But in the beginning it was not proposed to use steam in connection with the cylinder and piston which now really constitutes the steam-engine. Reverting again to the example of the gun, it was suggested to push a piston forward in a tube by the explosion of gunpowder behind it, or to repeat the Savery experiment with powder instead of steam. These ideas were those of about 1678-1685. The very earliest cylinder and piston engine was suggested by Denis Papin in 1690. These early inventors only went a portion of the way, and almost the entire idea of the steam-engine is of much later date. Mankind had then a singular gift of beginning at the wrong end. Every inventor now uses facts that seem to him to have been always known, and that are his by a kind of intuition. But they were all acquired by the tedious experience of a past that is distinguished by a few great names whose owners knew in their time perhaps one-tenth part as much as the modern inventor does, who is unconsciously using the facts learned by old experience. But the others began at the beginning. [Illustration: EARLY NEWCOMEN PUMPING ENGINE. STEAM-COCK, COLD WATER COCK AND WASTE-SPIGOT ALL WORKED BY HAND.] In 1711, almost a hundred years after the arrival at Jamestown and Plymouth of the fathers of our present civilization, the steam-engine that is called Newcomen's began to be used for the pumping of water out of mines. This engine, slightly modified, and especially by the boy who invented the automatic cut-off for the steam valves, was a most rude and clumsy machine measured by our ideas. There appears to have been scarcely a single feature of it that is now visible in a modern engine. The cylinder was always vertical. It had the upper end open, and was a round iron vessel in which a plunger moved up and down. Steam was let in below this plunger, and the walking-beam with which it was connected by a rod had that end of it raised. When raised the steam was cut off, and all that was then under the piston was condensed by a jet of cold water. The outside air-pressure then acted upon it and pushed it down again. In this down-stroke by air-pressure the work was done. The far end of the walking-beam was even counter-weighted to help the steam-pressure. The elastic force of compressed steam was not depended upon, was hardly even known, in this first working and practical engine of the world. Every engine of that time was an experimental structure by itself. The boiler, as we use it, was unknown. Often it was square, stayed and braced against pressure in a most complicated way. Yet the Newcomen engine held its place for about seventy-five years; a very long time in our conception, and in view of the vast possibilities that we now know were before the science. [Footnote: As late as 1880, the steam-engine illustrated and described in the "natural philosophy" text books was still the Newcomen, or Newcomen-Watt engine, and this while that engine