890 
effects are, in general, small. For example, in the Hartmann- 
Hilliar gauge the relief pressure from the circumferential vistons 
diminishes the pressure on the zero piston by less than 2%. Being 
small they can be found from the incompressive approximation, the 
error in which has been estimeted in (17.1) and (17.2). The gauge 
block itself 1s the source of rarefactione for two reasons: (1) 
it recoils as a rigid body and (2) the front face retreats because 
it is really soft rather than rigid. These effects have been esti- 
mated (paragraphs Pl. - 25.). The thickness of an “immovable" soft 
block is alwsys less than the thickness of an "immovable" rigid 
block (pseragranh 23). In general, the relief vressures due to 
gauge blocks in current use are pvrovortional to their velocities 
and are rather large, while corresponding vressures from the vistons 
are proportional to their accelerations and quite small. According 
to the Kirkwood criterion there should be no cavitation in front 
of the vistone which are now used. Their small diameter insures 
both no cavitation and also accuracy in describing them with the 
incomvressive anvroximation, Turbulence should be imnortant only 
if the vistons are of small diameter and travel long distances. 
In most geuges the vressure necessary to overcome viscosity and 
turbulence does not exceed 1 atmosphere. Finally the high velocity 
of the recoiling pistone introduces an error into a descrivtion 
of their motion by linear theory, and makes it necessary to con- 
gider the Bernoulli effect, (paragraph 27.). The Bernoulli 
pressure msy become of the order of 10 atmospheres. 
= Og s 
