VACUUM, in the strict sense ,a portion of space entirely devoid of matter. The "absolute vacuum," as thus defined, is still a mere intel lectual concept because no physical or chemical method has yet been devised for effecting the absolute removal of every last trace of matter from any portion of space having finite, meas urable dimensions. If a vessel of glass or other material impervious to air is filled with air or any other gas, and the contents of the vessel are then removed as perfectly as possible by means of a good modern air-pump, a degree of exhaustion can be attained which is so per fect that no residual matter can be detected in the exhausted space, save by the most refined methods of testing. Spaces which have been partially exhausted, but which still contain a quantity of gas great enough to he easily de tected, are called "partial vacuums," and those in which the process of exhaustion has been pushed to a very high degree are called, for the sake of distinction, "high vacuums." The first known method for producing a high vacuum was that of Torricelli, who filled a glass tube, closed at one end and more than 30 inches long, with mercury, and then inverted the tube so that its open end dipped beneath the surface of a vessel also containing mercury. The atmospheric pressure is capable of sustaining a pressure only equal to that due to a mercury column some 30 inches high (see BAROMETER) ; so that when the experiment here indicated is performed, the mercury sinks in the tube until it stands above that in the lower vessel by only this amount. The space at the upper end of the tube, from which the mercury thus re tires, was long considered to be the most per fect vacuum attainable. It is not absolutely perfect, however, because it contains a small amount of mercury vapor, as well as traces of air. The method of Torricelli is not a con venient one for the practical production of high vacuums, for it is often objectionable to intro duce mercury directly into the space to be ex hausted, and it is also difficult, in many cases, to fill such a space with mercury so completely as to exclude small bubbles of air. Hence, in the practical production of vacuums, it is usual to remove the air (or other gas) from the space to be exhausted, by means of some form of a pump. ' The earlier pumps that were em ployed for this purpose (and which are still employed when high vacuums are not essential) were similar in principle to the "suction pumps" that are used for drawing water. They con sisted of one or more cylinders, provided with tightly-fitting pistons, and valves as nearly per fect as possible. Each stroke of a pump of this kind removes a certain fraction of the mass of gas remaining in the vessel to be exhausted; but it is theoretically impossible to produce an absolute vacuum in this manner, not only be cause an infinite number of strokes would be required, but also because it is impossible to prevent a certain amount of leakage around the pistons of such pumps. Other grave mechanical difficulties also obtrude themselves when the vacuum attains to a moderate degree of per fection. For the production of the high vacu ums that are used in the manufacture of in candescent electric lamps': and for the far more perfect ones that are needed in modern scientific research, it is, therefore, necessary to employ some other means. Geissler, in 1855, invented a mercury pump, in which advantage is taken of Torricelli's principle for the production of a vacuum. The vessel to be exhausted is nct filled with mercury, but it communicates with a large bulb which can be alternately filled with mercury and emptied, by merely raising and lowering a reservoir of mercury, with which it is connected by means of a flexible tube. When the mercury is caused to run out of the pump bulb by lowering the reservoir, the air from the vessel to be exhausted expands into the bulb; and when the bulb is again filled by raising the reservoir, the air that the pump-bulb contains is caused to pass out through a special passage provided for that purpose, so that it does not return into the vessel that is being exhausted.
By means of a mercury pump working on this principle, it is possible to reduce the pres sure in the exhausted space to the 100,000th part of an atmosphere. About 1865, Spren gel invented an ingenious and still more per fect mercury pump, by means of which ex ceedingly high vacuums may be obtained. In Sprengel's apparatus a stream of mercury is caused to pass down a small tube called the in the form of a rapid succession of separate drops; a small quantity of air from the vessel to be exhausted being entrapped be tween every two successive drips. The mercury sweeps this entrapped air down through the "fall-tube,s which discharges, at the bottom, i cistern. The Spretigel pump is slow in its action, but by its aid it is possible to produce vacuums so nearly perfect that the residual pressure probably does not exceed the 400,000, 000th part of an atmosphere. In producing high vacuums, it is necessary to remember that glass and other solid bodies are usually covered by a thin coating of highly condensed air (or other gas), or of water vapor, which adheres to them like a thin film of varnish. This film is quite persistent under ordinary circum stances, and is not given off at once, even in a high vacuum. If the film were not removed during the exhaustion, however, it would grad ually leave the glass afterward, becoming dis seminated through the exhausted vessel, and so reducing the vacuum very seriously. To pre vent the vacuum from being injured in this manner, the exhausted vessel is heated, by means of a Bunsen burner or otherwise, while the pump is in action. The rise in temperature accelerates the motions of the molecules of the air-film to such an extent that these molecules fly off into the interior of the vessel, from which they are then removed by the pump. High vacuums are sometimes obtained by com bining the air-pump with other devices that arc suggested by chemistry and physics. Thus An drews produced high vacuums by filling the ves sel to be exhausted with carbon dioxide gas, re moving the greater part of this gas by means of an air-pump, and then absorbing as much as possible of the remaining quantity, by means of fragments of caustic potash that had been previously introduced. Advantage has also been taken, in a similar manner, of the absorp tive power of freshly-prepared charcoal, for improving the air-pump vacuums. The char coal is placed preferably in a side tube com municating with the exhausted vessel, and is kept heated during the exhaustion, in order to prevent it from absorbing the gas until after the pump has ceased working. The exhaustion having proceeded by means of the pump to as high a degree as is possible, or as is desired, the connection to the pump is sealed off, and the charcoal, upon being allowed to cool, then absorbs within its pores a considerable part of the gas still remaining in the vessel. The per fection of the vacuum in the finished tube can be regulated, within certain limits; for by heat ing the charcoal more or less a greater or lesser part of the gas that it contains can be temporarily expelled, so as to reduce the degree of the vacuum. Dewar successfully applied liquid hydrogen to the perfecting of vacuums; the vessel to be exhausted being provided with an auxiliary bulb which can be plunged into the liquid hydrogen after the exhaustion has pro ceeded as tar as possible by the direct action of the pump. The intense cold of the liquid hydrogen causes a considerable part of the residual gas in the vessel to condense upon the walls of the auxiliary bulb, which is then sealed of from the main bulb by means of a blow pipe.