The object of Mr. Girard is to get rid of all these drawbacks by the simple expedient of substituting water for oil. It would not avail to apply water precisely as oil is applied. Though any one's experience may tell him that two smooth pieces of metal will slide more smoothly on each other when they are wet than when they are dry, yet every one knows also that oil facilitates the movement much more perceptibly than water; and also, that in the case of oil there is no difficulty in maintaining the lubricating film, whereas water easily evaporates, and in case of the accident of even a moderate elevation of temperature, it would be expelled from the joint entirely. Mr. Girard proposes, therefore, to employ the water to act, first, by its pressure, to lift the Journal to be lubricated; and secondly, by its fluidity, to form a liquid bed or cushion between the journal and its box, on which the journal may rest in its revolution, without touching the metal of the box at all. The construction will be understood by referring to the figure. One of the journals is represented as removed, and in the cylindrical surface of the socket are seen grooves occupying a considerable part of the area exposed. These grooves communicate, by an aperture in the middle, with a tube which is represented externally, and which sends a branch to the other journal, through which water under a heavy pressure is introduced into the box beneath the journal. The effect of the hydraulic pressure is to lift the axle, opening a passage for the escape of the compressed water, which at the same time, because of its release from compression, loses the power to sustain the weight. If, therefore, by the first impulse, the axle is thrown upward to any sensible distance, it will immediately fall back again, once more confining more or less completely the water. After one or two oscillations, therefore, the axle will settle itself at length in a position in which, while the water will escape, it will escape but as a film of inappreciable thickness. In this condition the journal turns upon a liquid bed, and the resistance to its revolution is so excessively small that a slow rotation given by hand to a wheel sustained by it will be maintained for many minutes without perceptible retardation. In fact, the most striking illustration which can be given of the immense superiority of the _palier glissant_ over a support lubricated in in any other way, is furnished by placing two precisely similar wheels or disks side by side, weighing five or six pounds each, with a diameter of seven or eight inches, and journals of half an inch in diameter; one of them furnished with _paliers glissants_, and the other with boxes lubricated with fine oil. Give each of them a velocity of rotation of about one revolution in a second; the one lubricated with oil will come to rest before the other begins to give evidence of any sensible retardation; but if at any moment the