PRESSURE GAUGES 



25 



water (specific gravity of toluene '866 at 50 F. compressibility '000079 

 per unit volume per atmosphere) the error amounts to 1'02 per cent, at 20 

 atmospheres pressure and to 0'5 per cent, at 10 atmospheres pressure. 

 With kerosene compressibility '0000696 at 60 F. as the gauge fluid, this 

 error would be correspondingly less. A variation of temperature, by 

 altering the relative specific gravities of the two fluids, may alter the 

 gauge coefficient appreciably, as indicated below. The values of C are the 

 gauge coefficients for a gauge using water in connection with the particular 

 fluid. 



Over such a range of temperatures, the percentage error, corresponding 

 to a temperature range of 10 F., has a mean value of 2*4 per cent, with 

 sperm oil, 2*8 per cent, with toluene, and 2 P 4 per cent, with kerosene, 

 its magnitude increasing with the temperature. With water and 

 mercury, the coefficients of expansion are so nearly identical that the 

 error caused by this effect is practically negligible. 



Slight differences in the bores of the tubes and in their degree of clean- 

 liness also affect the readings by altering the form of the meniscus, while 

 where oil and water are used, the attraction between oil and glass causes 

 it to act in the gauge as though its specific gravity were higher than is 

 actually the case. In view of these facts, it is not sufficient to calculate 

 the gauge coefficient from a knowledge of the specific gravities alone if the 

 gauge is required for accurate work, and the only satisfactory way is to 

 calibrate it against a head of water under the same conditions as to 

 pressure and temperature as are likely to be experienced in use. 



In order that a piezometer may give an accurate record of the pressure 

 in a pipe containing water in motion, it is important that the surface of the 

 latter should be smooth in the neighbourhood of the piezometer opening, 

 and especially important that any burr or roughness produced on the pipe 

 wall by the drilling of the latter should be removed. Where a pipe is of 

 small bore, and where the pressure is to be measured at a point at some . 

 distance from one end, this is a matter of some difficulty, and in such a 



