BETWEEN THE FREEZING AND BOILING-POINTS. 
195 
Thermometer D. Box 1. 
Coils 640 + 160 = 800'030 units. 
8 units = 1° C., nearly. 
With 80 ohms external circuit. 
Date. 
Bridge-wire 
corrected. 
Total units. 
Corrected 
barometer. 
F.I. corrected to 
76 - 00 centims. 
July 21, 1898 . 
-4-679 
795-351 
75-806 
79-5924 
-4-655 
795•375 
75-816 
79-5913 
)) B . 
-4-614 
795-416 
75-826 
79-5925 
With 50 ohms external circuit. 
11 11 . 
- 4*651 
795-379 
75-816 
79-5917 
All these fundamental intervals, of course, only apply to thermometer P of each 
pair, or the one which is used to determine the rise of temperature in the water. 
No separate determinations are required when both the thermometers are at the 
same temperature over the scale between 0° and 100°. The correction is simply to 
thermometer P in its reading relative to thermometer C, when the water is heated 
through so many degrees in.the outflow end of the calorimeter. The “ cold” reading 
of the thermometers, when in place in the calorimeter, at any temperature of the 
water-jacket, will be the differential reading of the thermometers at the temperature 
indicated by the thermometer of the water-jacket. The effect of conduction from the 
ends of the calorimeter will appear as a slight change in this differential reading, but 
this is never more than '01° or ‘02~, and only comes in at the extremes of the range. 
In regard to the errors referred to in the first of this section, to he met with in the 
practical employment of platinum thermometers for very accurate work, the first one, 
due to a change in zero, can always be avoided by sufficient annealing and offers no 
difficulty. The second one is by far the most important, and is caused by the 
conduction of heat away from the air around the bulbs through the metal leads. 
This is rendered worse rather than better by the presence of the compensating leads, 
because of the greater amount of good conducting material introduced into the 
thermometer tubes. The employment of a small length of wire to connect the 
compensating leads cannot rectify it, nor will the prevention of convection currents of 
air in the glass tubes render it harmless. It can be reduced to a negligible quantity 
only by immersing a sufficient length of the thermometer tube, and can always be 
measured by withdrawing the thermometer tube more or less from the vessel or 
steam-jacket in which it is immersed, and determining the drop in temperature 
2 c 2 
