238 Mr. fr. Lane on Electric Conduction in Metals. 
14. A weak current was jars: used” to avoid gener 
as far as practicable, unequal heat in the mercury cylinders, 
lest the conducting power of the two should be unequally affect- 
ed. What degree of heat n ight: have been generated, could not 
be ascertained in any of my experiments, but it was never sensi- 
ble to the touch, though sometimes it appeared to be indicated 
by thermo-electric action, ($ 18.) The last experiment was re- 
peated with two pairs of plates, though not with a very strong 
charge, and the result gave 
U? Law : 1? L'w’: +586 ; 585. 
Thus we have good evidence of the truth of our second prop- 
osition. It might seem at first sight that the whole law, includ- 
ing both the propositions, was proved by this experiment alone, 
but it will be readily seen that this is not necessarily true. ‘The 
supposition, for instance, that in a given conductor the quantity 
is as the square root of the intensity, and in different conductors 
as the square root of the section, would explain as well as any 
other the result of the above experiment. It cannot, therefore, 
positively demonstrate one of the two propositions without the aid 
of the other, though it would certainly afford a strong probable 
support to both, and on this presumption, strengthened by these 
results with the metals, I ma perhaps rest in the experiments 1 
propose to make on conducting fluids, ($ 16.) 
15. Any one in a situation to make accurate experiments, will 
find it easy to examine this law in the manner here described, 
with a degree of precision very far exceeding what I have attain- 
ed. Conductors of very fine and uniform texture appear to be 
the principal desideratum. If these can be made perfect, I think 
that with a good galvanometer and considerable length of con- 
ductor, results may be obtained approaching even the exactness of 
geometrical measurements, and that an exceedingly minute devi- 
ation from the law, may, if it exist, be detected. Even if con- 
ductors cannot be mind absolutely tnifoath; the corrections that 
ought to be applied to them, can be ascertained beforehand by 
comparing them with many other similar conductors, and taking 
a mean of the results. Should circumstances permit, I shall pur- 
sue the subject further hereafter, unless the occasion for doing s0 
should be superseded. 
16. The experiments I have described still leave room for 
doubt as to the application of the law in question to the case 
