170 
MR. H. L. CALLENDAR OX THE PRACTICAL 
Mercury thermometers were used to check changes of temperature in various parts 
of the apparatus ; their readings are recorded in the column headed “ Thermometers.” 
S T gives the temperature on the stone slab under the gas furnace ; this, being care¬ 
fully screened, was not found to rise appreciably, even after the furnace had been full 
on for an hour. R gives the temperature indicated by a thermometer placed inside 
the resistance box. The temperature of the air was taken because variations in the 
temperature of the reading microscopes or their supports would introduce small errors ; 
as the variations of the air temperature were very small, these errors could be 
neglected. 
Table II., giving Results of Observations on Contraction of Hard Glass Tube, 
the Temperature being given by Platinum Wire. 
Temperature, 
pt. 
Expansion 
expressed in 
percentage of 
length. 
Expansion 
calculated by 
formula of 
interpolation. 
Mean 
coefficient of 
expansion by 
formula. 
° C. 
494 
•401 
•403 
•814 
165 
•122 
•120 
725 
136 
•098 
•097 
•717 
98 
•068 
•069 
•706 
60 
•043 
•042 
•696 
17 
•012 
•012 
•685 
481 
•391 
•389 
•810 
330 
•254 
•254 
•769 
222 
•165 
•164 
■740 
102 
•070 
•072 
•706 
19 
•013 
•013 
•685 
First 
series. 
I 
r 
Second series 
7 March. 30. 
J 
1 
Marcli 31. 
On the first occasion the permanent contraction of length which the glass tube 
underwent amounted to '052 per cent, of its length, and on the second to '026 per 
cent., although this glass does not begin to soften till about 700° C., and was 
subject to no strain whatever : with other materials (e.g ., porcelain) these changes 
would probably be more serious. They are a most important source of error, and 
place a limit to the accuracy attainable with an air thermometer at such temperatures, 
unless an arrangement is adopted to measure them. 
The volume of the bulb is also liable to change, owing to the variation of pressure 
to which it is subjected. At ordinary temperatures these changes are usually less 
than -Toooo P er atmosphere, and may be neglected ; but at high temperatures near 
the softening point of the glass they might become sensible, and the bulb might even 
undergo a permanent deformation. As the phenomenon in question cannot be readily 
measured at high temperatures, it is best to avoid the difficulty by so adjusting the 
mass of air enclosed that its pressure may amount to an atmosphere at the highest 
temperature to be reached, and the glass may thus be entirely freed from the 
uncertain effects of a strain at the high temperature. Such alterations of volume are 
