The errors which must be considered for accurate measurements of pressure with piston gauges are discussed. Harwood’s unique design of piston gauge which permits the operator to control the clearance between the piston and cylinder at any operating pressure is described. Instrument errors are thus substantially reduced, particularly those resulting from elastic distortion. A check list of errors inherent to piston gauges is presented.
These errors are to be considered when instruments are used to accuracies better than 0.5 %. The need for a check on the long-time variations of the piston gauge is supplied be the fixed points, changes of the piston gauge is supplied by the long-time variations of the piston gauge is supplied by the fixed points, changes of state of some very pure materials which take place at definite pressures. Such are the melting pressure of solid mercury at 0oC, the melting pressure of water at 30oC, the transition between crystalline states of bismuth. As an indication of its possibilities the experimental model of the controlled-clearance piston gauge was used to measure the melting point of mercury at 0Oc. Results were in satisfactory agreement with Bridgman’s value.