388 Proceedings of Royal Society of Edinburgh, [july 18 , 
May 1887) lias attempted to prove experimentally the vanishing of 
the Peltier effect at the neutral point in the case of certain alloys, 
whose neutral points with lead are at temperatures not much above 
the ordinary temperature of the air. In his method the Peltier 
effects produced by a current at the junctions of a thermopile (of 
the alloys A and B) are measured by the thermopile itself. Now 
at the neutral point of A and B, the thermoelectric power of the 
thermopile becomes zero, and hence the pile becomes, at that tem- 
perature, a wholly inefficient means of measuring any heating or 
cooling effect. Therefore Signor Batelli’s experiments by no means 
establish the vanishing of the Peltier effect at the neutral point. 
In the following experiments, measurements were made of the 
Peltier effect both above and below the neutral point, and these 
were compared with the values calculated from the directly ob- 
served neutral point, taking Professor Tait’s supposition a — Id. 
Thus if to be the absolute temperature of the neutral point of any 
two metals (whose lines are straight), then, according to the theory 
of the thermoelectric diagram, their Peltier effect at absolute 
temperature t is proportional to tit - to ). The neutral points of 
the specimens used were accordingly found in the ordinary way 
by heating up a junction in oil, and these were the only data 
necessary in finding the calculated values with which the observed 
values were compared. 
The apparatus used was a modification of that described by the 
writer in a paper read before this Society in 1882 ; it was con- 
structed in the following manner : — 
The ends of two strips of sheet iron (say) were soldered (or 
in the case of some metals, brazed) to the lower ends of an arch of 
sheet cadmium (say). The other ends of the iron strips had 
several copper wires soldered to them, leading to a rocking commu- 
tator connected with four cells of a Thomson’s “ tray-Daniell ” 
battery. In this circuit was also included a Helmholtz tangent 
galvanometer, which served to measure the battery current. The 
differences of temperatures of the iron-cadmium junctions were 
measured by means of an iron-German-silver thermopile (of 16 or 
20 junctions), bent into the shape of an arch, whose ends were 
inserted in the trenches between the iron and cadmium, being 
insulated from them by thin asbestos paper. The whole was packed 
