Trevithick's rough hand-sketch shows the steam-ring in the stuffing box and the steam-vessel; the particular use of the latter he has not described: probably it was because Cornish pumping engines, not having the controlling crank to limit the movement of the piston, are obliged to trust to the very admirable, but little understood, steam-cushion, without which the ascending piston would inevitably strike and break the cylinder-cover, while in the pole puffer-engine this danger was during the descent of the pole, and therefore the discharge-steam valve was closed, while the steam in the pole-case was still of ten or more pounds to the inch, so that by the time the pole reached the finish of its down-stroke, it had compressed this steam-cushion, filling also the steam-vessel, with a pressure approaching to that in the boiler, and equal to the weight of the pole and pump-rods. A comparatively small supply of steam from the boiler into the steam-vessel brought it up to the boiler pressure, sending the pole and pump-rods upwards with a spring. The steam-valve then closed, allowing the momentum of the great weight of pump-rods, together with the expanding steam, to complete the up-stroke. The discharge-valve was then opened for a moment, allowing a blast of steam to escape, reducing the pressure say to one-half. The weight of the rods caused their downward movement, raising the load of water in the plunger-pole pumps, and at the same time compressing the steam from the pole-case into the steam-vessel, equal at the finish of the stroke to the support of the pole and pump-rods. This most simple steam-engine combined in the greatest degree the two elements of expansion and momentum. The up-stroke began with a much higher pressure of steam than was necessary to raise the load; having given momentum to the rods, the supply of steam was cut off, and the stroke was completed by expansion. The down-stroke began with a comparatively low pressure of steam under the pole. The unsupported pump-rods fell downwards, setting in upward motion the column of water in the plunger-pole pumps. The discharge-valve was closed long before the completion of the down-stroke, and the momentum of the moving mass of rods and water compressed the steam driven from the pole-case into the steam-vessel up to a pressure equal to the support of the pole and pump-rods. The pole was, therefore, continually floating or rising and falling in steam of ever-varying pressure. Trevithick' s figures show the working power of the 33-inch pole as much greater than Watt's 72-inch cylinder engine, even when the steam pressure in the former was much reduced, and that Woolf's double-cylinder engine, of less power, cost ten times as much as the pole-engine. This sum probably included the costly buildings required for the beam-engines, which Trevithick's plan dispensed with.