gives a still more complete illustration of the wind-circulation in hurricanes, with a brief discussion of the application of the 8-point rule. Especial attention is called to the statement made thereon (referring, of course, to hurricanes in the North Atlantic, but no doubt true for the entire Northern Hemisphere) that

"although the 8-point rule is nearly true when the wind is anywhere from north to south by way of west (that is, generally speaking, in the navigable semicircle), it is liable to be a very poor guide when the wind is from any point in the first or second quadrant."

Also to the following, which is applicable to the Southern Hemisphere by the substitution of "to the left" for "to the right:"

"Perhaps the best general rule is that the center bears about eight points to the right of the direction from which the low clouds come, or, what is practically the same thing, eight points to the right of the wind at the moment of a sudden shift in a heavy squall; after such a shift the wind will remain steady in direction for a time, but the center is meanwhile moving along and the angle of bearing changes until the next shift, when it goes again to eight points, and so on."

Such diagrams, carefully prepared from complete and reliable data, are of far greater practical value to navigators than volumes of explanation: they appeal to the eye and will live in memory long after ideas conveyed by printed words have been forgotten.

HURRICANES IN THE NORTH ATLANTIC.—TYPICAL CIRCULATION OF THE WIND, FROM ACTUAL OBSERVATION.

[From the Pilot Chart of the North Atlantic Ocean, July, 1890.]

The above diagrams have been prepared from a large number of observations in order to illustrate the actual circulation of the wind in hurricanes, as a practical guide for navigators during the present hurricane season. The small chart that was presented on the Pilot Chart last month gave all the observations upon which the spiral lines were based for that particular hurricane (Nov. 25, 1888), and the same method has been followed here, only the observations themselves are omitted, for the sake of clearness. Perhaps the most important point to notice is that the surface wind blows in an inward spiral curve, and not circularly, except very near the center. The center therefore generally bears more than eight points to the right of the wind. Another very important point is the fact that although the 8-point rule is nearly true when the wind is anywhere from North to South by way of West (that is, generally speaking, in the navigable semicircle), it is liable to be a very poor guide when the wind is from any point in the first or second quadrant. With the wind from NE., for instance, the center may bear anywhere from South to SE.; with the wind East it may bear from SW. to South; and with the wind SE. it may bear SW., West, or even (in the tropics) W NW. Perhaps the best general rule is that the center bears about eight points to the right of the direction from which the low clouds come, or, what is practically the same thing, eight points to the right of the wind at the moment of a sudden shift in a heavy squall; after such a shift the wind will remain steady in direction for a time, but the center is meanwhile moving along and the angle of bearing changes until the next shift, when it goes again to eight points, and so on.

It will be noticed that the northernmost of these two hurricanes was moving very slowly during the two days selected for illustration: had it been moving faster, the in-draught (or departure from the circular direction) would no doubt have been somewhat less in advance and considerably greater in rear than what is indicated. It is exceptional also to find a storm in this region growing smaller, as this seems to have done on Sept. 10th; it died out altogether in a few days, instead of continuing its motion toward E NE., as is usually the case. In the tropics the usual progressive motion is about W. by N., and this, together with the steady increase in size, is well illustrated in the case of the Cuban hurricane; it should be noted, however, that the interval is here two days, and not one, as in the upper diagram.

Masters of vessels are earnestly requested to keep regular observations for this Office during the hurricane season, even if only position, wind, weather, and barometer, at noon, G. M. T., are noted. A single additional report often adds greatly to the completeness of the data used in preparing these diagrams.