The general scheme of Geisler’s pump is shown in fig. 1. A and B are pear-shaped glass vessels connected by a long narrow india-rubber tube, which must be sufficiently strong in the body (or strengthened by a linen coating) to stand an outward pressure of 1 to 1⅜ atmospheres. A terminates below in a narrow vertical tube c, which is a few ∣ inches longer than the height of the ' barometer, and to the lower end of this tube the india-rubber tube is attached which connects A with B. To· the upper end of A is soldered a glass two-way stop-cock, by turning which the vessel A can either be made to communicate through s and a hole in the hollow cock with the vessel to be exhausted (I., fig. 2), or through g with the atmosphere (II., fig. 2), or can be shut off from both when the cock holds an intermediate position. The apparatus, after having been carefully cleaned and dried, is charged with pure and dry mercury, which must next be worked backwards and forwards between A and B to remove all the air-bells. The air is then driven out of A by lifting B to a sufficient level, turning the cock into position II., and letting the mercury flow into A until it gets to the other side of the stop-cock, which is then placed in the intermediate position. Supposing the vessel to be exhausted to have already been securely connected with b, we now lower the reservoir B so as to reduce the pressure in A sufficiently below the tension in the gas to be sucked in, and, by turning the cock into position I., cause the gas to expand into and almost fill A. The cock is now shut against both a and b, the reservoir lifted, the gas contents of A discharged through a, and so on, until, when after an exhaustion mercury is let into A, the metal strikes against the top without interposition of a gas-bell. In a well-made apparatus the pressure in the exhausted vessel is now reduced to - 1 ⅛ or 1 ⅛ of a millimetre, or even less. An absolute vacuum cannot be produced on account of the unavoidable air-film between the mercury and the walls of the apparatus.

The great advantage of the mercurial over the ordinary air-pump is that it evacuates far more completely than the latter, that it affords direct and unmistakable evidence of the exhaustiveness of its work, and—last not least—that it enables one to transfer the gas sucked out to another vessel without loss or contamination, so that it can be measured and analysed. On account of this latter feature more especially, the instrument is highly valued as an auxiliary in gasometric researches. Without it the researches on which rests our present knowledge of the gases of the blood could not have been earned out. The actual instrument, as constructed for various kinds of work, has of course various complexities of detail omitted in the above description. For these the reader must refer to handbooks of practical physiology.

As it takes a height of about 30 inches of mercury to balance the pressure of the atmosphere, a Geisler pump necessarily is a somewhat long-legged and unwieldy instrument. It can be considerably shortened, the two vessels A and B brought more closely together, and the somewhat objectionable india-rubber tube be dispensed with, if we connect the air-space in B with an ordinary air-pump, and by means of it do the greater part of the sucking and the whole of the lifting work. An instrument thus modified was constructed by Poggendorff (see his Annalen, vol. cxxv. p. 151, 1865), and another, on somewhat different principles, by Prof. Dittmar (see the “Challenger" Reports).

Even a Geisler’s stop-cock requires to be lubricated to be absolutely gas-tight, and this occasionally proves a nuisance. Hence a number of attempts have been made to do without stop-cocks altogether. In Töpler’s pump this is attained by using both for the inlet and the outlet vertical capillary glass tubes, soldered, the former to somewhere near the bottom, the latter to the top of the vessel. These tubes, being more than 30 inches high, obviously act as efficient mercury-traps; but the already considerable height of the pump is thus multiplied by two. This consideration has led Alexander Mitscherlich {Pogg. Ann., cl. 420, 1873), and quite lately F. Neisen {Z. f. Instrumentenkunde, 1882, p. 285) to introduce glass valves in lieu of stop-cocks. As glass floats on mercury, such valves do not necessarily detract from the exhaustive power of the pump.

2. The Dynamic Pump.— This was invented in 1865 by H. Sprengel. The instrument, in its original (simplest) form (fig. 3), consists of a vertical capillary glass tube a of about 1 mm. bore, provided with a lateral branch b near its upper end, which latter, by an indiarubber joint governable by a screwclamp, communicates with a funnel. The lower end is bent into the shape of a hook, and dips into a pneumatic trough. The vessel to be exhausted is attached to b, and, in order to extract its gas contents, a properly regulated stream of mercury is allowed to fall through the vertical tube. Every drop of mercury, as it enters from the funnel, entirely closes the narrow tube like a piston, and in going past the place where the side tube enters entraps a portion of air and carries it down to the trough, where it can be collected. If the vertical tube, measuring from the point where the branch comes in, is a few inches greater than the height of the barometer, and the glass and mercury are perfectly clean, the apparatus slowly but surely produces an almost absolute vacuum.

The great advantages of SprengeΓs pump lie in the simplicity of its construction and in the readiness with which it adapts itself to the collecting of the gas. It did excellent service in the hands of Graham for the extraction of gases occluded in metals, and since then has become very popular in gas-laboratories, especially in Britain. Many improvements upon the original construction have been proposed. One of these which deserves mention is to pass the mercury, before it enters the “falling” tube, through a bulb in which a good vacuum is maintained, by means of an ordinary air-pump or a second ‘ t Sprengel.” (W. D. )