739 
of Edinburgh; Session 1873 - 72 . 
The following arrangement of apparatus was employed :— 
1. A long iron bar, one of those used by the late Principal 
Forbes in his experiments on the conduction of heat, was heated 
at one end in the usual manner. This formed the source of heat at 
.once steady and graduated, so that, by contact with it at various 
parts, the platinum wire experimented with could be kept at any 
required temperature. 
2. Small glass tubes were fitted into holes in the bar at regular 
intervals, and turned over a little at the edge in the form of a lip. 
These served the double purpose of preventing metallic contact with 
the bar (and thus introducing ordinary thermo-electric currents), 
and also served as guides to the same point of contact in each 
experiment. 
3. A small iron bar kept at the temperature of the room. 
4. A reflecting galvanometer (with somewhat massive mirror 
and magnet, so as to “ integrate”), with a scale placed at the dis¬ 
tance of six feet, so that the smallest deflection of the needle could 
be readily observed and measured. 
5. Two pieces of the same platinum wire connected with the 
galvanometer in the usual manner. 
The mode of working was as follows :—The free end of one of the 
platinum wires rested on the small bar, and was thus kept at the 
temperature of the room. The free end of the other wire was 
placed in one of the glass tubes on the heated bar, and, while in 
that position, and after it had attained the temperature of the bar 
at that particular spot, the wire from the small bar was brought 
into contact with it, and the sudden deflection of the galvanometer 
needle noted. 
With this arrangement very good and steady results were 
obtained when care was taken to keep the wires perfectly clean, 
and to apply the same amount of pressure in making contact in 
every experiment, because any deficiency of contact increased the 
resistance so as greatly to affect the currents. 
The results show that for platinum wire the current, as indicated 
by the deflection of the galvanometer needle, is exactly as the dif¬ 
ference of temperature between the two wires. 
To show the steadiness of the results, I give the details of one 
experiment— 
b M 
VOL. VII. 
