314 PROCEEDINGS OF SECTION C. 
specially made, or a stock instrument opened, and its mer- 
cury contents adjusted and then calibrated. The ther- 
mometer is removed from the metal case in which it is 
bought, and mounted on corks in an outer-sealed glass tube, 
as before. All the thermometers should, of course, be com- 
pared against standard instruments, and their behaviour 
noted under both varying and steady temperatures. 
Electrical thermometers present advantages, m that, once 
inserted, they can be left mm sitdé indefinitely, and be 
cemented in, so that: the changes of temperature that take — 
place at the base of the hole many feet from the face of 
the rock can be watched, and measured time after time, 
. without any disturbance, and the rate noted at which the 
rock cools under the action, say, of a ventilating air current, 
in mines and tunnels. : 
Siemen’s resistancethermometer is naturally far more 
suited to the conditions of these particular experiments than 
is any method involving the use of thermo-junctions, and the 
author has employed both the original form of circuit and 
the one with separate compensating leads, as used by Cal- 
lender. Against the slight theoretical advantage of the 
latter form must be set the disadvantage of the rather more 
complicated circuit, to make which several pieces must be 
put together in a dirty mine drive in a bad light. The 
thermometer coils that the author has latterly used have 
been of platinum, and of about 25 ohms each, with the 
leads of 16 or 18 B.W.G. copper. The bridge should, of 
course be worked with equal ratios to the arms, on account 
of the varying resistances of the compensating leads; these 
arms have been made of 10 ohms resistance. 
The difficulties met in the insulation of a thermometer 
coil intended for use at high temperatures are not met in 
the present case, and the important thing to consider is, 
perhaps, the fixing of the coil to the leads, so that deforma- 
tion does not take place subsequent to the calibration. A 
coil wound upon a groove in a glass cylinder, or on a mica 
one, is thus better than one upon a mica plate; but in some 
of the author’s instruments a grooved ebonite reel was used, 
and the very stout copper leads (No. 10 B.W.G.) were made 
into rigid terminals, passing through the reel, and on which 
the coil could be adjusted to its resistance. 
There are two good methods by which the resistance can 
be read off, and the reading made. One is by the ordinary 
Wheatstones bridge (as used by Calender), with two 
equal ratio arms, and a set of resistance-coils that can be 
adjusted to the resistance of the thermometer. For work 
