Magnetism on a permanent Electric Current. 313 



thicker plates. Arguing from these facts alone, it would appear 

 that the transverse effect in thin leaf gold is relatively much 

 smaller than the effect in strips of sensible thickness ; but this 

 is hardly a safe conclusion. Three objections to the above 

 method of determining the thickness by means of the resist- 

 ance are evident : — 1st. Gold leaf so thin as to be transparent 

 is by no means continuous, but is perforated by a multitude 

 of small holes ; so that the electricity is, as it were, obliged to 

 wind or zigzag its way through the strip, thereby having a 

 longer path and meeting a greater resistance than if it could 

 pursue a direct course. 2nd. Gold leaf is an alloy about twenty- 

 three carats fine; and the resistance of such alloys is often much 

 larger than that of either of the pure metals. 3rd. It is difficult 

 to secure good contact at the ends of the strip. In the plate 

 under consideration the contact was probably very bad, and may 

 have been many per cent, of the whole resistance of the plate 

 as measured. 



All these sources of error affect the result in the same way. 

 To compensate, it would be necessary to diminish the resistance 

 as measured, and then, in deducing the thickness, use a specific 

 resistance higher than that belonging to gold. In using thin 

 silver plates, I have in a rough way made a correction for the 

 error due to contact-resistance ; but the gold leaf is in several 

 respects so unsuitable for any thing like accurate work, that it 

 does not seem worth while to spend any more time upon it at 

 present. In fact I would in the present article dismiss the sub- 

 ject of gold-leaf strips with a very few words, were it not the 

 case that, in a matter of this kind, it seems proper that the 

 public should be informed of any facts that have the slightest 

 suspicious appearance. 



The gold plates which are now to be described were of com- 

 paratively thick metal, such as is used by dentists. The metal 

 in this shape is said to be very pure; and the thickness was so 

 considerable as to make it possible to weigh the strips with 

 sufficient accuracy. The determination of the thickness in 

 this way involves the assumption that the specific gravity is 

 that given by the tables ; but the error from this source must 

 he very much smaller than the sum of those introduced by 

 employing the resistance method. 



Gold used by dentists is classed under various heads, accord- 

 ing to the manner of tempering. The kinds I have used are, 

 I think, " soft " or " semicohesive," and "hard" or "cohe- 

 sive." I noted the varieties, thinking that specific peculia- 

 rities might possibly appear in their behaviour. The number 

 attached to each plate is the commercial number of the spe- 

 cimen, and indicates approximately the number of grains in a 



