of other air on each side of it, and it is quite possible that the flow of the southwest monsoon, at any place and time, may be affected by something that occurred long before in a very distant part of the globe. At first it was thought that the condition of the snow lying on the ground in the Himalaya Mountains would determine the movement of the monsoon and the amount of rain in the lowlands, but, as I have elsewhere stated, the supply of air to the southern Indian Ocean must ultimately come from the great westerly winds of the roaring forties, and therefore the Asiatic circulation must be affected by the condition of affairs in south Africa and the south Pacific and even the south Atlantic oceans.
On November 14, 1896, simultaneous balloon ascensions were made from St. Petersburg and Munich and from intermediate cities, in the midst of an area of high pressure that was moving slowly eastward over Europe. My study of these observations in the light of my maps of high-level isobars for the northern hemisphere gave occasion for the following long range forecast which was made early in December, and, of course, long before we received any reports from India, "As a result it is quite possible that this area may have brought to upper India light snow followed by cold dry weather about the first of December, 1896."
As this prediction was abundantly verified, as shown by the reports which we received a year later, it may be worth saying that the study of these upper isobars explains why the areas of high pressure over North America usually move at first from the northwest; subsequently their velocity diminishes while the path turns more nearly eastward; on the other hand, similar areas of high pressure and cold air in Europe are apt to come from the northeast before they turn southeastward.
It seems certain that the atmosphere is so mobile that whatever happens on one side of the globe will soon be known by its results on the opposite side. Whatever happens in the atmosphere fifteen miles above the earth will soon produce results at the earth's surface far away. Meteorology must embrace the whole atmosphere above and below, north and south, east and west.
I suppose that the most important problem of the present time is to attain a clear idea of the mechanics of the earth's atmosphere as a whole. We separate this problem into three closely related divisions. First we treat the atmosphere as we would a liquid of very rare but uniform density. Next we introduce the idea of a gas which enlarges or contracts in volume with every change of pressure or temperature.
Finally we pass from simple dry air and moisture to study the
- See Monthly Weather Review, November, 1896, p. 415.
- Published as chart VII. in that number of the Review.
- See p. 420 of that same number of the Review.