

THE SIPHOX-GAUGE 277 



artially filled with it, to be placed under the receiver 

 f the air-pump, while the atmosphere is still at the 

 amnion pressure and communicates freely with the 

 pen end of the gauge, the pressure of the atmosphere 

 dll sustain the mercurial column in the closed branch, 

 r hich is shorter than a common barometer, and the 

 bsed leg of the tube will remain filled ; and no change 

 ill happen in the height of the mercury in both legs, 

 yen after several strokes of the piston, as long as the 

 ressure of the air in the receiver is capable of support- 

 iig a column of mercury equal to difference in height 

 etween the surface of the mercury in the closed leg, 

 nd the surface of it in the open branch. But when the 

 ressure of the air decreases, the weight of this column 

 'ill predominate, and the mercury in the closed 

 ranch will fall, so that the difference of the levels in 

 IG two legs will always be equal to the pressure of 

 le rarefied air in the receiver. Thus, if the surface 

 f the mercury in the closed leg is l cm higher than 

 hat it is in the other, the pressure of the air in the 

 reiver is the seventy -sixth part of the common atmo- 

 :>heric pressure. A good air-pump should be capable 

 { diminishing the pressure in the receiver until the 

 ,iuge indicates only a difference of level of l mm . 



(The gauge must be perfectly free from air, or otherwise the 

 jessure of the air contained in the closed leg of the tube may be 

 Beater than that of the air in the receiver, in which case the sur- 

 ijje of the mercury in the open branch will be higher than in the 

 Msed ; such a gauge would evidently be quite useless. 

 iCare is necessary in letting the air again into the exhausted 

 i^eiver, or the mercury in the gauge will strike with much force 

 siainst the top of the closed leg, and may possibly break it ; the 

 *->p-cock should therefore only very gradually be brought back 

 ain into the position I. 



