152 
Bunsen’s pairs connected for quantity. The same day, in Mr 
Kemps laboratory, I laid a brass “ Trevelyan” bar or rocker on the 
edge of the brass plate, forming the outer rail of Mr Gore’s ma- 
chine, and connecting the rail with one pole of the same battery, and 
the bar (by means of a globule of mercury inserted in a cavity in its 
upper surface) with the other, energetic vibrations commenced quite 
resembling those occasioned by heat in the ordinary form of the ex- 
periment on a leaden support. 
“ I have since found, among other results, — 1. That the vibration 
goes on in whichever direction the electric current passes. [At first 
I thought that there was a superior effect when the current passed 
from a good to an imperfect conductor, but this has not been con- 
firmed, as far at least as I have gone]. 2. The vibrations take 
place both between metals of the same kind and heterogeneous me- 
tals. 3. When heat is applied to a brass bar vibrating on cold 
lead, and then electricity is applied as before, the effects are super- 
added to one another whichever way the current passes, the vibra- 
tions becoming more energetic, and if there be a musical note it be- 
comes graver [owing, it is assumed, to the increased arc of vibra- 
tion], 4. When a bar of brass was placed so as to vibrate on two 
parallel upright plates, also of brass, respectively connected with the 
poles of a battery, the vibrations continued, when the whole was im- 
mersed partly or wholly in water, and even when flooded by a power- 
ful continued stream of cold water from a five-eighth inch pipe under 
considerable pressure. From this experiment I conclude that the 
effect of the heat developed by the electrical current in the thin up- 
right plates may be fairly considered to be reduced so low as to be 
incapable of producing a sensible result (if such were ever the case). 
Indeed, allowing for the resistance and friction of the water tending 
to diminish the vibration, there is no ground for thinking that the 
action was less energetic in the one case than in the other. It is 
consequently reasonable to conclude that the effect in question is due 
to the repulsive action of the electricity in passing from one con- 
ducting body to another, and not to its effect in producing expansion. 
Now this is precisely the effect which I attributed to heat in the 
paper of 1833 already referred to. I therefore consider it a strong 
confirmation of the opinion I then expressed, from which I have 
never swerved, although it has not in general been received with 
much favour. The importance which I attach to this new con fir- 
