brittleness of the material, will be very liable to produce fracture.

In the form of carriages with four wheels, when springs are not used, the weight upon any one of those wheels is far from being regular. The frame of the carriage is rectangular, and the centres of the axles are, as nearly as possible, in the same plane; now, when the bearing surfaces of the rails do not lie in a plane, exactly parallel to the plane of the axles, (and this, in practice, is seldom the case,) the weight of the carriage will rest upon three of the wheels only. If there be a want of correspondence in the undulations, or in the yielding of the rails, the weight may change its points of support, and will, during the instant of transmission, be sustained by two of the wheels alone, and those will be diagonally on opposite sides of the carriage; the transition of the weight, from one wheel to another, will, therefore, produce a continual succession of blows or shocks to the rails, which will be productive of considerable injury, and occasion breakage.

From these causes, and others, which it is not necessary to mention, we find, that, in practice, it is not advisable to subject the rails to a greater load than is considerably within the limit of their absolute strength. In the preceding experiments, the least weight borne by the rails, formed of a mixture of metals, is seven tons, and by the unmixed rails five tons; and the rails were of the size and weight of a railroad, on which the carriages that passed upon it were intended to be four tons, supported upon four wheels.

In extreme cases, when the inequalities of the road throw the weight from one wheel to another, the greatest strain upon any one rail cannot amount to more than two tons; therefore, the proportion, which the load that can be carried with perfect safety, bears

to the absolute strength of the rails, is, in the one case, 3,5 1, and in the other, 2,5 : 1. But, as it may be supposed, that a mixture of metals will be mostly used; we may, in practice, say, that the strain to which the rails of any railroad should be subjected by the load, ought not to amount to more than about one third of their absolute strength, or of that weight which would produce fracture.

We may thus find the strength of any section of rail suitable for a given weight of carriage.

Let w the utmost strain to which the rail is subjected, in lbs. ;

the distance between the supports, in feet;

b the extreme breadth of section, in inches; d the extreme depth of section, in inches; qb-the difference between the breadth in the middle and the extreme breadth; and pd=the depth of the narrow part or rib in the middle.

bd (1-qp3). See Tredgold, on Cast

Then

Iron.

§ 8.-Comparative Durability, of Cast and Wrought Iron Railway Bars.

In the formula for calculating the strength and rigidity of wrought-iron bars, we have given those necessary for cast-iron rails; and we now come to the question of comparative durability, between cast and malleable iron rails. The introduction of the latter being comparatively recent, the opportunity of subjecting them to the

test of experiment has not existed sufficiently long, to produce any very conclusive decision from that source; the opportunities of doing so are, also, not numerous ; and many people are more disposed to concur in, and yield to, the general or current opinion, than either wait the result, or submit to the tedious operation of experiment. In this case, also, no experiment can be decisive, unless acted upon for a number of years; we are, therefore, almost obliged to act upon speculative opinion, until sufficient time has elapsed to produce conclusive evidence, in favour of one mode or other.

Independent of economical considerations, with respect to the durability of wrought-iron rails, their safety, compared with cast-iron, upon public lines of road, has already produced a general concurrence in their favour ; and, therefore, perhaps, the question of durability becomes of less importance. Still, as in some cases, where rapidity of transit is not necessary, relative economy may become an object, we shall, therefore,-such as our experience enables us, speak of the comparative durability of cast and wrought iron.

Experiments are going on at present, where both kinds of rails, accurately weighed, are laid down, and subjected to the passage of the same quantity of traffic over them; the result of these, so far as they have gone, is in favour of wrought iron. In the operation of making the cast-iron rails, the surface is partially casehardened in the casting; this may be seen in all castiron rails, extending to a certain depth from the surface. Any experiment, shewing the comparative wear, must, therefore, be continued until after the outer hardened surface be worn through; and, it is presumed, that sufficient time has not yet elapsed to furnish this. We have, therefore, been obliged to reject the data founded on this mode of experimenting, and shall give the result

of a different sort of test, more severe, and which, it is trusted, will be deemed sufficiently approximate, to justify its presentation to the reader.

Upon the Killingworth railway, we had originally cast-iron wheels upon the locomotive engines; about four years ago we adopted wrought-iron tires. Now, as we have, in this way, the relative wear of cast and wrought iron upon the wheels, which run upon the rails; and as the nature of the action will operate nearly alike, whether upon the surface of the rails or of the wheels, we shall, by that means, have a pretty near approximation to the relative wear upon the rails. In this way, we have a considerably more severe test; as, if we take the quantity of traffic equal to 2000 tons passing along the railway daily, and suppose the carriages to convey three tons each, with three-feet wheels, the relative wear of the wheels and rails is as 53: 1, nearly.

The average wear of the cast-iron wheels was above half an inch in nine months; and, with the wroughtiron tire, the wear of one pair of wheels has been a quarter of an inch in three years, and with three other engines one eighth of an inch in twelve months; making the wear, at least, as five to one in favour of wrought-iron. The actual wear of the rails will not be to the same extent as this, as the engine wheels sometimes slip round, or slide upon the rails, in bad weather. The wear of the wheels of the common carriages will not be so much, for the same reasons; but although it should be observed, that, from this, we ought not to deduce the actual duration of wrought-iron rails, as, their surfaces being narrower than the wheels, the wear will be, perhaps, more than proportionably greater, yet the relative wear should, however, remain the same. We now give the following experiment made on the Stockton and Darlington railway:

Cast-iron rails, four feet long, over which waggons only pass, weighing four tons each, when loaded: 86,000 tons passed over in a year, exclusive of waggons: weight of rail, 63 lbs. loss of weight in twelve months, 8 oz. The loss of weight in malleable-iron rails was, in the same period, 8 ounces for 15 feet length: the same quantity of goods, 86,000 tons. This will give the difference of wear 15: 4 in favour of wrought-iron rails.

These experiments shew, that if rails of the proper degree of strength be used, the durability is decidedly in favour of wrought-iron rails; and we have before observed, that in rails properly manufactured, none of the exfoliation or oxidation, originally dreaded, exists. § 9.-Comparative Resistance to the Carriages, of Cast and Wrought Iron Railway Bars.

The next enquiry, is, whether the resistance is greater upon wrought than upon cast iron rails. As in the case of the cast-iron, when first introduced, the wroughtiron rails were made far too slight; and observation shewed, that, to a certain extent, the resistance appeared greater than upon cast-iron; and this, as has been shewn, was owing to the bending.

Referring to the experiments, on the comparative resistance of cast and malleable iron rails, these experiments, it will be seen, were made with the surfaces of the rails quite dry, and free from dust; to ascertain the increase of resistance by the occurrence of any extraneous matter on the rails, the surfaces were watered pretty freely. The number of vibrations, before attaining a state of rest, when dry, was 540 and 570 respectively, and, when watered, 375. Upon wroughtiron rails, when dry, with a less extent of vibration, the number of oscillations was 404 and 412, while, when chalked, the number was 230. With cast-iron, when