11

(3) The direction of the center exercises an influence upon the variations of the angle betweenthe direction of the wind and the radius vector. The following factors influence the variations inconvergence:

(a) The direction of the prevailing wind.

(&) The topographical conditions.

(c) The progressive movement.

( d ) The shape of zone D.

The general winds which prevail on the edge of the cyclones make their influence felt only atplaces which are far distant from the center, and in such a way that southerly winds modify theconvergence of the winds on the south side, but not of those on the north side. In a similar waynortherly winds affect the convergence of cyclonic winds on the north side only, but have no effecton those of the south side. The cyclonic winds of the south side lie between the points northwestthrough south to southeast, those of the north side between northwest through north to southeast.It results from this for the Archipelago, that, owing to the general winds on the outside of cyclones,the outer cyclonic winds of the southern side—southwest, south, and southeast—show a stronginclination toward the center during the months of the second and third group; during the monthsof the first group the winds of the northern side are less convergent.

The directions of the wind for zone A of the cyclonometer are more convergent than those ofthe other zones. It is, of course, impossible to express on the wind disk of the barocyclonometerthe influence which the prevailing winds have upon the convergence of cyclonic winds.

The influence which topographical conditions may exercise needs hardly be considered by thesailor. An observer at a fixed observatory will soon fin'd out these influences (as they are exerted,for example, by a mountain chain close at hand).

'The wind directions will be less convergent on the front side, owing to the progressive move-ment of the cyclone, so that they will more approach a circular form than is the case in the rear.The directions of the wind on the two sides of the track undergo, owing to the direction of thetrack and the intensity of the progressive movement, changes which are all the greater the greaterthe velocity of the cyclone.

In drawing the wind disk we have taken 45° to be the value of the mean inclination of thewind. 1 The shape of the central region of cyclones changes the wind direction but slightly. Asthe shape changes a great deal this can not be taken into account when drawing the wind disk.We have preferred to represent it as circular, and this representation can not be a source of seriousmistakes.

There is no mention of rain on the dial, because its relation to pressure depends entirely onits causes, and precipitation may occur with different barometric heights.

WORKS WJLIC.LI MAY BE CONSULTED IN CONNECTION WITH THE PRECEDING.

Barometer Manual for the Use of Seamen. London, 189G.

Comte. Thcorie des Variations Brusques que Presenteut les Curves du BaromCtre Enregistreur Pendant lesOrages. Ann. de la Soe. Met. de France, 1884.

Dallet. La Provision du Temps et les Predictions MfitSorologiques. 1887.

Hermoso, Leon. El Barometro Aneroide y Modo de Graduarlo para la Prevision del Tienipo. Boletin Meteorol.de Neherlesoom.

Pernter. Reglas para el Uso del Barometro Aneroide. 1881.

Plumandon, J. R. Le BaromGtre Applique a la Provision du Temps en France. Paris, 18S3.

Wiiymper, E. A New Mountain Aneroid. Amer. Geo. Soc., vol. 31, p. 75.

Ferrel, “A Popular Treatise on the Winds,” p. 304.