0?i the Luminous Phenomena of the Voltaic Battery. 539 



the two points are withdrawn from one another, or the point 

 from a plate. The repulsive force being less, I saw that as 

 soon as the contact took place the points fused ; that the liquid 

 matter,— the melted metal, followed the receding movement of 

 the point, and remained adhering to the point and to the plate, 

 whilstitwas pushed from one of the poles to the other, — a move- 

 ment which was perceivable enough by means of the micro- 

 scope; at the same time the metal burnt, producing the well- 

 known light of the galvanic arc. 



" This phenomenon is not that of the transfer of the matter 

 between two poles placed at a distance from one another. 

 You may, it is true, by suddenly withdrawing the points, break 

 the thread of melted metal by which the current is conducted 

 in this experiment, and the transfer of the matter will then be 

 made through the air ; but it seems tome that there will always 

 be some difficulty in ascertaining which of the two phaenomena 

 takes place. It cannot always be decided whether the transport 

 of the fused matter, forming an uninterrupted thread, takes 

 place, or if it is effected through theair without the points being 

 connected by any \continuous liquid?'] matter whatever. 



" The phaenomenon which is observed after a previous con- 

 tact seems to me to be of a complex nature, and to lead ne- 

 cessarily to uncertain results. 



" Fortunately I could resort to a contrivance which enabled 

 me to establish the current with transfer of the particles, when 

 the electrodes were placed at a sufficient distance from one 

 another, and without there being the least contact. This con- 

 trivance consists in passing an electric discharge from point to 

 point, or from a point to a plate arranged previously at the 

 distance at which the action of the current permits the transfer. 

 As soon as the electric spark passes, either from the positive 

 pole to the negative pole, or in the opposite direction, the cur- 

 rent is immediately established, and the transfer of the particles 

 commences*. 



" After having thus removed a first cause of error, which 

 the previous contact seemed to me liable to introduce into the 

 results, 1 observed that there still remained another, namely 

 the combustion of the metallic particles, the transfer of which 

 does not take place without a great development of heat. 



" It was necessary for me to render this combustion impos- 

 sible by making my experiments in vacuo ; an apparatus con- 

 structed for this purpose by one of our most skilful engineers, 

 M. Logeman, and which 1 shall describe in detail when my 

 experiments are completed, fully answered my wishes. 



* This had been previously done by Sir J. Herschel. See a paper by 

 Mr. Daniell in the Philosophical Transactions. — Ed. Phil. Mag. 



2 O 2 



