| Month. | United States. | Atlantic Ocean Middle Latitudes. | Europe. |
|---|---|---|---|
| January | 33.8 | 17.4 | 17.4 |
| February | 34.2 | 19.5 | 18.0 |
| March | 31.5 | 19.7 | 17.5 |
| April | 27.5 | 19.4 | 16.2 |
| May | 25.5 | 16.6 | 14.7 |
| June | 24.4 | 17.5 | 15.8 |
| July | 24.6 | 15.8 | 14.2 |
| August | 22.6 | 16.3 | 14.0 |
| September | 24.7 | 17.2 | 17.3 |
| October | 27.6 | 18.7 | 19.0 |
| November | 29.9 | 20.0 | 18.6 |
| December | 33.4 | 18.3 | 17.9 |
| Year | 28.4 | 18.0 | 16.7 |
To find the actual speed of the wind at a place, of course, the linear velocities of whirl and of translation must be combined; or, vice versa, if one of these be known it can be graphically subtracted from the observed wind velocity to find the other. This combination of two wind components to find their resultant, or, vice versa, can easily be done by laying off on paper, arrows of suitable length and direction to represent the two known velocities, placing the head of one arrow to the tail of the other, then completing the triangle, and taking its third side to represent the required wind velocity, in magnitude and direction. Obviously if the cyclone moves eastward, whirling oppositely to the hands of a watch, the swiftest wind is on its right side, which consequently is known as the dangerous side. In the northern hemisphere, therefore, the rule for dodging a great whirlwind is to run north, if that be practicable.
Stationary cyclones occur under favorable conditions. At least that name has been applied to columns of hot air streaming up from a fixed base, more or less circular. Every island in the ocean generates such a vortex on a clear, hot summer day, since its temperature far exceeds that of the surrounding water. All day long this uprush continues whatever be the humidity. And if the soil slopes upward steeply, the vortex is so much the stronger, particularly if the island be in a calm region. Above such a tract the gulls and vultures, and possibly even man, might soar all day without motive power. This condition and its interesting possibility deserve investigation.
Cyclones may occur at any season, but in general they are most abundant when the greatest temperature disturbances occur. The relative frequency of tropical cyclones for various localities and for the twelve months of the year is seen in the following table[69]:
The Yearly Periods of Cyclone Frequency in Several Seas
| Arabian Sea. | Bay of Bengal. | S. Indian Ocean. | Java Sea. | China Sea. | Havana. | |
|---|---|---|---|---|---|---|
| No. of years | 234 | 139 | 40 | ... | 85 | 363 |
| No. of cyclones | 70 | 115 | 53 | 12 | 214 | 355 |
| Authority. | Chambers. | Blanford. | Piddington Thom and Reid. | Piddington and Thom. | Schuck. | Poey. |
| Jan. | 6 | 2 | 17 | 25 | 2 | 1 |
| Feb. | 4 | 0 | 25 | 42 | 0 | 2 |
| Mar. | 3 | 2 | 19 | 8 | 2 | 3 |
| April | 13 | 8 | 15 | 8 | 2 | 3 |
| May | 18 | 16 | 7 | 0 | 5 | 1 |
| June | 29 | 9 | 0 | 0 | 5 | 3 |
| July | 3 | 3 | 0 | 0 | 10 | 12 |
| Aug. | 3 | 4 | 0 | 0 | 19 | 27 |
| Sept. | 4 | 5 | 2 | 0 | 27 | 23 |
| Oct. | 6 | 27 | 2 | 0 | 16 | 17 |
| Nov. | 14 | 16 | 7 | 0 | 8 | 5 |
| Dec. | 3 | 8 | 6 | 17 | 3 | 2 |
The tornado is a slender cyclone or hurricane. It is usually but a few yards or rods in diameter, and seldom exceeds one mile across its active column, whereas a cyclone may cover an area of any size from fifty to one or two thousand miles in diameter. Moreover, the cyclone requires for its inception an extensive pressure gradient marked by closed isobars, and once generated may last several days. A tornado per contra may spring into action where the lateral pressure is uniform, spend its force in a few moments, and leave a uniform barometric field in its wake. In shape the tornado is usually of greater height than width. The cyclone is far-flung laterally, but in height may not exceed the narrow tornado, since both must terminate beneath the isothermal layer, and commonly do not extend so high. Both vortices are caused by the ascensional force of hot air. In both the air spirals in and upward at the bottom, out and upward at the top, constantly cooling by expansion, and finally descends on the outside to complete the closed circulation. In general the tornado is the more violent and destructive, though limited to a brief and narrow path. More aptly, perhaps, the tornado may be called a slender hurricane of brief duration; both of them being small cyclones, or aërial vortices, of minor size and concentrated intensity. The relation of the tornado and cyclone has been defined as follows, by Professor Moore:
“The cyclone is a horizontally revolving disk of air of probably 1,000 miles in diameter, while the tornado is a revolving mass of air of only about 1,000 yards in diameter, and is simply an incident of the cyclone, nearly always occurring in its southeast quadrant. The cyclone may cause moderate or high winds through a vast expanse of territory, while the tornado, with a vortical motion almost unmeasurable, always leaves a trail of destruction in an area infinitesimal in comparison with the area covered by the cyclone.”
Two initial conditions seem essential to the genesis of a substantial tornado. In the first place, the atmosphere of its immediate locality must have appreciable gyration. Of course, in all extra equatorial regions the air has some incipient whirl due to the earth’s rotation, and this whirl is magnified as the fluid is sucked into the vortex. But the magnification may be slight owing to the brief lateral displacement of the air feeding the tornado. If, however, the fluid be drawn from a considerable distance, and have from local conditions some additional whirl superadded to that due to the earth’s rotation, the gyratory flow in the medium near the vortical axis may be very swift. On the other hand, the additional whirl, due to local conditions, may tend to neutralize that due to the earth’s component, thereby leaving a very feeble gyration, if any. But in general the rotation of tornadoes is observed to be in the direction of the earth’s component; to the left north of the equator, to the right south of it. This observation is doubtless the more striking because when the accidental local spin conspires with the permanent terrestrial one, the resultant whirl is intensified, while in the opposite case it is so enfeebled as to attract scant, if any attention.
In the second place, the genesis of a tornado requires unstable equilibrium in the local atmosphere. This instability, as in cyclones, may arise from abnormal temperature gradation. Thus, if along any vertical the temperature falls more than six degrees Centigrade for one thousand meters ascent, a mass of air started upward will continue to rise, since it cools less rapidly than the environing medium. In this way there will ensue a continuous uprush of air so long as the unstable state endures; and the action may be very vigorous if a large stratum of air is greatly heated before it disrupts into the cold upper layers. In general, the loftier the tornado the more violent it is, just as the taller flue generates the stronger draft with the same temperature gradient.