IO 



NATURE 



[May 4, 1882 



tendencies — gyration and inflow, or outflow according as 

 the air is in the interior or exterior part — the air near 

 the surface takes a middle course, and flows spirally 

 around and toward the centre from the zone of maximum 

 pressure on the one side, and on the other in a contrary 

 spiral outwards from the centre to the outer limit of the 

 anticyclone. 



It is important to observe that the author explains the 

 accumulation of air with its maximum at the dividing 

 line between the interior and exterior districts (cyclone 

 and anticyclone, as they are termed elsewhere throughout 

 this work) as " due at the start mostly to the gyrations in 

 the upper part of the atmosphere," which, being less influ- 

 enced by friction, are in consequence more circular than 

 those below ; the pressure from this accumulation tending 

 to force the air near the earth's surface out from beneath 

 it on the one side toward the centre of the cyclone, and on 

 the other toward the outer limit of the anticyclone. 



The difference of pressures or gradient between the 

 regions of high and low pressure in a c> clone, is thus 

 shown to be, not so much the cause of the wind, as the 

 mechanical result of the deflecting force of the earth's 

 rotation and the centrifugal force engendered by the 

 gyrations. 



It should, however, be borne in mind, that the forces just 

 mentioned, are by no means to be regarded as causing the 

 cyclone in the sense of being independent sources of 

 energy. They can only arise in consequence of some 

 initial motion of the air, which must itself be due to a 

 small difference of pressure, and unless such primary dis- 

 turbing cause be continu illy maintained by external 

 influences, the entire system of motion will shortly come 

 to rest. 



The preceding view of cyclone generation has already 

 made some way since Ferrel first enunciated its leading 

 characteristics in his previous writings. It lies midway 

 between what is sometimes called the in-blowing or 

 ascension-current theory of Reye and Espy, which regards 

 the central depression as the main cause of the wind, and 

 thad held by Thorn, Meldrum, Willson, and Loomis, 

 according to which the central depression is mainly due to 

 the centrifugal force generated by two pre-existing currents 

 passing one another in opposite directions. A third theory, 

 held by Blanford and Eliot, and evolved chiefly from a 

 study of the cyclones in the Bay of Bengal, makes the 

 condensation of vapour the primary cause of disturbance, 

 but allows the greater part of the subsequent depression 

 of the barometer to be due to the causes adduced by 

 Ferrel. This latter theory, in fact, only differs from that 

 put forward by the author in the part plajed by condensa- 

 tion of vapour in giving rise to the initial motion of the air, 

 which Ferrel considers to be considerably less than that 

 exeicised by a difference of temperature. Among the 

 conclusions arrived at by the author, and which are gene- 

 rally confirmed by the results of observation, may be 

 noticed the following ; but these, it must be remarked, are 

 only strictly true for a regular, symmetrical, and stationary 

 cyclone: — 



(a) " The wind inclines towards the centre from the 

 direction of the tangent, and the amount of inclination is 

 nearly in proportion to the friction (mainly of the air 

 against the earth's surface)." 



(b) "The inclination diminishes with the altitude, and 

 therefore at some distance from the earth's surface the 

 gyrations are more circular than near it." 



(c) " Toward the centre of a cyclone, where the gyratory 

 velocity is greater, the inclination is less, and therefore 

 the path more nearly circular." 



(d) " The inclination increases with decrease of latitude, 

 attaining its greatest value at the equator, where the air 

 should flow directly towards or from the centre, and there 

 should be no g) rations." 



(e) " As the motion of the air below in cyclones is /award 

 the centre, in the upper regions of the atmosphere it must 



be nearly circular, but inclined to the tangent a little/raw 

 the centre." 1 



Inertia comes into play where a c> clone is increasing or 

 diminishing in violence, and its effect is to increase the 

 inclination in the former case and diminish it in the latter, 

 but in general the amount is found to be insignificant. 



It was stated in the above brief sketch of the theory, 

 that outside the annulus of high pressure surrounding a 

 cyclone the air should move outwards anticyclonically. 

 Ferrel subsequently puts the matter thus : " Every cyclone 

 is accompanied by a corresponding anticyclone, and the 

 former cannot exist without the latter." 



The words cyclone and anticyclone are here used quite 

 apart from the question of barometric pressure, and 

 simply mean districts in which the motion of the air is 

 spirally in towards the centre, or out from the centre 

 respectively. Guldberg and Mohn likewise adopt this 

 definition, which is obviously far more scientific than the 

 too common habit of referring to them as regions of low 

 and high pressure. 



Mr. Ferrel, however, differs from all previous investi- 

 gators in thus linking together the cyclone and anticyclone 

 as mutually dependent parts of the same phenomenon. 

 They have hitherto been treated separately, at least in 

 practice, and though the author's conclusion sounds like 

 a simplification, because it makes one out of two, we 

 scarcely think he has proved the converse to his proposi- 

 tion, viz. that every anticyclone is accompanied by a 

 corresponding cyclone, and cannot exist without it. 



For example, in the case of such an anticyclone as every 

 winter forms over Central Asia it would be difficult to 

 point out exactly its corresponding cyclone or cyclones, 

 though it is possible, as the author says, that it may be 

 partly due to the overlapping of the anticyclones, which 

 should surround the permanent North Atlantic and Pacific 

 cyclones at this season. 



The relation between the barometric gradient and the 

 velocity of the wind in a symmetrical cyclone is given by 

 the following equation : — 



-_ io76'4(2« cos ij/ + i>)s P 

 cos i{ 1 +-004/) P'' 



where v = s cos ' ', and G is the gradient in millimetres 

 r 



per sixty geographical miles, s the velocity of the wind 

 in metres per second, n the earth's angular velocity of 

 rotation per second, i^ the co-latitude, i the inclination of 

 the path of the wind to the isobar, P P l the barometric 

 pressure at the given elevation and at the earth's surface 

 respectively, t the temperature in degrees Centigrade, and 

 r, the distance from the low centre in kilometres. Where 

 the gradient is given, the velocity of the wind can be 

 conveniently found from the equation 



5 = ._ J a ± N '| a- + b G, 

 where, if the ordinary English units of space and time 

 are used, viz. a mile and an hour, and the gradient is ex- 

 pressed in inches per sixty geographical miles, we have — 

 t 0-52505 r C0!»\)/ 



) a= c^TT 



) r(i+-oo4l) P 



1° ~ -005262 P 1 

 The equation for the gradient in terms of the wind's 

 velocity is substantially the same as that already given 

 by the author in Silliman's Journal for 1874, with the 

 exception of the temperature correction, which was there 

 simply referred to in the text. 



As the whole question of the author's formula for the 

 gradient has been thoroughly ventilated in his previous 



1 This conclusion only applies to districts within a moderate distance from 

 the centre. At great distances fr u m the centre the radial component pre- 

 dominates, and the air fl^ws nearly directly towards the centre below, and 

 from it above. 



