4 THE ANTARCTIC MANUAL. 
holds. Thus, in some preliminary observations with one of the 
pendulums of the Expedition a coincidence period of 28°750 secs. 
was observed. This gives a period of 0°50884 sec. To this period, 
however, so determined, various corrections are required, and the 
apparatus is arranged to enable these to be obtained with ease. 
Temperature Correction.—The dimensions of the pendulum vary 
with the temperature. The observed time of swing, therefore, 
must be corrected to some standard temperature for each degree of 
temperature. The amount of this correction will be determined at 
the National Physical Laboratory before the Expedition starts. In 
order to determine the temperature accurately a fourth dummy pen- 
dulum is provided, which contains a thermometer embedded in the bulb. 
This thermometer can be read through a glass panel in the outer 
case while the pendulums are being swung, and frequent readings 
should be taken during the observations, the time of the readings 
being noted. From these readings the mean temperature during the 
observations is obtained, and it is this which is to be used in the 
correcting term. 
It must be noted, however, that the range of temperature over 
which it is possible to determine this correction in England is very 
small compared with the range to which the pendulums will be 
exposed, and a determination of the correction during the Expedition 
will be of real value. If it should be found that the temperature of the 
observatory hut varies greatly during the 24 hours, it would be of ser- 
vice to keep a thermograph running during the observations. In any 
case the temperature of the air should be recorded from time to time. 
If we call & the change in period for 1° of temperature, T the 
standard temperature, and ¢ the mean temperature during an observa- 
tion, then the correction to the observed time of survey from this 
cause is — k(t — T). 
Pressure Correction.—The period depends, to some extent, on the 
pressure of the air in which the pendulum swings. The simplest 
way of avoiding trouble from this cause is to keep the pressure con- 
stant during an observation. Moreover, the swing is damped by 
the presence of the air, and as it is desirable that the swings should 
continue for a long time, the air pressure should be reduced. 
To secure these ends the swinging pendulums are enclosed in an 
air-tight case with glass windows, through which the observations 
are made. A gauge is fitted to this case, by means of which the 
pressure inside can be read, and an air-pump is supplied to reduce 
the pressure to some definite value. 
{n the U.S. survey a pressure of 60 mm. at 0° C. was adopted as 
