The engine was started at 80 revolutions per minute. This was the same speed at which their old engine was supposed to run, but practically its speed had always fallen to 60 revolutions whenever two passes were in the rolls together. I should say here that the new engine was set at the opposite end of the train from the old one, and the only change made was disconnecting the old engine and connecting the new one. The advantage was found in the fact that with the new engine four or even five passes could be in the rolls simultaneously and the speed of the engine never fell sensibly below 80 revolutions per minute. The result was that the first week the train turned out 2400 tons of rails instead of 1200 tons, which was the former limit. This latter was a product of which they had been quite proud and which they claimed exceeded that of any other mill. Mr. Daniel N. Jones, their chief engineer, increased the speed of the engine five revolutions per minute each week for four successive weeks by changing the governor pulley for a larger one. This he did every Sunday when the mill was idle, increasing the speed finally to 100 revolutions per minute and the production to 3000 tons per week. He prided himself on doing this without the men at the hooks finding it out, which if they had done might have made trouble. This seems a very small thing to say when for many years the output of a rail-train has been 3000 tons a day without the aid of human hands; but at that time it was considered an immense achievement. It was also a remarkable thing for the company financially, as directly after a greatly increased demand for steel rails appeared and the price rose to $60 per ton, at which it was maintained for some time. [Illustration: DANIEL N. JONES] This thoughtful act of Mr. Jones was an example of his magnificent co-operation with me in all my work. Mr. Jones had insisted that the cylinder should have a support at the back end, as he felt sure that without it the running of the piston, weighing 3600 pounds, would produce a deflection; so a support was built under the end of the cylinder, which was cast with a corresponding projection underneath. These surfaces were planed parallel with each other, but I took pains to secure a space between them sufficient to admit a sheet of paper, and when the engine was running I was able to draw a sheet of paper through that space without its being seized, showing the support of the cylinder from the bed to be sufficient, as I had claimed it would be. Mr. Jones laughed. [Illustration: Connection of Arms and Rim in Mr. Fritz’ Fly-wheel]