

Mr. H. Darwin on an Electric Thermostat. 409 
The main circuit includes “ series-coils ’ which are situated 
outside the oil-bath, and whose function is to modify the 
supply of current to the permanent heating-coil. _ Inter- 
mittently the shunt-cireuit is closed. the total heating effect 
being thus augmented so long as the shunt-circuit is complete. 
There are three respects in which the operation of the heating- 
coils may be automatically adjusted. Thus an increase in 
the heating effect is produced by a more frequent closing of 
the shunt-circuit, by increasing the duration of each such 
closing, and by diminishing the resistance which the series- 
coils interpose in the main circuit, 
Suppose now, as a particular case, that the thermostat has. 
been running continuously for a considerable time, the 
temperature of the atmosphere having been fairly constant 
meanwhile, so that something approximating to a steady 
state has been reached: that is to say, though the actual rate 
of heat production varies abruptly, as often as the shunt- 
circuit is closed or opened, yet the mean heating effect of 
the current remains approximately constant from one half 
hour to another. If a fall now takes place in the temperature 
of the surroundings, so that there is a greater demand for 
heat, the necessary Yeadjustment will be effected as follows :— 
The increased demand will be immediately met by a more 
frequent closing of the shunt-cireuit (containing the inter- 
mittent heating-coil), and the same mechanical movement 
which produces this more frequent closing of the shunt- 
circuit will cause the closing to become progressively of 
longer duration, and will also in time decrease the resistance 
of the series one so that by degrees the closing of the 
shunt-circuit ceases to be required so frequently. 
The controlling-bridge B has one pair of opposite arms of 
copper, these two arms “being i immersed in the oil-bath. The 
remaining arms are of manganin, the connexion to battery 
and galvanometer being made in the usual way. It is thus 
evident that if the bridge is balanced at some definite 
temperature, a change of temperature will disturb the balance ; 
for the product of the resistances of the two. copper arms 
changes rapidly with change of temperature, while the 
corresponding product for the two manganin arms _ is 
practically constant, so that for any determinate adjustment 
of the bridge, equality of these two products can only obtain 
when the copper arms are at one particular temperature. 
By adding to the resistance of one of the manganin arms the 
temperature of balance is raised; while by adding to the 
resistance of one of the copper arms it is lowered. 
One manganin arm consists of two coils in series, and one 
