(1) To make a cheap. Obtain a straight fine glass tube, about 33 in. long, and with a clean interior, sealed at one end, and having an even uniform bore of about 2 1/2 lines diameter. The mercury to be used should be perfectly pure, and free from all air and moisture. This latter requisite may be assured by heating the mercury in a porcelain dish to nearly the boiling-point, previous to using it. The tube is then held securely, with the open end uppermost, and carefully filled with the liquid metal. The open end of the tube is then securely covered with the finger, the tube is inverted, and the end is covered by the finger plunged below the surface of a little mercury placed in a small vessel to receive it. The finger is then removed, when the mercury in the tube will immediately fall to a level of about 30 in. above the surface of that in the small reservoir below. In order to attach the scale correctly, it will be necessary to compare the indications with those of some good instrument.

(2) Exhaustion of tubes without application of heat. The defects inherent to the methods at present employed for exhausting barometer tubes induced Klobukow to study the various methods, especially that of Bogen, adopted in meteorological observatories. By a modification of the latter he succeeds in producing a complete vacuum by means of a mercury column without application of heat - Bo gen's method. A well-dried barometer tube is charged with pure mercury, the imprisoned air is removed by shaking, and the orifice is closed with the finger; it then is inverted over a mercury trough, partly immersed in the metal, placed in a vertical position, and the surplus of mercury allowed to escape. The tube is again closed by pressing the fingers on to the open end, brought in a horizontal position, and gently shaken for a short time. After a portion of the air embedded in mercury has entered the vacuum, the tube is transferred to the trough, manipulated as before, and the operation is repeated twice or three times till the mercury is freed from the adhering air.

When the sectionof the tubing is reduced to that of a capillary tube, the filling with mercury by this method is impracticable; this difficulty is overcome by the following modification.

The barometer tube b being filled with mercury in the usual way, is connected by a rubber tubing of 2-3 in. long to a glass tube, open at both ends, of the same diameter and length as b; the tubes are placed in the position as shown in Fig. 358, and a is filled with mercury. Care being taken that no air remains confined in the bend c, a few taps should be given to the rubber tubing by the hand, or pressed by the finger with sufficient force to cause the emission of a few drops of mercury. By introduction of an air bubble into the tube a, and partial closing of the orifice with the finger, all air confined in the rubber bend is expelled by compression of the rubber. The compression being maintained for a few seconds till a small portion of the mercury has been forced out and the tube hermetically closed, when the closed end is immersed in mercury, and the finger withdrawn from the orifice, Fig. 359.

The following operation consists in the formation of a vacuum: - By raising the barometer tube 6, Fig.360, it then is lowered while a is moved in a nearly vertical position, Fig. 361, which causes a migration of the previously produced vacuum fg to the opposite portion hi of the tube. The exhaustion of the tube is repeated twice or three times, during which the tube is gently moved until the characteristic sound of the metal, on lowering and raising of the tube, becomes audible. After complete exhaustion, the tubes are disconnected, b is closed and immersed in mercury.

The charging of a with mercury is effected by means of a small funnel, Fig. 358, and a complete exhaustion of the tubing is attained within 20-30 minutes.

The method is applicable to siphon and cistern barometer, and can also be used for the exhaustion of capillary tubing; the employment of an air pump and application of heat in exhausting the latter will greatly accelerate the operation. The manipulations which the method requires are simple; they can be performed with good results by persons unaccustomed to experimental work, always giving an excellent vacuum.

(3) How to fill a tube with quicksilver. Having got the tube which is open at one end (narrow end), heat the bulb in a flame; in doing this the air in the bulb expands; but the other end of the fine tube being open, the expanded air gets out through this end. Next, before the air has had time to cool, plunge the open end of the tube below the surface of a vessel containing mercury. As this air cools it shrinks into less bulk, and the pressure of the air from without drives the mercury to occupy the vacant space. Part of this mercury will therefore be driven into the bulb. We next take the bulb with the mercury in it, and heat it well above the flame of a lamp, bulb, tube, and all. The mercury will soon begin to boil, and its vapour will be driven out; and the air before it, until bulb and tube will both be filled with vapour of mercury. When this is done, we plunge the open end of the tube once more into a vessel of mercury. As there is only vapour of mercury in the tube, when this cools it will condense, and the mercury in which the instrument is plunged will go into the bulb and tube, and it will be filled.

Be very careful not to inhale the mercury fumes. (R. W.)

(4) Application of Wright's apparat as for distilling, to the filling of barometer tubes. In Fig. 362, a is a vessel full of impure mercury; 6 a tube about 30 in. long; c an enlargement of b;d and e tubes inclined in opposite directions; / an arm for connecting with a Sprengel pump; g a tube a little over 30 in. long; h a reservoir with an outlet to the air; h is filled with pure mercury.

Exhausting barometer tubes.