364 Proceedings of the Royal Society 
material alteration of the surface produced by this change of tempera- 
ture — oxidation, for example. That this is the most probable view 
is borne out by known facts, and by certain results which I myself 
obtained relating to the subject of contact electricity. Hankel long 
ago showed that, with the great majority of metals, there was a 
negative-growth in time — for example, iron recently polished or filed 
was electrically positive to iron which had been left for some time 
in the air. This was probably due to oxidation ; and it became a 
question of interest to compare this “ time-variation ” with the 
“ temperature-variation ” discussed above, A series of experiments 
were made very similar to those described above and differing only 
in this, that the lower surface was permitted to vary through time, 
without any alteration in temperature. The curves obtained by 
plotting the electrometer deflections against the time were very 
similar to the ordinary curves of cooling — somewhat logarithmic in 
appearance; and markedly dissimilar to the curves showing the 
“ temperature-variation ” of the same metals. Further, that metal 
varied in time fastest, which was the most positive : aluminium, 
zinc, iron, and copper being their order, taking first that one whose 
curve of time- variation was steepest. This result accords well with 
the theory given by J. Brown, Esq., of Belfast, in the “Philosophi- 
cal Magazine” (1878-79), to the effect that the position of the 
metals in Volta’s contact series depends mainly, if not entirely, upon 
their chemical affinity for air, the most positive being that which has 
the greatest affinity. That the most positive (aluminium, namely) 
should also be that which varies fastest in time is extremely pro- 
bable ; and it is also a plausible enough hypothesis that the most 
positive should also have the most rapid negative-growth with tem- 
perature. Now, as far as these experiments go, this is really the 
case. Zinc, iron, copper, are in the order of magnitude for tempera- 
ture-variation, and also for time- variation ; and the same order holds 
in Volta’s contact series beginning with the most positive. It would 
thus appear that for any one of the metals zinc, iron, copper, and 
(we may add) tin, there corresponds a definite surface condition to 
every temperature — a condition which is permanent even after the 
surface has cooled, which has the effect of making the surface electri- 
cally negative to its original self, and which no amount of polishing 
can alter as long as the temperature is kept constant. Hence we 
