In the instance before us, the two 3-inch cylinders already alluded to in the proving-frame took the place of the larger cylinder of the ordinary apparatus; and they were connected with the forcing-pump by a strong metal tube. When a girder had been fixed in the frame for proving, the force-pump was worked till the pistons underneath the girder carried it off its lower bearings and pressed it upwards against the "shoulders," by which it was firmly held, and the pressure was then continued until the amount previously fixed upon as necessary for proof had been obtained. This was ascertained by means of a self-adjusting apparatus attached to the hydraulic press. An iron cylinder 1½ inches diameter was placed in communication with the pipe connecting the pump and the press, so that the pressure obtained in it was, in proportion to its diameter, the same as that in the large cylinder; and it was fitted with a piston-rod, working in a vertical direction. This piston-rod was connected with a lever, from the end of which a scale-pan was suspended, at a distance from the fulcrum ten times greater than that of the point of attachment of the piston from the same. The weight of the scale-pan and lever were balanced by a large mass of iron at the other end. In the scale-pan a certain weight was placed, proportioned to the proof desired to be obtained; and the action of the pump was continued until the water, rising in the iron cylinder just described, forced up the lever, and with it the weight attached; and thus indicated that the pressure to which it was desired to subject the girder had been reached. The weight to be placed in the scale-pan was thus determined: the diameter of the lever cylinder being 1½ inches, and that of each of those in the proving-frame three inches, the pistons or "rams" in the latter presented together eight times the surface of that in the lever cylinder; which being multiplied by the difference of length of the two parts of the lever, determines the weight for the scale-pan to be one-eightieth of that to which it was desired to prove the girder. The ordinary gallery girders were tested with a pressure equivalent to a weight of fifteen tons; but it was calculated that, when fixed, the greatest weight they would have to sustain would be seven-and-a-half tons. In one instance, for the sake of experiment, the pressure was continued beyond the proof weight of fifteen tons, to see what amount of strain the girders would bear without fracture, and it was found that a strain of thirty tons produced no injurious effect; but the girder broke with an additional weight of half a ton. Roof of Transept.