I had an order from Mr. Mathieson, manager of the works of the National Tube Company, at McKeesport, Penn., for two engines, 28 and 32 inches diameter, with 48 inches stroke. The interest of this story centers in the former of these engines, which made 125 revolutions per minute. One day the governor spindle stuck fast in its column, an accident I never knew to happen before or since, whether caused by a tight fit or for want of lubrication I do not know. Of course the engine ran away like mad. Mr. Mathieson and I were in the engine-room; the last I saw of him his coat skirt was nearly horizontal as he rushed through the door. The engineer ran to screw down the starting-valve. I thought that would be too long a process and ran in front of the fly-wheel to unhook the gab. On the instant, however, I feared what might be the possible effect in the cylinder of instantly arresting the motion of the admission valves at an unknown point in the stroke at that speed, and I did not do it. In a few seconds the engineer had the valve closed, and the engine soon slowed down. The fly-wheel, which was 20 feet in diameter, did not burst, and I was confident it would not. I never had an accident to a fly-wheel, but this was the most severe test to which my fly-wheels were ever subjected. I have heard of many accidents to fly-wheels, in which it was evident that they were so carelessly made it seemed as if they were intended to burst on a moderate acceleration of their speed. This fly-wheel was necessarily made in halves in order to transport it, and the joints were so made as to be as strong as the section of the rim. As the accompanying drawing will show, they were held together by two steel loops opened out of the solid and shrunk in. It will be seen that any section of cast iron at this point was equal to the section of the rim, while the steel loops were stronger. The halves of the hub were held together by bolts and steel rings.