406 Bourdon, Tait, and Amagat High-Pressure Gauges. 



tube) is constructed of metal. Furthermore, choice must be 

 made of as thin-walled a steel measuring-tube as possible, in 

 order that the volume-increase due to pressure may be large 

 relatively to the simultaneous fluctuations of temperature. 

 This can be accomplished by connecting the Tait gauge and 

 the compressor by an interposed piston libre. In the annexed 

 diagram let AB be the cylindrical piston, consisting of a 



single piece of which one end is larger in diameter than the 

 other*. Let this piston fit the hollow cylinder, C C C, 

 accurately, so as to be capable of motion to and fro, or of 

 rotation, with the minimum of friction. Let the tube D be 

 in connexion with the compressor, the tube E with the 

 Tait gauge, and let the whole apparatus be filled with oil (§ 1). 

 Then it is clear that exceptionally high pressure acting at D 

 can be counterbalanced by relatively low pressure at E. 

 Hence a Tait gauge reading to only 1000 atm. may be made 

 to indicate pressures two, three, or many times as large as 

 this, by selecting a suitable size of piston ; and hence an 

 exceptionally thin-walled steel tube may be inserted in the 

 gauge. 



I am far from underestimating the importance of Amagat's 

 manometre as a standard instrument leading at once to abso- 

 lute results ; but it is certainly uncomfortable to be obliged 

 to make readings on a column 4*6 metres high even under 

 favourable circumstances. ^Nor would it conduce to con- 

 venience and certainty if the registration were made electri- 

 cally. On the other hand, in the case of the Tait gauge, the 

 pressures are read off on a horizontal tube about one metre 

 long, and as far as 2000 atm., without pistons libres, and 

 therefore quite without unavoidable leakage f and continually 

 decreasing pressure. 



* My colleague Dr. William Hallock devised other high-pressure 

 apparatus acting on the above principle, some time ago. 



f Regarding high-pressure gauges rigorously free from leakage, let me 

 say that I have the apparatus ready for expressing pressure in terms of 

 the isothermal diminution of the electric resistance of pure mercury. Cf. 

 American Journal, xl. p. 219 (1890). 



