RADIATION IN ABSOLUTE MEASURE. 
597 
the plane of the mirror, for the purpose of bringing the zero of the scale to the cress 
wire of the telescope. This is not often required, but occasionally, wdien the sun 
shines strongly on the iron frames of the Glasgow University windows, the zero 
may be pulled round by a few divisions ; and, occasionally, also a movement of 
some iron in the laboratory slightly displaces the zero. The “ zero magnet ” 
being moved slightly, with its axis along a line perpendicular to the plane of the 
mirror, adjusts the zero without altering the field of force. I find it an excellent 
plan to put a little “ soft wax on the zero magnet, and press it on to the table in 
front of the galvanometer shelf. By cautiously pressing on the magnet it may be 
made to slide very slowly, with viscous yielding of the wax, for the purpose of 
setting the zero of the galvanometer. 
I have mentioned already that I have made, with the assistance of my friend, 
Mr. A. Tanakadate, a determination of the thermo-electric value of a platinum- 
platinoid couple. We considered it necessary, however, to make special determi¬ 
nations of the values of the actual couples used in these experiments, and this was 
done each time a fresh soldering of the platinoid and platinum was made. Platinum 
wire is so variable in its qualities that it is not safe to assume that the physical 
properties of difiPerent portions of the same hank even are the same.t 
The standardizing of the thermo-junction is performed with the circuit precisely as 
it is to be used during the radiation experiments, with every joint soldered, and with 
the field of force of the galvanometer suitably adjusted and rendered unchangeable, 
as described above. Along with the thermo-electric wires, the leads, and the galvano¬ 
meter, a platinoid resistance of about 12 ohms was introduced into the circuit. The 
sensitiveness of the arrangement was quite suflficient to permit of this being done ; and 
it was of great advantage to introduce this added resistance, as it rendered insensible 
small variations of resistance (due to temperature) in the long platinum thermo¬ 
electric wire, which in early experiments gave rise to curious and embarrassing 
variations in the galvanometer readings. 
From the standardizing, or calibration, as it is often called, of the thermo-electric 
circuit in use, a formula is obtained, from which is calculated, from the '■ corrected 
deflection of the galvanometer, the difference of temperatures of the two junctions 
in terms of centigrade temperatures on the air thermometer.§ Every temperature 
* The material used by chemists for making diachylon sticking plaster answers admirably. 
t The best makers will not, I believe, guarantee to .supply uniform platinum, in the physical sense of 
the word uniform, knowing that it cannot be done, at present at all events. 
t The deflection is read off on a plane scale, and is corrected for the fact that the angle turned through 
by the light ray is twice the angle of deflection of the magnet and mirror. In order to apply this 
correction with facility, a table of numbers was constructed. The correction may now, however, be 
easily applied by the use of the recently published “ Reductions-Tabellen zur Gauss-Poggendorffschen 
Spiegelablesung,” by Dr. Paul Czeemak. J. Springer, Berlin. 
§ Sir William Thomson’s article “Heat,” ‘ Encyc. Brit.,’ and J. T. Bottomley, ‘ Proc. Roy. Soc., 
Edinburgh,’ 1888, and ‘ Phil. Mag.,’ August, 1888. 
