440 SHORT MEMOIRS ON METEOROLOGICAL SUBJECTS. 



the opposite direction, aud tbo component v cos i of the rotatory move- 

 ment must be negative, and therefore /I i^ will be also negative. At a 

 certain distance from the center of a perfect cyclone, between the center 

 and tbe extreme limit, the pressure will be a maximum, and A B = 0. At 

 this distance, ^ = 0, or calms must also prevail. Therefore an area of 

 high pressure must, in general, be attended by calms. An easy compu- 

 tation shows that for the same gradients the wind-velocity must sensibly 

 diminish toward the center. Unfortunately, the observations at hand do 

 not suffice to test the relation between observed gradients and velocities. 



The empirical law of Buys Ballot is, says Ferrel, included in the pre- 

 ceding expression for AB. This expression is, however, correct for all 

 latitudes and inclinations of the wind to the center, where Buys Ballot's 

 law no longer applies. With reference to the force or velocity of the 

 wind, we perceive by a simple inspection of the formula that the velocity 

 in all parts of a cyclone, for the same constant geographical latitude, is 

 not alone proportional to AB, and, further, that in different latitudes 

 the value of v for the same gradient is nearly inversely proportional to 

 the sine of the latitude, especially at a considerable distance from the 

 center. Hence, also, for the same gradients the velocity of the wind is 

 much greater in the tropics than in higher latitudes. We may add 

 that Toynbee (see this journal, vol. ix, p. 79, and vol. x, p. G3) had 

 also previously called attention to this fact, aud Blanford had drawn 

 the same conclusion from the isobars for India. The form in which 

 we have presented Ferrel's equation further shows that for equal wind- 

 velocities the gradient diminishes also with the temperature, but to a 

 very slight degree. The same result is furthermore brought about to 

 a very considerable extent in the higher regions of the atmosphere by the 

 diminishing atmospheric pressure, indeed, the barometric variations 

 actually do diminish with the altitude, while the wind-velocities do not. 



All irregular barometric variations depend almost entirely upon 

 cyclonic action and are in general caused by the passage of a cyclone 

 near the place of observation ; hence, at the equator, where no cyclone 

 can be formed, the irregular barometric variations are scarcely sensi- 

 ble. Ou this point, Ferrel quotes the same evidence from Sykes and 

 Humboldt, as given in Herschel's Meteorology, that we have ourselves 

 reprinted on page 12, volume x, of this journal. At the equator, no 

 cyclone can arise; and therefore in storms the movement of the air at 

 the earth's surface is directly toward the center of the area of dimiu- 

 ished density of the atmosphere, aud the inertia of the air, as well as the 

 friction, both overcome by the force that results from an almost imper- 

 ceptible gradient. Hence, Ferrel also considers that his neglect of the 

 inertia in the deduction of his formula for the gradient can in any case 

 only give rise to a very slight error. In a cyclone, tbe force that results 

 from the gradient is almost entirely held in equilibrium by the centrifugal 

 force that arises in consequence of the rotatory motion and the earth's 

 diurnal rotation, and so great is the mobility of the air and so slight the 



