398 EXPERIMENTAL INVESTIGATIONS. 



not having a large battery, I did not go beyond ten cells, 

 which I found would well bear a jar of one square foot coated 

 surface. 



By the aid of Mr. Gassiot's more powerful apparatus, I 

 have, with him, used thirty cells of the nitric acid battery, two 

 inches by four, and five square feet of coated surface ; the 

 effects were very striking a roar of voluminous discharge of 

 O'6 of an inch long increased to i'5 inch when the flame of a 

 spirit-lamp was placed between the terminals. I have never 

 witnessed such a torrent of electrical discharges ; the noise 

 could not be borne long without great discomfort. 



With the same voltaic battery, and an additional square 

 foot of coated surface, the effect was somewhat diminished. 

 Mr. Gassiot had not more than thirty cells available at the 

 time of our experiments, so that I have not yet ascertained 

 the limit to which this increase of power can be carried. I 

 presume, however, there is a limit, for reasons which will be 

 presently apparent. 



The following precautions are essential to the success of 

 the experiment : 



ist. The wire proceeding from the outer extremity of the 

 secondary coil must be connected with the inside or insulated 

 coating of the Leyden battery, if the battery is not wholly 

 insulated. The reason of this is, that the outer extremity of 

 the coil is the better insulated portion, and also that to which 

 electricity of tension flies off; a good spark can, under ordi- 

 nary circumstances, be obtained from the outer, but scarcely 

 any from the inner terminal of the coil. 



2nd. The distance between the hammer of the contact- 

 breaker and the soft iron core should be made as great as 

 practicable, at least one-eighth of an inch ; this is an impor- 

 tant point as to the theory and experimental results of the 

 Ruhmkorff coil. Time, as is well known, is necessary for the 

 development of electro-magnetism ; and M. Matteucci, in his 

 recent valuable book on induced electricity, has shown some 

 remarkable results flowing from this fact. If the hammer be 

 too near the core, the former is raised before the latter has 

 time to be fully magnetised ; and when a Leyden condenser is 

 used ; farther time is required for this to be charged. This 



