question; and, we are the more inclined to rely upon this mode of settling the relative rigidity, inasmuch as it is not improbable, that the mode of manufacturing the elliptic rails may have some effect, in altering their rigidity, in different parts of the rail, as compared with what theory would deduce.

§ 4.-Experiments on the Comparative Rigidity of different Sections of Rails.

The following are experiments, made on the flexibility of different sections of rails. The weights were applied by means of a steelyard, and the deflection was measured by a micrometer screw, seventeen threads to an inch, with a dial-plate, shewing the sixty-fourth part of a revolution of the screw, and thus exhibiting the 1088th part of an inch deflection.

In conducting the experiments, a strong framing of timber was erected, the chairs of the rails were bolted firmly to the framing, and the rails fastened to the chairs, by the same keys which were intended to be used, in laying them for permanent use.

In all the experiments, the distance between the centre of the supports was, precisely, three feet; but as each end of the rail rested upon the chairs, intended for their use, the width of the base of the chair is stated, in the account of each experiment, thereby giving the length of clear bearing. They are stated, either to have been fixed or loose; when loose, the two ends, only, of the three-feet length, subjected to experiments, were firmly fixed to the two chairs, forming its support; but, when stated to be fixed, the entire rail was firmly keyed down to each chair, and all the remaining divisions of the rail, or all the chairs, at

every three feet, were firmly screwed down to the timber framing.

The first column of the tables shews the weight, applied to the middle of the rail, in cwts.; the second column shews the deflection in the middle of the rail, while the weights were resting upon it; and the third column shews the deflection, after the weights were removed. The deflection, after the weights were removed, and again applied, is, also, shewn in the tables. Thus, in Table I., weights were successively applied, amounting to sixty-eight cwt., when the deflection was found to be 0386 of an inch; they were then taken off, and no permanent deflection observed. The weights were applied again, and the deflection found to be the same as before. This was the case with five weights; but on six being applied, equal to 102 cwt., the deflection was ⚫0615; they were then removed, and a permanent set of 0027 had taken place; the weights were again applied, and the deflection was found to be greater than when the same weights were applied, or 00624, shewing that the rail had been injured. The third column, in each set of experiments, gives the deflection, corresponding to each seventeen cwt. of insistent weight.

TABLE I.

Wrought-iron rail, Fig. 7, Plate II., 14 feet 11 inches long, of five lengths, weight 167 lbs., 33 lbs. per yard. Breadth of top 2 inches, depth 1 inch; depth of rib 1 inches, breadth inch; depth of keel inch, breadth inch; extreme depth in middle 3 inches, tapering away, in a semi-elliptical curve, to 23 inches at each end. Breadth of chair 34 inches.

TABLE II.

Wrought-iron rails, Fig. 10, Plate II., 14 feet 10 inches long, of five lengths, weight 157 lbs., 32 lbs. per yard. Breadth of top 24 inches, depth inch; depth of rib 1 inches, breadth inch; depth of keel 1 inches, breadth 1 inch; extreme depth in middle 3 inches, tapering away, in a semi-elliptical curve, to 24 inches, within 14 inch of each end.* Breadth of chair 34 inches.

In this and the two following rails, a convex projection at the end of each, 3 feet length, about inch deep, and 3 inches long, is rolled upon the under side, which fits into a corresponding cavity in the base of the chair.

each 17 cwt.

TABLE III.

Wrought-iron rail, Fig. 10, Plate II., 14 feet 8 inches long, of

Breadth of inches, breadth

inch; extreme

five lengths, weight 162 lbs., 33 lbs. per yard. top 2 inches, depth 1 inch; depth of rib 1 inch; depth of keel 1 inches, breadth depth in middle 3 inches, tapering away, in a semi-elliptical curve, to 2 inches, within 14 inches from the end. Breadth of chair 34 inches.