64 ROYAL SOCIETY OF CANADA 
structed mirror galvanometer, reading with a telescope and scale. The 
apparatus was adjusted in such a way that the scale reads direct in 
degrees Fahrenheit at the rate of 1 cm. to 1°. The scale is divided into 
millimetres on opal glass, and can be read by estimation to ;}, of a 
degree F. 
Meruop or TAKING OBSERVATIONS. 
The thermometers are connected to. numbered plugs, any one of 
which can be placed in a hole connected to the indicator. The current 
is turned on by a suitable switch, and the scale-reading taken and 
recorded to ‘01° F. The readings thus taken require certain small 
corrections to reduce them to the true scale of the air-thermometer, but 
owing to the manner in which the galvanometer and the balancing 
resistances have been compensated for change of temperature, and in 
consequence of the accuracy of the adjustment of the thermometers, 
these corrections do not amount on the average to more than one-tenth 
of a degree Fahrenheit. 
The observations have, in general, been taken by one of the obser- 
vatory assistants ; occasional observations have also been taken by Prof. 
McLeod and by Prof. Callendar, as a check upon the accuracy and as a 
test of the condition of the apparatus. From the various tests applied 
from time to time, it appears that the observations of different observers 
agree very closely, and that the probable error of a single observation is 
less than one-fiftieth of 1° F. 
The apparatus is so arranged that the condition of insulation of each 
thermometer can be tested at any time, without disturbing it, by making 
a very simple change in the connections. No defect of any consequence 
has so far been observed. The insulation of nearly all the thermometers 
is about one thousand times as good as is actually required. 
Location oF THERMOMETERS, 
A trench about 3 feet wide and 9 deep was dug in the soil of the 
level ground of the gardens, exposing a vertical face on one side. Into 
this face of undisturbed soil holes were bored with a half-inch iron rod, 
kept in a horizontal position by means of a spirit level, at various pre- 
determined depths. The thermometers were inserted into these holes to 
a distance of nearly 3 feet from the vertical face, beginning from the 
bottom. The lowest thermometer was inserted at 9 feet, the next at 5 
feet 6 inches, the others at 40”, 20”, 10’, and 4’’, respectively. The wires 
leading from the various thermometers to the building were buried side 
by side in a trench about 3 feet deep. 
The ground where the thermometers are buried is covered with a 
layer of turf, and consists of loose light-brown sand of fairly uniform 
