of Edinburgh, Session 1883-84. 
431 
result for gold, when caoutchouc was 
the rubber, is doubtless due to an error 
in entering the result. But with sealing 
wax as the rubber we have varying 
signs. When the electricity of the 
rnetal was negative, there was no red 
scratch (a very little in the case of 
silver) left on the metal ; when the 
electricity of the metal was positive, 
there was a red scratch, excepting 
in the case of nickel. I conclude 
that the electricity of the metal is 
normally negative, but that when the 
wax is abraded it is changed to 
positive. 
By taking the average of all the 
averages, we get the values in the 
bottom row of Table I. ; they are 
exhibited graphically on the accompany- 
ing scale. The order of the metals here 
exhibited is not an arbitrary order ; it 
agrees pretty well with the order in 
which the chemists arrange them, with 
respect to their affinity for oxygen. 
This was observed by Professor Tait at 
an early stage of the experiments. The 
metals of the iron group are found close 
together, and their order among them- 
selves is also significant. The metals of 
the copper group are found together, 
but of these lead is at a considerable 
distance from the others; the entries 
for that metal, however, vary consider- 
ably. Platinum and gold are next one 
another ; the value for the former is as 
yet only approximate. The chemist 
classes tin along with antimony and 
bismuth; the two latter are found to- 
{ 181 All - 
l=S 
O 
fH 
o 
g 
.S 
136 Pt-- 
126 Sn-- 
102 
100 
p 
p 
O 
O) 
jp 
Ph 
O 
Q 
Ag-- 
Cu-- 
I 62 Pb - 
f 59 Ni- 
56 Fe- 
o 50 Al- 
I 45 Zn - 
L 42 Mg - 
r 38 Sb- 
S ^ 1 
.g p ^ 32 
to I 
<1 L 22 Bi- 
- Brass. 
-G.S. 
0 
ELECTEO-FRICTIONAL 
SCALE. 
