VACUUM, in physics, a portion of space void of matter. The possibility of the existence of a perfect vacuum has been a favorite subject of discussion among metaphysicians; and that the neg ative view was popular is indicated by the historical phrase "the vacuum which nature abhors." Descartes asserted that if the contents of a hollow vessel were taken out without anything entering to fill its place the sides of the vessel would be in contact. This assertion, however, is based on the dogma that the only essential property of matter was its extension; that, in other words, mat ter was indistinguishable from space. Such inquiries, however, can lead to nothing definite. The experimental phil osopher is obliged to recognize the prac tical impossibility of obtaining a per fect vacuum. The so-called Torricellian vacuum, which exists above the mer cury column of a barometer, is really filled with mercury vapor probably mixed with a small percentage of air. The very action by means of which the air pump produces its vacuum shows that there can never be an absolute void—a little air always remaining behind. Andrews, fol lowing up a suggestion of Davy, ob tained remarkable vacua by first pump ing in carbonic acid so as to expel as far as possible the air, and then after ordinary exhaustion leaving the carbonic acid to be taken up by moistened caustic potash which has been previously placed in the receiver. In this way Andrews
removed all but s..'rra-5 of the gas originally present, whereas by ordinary air pump action only Ill of the gas can be removed. Frankland, Gassiott, Crookes, Dewar, and others, have de vised various improvements. Dewar takes advantage of the property of the gases to condense on the surface of solids in a remarkably ingenious way. A piece of carbon, placed in the vacuum ves sel, is kept heated while the vessel is being exhausted. After the exhaustion is completed as far as possible, the car bon is allowed to cool so as to permit the small quantity of gas present to con dense in its pores. A gentle heating of the carbon at once releases a quantity of the gas, so that this method is inval uable in the study of electric discharges through rarefied gases, since the rarefac tion can be so easily altered at will.