It will be remembered that in 1808 Sir Humphry Davy constructed his battery of 2,000 cells, and thus succeeded in exalting the tiny spark obtained in closing the circuit into the luminous sheaf of the voltaic arc. He also observed that the spark passed even when the poles were separated by a distance varying from 1/40 to 1/30 of an inch. This appears to have been subsequently forgotten, as we find later physicists questioning the possibility of the spark leaping over any interpolar distance. Mr. J. P. Gassiot, of Clapham, demonstrated the inaccuracy of this opinion by constructing a battery of 3,000 Leclanché cells, which gave a spark of 0.025 inch; a similar number of "de la Rue" cells gives an 0.0564 inch spark. This considerable increase in potential is chiefly due to better insulation. The great energy of this battery was illustrated by a variety of experiments. Thus, a large condenser, specially constructed by Messrs. Varley, and having a capacity equal to that of 6,485 large Leyden jars, was almost immediately charged by the current from 10,000 cells. Wires of various kinds, and from 9 inches to 29 inches in length, were instantly volatilized by the passage of the electricity thus stored up. The current induced in the secondary wire of a coil by the discharge of the condenser through the primary, was also sufficiently intense to deflagrate wires of considerable length and thickness. It was with such power at his command that Dr. De la Rue proceeded to investigate several important electrical laws. He has found, for example, that the positive discharge is more intermittent than the negative, that the arc is always preceded by a streamer-like discharge, that its temperature is about 16,000 deg., and its length at the ordinary pressure of the atmosphere, when taken between two points, varies as the square of the number of cells. Thus, with a battery of 1,000 cells, the arc was 0.0051 inch, with 11,000 cells it increased to 0.62 inch. The same law was found to hold when the discharge took place between a point and a disk; it failed entirely, however, when the terminals were two disks. It was also shown that the voltaic arc is not a phenomenon of conduction, but is essentially a disruptive discharge, the intervals between the passage of two successive static sparks being the time required for the battery to collect sufficient power to leap over the interposed resistance. This was further confirmed by the introduction of a condenser, when the intervals were perceptibly larger.