It must not be inferred from what has been said above that the mystery of the lightning flash is now fully resolved. This is far from being the case. It is not at all clear how an electrical discharge can break down the resistance of the air along a path a mile or more in length, as commonly happens in the thunderstorm. It was formerly stated, on good authority, that the difference of potential required to produce such a flash would amount to upward of 5,000,000,000 volts. Certain facts have lately been adduced to show that such great differences of potential need not be assumed. Moving-camera photographs of the sparks produced by electric machines show that such sparks begin with small brush discharges which gradually ionize the air and thus build up a conductive path for the complete discharge. Something of this kind may occur in the atmosphere. Streaks of air already strongly ionized and more or less continuous sheets of rain would also help to provide conductors for a discharge. If lightning does build up its path somewhat gradually, the process might, in certain cases, be so slow as to account for the deliberate movement of rocket lightning, and also, perhaps, furnish a clue to the hitherto unsolved mystery of ball lightning. Humphreys has tentatively suggested that all genuine cases of ball lightning are “stalled thunderbolts”; i. e., lightning discharges that have come to a halt, or nearly so, in their progress through the air. As to the visibility of lightning Humphreys says, in his “Physics of the Air”: “Just how a lightning discharge renders the atmosphere through which it passes luminous is not definitely known. It must and does make the air path very hot, but no one has yet succeeded, by any amount of ordinary heating, in rendering either oxygen or nitrogen luminous. Hence it seems well-nigh certain that the light of lightning flashes owes its origin to something other than high temperature, probably to internal atomic disturbances induced by the swiftly moving electrons of the discharge, and to ionic recombination.” A few attempts have been made to measure the strength of current in a lightning discharge. Many substances become magnetized when an electric discharge occurs in their vicinity, and it has been pointed out by F. Pockels that when basalt rock is magnetized in this way the amount of magnetism is an indication of the greatest strength of current to which it has been exposed. Pockels examined specimens of basalt from the top of Mount Cimone, in the Apennines, where lightning strokes are common, and found many of them more or less magnetized. He also exposed blocks of basalt close to a branch of a lightning rod in the same region. He thus obtained values for the strength of current in lightning discharges ranging from 11,000 to 20,000 amperes. Humphreys, from the crushing effect of a lightning stroke upon a hollow lightning rod, has computed the strength of current in the case examined to be about 90,000 amperes.