This is always alluded to as a Torrie•llian vaenum.
and is found ill the ordi nary barometer. in Isfirt, Geissler invented a inercu rial air pump in which the varitum is produced by connecting a receiver with a Torricellian vacuum. The original form of tdssler's pump is shown in the accompanying • diagram, which will s) rye to illus trate the principle of the operation of pumps of this class, though they have re •eived numerous modifiea lions and improvements.
Lt umst m•reury pumps the parts are made of glass, the connections be ing made with rubber tub ing. to the diagram, .1 is a large bulb, B is a tube about :3 feet long, (' a rubber tube uniting the lower end of 11 with the vessel D, which is open On top. .1 C:111 be connected with either of the tulles G or but nut with both at (owe. or it ran be shut °IT from omit. The receiver to be exhausted is connected with G, and I' leads to the open air. Enough mercury is w-ed to fill .1, B. and D. as shown, and the vessel I) is capable of being raised or lowered. The operation of the pump is as fol lows: Suppose the vessel D is raised it little higher than A, as in the figure. The mercury will flow into the bulb A, which it tills if the cock h' is turned so as to vonnect .1 with the outside air, The cock is then turned so as to connect .1 through the tube G with the vessel to be exhausted. the air in Mild] at this stage is at atmospheric pressure. 1) is then lowered, and the level of the mereury in A is lowered in con sequence, the mercury running down B and C to D. As the mercury in A descends. air is drawn from the receiver through G into so when the mercury has descended below A the whole space is filled with the air drawn through G. which, having expanded from the receiver at tached to G, is at. less than atmospheric pressure. The cock E is then turned so as to cut off corn munieation between A and G. D is then slowly raised, and the mercury flows gradually baek into .1, compressing the air above it until it is at atmospherie pressure. At this point the cock P should he turned to connect .1 with the out side air •', and as I) continues rising the mer cury continues to drive out all the air at P, until the bulb .1 is filled with mercury to the cock E. which is then closed so as to cut off all commu nication with A. When I) is again lowered, the mercury does not begin to fall in .1 until 1) is about 30 inches below I. It then begins to descend, leaving a Torricelliau vacuum above it, and 1) is lowered until .1 is empty. The cock is then turned so as to connect .1 with the re ceiver through G. and the remaining air in that vessel expands and fills A. The cock E is net turned off, I) is raised, and the mercury rising in A compresses the air above it until it is let out at P by turning the cock. fly repealing this operation a suffieient number of times a vacuum is gradually produced in the receiver connected to G. %Viten the operation is nearly finished great can' must be taken not to raise the ves sel b too rapidly, ot the impact of the mercury against the top of the bulb .1 will break the apparatus. It will also be seen that when the vaenum is nearly reached the mercury in A will be at the top of the bulb when I) is about 30 inches below. If the valve should be turned to 1' at this point, the inrush of air would drive the mercury down. Therefore, no communica f between A and 1' must be made until I) has been raised on a level with E, and no com munication between G and .1 nmst be made until D is lowered '30 inches again, otherwise mercury it i11 run through G into the receiver rvllieL is Th, just deseribed may be taken as the type of mercury pumps, which are classi fied as upward driving, and, while a number of improvements in details have been introduced, making them of a more prat-that type for fac tory use, these pumps all operate on the prin ciple of eonneeting the receiver to be exhausted with Torrieellian vacuum.
Sprengel brought out his well-known form of mercury pump in ISO, and the diagram shows it in its simplest form. The Sprengel pump is a general type of what are elassified as downward driving pumps. A is a funnel having a stop-cock C. and B is a tube of small bore, called the shaft or fallAnbe. The receiver to be exhausted is conneeted to the tube G, which branches off from near the top of the shaft. The tithe It terminates very close to the bottom of the vessel D, which is provided with a spout P, as shown, leading to the cup E. The distan•e from the branch G 1.o the top of the mercury in the vessel P must be at least three feet. A is filled with mercury, which flows down the shaft B, the rate of flow being regulated by the cock U. so that it very small stream is allowed to fall. This mercury in fall ing breaks up into short lengths, between which are small eolutuns of air which How in at the ,junction of G with the shaft B. The weight of the mercury forces these short col umns of air down the shaft. It to the
mercury in D, from the surface of which they eseape, nw mercury as it runs into the cup II must he poured back into the funnel A. This operation eontinnes until no more air is Parried down with the mer cony. When the vacuum is nearly cionpleted, the mereury in the fall tube will fall with a sharp, rattling noise, showing that there is not enough air carried down with it to act as a cushion. With all kinds of mercury pumps. however. it is necessary to eontinue the operation for a considerable time after the re ceiver is apparently exhausted. Even when no more air appears to lie car ried on by the pump. the vacuum will improve as the operation con tinues. The reason for this is that the air sticks to the surface of the glass, forming a sort of coat ing, which is swept off the surface by the pump. but very slowly. The simple form of Sprengel pump is better than the simple Geissler pump, but is not well suited to fac tory work on aeeount of the slowness of its action. The drawback is overcome, to a great extent, by supplying the pump with a number of fall-tubes, which act together as a single one. For example, if six fall-tubes are used, the work of removing the most of the air is done in one sixth of the time required by a single pump. After the greater part of the air is removed, how ever, the time taken to produce a good vacuum is not nearly so much reduced, and it is chiefly in the early part of the operation where the saving of time is effected. Another drawback to all mercury primps is their liability to break age, even with the most careful usage. In the Sprengel pump. owing to the continual hammer ing of the mercury, the fall-tubes are very often broken, even after only a very short usage. A method is in use with both of these forms of pumps which consists of exhausting into a par tial vacuum instead of into the atmosphere. This is accomplished by inclosing the part of the apparatus where the air is expelled in a cham ber which is kept at a partial vacuum by means of a mechanical or water air pump. By this means the mercury pump will work against a pressure much less than the atmospheric pres sure, and consequently the fall-tubes and the height to which the mercury must be raised can be very much reduced, while the air is much more readily drawn down and out of the fall tubes. In factory work the raising of the mer cury from the lower to the upper level of the pumps is done mechanically and not by hand. It may be raised by a force-pump, or in small buckets on an endless chain, or by air pressure. The latter may be simply atmospheric pressure, and the mercury is raised by being broken up into small lengths with air spaces between, like a Sprengel pomp working upward into a vac uum chamber. The illustrations show modern forms of mercury pumps. In an improved form of Sprengel pump designed by G. W. A. Kahl baum a steel gun-barrel replaces the glass fall-tube. This avoids the electrification of the glass by the friction of the falling mercury, and with the other improvements introduced enables a higher vacuum to be attained than ever previously, In this way, in ,1901, he was able to obtain a degree of exhaustion eorre spondiug to a pressure of .000001S millime ters of mercury, which is considered the best on record.
The degrees of exhaustion reached by the vari ous mumurial air pumps may be seen from the following table adapted from Mfiller-Ponillet, Lehrbnch• der l'hycik: The water pump invented by Bunsen is a simple form of apparatus that is found generally in physical and chemical laboratories, and ade quately answers when too high a degree of ex haustion is not required. It consists of a tube attached to a faucet or other supply of water under pressure. through which water empties into a chamber provided with two outlets. From one of these the water flows out, carrying with it the air from the vessel to be exhausted, which is connected with the second tube. In its orig inal form this piece of apparatus was made of glass and rubber tube, but with metallic parts that allow it to be connected to an ordinary fau cet. It is extensively used in laboratories. The page illustration 3) shows one form of such a water pump. Descriptions of air pumps of various forms are to be found in all the large treatises on physics, including those of Ganot. Desehanel, and 1\liiller - Pouillet (Brunswick, 1886), the latter (in German) giving a complete account of the most important types of apparatus of this Blass. In the Journal of the Society of Arts, Volume XXXVI. (London. ISM, there is an interesting and valuable article on 'The De velopment of the 1\lereurial Air Pump," by S. P. Thompson, in which the various forms of this instrument are described. This has been re printed in book form. The reader is also re ferred to the columns of the der Physik and Chemie and the American Journal of Sci ence, in which are described many forms of air pumps and vacuum apparatus.