398 BOTANICAL GAZETTE [NOVEMBER 
When the temperature of the oven falls below the required point, 
the thermostat again functions and the switch is instantly closed. 
This action continues automatically as long as the current is 
supplied, or until the mercury in the cups is all oxidized. In one 
instrument it was found that a difference of o.01° was sufficient 
to actuate the mechanism. As has been said before, turning the 
adjusting screw to the right adjusts for higher temperatures, and 
vice versa. It will be seen that the two circuits are absolutely 
independent of each other, and that when the oven is at the required 
temperature both circuits are broken; that the electricity is only 
on when the temperature is below the required point. 
Because the switch closes so sharply when the thermostat 
functions, it is advisable to drill a small hole in the core of the mag- 
net nearest the switch and insert a rubber plug for a buffer. If the 
mercury splashes too much, rubber corks, with a hole in the center, 
may be put in the tops of the cups. The flash which occurs when 
the switch points leave the mercury is bright, but it does no harm. 
It is possible to use an alternating current, but it is very difficult 
to adjust the armature spring so as to avoid the unpleasant hum- 
ming. It is planned to try a weighted or balanced armature with 
alternating currents. A battery may be used to actuate the switch 
if magnets of lower resistance are used, and the lamp rheostat 
replaced with a plug; then the alternating current may be used to 
heat the coil in the oven. 
If preferred, the lamp used for a rheostat may be placed in the 
oven and connected to the socket on the switchboard with a cord 
and plug. This arrangement is very convenient, for the instant 
the door of the oven is opened the lamp is lighted. Also the heat 
of the lamp increases the efficiency of the apparatus. 
Anyone at all familiar with tools can construct the apparatus 
at a cost of about $3.75 for material. Of course he must have 
access to an engine lathe to wind the coil so that the wires will not 
touch each other. After the first experimental instrument was 
found to work satisfactorily, the interest of a very expert mechani- 
cian, Mr. A. W. Srrickrer, 1311 E. 57th St., Chicago, Ill., was 
enlisted. He suggested many improvements, always having in 
mind increase of efficiency and lowering of cost. It is largely due 
