of Edinburgh, Session 1885 - 86 . 
691 
the next column gives the temperature of the alloy at that instant ; 
and the last column gives the deflection reading on the scale of 
the galvanometer; which, as the angle was always very small, is 
approximately enough proportional to the current in the galvano- 
meter circuit, and therefore to the resistance of the alloy. 
At first the alloy was always somewhat above its melting point, 
and the observations were continued for a short time after the alloy 
appeared to be perfectly solid. 
Fig. 1 is a diagram to illustrate the arrangement of the circuit 
in the experiments with the trough. B represents the Leclanchd 
cell, Bj the resistance inserted in the battery circuit, K the key in 
the same, T the trough, G the galvanometer inserted in the shunt 
circuit taken from two points in the trough, and R 2 the resistance 
inserted in this circuit. 
Fig. 2 represents the relation between the resistance and temper- 
ature of tin as it solidifies. The resistances are measured along 
OB, and the temperatures along OT ; these become lower as they 
recede from the point 0. 
Fig. 3 shows the same for fusible metal, except that in this case- 
time, not temperature, is measured along OT. 
Fig. 4 gives the curve for the alloy SnPb 3 , and fig. 5 for Sn 3 Pb,. 
resistance being measured along OR, and temperature along OT. 
The points on these curves at which observations were taken are 
marked by small crosses. As these observations were made at 
regular intervals of 30 seconds, the projection of these points on 
the temperature axis in figs. 2, 4, and 5, indicate the rate at 
which the temperature of the cooling alloy changed. The more 
closely these projections are crowded together the more slowly does 
the temperature change ; and it will be observed that it is at the 
very points where the falls in resistance occur that this takes 
place. This gives additional corroboration to the accuracy of the 
observations. 
These experiments have been performed in the Physical Labora- 
tory of the Edinburgh University, through the kindness of Professor 
Tait, to whom, and to Professor Chrystal, at whose suggestion the 
subject was originally untertaken, the authors are deeply indebted 
for valuable advice. 
