CONVECTIVE CURRENTS IN GOOD CONDUCTORS. 153 



working of the machine could be observed ; but at the same time there was not the 

 smallest indication of electrical charge about the conductor of the machine, so com- 

 plete was the discharge. 1 conclude that the quantity of electricity passed in a given 

 time had been too small, when compared with the conducting power of the fluid, to 

 produce the desired effect. 



1608. I then charged a large Leyden battery (291.), and discharged it through 

 the wire^, interposing, however, a wet thread, two feet long, to prevent a spark in 

 the water, and to reduce what would else have been a sudden violent discharge into 

 one of more moderate character, enduring for a sensible length of time (334.). I 

 now did obtain a very brief elevation of the water over the end of the wire ; and 

 though a few minute bubbles of gas were at the same time formed there, so as to 

 prevent me from asserting that the effect was unequivocally the same as that obtained 

 by Davy in the metals, yet, according to my best judgement, it was partly, and I 

 believe principally, of that nature. 



1609. I employed a voltaic battery of 100 pair of four-inch plates for experiments 

 of a similar nature with electrolytes. Tn these cases the shell-lac was cupped, and the 

 wire h 0*2 of an inch in diameter. Sometimes I used a positive amalgamated zinc 

 wire in contalct with dilute sulphuric acid ; at others, a negative copper wire in a so- 

 lution of sulphate of copper ; but, because of the evolution of gas, the precipitation of 

 copper, &c., I was not able to obtain decided results. It is but right to mention, 

 that when I made use of mercury, endeavouring to repeat Davy's experiment, the 

 battery of 100 pair was not sufficient to produce the elevations*. 



1610. The latter experiments (1609.) may therefore be considered as failing to give 

 the hoped-for proof, but I have much confidence in the former (1605. 1608.), and in 

 the considerations (1603.) connected with them. If I have rightly viewed them, and 

 we may be allowed to relate the currents at points and surfaces in such extremely 

 different bodies as air and the metals, and admit that they are effects of the same kind, 

 differing only in degree and in proportion to the insulating or conducting power of the 

 dielectric used, what great additional argument we obtain in favour of that theory, 

 which in the phenomena of insulation and conduction also, as in these, would link 

 the same apparently dissimilar substances together (1336. 1561.) ; and how com- 

 pletely the general view, which refers all the phenomena to the direct action of the 

 molecules of matter, seems to embrace the various isolated phenomena as they suc- 

 cessively come under consideration ! 



1611. The connection of this convective or carrying effect, which depends upon a 

 certain degree of insulation, with conduction; i. e. the occurrence of both in so many 



* In the experiments at the Royal Institution, Sir H. Davt used, I think, 500 or 600 pairs of plates. Those 

 at the London Institution were made with the apparatus of Mr. Pepys, (consisting of an enormous single pair of 

 plates), described in the Philosophical Transactions for 1823, p. 187. 

 MDCCCXXXVIII. X 



