In 1782 came Watt's last steam invention, an engine that used steam _expansively_. This was an immense stride. He was also at the same time the inventor of the "throttle," or choke valve, by which he regulated the supply of steam to the piston. It seems a strange thing that up to this time, about 1767, an engine in actual use was started by getting up steam enough to make it go, and waiting for it to begin, and stopped by putting out the fire. Then he invented the "governor," a contrivance that has scarcely changed in form, and not at all in action, since it was first used, and is one of the few instances of a machine perfect in the beginning. Two balls hang on two rods on each side of an upright shaft, to which the rods are hinged. The shaft is rotated by the engine, and the faster it turns the more the two balls stand out from it. The slower it turns the more they hang down toward it. Any one can illustrate this by whirling in his hands a half-open umbrella. There is a connection between the movement of these balls and the throttle; as they swing out more they close it, as they fall closer to the shaft they open it. The engine will therefore regulate its own speed with reference to the work it has to do from moment to moment. [Illustration: THE GOVERNOR.] Through all these changes the original idea remained of a vacuum at the end of every stroke, of indispensable assistance from atmospheric pressure, of a careful use of the direct expansive power of steam, and of the avoidance of the high pressures and the actual power of which steam is now known to be safely capable. [Footnote: In a reputable school "philosophy" printed in 1880, thus: "In some engines" (describing the modern high-pressure engine, universal in most land service) "the apparatus for condensing steam alternately above and below the piston is dispensed with, and the steam, after it has moved the piston from one end of the cylinder to the other, is allowed to escape, by the opening of a valve, directly into the air. To accomplish this it is evident that the steam must have an elastic force greater than the pressure of the air, _or it could not expand and drive out the waste steam on the other side of the piston, in opposition to the pressure of the air_." According to this teaching, which the young student is expected to understand and to entirely believe, a pressure of steam of, say eighty to a hundred and twenty pounds to the inch on one side of the piston is accompanied by an absolute vacuum there, which permits the pressure of the outside air to exert itself against the opposite side of the piston through the open port at the other end of the cylinder. That is, a state of things which would exist if the steam behind the piston _were suddenly condensed_, exists anyway. If it be true the facts should be more generally known; if not, most of the school "philosophies" need reviewing.] Then an almost unknown American came upon the scene. In