RELATION OF GRAVITY TO ELECTRICITY. 
3 
diminution or cessation of motion either way, or whether the rising and the falling 
were made to occur isochronously with the times of vibration of the galvanometer 
needle. So, though no effect of gravity appeared in the helix itself, still no source 
of error appeared to arise in this mode of using it. 
2706. A solid cylinder of copper, three-fourths of an inch in diameter and 7 inches 
in length, was now introduced into the helix and carefully fastened in it, being bound 
round with a cloth so as not to move, and this compound arrangement was allowed 
to fall as before (2705.). It gave very minute but remarkably regular indications of 
a current at the galvanometer ; and the probability of these being related to gravity 
appeared the greater, when it was found that on raising the helix or core, similar 
indications of contrary currents appeared. It was some time before I was able to 
refer these currents to their true cause, but at last I traced them to the action of a 
part of the connecting wires proceeding from the helix to the galvanometer. The 
two wires had been regularly twisted together, but the effect of many falls had opened 
a pai t near the middle distance into a sort of loop, so that the wires, instead of being 
tightly twisted together like the strands of a rope, were separate for 3 feet, as if the 
strands were open. In falling, this loop opened out more or less, but always in the 
same manner ; and the consequence was that the part of it representing the transverse 
opening, which was furthest from the galvanometer, travelled over a larger space 
than the corresponding part nearest the galvanometer. Now had they travelled 
through equal spaces, the effect of the magnetic lines of force of the earth upon them 
would have been equal, and no effect at the galvanometer would have been produced; 
as it was, currents in opposite directions, but of unequal amounts of force, tended to 
be produced, and a current equal to the difference actually appeared. Such a case 
is described in my earliest researches on terrestrial rnagno-electro induction (171 •)• 
It is evident that the current should appear in the reverse direction, as the helix and 
wires are raised in the air, and thus arose the reverse effect described above. There 
fore no positive or favourable evidence was supplied in favour of the original assump- 
tion by this use of a copper core in the helix. 
2707. The copper was selected as a heavy body and an excellent conductor of 
electricity. On its dismissal, a bismuth cylinder of equal size was employed to 
replace it as a substance eminently diamagnetic, and a bad conductor amongst 
metals. Uncertain evidence arose ; but by close attention, first to one point and then 
to another, all the indications disappeared, and then the rising or falling of the bis- 
muth produced no effect on the galvanometer. 
2708. An iron cylinder was also employed as a magnetic metal, but when made 
perfectly secure, so as to prevent any motion relative to the helix, it was equally in- 
different with the copper and bismuth (2706. 2707.). 
2709. Cylinders of glass and shell-lac were employed as non-conducting substances, ‘ 
but without effect. 
2710. In other experiments the helix was Jixed, and the different substances in the 
B 2 
