than it does the lighter moist air. For both reasons, therefore, the lighter is raised up by the denser air, and overflows toward the pole. Now a body on the earth's surface and in motion relative to it, while at the same time rotating with it, will appear to an observer on the earth to be deflected toward the right hand as it moves forward in the Northern Hemisphere, but to the left hand in the Southern Hemisphere. By virtue of this deflection the winds that are blowing toward a region of low pressure acquire, each for itself, a deflection toward the right or the left respectively, so that instead of meeting at the centre they whirl around it in an inflowing spiral curve. By reason of this circulation an outward centrifugal pressure is produced, and the barometric pressure in the central region is much smaller than it would be if the winds flowed directly to the centre, without any spiral circulation. In extensive storm areas this general tendency of the lower winds to circulate around a centre may exist over a region a thousand miles or even more in diameter.
The mechanical and thermal problems connected with the generation and maintenance of storms are set forth without technical mathematics by Prof. William Ferrel in his Popular Treatise on the Winds (1st ed., New York, 1889). They are discussed most elaborately in a technical manner by Prof. F. H. Bigelow in his Report on International Cloud Work (published as vol. ii. of the Report of the Chief of the Weather Bureau for 1898). A full presentation of the work that has been done upon this subject by all students during the past century will be found in the Lehrbuch der Meteorologie, by Prof. Dr. Julius Hann (Leipzig, 1901), of which an English translation is promised in the near future.
.jpg)
Fig. 2. TRACKS OF CENTRES OF LOW AREAS, JANUARY, 1901.
As far as possible the paths pursued by storm centres during past years are platted upon charts, then they are classified according to their general characteristics and studied with reference to their relations to the topography of the continents and the general distribution of the barometric pressure, temperature, and moisture, and especially to the so-called general circulation of the atmosphere. The frequency of storms and the general types of storm paths are, of course, best known for the continental portions of the Northern Hemisphere, but there is also sufficient data for the oceanic regions to justify an attempt at presenting the accompanying sketch chart of the Northern Hemisphere showing the general character of the storm paths and the locations where they most frequently occur. (See Fig. 1.) This chart shows by its numerous lines and arrow heads the occurrence and the direction of motion of storm centres along the lines of greatest frequency, so far as now known. The Arctic region. Northern Africa, and Central Asia must be considered as blanks; we have no daily maps for these regions, and only know that general cyclonic storms are infrequent in Africa and Asia. The chart shows that storm centres move slowly westward when within the tropics and also slowly toward the pole, but move more rapidly eastward between the tropics and the Arctic Circle, as well as more rapidly northward. The zone of greatest storm frequency lies between latitudes 45° and 55°. In general the path of any observed storm may be predicted on the basis of a simple study of this chart of storm tracks. But individual temporary departures from average conditions are so great that in actual weather forecasts it is necessary to allow these general maps of types and averages to have only a very slight influence upon the work. It is, in fact, always necessary to consider what the special individual storm has been doing during the preceding few days, and to what extent it is normal, and to what extent it is abnormal, as to the direction and velocity of its motion, and
