2GO LECTURE XXIX. 



The density and pressure of the air may be diminished, or the air may be 

 perfectly or very nearly withdrawn from a given space, either by means of 

 a column of mercury, or by the air pump. The ancients sometimes ex- 

 hausted a vessel imperfectly by the repeated action of the mouth, and 

 preserved the rarefaction by the assistance of a stopcock. The Torricellian 

 vacuum, obtained by inverting a receiver filled with mercury, and fur- 

 nished with a descending tube at least 30 inches long, is the most perfect 

 that can be procured ; but there is generally a portion of air adhering to 

 the vessels, and mixed with the mercury, which may often be considerably 

 diminished by agitation, but can only be completely expelled by boiling 

 the mercury for some time in the vessel and its tube, previously to their 

 inversion. (Plate XXIV. Fig. 324.) 



The construction of an air pump greatly resembles that of a common 

 sucking pump for raising water ; but the difference in the operation to be 

 performed requires a difference in several particular arrangements. The 

 objects are, to rarefy or exhaust the air as completely, as expeditiously, and 

 as easily, as possible. In order that the exhaustion may be complete, it is 

 necessary that no air remain in the barrel when the valve is opened, and 

 that the process be very long continued. For, supposing all the parts of an 

 air pump to be perfectly well fitted, and the exhaustion to be carried on for 

 any length of time, the limit of its perfection will be a rarefaction expressed 

 by the proportion of the air remaining in the barrel, when the piston is down, 

 to the whole air that the barrel is capable of containing ; for such will be 

 the rarity of the air in the barrel when the piston is raised. It becomes, 

 therefore, of consequence to lessen the quantity of this residual air as much 

 as possible : and at the same time to take care that the valve may be capable 

 of being accurately closed and easily opened, or that a stopcock may be 

 occasionally substituted for it, which may be opened and shut by external 

 force, when the elasticity of the air remaining is too small to lift the valve. 

 In pumping water from a well, we raise an equal quantity at each stroke, 

 but in the air pump, we withdraw at most only equal bulks of the air dif- 

 ferently rarefied, so that the quantity extracted is continually diminished 

 as the operation proceeds. Thus, if one tenth of the air were exhausted by 

 the first stroke, only nine tenths as much, that is, one tenth of the remain- 

 der, would be drawn out by the second ; hence, in order that the process 

 may be expeditious, it is of importance to have the barrel as large as pos- 

 sible in proportion to the receiver. In cases where the presence of aqueous 

 vapour would be of no consequence, the exhaustion might be made very 

 rapidly by filling the whole apparatus with water, which was the me- 

 thod first employed by Otto von Guericke, the inventor of the modern 

 air pump. 



In order to lessen the labour of the operation, two barrels may be em- 

 ployed, and so connected as to work alternately ; in this manner the 

 pressure of the atmosphere, acting on both pistons at once, opposes no 

 resistance to their motion in either direction. In Smeaton's pump* a 



* Ph. Tr. 1751-2, xlvii. 415. See also the Dutch translation of Dr. Priestley's 

 Observations and Experiments on different kinds of air, vol. ii. 1781. Cavallo, Ph. 

 Tr.1783, p. 435. 



