GENERAL MEETING. 27 
it will regain its ordinary freezing point reading in one month. 
The change in the first part of this period is much more rapid than 
towards the end. To recover the depression caused by 50° re- 
quires only two days. The older a thermometer the more quickly 
it gains its freezing point corresponding to ordinary temperatures. 
An instrument forty years old will regain its freezing point after 
exposure to 100° in one week while an instrument three years old 
requires a month. 
The more alternations of temperature a thermometer is subjected 
to the more quickly its freezing point rises. 
A thermometer subjected to a very high temperature, as 350°, will 
will have its freezing point raised from 12° to 20°. This rise is © 
not due to softness of the glass at the high temperature and a con- 
sequent diminution in the volume of the bulb by the atmospheric 
pressure. This is shown by experiments with weight-thermometers. 
In these the tubes are open to the pressure of the air and there is 
as much pressure inside as outside the bulb. 
As heating to 100° depresses the freezing point while heating to 
350° raises it there must be some intermediate temperature for 
which there is no change. This point is usually at the tempera- 
ture of about 160° to 180°, but varies widely with thermometers 
made of different kinds of glass. 
When a thermometer is subjected to a very low temperature a 
temporary rise in its freezing point is produced. To produce an 
appreciable rise requires a long-continued exposure. After being 
kept twenty-four hours at — 30° the freezing point is found to be 
about 0.05° higher than at first. 
One hundred degrees on the centigrade scale is taken as the tem- 
perature of steam from pure water boiling under a normal barometric 
pressure equal to 760 mm. of mercury. A variation of 1 mm. 
in the pressure will change the temperature 0.04°. 
Zero is taken as the temperature at which pure ice melts when 
subject to an atmosphere of pressure. An increase of a whole at- 
mosphere lowers the temperature of the melting point of ice 0.008°. 
This is to be distinguished from the effect of an atmosphere of 
pressure on the reading of a thermometer ; by compressing the bulb 
it causes the thermometer to read about 0.°2 higher than if there 
was no pressure. 
The fundamental distance on a normal thermometer is taken as 
the reading it would have at a true temperature of exactly 100° 
