54 L. R. INGERSOLL 
\action” a reasonably rapid action of the thermometer, once the bulb starts 
to change in temperature, as.otherwise it would occupy too much time in 
coming to the temperature of the hole. 
After some calculation and much experimentation a very satisfactory 
solution of the problem has been reached, involving insulating the bulb in 
vulcanite, wax and dynamo paper with waterproof lacquer as shown in Fig. 
1. If such a thermometer is subjected to even such a radical temperature 
change as is involved in plunging into ice water, it is a full minute before the 
mercury thread shows a change of 1/10°. Having started to move, however, 
it falls several degrees a minute and comes down (asymptotically) to the 
Bakelite Vulcanite 
tube Paper \Wwax Cork 
ei’ VLLITLZI NL ZLLLLLL LL | QLZZILLLZL LT LL 
ELIE LLLLE IL 
: ERTS SAN | ———> 
NS SS 










& 
Fig. 1. Section of mounted thermometer (Overa!! length, 35 cm). 
temperature of the bath—within 0.02°, say—in about half an hour. The 
final design of these thermometers has proved very satisfactory indeed for 
the work in hand. Difficulties experienced with the early ones over the se- 
paration of the mercury thread have been entirely overcome in the later ones 
by having them made with about one atmosphere of inert gas pressure over 
the mercury. This has the further advantage of relieving the bulb of practic- 
ally all pressure, inside or outside, and doubtless tends to render the zero 
still more stable. 
Temperature readings 
As the holes are drilled in the usual manner with air-operated machines, 
with a flow of water through the drill steel, there is almost no heating effect 
due to drilling. Temperature readings taken immediately after the hole is 
completed are sometimes a degree high, but after a few hours all evidence of 
drilling heat has usually disappeared. To be on the safe side, however, litiie 
reliance is placed on readings made less than 24 hours after drilling, and the 
whole question of possible small errors due to heat of drilling is one on which 
considerably more study is to be expended. 
Thermometers are used in tandem groups of two or three, run into the 
hole with a stout wire, with the last one carrying a rubber gasket. The mouth 
of the hole is stopped with a wooden plug. While the thermometers reach 
equilibrium in half an hour they are leit for-at least an hour and then quickly 
withdrawn and read. They are then replaced in the hole, to be read again 
about every two hours until several sets have been taken. This procedure is 
repeated a couple of days later, with occasional readings continuing for sev- 
eral weeks or months. In addition, air, rock face and psychrometer readings 
are occasionally taken in the drift to furnish general information as to the 
thermal conditions at the station. 
Plots of these temperatures with time usually give an approximately 
198 
