1896 - 97 .] J. E. Erskine Murray on a New Thermometer . 301 
of gases is very misleading to the student, who is apt to look 
upon it as a factor of a different nature to the observed pres- 
sure in the ordinary air thermometer, and hinders his realisation of 
the fact of the simple proportionality of temperature and pressure 
or volume. 
§ 6. The length of a degree on this thermometer depends only on 
the total pressure at some given temperature ; hence it is clear that 
the sensibility of the thermometer may be altered to suit any pur- 
pose by choosing a proper pressure at which to fill the bulb with 
air at the standard temperature. Thus, if great sensibility be 
required, the air when immersed in melting ice should be under 
great pressure, making the height h, which is equivalent to 
273°, large; while, if a great range of temperature be required 
without great sensibility, it is only necessary to partially exhaust 
the bulb, at the same time placing the reservoir beloio the volume 
mark, so that the atmospheric pressure may be partially counter- 
balanced. 
The length of the pressure-gauge tube of the thermometer may 
thus be indefinitely reduced, though of course at low pressures the 
open reservoir would require to be placed at nearly the barometric 
height below the volume mark. 
The easiest way to obtain air at 0° C. at a pressure somewhat 
below that of the atmosphere is to fill the bulb with air at a high 
temperature, balancing the reduction of pressure and maintain- 
ing the volume constant as it cools by lowering the reservoir. 
When at 0° C. the air will now have a pressure below that of the 
atmosphere. 
§ 7. In order to obviate the necessity for a correction for capil- 
larity on account of the difference in the diameters of the pressure- 
gauge tube and the volume-gauge, the constant volume mark is put 
at the end of the fine tube of the stem, just where it joins the 
wider tube which forms its downward continuation. At this point 
the tube is conical, and its walls make an angle of about 45 degrees 
with the horizontal. Hence the mercury surface when at this 
point will be a plane, and its surface tension will act entirely in a 
horizontal direction. 
