NATURAL PHILOSOPHY. 213 



retain the exhaustion perfectly for any length of time, two open 

 vessels are introduced, one of which may be conveniently placed 

 above the other ; the lower vessel containing concentrated sulphuric 

 acid, the upper a thin layer of a solution of caustic potash, which has 

 been recently concentrated by ebullition. The precise quantities of 

 these liquids is not a matter of importance, provided they are so 

 adjusted that the acid is capable of desiccating completely the potash 

 solution without becoming itself notably diminished in strength, but at 

 the same time does not expose so large a surface as to convert the 

 potash into a dry mass in less than five or six hours at least. The 

 pump is in the first place worked till the air in the receiver has an 

 elastic force of 0-3 or 0-4 inch, and the stop-cock below the plate is 

 then closed. A communication is then established between the tube 

 for admitting air below the valves and a gas-holder containing carbonic 

 acid, which has been carefully prepared so as to exclude the presence 

 of atmospheric air. After all the air has been completely removed 

 from the connecting tubes by alternately exhausting and admitting 

 carbonic acid, the stop-cock below the plate is opened and the carbonic 

 acid allowed to pass into the receiver. The exhaustion is again quickly 

 performed to about half an inch or less. If a very perfect vacuum is 

 desired, this operation may be again repeated ; and if extreme 

 accuracy is required, it may be performed a third time. It is not 

 likely that anything could be gained by carrying the process further. 

 On leaving the apparatus to itself, the carbonic acid which has dis- 

 placed the residual air is absorbed by the alkaline solution, and the 

 aqueous vapor is afterwards removed by the sulphuric acid. The 

 vacuum thus obtained is so perfect, that even after two operations it 

 exercises no appreciable tension. 



To give a clear conception of the progress of the absorption, I will 

 describe in detail one observation in which the tension was measured 

 simultaneously by a good syphon-guage and by a manometer, formed 

 of a barometric tube 0-5 inch in diameter, inverted in the same reso- 

 voir of mercury as a similar tube communicating with the interior of 

 the receiver. The barometer had been carefully filled, and the 

 depression of the mercury estimated by the method already described 

 at less than 1 1 Q ^ of an inch. 



Previous to the admission of the carbonic acid, the exhaustion was 

 carried only to 0-4 inch ; it was again carried to 1 inch ; and a third 

 time to 0-5 inch, after which the apparatus was left to itself. The 

 manometer indicated a pressure in 



15' of 0-25 inch. 



30' " 0-17 " 



80' " 0-10 " 



200' " 0-02 ' 



In twelve hours the difference was just perceptible, when a per- 

 fectly level surface was brought down behind the tubes till the light 

 was just excluded. In thirty-six hours not the slightest difference of 

 level could be detected. The vacuum has remained without the 

 slightest change for fourteen davs. 



