the warmer air above it. A second “nose” appears about
9.00, when clearing gusts marked the end of the storm.
Forecasting Thunder-storms.—From the nature of the case, the general forecasts made by the Weather Bureau cannot designate the loci of possible thunder-storms, because the general forecasts are made too far ahead, and also because such storms are local.
The meteorologist in charge of the local station is able to forecast more definitely; and, where stations not far apart are fortunately situated, the formation of thunder-storms may be indicated with a fair probability of verification. With warm, moist air on the south side of a low, thunder-storms may be expected; and if one has formed, its path may be predicted with reasonable exactness. In the hands of a trained observer a barograph is a most useful aid. With the aid of the daily weather map, the local conditions of temperature and humidity, and the barogram, at least two hours’ notice may be given.
The layman also may forewarn himself with a reasonable degree, if not of certainty, at the least, of probability. An aneroid barometer, if watched closely, may be serviceable; unless intelligently used it is of doubtful service to any but a trained observer. Nevertheless, there are indications that should warn even a casual observer who bears in mind that the thunder-storms disastrous to crops occur mainly in June, July, and August, and also that almost always they occur between mid-afternoon and sunset.
Warm and moist air is necessary to the formation of a thunder-storm; moderately quiet air is also necessary. A thunder-storm is not likely to form where a stiff wind is blowing. Cumulus clouds may be regarded with suspicion; indeed the
