1887.] 
Mr W. Peddie on Transition Resistance. 
223 
xnents, and Plate VII.). This was most probably due to alteration 
of resistance or to alteration of capacity of the electrodes regarded as 
condensers. To settle the point, I used a Helmholtz galvanometer 
capable of indicating a current of the one ten-millionth part of an 
ampere. It was necessary to use an instrument of such sensibility ; 
for the current, although it started with a comparatively large value, 
very rapidly fell to an exceedingly small fraction of its original 
strength because of polarization. Evidently the initial value of the 
current-strength is independent of polarization; and this constitutes 
the great advantage of the present method. In order to bring the 
first deflection of the index on the scale, it was necessary to shunt 
the galvanometer. Of course, the deflection could not be read with 
accuracy for some time after joining the battery in the circuit 
because of oscillations of the galvanometer needle; but, by proper 
shunting, the interval could be made as small as 10 seconds when 
required. Readings of the deflection (to which the current-strength 
is proportional) were taken at short intervals, and the results exhibited 
graphically. The curve drawn through the points so obtained could 
easily be continued backwards to cut the axis parallel to which 
current was measured, thus giving the initial value. From this the 
amount of resistance in the circuit could be obtained. I did not, 
however, actually determine the resistance in this way. I obtained 
the curve for the case when the electrodes had been left unheated for 
some time ; then I washed the plates and heated them to redness, and 
repeated the experiment under such conditions as to obtain a curve 
coinciding practically with the former. The results showed that 
in all cases a considerable additional resistance had to be placed 
in the circuit to produce coincidence. This shows conclusively that 
a transition resistance exists; and also gives an estimate of its 
amount, for the transition resistance must be equal to the resist- 
ance added in the second case. It is necessary to remark that the 
resistance of the electrolyte was the same in both cases; and, 
further, its total amount was very small in comparison with the 
resistance of the rest of the circuit. Since only one Daniell cell 
was used, the liquid was not decomposed, and the current-strength 
after two or three seconds was only of the order of thousandths or 
ten-thousandths of an ampere. Hence, although the temperature 
coefficient of the resistance of the liquid is large, it was quite im- 
VOL. XIV. 19/10/87 
p 
