The marginal sketch will shew this mode of experi menting. abc is the wheel, in different positions; fg the scale for measuring the extent of the vibrations; w w w the weight placed near the periphery. Thus, at the commencement, the wheel was rolled by hand into the position b or c, when the extent of the vibrations, from the point of rest, was five or seven inches; the wheel was then left at liberty, and made so many vibrations, before the distance from the point of rest was diminished an inch, or until the extent of vibrations from the centre was four or six inches; and so on until the wheels became at rest, in the position a. M a The cast-iron rails were three feet nine inches long, and weighed fifty-six pounds; section shewn in Fig. 3. Plate II. The wrought-iron rails weighed twentyeight pounds per yard, the bearings three feet apart; section same as experiment on the strength of rails. These experiments will shew, in the first place, that the resistance was the same, with malleable and cast-iron rails, until the incumbent weight reached 30 cwt. With 40 cwt. the number of vibrations from 7 inches to 3 inches, in cast iron, was 146; while, upon wrought iron, it was 134 and 138 respectively; shewing a trifling increase of resistance with that weight. The rails were then wedged up on the under side, to prevent them from bending, when the number of vibrations of cast iron was 148; shewing that scarcely any resistance was owing to the want of stiffness. With the wrought iron, the wedging had an increase of effect, and brought up the resistance nearly equal to that of cast iron; and thus proving, that, when no bending takes place in the wrought iron, the resistance is precisely the same as with cast iron. We, however, find, by these experiments, that when the deflection is to a certain extent, an increase of resistance takes place, and that the rails of wrought iron, upon which these experiments were made, were only sufficient for carriages of four wheels, weighing three tons. When the wheels were loaded with 30 cwt., no increased effect took place on wedging up the middle of the rail; but when loaded with 40 cwt., the number of vibrations was increased by making the rail perfectly rigid. The deflection of the wrought-iron rail, as ascertained by experiment with 30 cwt., was '032 inch, and with 40 cwt. 043 inch; so that we find it will not be advisable to use rails, the deflection of which, when loaded, amounts to '032 inch, especially when the rails are laid down; as, when they become worn, the deflection being greater, an increase of resistance will take place. With carriages mounted upon springs, which yield to all the inequalities and want of parallelism of the road, we may suppose the whole weight of the carriage divided equally upon the four wheels. But with carriages without springs, as before stated, the weight of the carriage, very frequently, rests upon two wheels only; and, in practice, therefore, we must, with these kind of carriages, suppose the whole weight acting upon two wheels. To ensure, therefore, no increase of resistance in the use of wrought-iron rails, it would appear, from these experiments, that their rigidity should be such as, by carriages with springs, one fourth, and by carriages without springs one half, of the weight will not cause a deflection in the middle of the rail, equal to 032 parts of an inch. In the experiments previously detailed, the weights were applied carefully upon the rail, and, therefore, the rail was not subjected to the shocks and blows which occur in practice. It became, therefore, almost necessary that experiments should be made to ascertain the difference between the strain, to which railway bars were exposed in use, with the consequent deflection, and that which experiments carefully made presented. Professor Barlow, in his enquiries on the subject of the best form of rails for the London and Birmingham railway company, made some experiments on the Liverpool and Manchester railway, to ascertain the amount of deflection of different sections of rails, while the carriages passed over them. The locomotive engines are the heaviest species of carriage upon that railroad, and, therefore, one of these machines was taken to perform the experiment. The deflection was measured by an instrument placed underneath the middle of the rail, equi-distant from the points of bearing, and the extent of the deflection was ascertained by a multiplying index. The Speedwell, and Swiftsure engines were employed, the weight on the driving wheels of the latter being 5 tons 16 cwt.; and the velocity was varied from a slow rate, to that of twenty miles an hour. The following Tables shew the result of these experiments : TABLE XXVIII. No. 1. experiment, Speedwell engine. Nos. II. and III., Swiftsure. Grand Junction, 62 lbs. rail, bearings 3 feet 9 inches. Swiftsure engine, same railway bars, with bearing lengths of |