ON THE MECHANICAL EQUIVALENT OF HEAT. 
383 
The altitude of the Owens College, Manchester, has no appreciable effect on the 
value given by the above formula. 
I give below the table of steam pressures used in the calibration of the scale of 
the thermometer Pi. 
Temperature on 
Ceutrigrade scale. 
Temperature on 
Fahrenheit scale. 
Pressure of steam in 
millims. of mercury 
reduced to 0° C. and sea 
level in lat. 45°. 
Pressure of steam in inches 
of mercury reduced to 0° C. 
and sea level in latitude of 
Manchester. 
O 
99 
210-2 
733-305 
28-849 
100 
212-0 
760-000 
29 899 
101 
213-8 
787-590 
30-984 
102 
215-6 
816 010 
32-102 
24. The general arrangement of the apparatus used to check the scale of the 
thermometer Pi will be gathered from the annexed sketch (fig. 8), and from Plate 6 
attached to Professor Reynolds’ paper, (Part I., par. 48). 
A is an ordinary copper boiling-point apparatus, the steam from the boiling water 
passing up an inner tube in which the thermometer to be tested is hung, 
and then flowing down again so as to jacket this tube, finally escaping into 
the atmosphere through the cock shown. The top of the inner tube is 
closed by a cork having two holes, in one of which is fitted a half-inch brass 
tube for connection with the manometer, the other carrying the thermometer. 
B is a glass flask containing an artificial atmosphere, of which the pressure is 
under control. 
C is the combined barometer and manometer used to measure the pressure in 
A and B. 
D is the tin receiver previously described, the pressure in which is kept at about 
18 inches of mercury, as measured on a U-gauge. This receiver is in free 
communication through a capillary glass tube with the tube connecting the 
flask B and the manometer C. 
The bore of the capillary tube just mentioned is just sufficient to admit a very 
small stream of air from the receiver through the flask B, and so out into the atmos¬ 
phere by way of the cock on the boiler. The object of this stream of air was to 
counteract the tendency of the steam in the boiler to diffuse down the connecting 
rubber tube into the flask, where condensation would occur, and possibly some w^ater 
might get into the barometer, it having been found quite impossible to keep a steady 
pressure in the apparatus whenever the steam made its way as far as the glass flask, B. 
The boiler was well lagged and protected as far as practicable from draughts. A 
