13

touched; we can then by the aid of Fournier’s rule find hy calculation the direction of the center’sadvance with sufficient accuracy.

According to Commander Fournier the following proportion expresses the relation betweenthe descent of the barometer within a cyclone and the distance of the center—

P-P D rP-P, D 2

Here P is the atmospheric pressure at the outskirts of the storm, P, and P 2 are the heightsof the barometer at two different times of observation, D, and D 2 the distances of the vortex atthe respective moments. P may be taken from the table given on page 7, while 1-^ will be thebarometer reading at the time of setting the double needle for third approximation, corrected, ifnecessary for daily oscillation, and P 2 the reading when the graduated needle is set eventuallysimiliarly corrected. P—P, and P—P 2 are consequently easily found. But in order to solve theproblem we must know a third term. Not necessarily, because the absolute values of the distancesdo not affect the directio-n in which, the center is moving; this direction depends upon their relativevalues and the included angle. But we know—

A

D.

since the terms of the left side are given. In order to arrive at numerical results we assume D,divided into 100 equal parts, thus making it a numerical constant, though the unit by which it ismeasured may vary within very wide limits. The solution of the proportion will now give us theproportional value of D 2 expressed in a number of the equal parts of D„ which number may hegreater or less than 100.

In the apparatus Dj is represented by the distance of the pivot of the small needle fromthe center of the wind disk, which distance, as will be remembered, equals the graduated portionof the other needle=100. Having therefore calculated D, from the above proportion, all that isnecessary is to set the small needle in such manner that it points to that division line of the grad-uated needle whose ordinal number is equal to D 2 , then the small needle will be parallel to thetrajectory of the cyclone.

Corresponding to this direction we turn the disk of the cyclonometer until the central arrowis parallel to the small needle. Should the barometer continue to fall rapidly, it will be veryadvisable to repeat the whole operation, taking now into consideration the wind directions givenfor zone C or D. But there is an imperative necessity of repeating the third approximation asoften as the prevailing wind veers; it may prove disastrous to wait until the wind has run throughtwo or three points of the compass.

It may be well to point out the meaning of the whole procedure. The problem of determin-ing the direction of the trajectory resolves into the following: Given one angle of a triangle andthe proportion of the two including sides: find the direction of the third side. The two determina-tions of the center’s bearing, the first of which is laid down on the wind disk by the position of thedouble needle and the second hy that of the graduated needle, give us the included angle, while thefraction—

p -P 2

P-P,

furnishes the proportional lengths of the sides. D„ represented hy two-thirds of the half-lengthof the double needle, having a constant numerical value equal to 100 (whatever its linearmagnitude may be), we calculate D 2 and lay it off on the graduated needle by pointing thesmall needle at the corresponding division of the graduated end, which manipulation brings thesmall needle into a position parallel to the line whose direction is sought—that is, the trajectoryof the typhoon.