but generally from one-third to one-half is found to consist of detritus under three-eighths of an inch in diameter, and there is a very constant proportion of about one-fifth of mud and detritus under one-thirtieth of an inch in diameter. This appears to be the amount necessary to fill the voids between the fragments of stone when compacted together. In an ill-kept road, from which the mud is not removed, the proportion of detritus is much higher, and mud may constitute nearly one-half of the coating. In proportion as the detritus and mud are kept down to the minimum by constant removal from the surface, so will the road be able to resist the action of wet and frost and the wear of the traffic.

The wear of materials, resulting in their gradual reduction to detritus, is due to the joint action of the traffic and the weather, which cause surface wear, wear arising from cross breaking, and from rubbing of the stones together. When there is no movement in the body of the road, and the wear is confined to the crushing and grinding at the surface, it is the least possible; but, when a road is weak from insufficient thickness or solidity on a yielding foundation, bending and cross breaking of the coating take place under passing loads in addition to the surface wear, and the effects are aggravated by the softening action of water finding its way into the road through cracks formed in the surface and by the disintegrating action of frost. The wear and waste are thus far larger than on roads of sufficient strength, properly maintained. The destructive effect of wheels is greater as the diameter is less, and to a much greater degree as the tire is narrower. On hard and strong roads no greater width of wheel than 4⅜ inches is useful, as a wider tire does not bear evenly, but on yielding roads a greater width is of some advantage, though it does not prevent damage from bending and cross breaking of the whole coating under excessive loads. A good deal of attention has been given by French engineers to the measurement of traffic, wear, and the consumption of road materials. Without a knowledge of the amount of traffic accurate comparisons of wear are impossible, and an account of the traffic on the roads of France is taken periodically in “collars” or horses drawing loads, and in the weight drawn. Traffic as measured by weight drawn has of late been...