In pursuing the experiments on the single axle, I adopted the contrary mode. I first of all put the axle loaded with weights, into rapid motion, and ascertaining the time occupied, until it assumed a state of rest; I thus obtained more accurate results, than it was possible to do by pursuing the contrary method. To effect this, the axle was placed upon two bearings, at such a height from the ground, as would allow a weight to descend thirty feet; a wheel was fixed in the middle of the axle, two feet diameter, around which a cord was wound, to the end of which a weight was attached; and rings of lead were fastened upon the axle, to vary the weight. In each experiment, the cord was wound around the wheel, and the weight was thus elevated precisely thirty from the platform; by withdrawing a pin, the weight was then let free, and, falling thirty feet, unwound the cord, and put the axle and lead weights into rapid motion; the cord then detached itself, and left the axle to turn freely round, until the friction of the axles brought it to rest. By a proper apparatus, the time occupied, during each ten revolutious, of the axle, was measured, as also, the whole time, until it came to rest; by which means, not only the absolute amount of friction was obtained; but, also, the friction at different velocities; and, by varying the weights, from 1331 lbs. to 4140 lbs., the relative resistance with different weights, was also ascertained. The principal object, however, of instituting this set of experiments,, was to ascertain if the friction varied with the surface of bearing; and, if there was any, and what size of bearing, subjected to a given pressure, produced the least resistance. With this view, bearings of three, four and a half, and six inches, respectively, were used; the diameter of the axle in each case being three inches; and, on each of

which, the successive weights of 1331, 2465, 3622, and 4140 lbs., were placed. With these variations, the number of experiments made were more than 600, and they were repeated in every possible way, to leave no doubt, as to the accuracy of the result; the weight, in each experiment, falling precisely thirty feet.

It would be carrying the subject too far, to give detail of those experiments, which were so numerous; though, when we consider, that the friction of attrition forms so prominent, and, indeed, almost the sole resistance of carriages, any experiments, tending to elucidate the subject, must be of importance. Having, however, previously brought these experiments into the shape of a separate treatise, and finding them too voluminous to embody in this work, they will, probably, be presented to the notice of the reader, in some way or other; we shall, therefore, at present, only give the result of those, which seem to apply more particularly to the subject of our inquiries, as regards railroad carriages.

As before stated, the weight which put the axle in motion, was made to fall in every experiment, precisely thirty feet; when the weight of axle and lead, and the size of bearing were not varied, in every experiment, the number of revolutions, by the same moving power, ought to have been the same. But this, I found, was not the case; the difference, perhaps, was not of that magnitude, to render it of any importance, in practice, but still there was a variation in the result of each experiment, which could not be overlooked; and the degree of smoothness of the axle, and bearing, precluded the idea, that the variation proceeded from any asperities on the surface; and, if it did, the variation should have been uniform, and progressive, whereas, it

was quite irregular; at last, I found the irregularity was owing to the quantity, and mode of applying the oil; which, as it appears of great consequence, in diminishing the friction, I shall endeavour to explain. The axle rested upon the chairs, without any cap, or cover, as here shewn ;

B

where A, represents an end view of the axle, and в, the chair. At the commencement of each experiment, the axle was oiled in the usual way, with fine neat's foot oil; but it was found, that unless the oil was continually feeding upon the axle, as it turned round, the result was never the same; unless the oil was supplied in such quantities, that when the axle turned round in the direction of the dart, the oil was heaped up against the axle, as shewn at c, and thus kept up a continual supply to the axle, the result was not uniform. When that was not the case, although the axle was well oiled, yet, unless the oil was kept constantly feeding upon the axle, as it turned round, a maximum effect did not take place.

The following Table, being one of the series of experiments, will shew the effect:

TABLE X.

Weight of axle and lead, 2465 lbs., bearing 4 inches.

Axle well oiled in the four

first experiments.

Axle well oiled in the following experiment.

experiments.

Number of Vibrations of Number of Number of Vibrationsof Number of pendulum. revolutions. experiments. pendulum. revolutions.

In conducting these experiments, the first four were made with the axle oiled, so as to keep it constantly feeding on, as shewn in the figure. The weight, being drawn up, was liberated, and falling thirty feet, the respective number of revolutions in the third column, were made before the axle came to a state of rest; the second column, being the number of oscillations of a pendulum, vibrating 300 times in 157 seconds. At the end of experiment 277, the oil, which was resting upon the bearing, heaped up, as shewn in the figure, was merely removed, as cautiously as possible, so as to allow that which surrounded the axle, to remain; the weight, then, was drawn up, as before, and falling precisely the same distance, the number of revolutions was, in that experiment, 189. No additional oil being applied, the weight was successively drawn up and liberated, as before, and the number of revolutions in each experiment, before the axle came to rest, are shewn in the Table; in the 300th experiment, the number of revo lutions, by the same moving force, was only thirtyseven, during the whole of which period the axle was never touched, no oil was applied, or none removed. At the end of the 300th experiment, the axle was again copiously oiled, so as to feed on during the whole of the 301st experiment, as shewn in the diagram, when the number of revolutions were 265. The oil was then removed, as before, when the number regularly diminished, until the 323rd experiment, when it was again reduced to thirty-six; and when, in the next experiment, the oil was applied as before, the number was increased to 278; by the same weight, falling precisely the same distance, which, in the previous experiment, only produced thirty-six revolutions.

On examining the Table, it will be found that the