The observations confirm Wilson's remark that "of the three kinds of electric current which may accompany precipitation, the convection current of lightning discharges and continuous currents due to the intense electric fields, it is quite possibly the last which contributes most to the interchange of electricity between the earth and the atmosphere." The thunderstorms studied by Schon land produced little rainfall, but the preponderance of positively charged rain in his experiments is consistent with general expe rience. If the rainfall over the area of his hypothetical storm, 20 had been at the rate appropriate to heavy thunderstorms, say I mm. per minute, and if the rain had been highly charged so as to carry four electrostatic units per cubic centimetre of water, the average current would have been of the order amperes. Other observers have found strong positive gradients as frequent as strong negative ones during storms.
If the conclusions of Schonland can be accepted as generally true, it will follow that the net flow of positive electricity from ground to atmosphere in thunderstorms is adequate to supply the electricity needed to maintain the air-earth current over the parts of the globe enjoying fine weather. According to Wilson, positive electricity and negative are separated in the cloud, the positive charge being uppermost.
Between this upper charge and the Heaviside layer there is a conduction current, facilitated by the comparatively high con ductivity of the air. The Heaviside layer may be regarded as a perfect conductor. The current between this layer and the ground in the fine-weather regions is downwards.
The current from the air to the ground is known to be nearly always positive, and Wilson's theory provides an acceptable explanation.
on the observation that, when water-drops are broken by an air blast the spray is electrified positively, the air negatively. In a cumulo-nimbus cloud there are regions in which the upward cur rents are powerful enough to prevent any rain falling through them. Simpson contends that such currents are so persistent that the water accumulates in them for several minutes. As the rain drops are continually breaking up and recombining, they acquire large positive charges. The air currents convey the negative charge to less turbulent regions, and the droplets which form in such regions become negatively charged. Thus negatively charged rain is to be expected in the greater part of the cloud.
Simpson has applied numerical tests to prove that the causes are adequate to produce the observed effects. In his typical thunderstorm, the region in which water accumulates is equivalent to a sphere 2 km. in diameter, and is about one-hundredth part of the whole cloud. It is supposed that, after a quarter-of-an hour, there is enough water to make to cm. of rain on the ground immediately below. This estimate is arrived at by consideration of the volume of air carried up by a current of 8 metres per second, and of the amount of water vapour contained in the air at the temperatures prevailing at 2 km. and 4 km. above the ground. The accumulated water is equivalent to 4.4X drops 5 millimetres in diameter. If all these drops were broken into spray, about 7 coulombs of positive electricity would remain on the spray, and an equal quantity of negative electricity would be carried away by the air currents.
To explain the great strength of the electric field necessary to start a lightning flash, Simpson supposes that the density of the electric charge in one part of the region of separation is very great. A sphere Soo metres in diameter would occupy one sixty-fourth of the whole region in which electrification was taking place. If the water drops in the sphere were broken up 256 times a charge of 28 coulombs would be developed, and if the process took 4 minutes, the strong air-currents of 8 metres per second would remove the complementary negative charge to a distance averaging about a kilometre.