ON ELECTRIC CONDUCTIVITY. 417 
be expected. It is to this elastic part of the phenomenon that the value of o 
given below is related. 
The following is an instance of a set of experiments :— 
Weights: on each side 8 lbs. ; distance of mirror from scale, 258 cms. 
Weights off, Weights on. 
D, cms. Reading on Scale. D, cms. | Reading on Scale. 
| 
52-2 / 50:0 46°3 | 62:4 
50:9 | 50:3 46-0 62:5 
50-4 | 50°3 45'9 | 62°5 
504 50°3 45°8 | 62°4 
50°3 | 50°3 45°9 62:5 
50°4 50:3 45°9 | 62°5 
50-4 50:3 45:9 62:5 
For the distances D, and D, the numbers 50°4 and 45:9 being adopted, we 
calculate 6=0-02409 and o=0-155. 
The weights were varied in different experiments from 4 to 28 lbs., and the 
proportionality of change of conductivity and strain, assumed in the above 
formula, was between these limits sufficiently proved. 
The main average value obtained for the constant o and the used specimen 
of brass is 
o—0158. 
Now, the way followed usually in determining coefficients of elasticity 
might be used also in the present case, and the other constant alluded to above 
determined by observing the change of resistance of a stretched wire. 
I find for that resistance the approximate formula, 
r=R(1+sr) 
R being the resistance of the unstrained wire, 7 that of the strained, A the 
longitudinal strain, and s expressed as follows :— 
+ 
s=1+4+2p+5 ol +p) +p(l—2n). 
Here p denotes the “ Poisson’s ratio,” o the constant forming the subject of the 
experiments described above, and p is to be defined as follows :—Take a 
VOL. XXX. PART. I. oT 
