1887.] Mr A. Campbell on Measurement of Peltier Effect. 395 
and convenience, the high temperature thermometer was always read 
with a telescope. 
At first it was thought desirable, instead of employing the Helm- 
holtz galvanometer, to measure the battery current in all cases by 
the Peltier effect in a standard set of iron-German-silver junctions, 
which would, it was hoped, integrate the current during the period 
of time for which it was run. Although this arrangement was sub- 
sequently discarded as a current-measurer, some interesting results 
were obtained from it. It consisted of 23 squares of sheet iron and 
German silver (each 3 cms. square), soldered, three by three, into 
seven strips, which were then soldered to one another in zigzag form. 
Along the middle junctions were placed the ends (100 in all) of 7 
iron German silver thermopiles, insulated from the junctions by 
thin paper. The whole was wrapped tightly in wadding, and placed 
in a tin box surrounded by cold water. A thermometer was inserted 
with its bulb touching the metals. The current was sent through 
the iron-German-silver zigzag, and the Peltier effects measured by 
the thermopiles. 
§ 2. Time Curves of Peltier Effect. 
In order to investigate the rate at which the growing Peltier 
effect temperature-difference showed itself by the thermopile de- 
flection, the following simple chronographic method was adopted. 
In this a number of observers were made use of. Observer A held 
a watch to his ear, and counted half seconds aloud in exact time 
with its beats, also making and breaking the battery circuit at 
exactly the proper times. Observer B, watching the galvanometer 
scale, said sharply the syllables, “ Tic, tac, to, tee, tic, tac, to, tee,” 
and so on, as the light-spot passed over certain prearranged divisions 
of the scale. Eight other observers noted down the numbers after 
which they heard the tic, tac, to, or tee, one of these words being 
allotted to each pair of observers. This method p roved wonderfully 
accurate. The pairs of observers very seldom disagreed by half a 
second, although the numbers had to be noted very rapidly. The 
results of some of the measurements are given in Tables VI., VII., and 
VIII., which are for three different current-strengths. In these the 
current was put on for 100 half-seconds, broken for 100, put on in 
