AIR PUMP. An instrument for removing the air from a vessel. These pumps may Inc divided into two classes, mechanical air pumps and mercurial air pumps. The mechanical air pump was invented by Otto von Guericke about 1654, and a specimen of his early apparatus is shown in Fig. 1 of the accompanying page illustration. ln Fig. 2 is illustrated a modern simple air pump whose essential part is a hollow brass or glass cylinder, in which an air-tight pis ton is made to move up and down by a rod. From the bottom of the cylinder a connecting tube leads to the space which is to be exhausted, which is usually formed by placing a bell-glass, called the receiver, with edges ground smooth and smeared with lard, on a flat, smooth plate or table. When the piston is at the bottom of the barrel and is then drawn up. it lifts out the air from the barrel, and a portion of the air un der the receiver, by its own expansive force, passes through the connecting tube and occupies the space below the piston, which would other wise be a vacuum. The air in the receiver and bar rel is thus ra reliod. The piston is now forced down, and the effect of this is to close a valve placed at the mouth of the connecting tube and opening inward into the barrel. The air in the barrel is thus cut off from returning into the receiver, and as it becomes condensed forces up a valve in the piston, which opens outward, and thus es capes into the atmosphere. When the piston reaches the bottom and begins to ascend again this valve closes; and the same process is re peated as at the first ascent. Each stroke thus diminishes the quantity of air in the receiver; but from the nature of the process it is evident that the exhaustion can never be complete. Even theoretically there must always be a portion left, though that portion may be rendered less than any assignable quantity; and practically the process is limited by the elastic force of the re maining air being no longer sufficient to open the valves. The degree of rarefaction is indi cated by a gauge, on the principle of the barom eter. As this air pump only withdraws the air at the rate of one cylinder full for a double stroke of the piston, pumps with two barrels are frequently used, in which case the pistons are each attached to the same handle but each moves in an opposite direction to the other, the object being to double the work done at each stroke of the handle. Such zi pump is illus trated in Fig. 5 of the page plate. A large
number of modifications of this type of pump have been invented, all of which are the same in general principles. There are several rea sons why such pumps do not continue the process of rarefaction indefinitely, but after a certain stage their effects cease and the tension of the air undergoes no further change. Leakage at various joints in the pump is one limiting cause to the action of the machine. It is impossible to prevent leakage entirely, and at the beginning of the operation the quantity of air which enters the receiver through leakage is very small in comparison with the amount pumped out. But as the exhaustion proceeds the leakage is faster on account of the reduced pressure in the re ceiver, and finally a limiting point is reached when the inflow and outflow are equal and no reduction in the tension of the air takes place. Another limit to the action of this machine is caused by the fact that there must always be some space between the bottom of the piston and the lower end of the cylinder, which is untra versed by the piston. At the beginning of the operation tills space contains air at atmos pheric pressure, which is rarefied at each stroke of the piston; but some tension always remains there, and when the air of the receiver reaches the same tension no further effect will be produced by the pump. Perhaps the most important trouble, however, with this type of air pump, as well as the most difficult one to remedy, is the absorption of air by the oil used for lubricating the pistons. This oil finds its way in a greater or less quantity to the bottom of the cylinder, where its absorbed air is partially given up at the moment the piston begins to rise. This class of pumps is not good enough for the manufacture of incandescent lamps and vacuum tubes, and recourse is had to the mercurial air pump, by means of which a much greater degree of exhaustion is obtained. The principle of the mercurial air pump was first known in the seventeenth century, when To•ricelli showed how to produce a vacuum by filling a tube over 30 inches long and closed at one end, with mercury, and then inverting the tube, with the open end temporarily closed, in a vessel containing the same liquid. The mercury in the tube then descends to a height equal to that of the barometer above the level of the mercury in the lower cup, and a vacuum is left in the top of the tube.