August 13, 1891] 



NATURE 



349 



too well known to render necessary any detailed descrip- 

 tion of them. An account given by Mr. Archibald in 

 vol. xxxviii. of this journal (p. 149) is quoted without dis- 

 approval by M. Faye in his latest publication in the 

 Comptes rendus, and may therefore be accepted as just. 

 Its essential points are that cyclones are generated as 

 great eddies in the higher regions of the atmosphere, 

 and that there is a downrush of air in the vortex. " Dans 

 ces tourbillons, tout semblables h ceux qui se forment 

 dans les cours d'eau, les spires, d'abord tr^s larges, 

 iront en se retrdcissant par en bas, et leur girations pro- 

 gressivement accel^rees, en vertu d'une loi bien connue 

 de mecanique, amcnent au contact du sol, et y concen- 

 trent sous une aire bien plus ^troite que celle de leur 

 embouchure les energies continuellement renouvellees du 

 fleuve atfrien jusqu'a ce que son elargissement croissant 

 aboutisse a la decomposition du cyclone." 



Further on, with respect to the descending current in 

 the vortex, he remarks : " L'air envoyd en bas sera en 

 petite quantitd mais animd d'une vitesse de rotation 

 enorme." 



I leave aside for the present any criticism of the phy- 

 sical and mechanical actions which M. Faye conceives 

 to take place in these unfortunately inaccessible vortices 

 of the higher atmosphere, and which I, for one, am unable 

 to reconcile either with the results of direct observation 

 or with well-established physical laws. For the moment 

 I wish to concentrate attention on the question of fact, 

 whether there is an indraught of air to the cyclone vor- 

 tex at the earth's surface, and therefore necessarily an 

 ascending current over it, or, on the contrary, an outflow 

 from a descending current. This is the crucial point of 

 the controversy, and by the answer M. Faye's theory 

 must stand or fall. Indeed, M. Faye seems to recognize 

 this, since he says : — 



" L'argument le plus solide, celui qu'on m'opposait 

 toujours pour prouver que l'air etait ascendant dans les 

 cyclones, k savoir le fait que les isobares dtaient partout 

 et toujours coupds sous un angle assez notable par les 

 fleches des vents, de mani^re ^ accuser une tendance 

 nettement centripete, &c." 



He admits, too, that in certain cases there is really an 

 indraught and ascent of air ; only, on his view, these are 

 not cyclones. 



In order to forestall any objection on this score, I will 

 take as the subject of inquiry the cyclones of the Bay 

 of Bengal, the typical cyclones to which Mr. Piddington 

 first applied the name, however etymologically incorrect. 

 I trust, by this restriction, to escape ignominious dismissal 

 from court on the plea that my witnesses are impostors 

 — merely " prdtendus cyclones " — and that their evidence 

 is consequently irrelevant. 



My first experience of a great tropical cyclone was the 

 memorable storm that devastated the port and city of 

 Calcutta on October 5, 1864. Up to that time, my 

 acquaintance with cyclones was, like M. Faye's, " aca- 

 demic"; and under the impression that Reid's and Pid- 

 dington's description of the winds, as blowing in circles 

 or at right angles to the radius vector of the vortex, was 

 an established scientific fact, on the evening of that day 

 I sketched out, for the information of some friends, the 

 probable course of the storm that was then passing away, 

 having swept the port of its shipping, and left half the 

 houses around us more or less wrecks. Having no other 

 guide at the moment than the changing directions of the 

 hurricane as experienced at Calcutta, on the supposition 

 that the centre lay at right angles to these directions, I 

 inferred that the storm had reached us from the north- 

 east corner of the bay, and had followed a north-west or 

 west-north-west course past Calcutta. What was my sur- 

 prise, then, when accounts began to come in from other 

 places in Bengal, showing that the course of the storm 

 had been almost due north ; and when, further, on plot- 

 ting down the wind directions reported from other sta- 



NO. 1 1 37. VOL. 44] 



tions according to the hours at which they had been 

 observed, I found that, instead of being at right angles 

 to the radius vector, they were strongly inclined inwards ; 

 and such as, after making all allowances for their being 

 only estimated directions and perhaps, therefore, a point 

 or two in error, could be reconciled only with a 

 sharp spiral indraught to and up to the central calm. 

 Later on, when I obtained copies of the logs of ships 

 that had been involved in the storm in its passage up the 

 bay, I found that their wind observations, equally, were 

 compatible only with spiral directions. Unlike M. Faye, 

 I had no theory to support, and I submissively accepted 

 the teaching of the evidence which lay so plainly before 

 me. 



This evidence is set forth on Plates I. and II. of the 

 Report drawn up by Colonel Gastrell and myself, which 

 was widely distributed at the time to scientific bodies, so 

 that, in all probability, a copy must exist in the library of 

 the Academie des Sciences. 



Since then, many other storms in the Bay of Bengal 

 have been carefully investigated, and their full details 

 embodied in Reports drawn up by Messrs. Wilson, Eliot, 

 Pedler, and myself. Without a single exception, the 

 evidence thus accumulated has been to the same effect 

 as that of the cyclone of 1864, and these gentlemen have 

 all arrived at conclusions similar to mine. Thus, Mr. 

 Wilson says ^ : — "The following rule may be used to 

 determine the approximate bearing of the centre with as 

 much accuracy as it seems to be possible to arrive at : 

 In the northern hemisphere, with the face to the wind, 

 the direction of the centre is from ten to eleven points to 

 the right-hand side " ; and, to quote only one of Mr. 

 Eliot's numerous references to this subject,'-' " The air is 

 drawn into the centre [of a cyclone], but is not drawn 

 directly to it. The particles move by a kind of spiral 

 path to the centre." And he gives a diagram, followed 

 by charts of the Balasore cyclone of May 1886 and the 

 Madras cyclone of November of the same year, as illus- 

 trative examples. And Mr. Pedler, in summing up the 

 evidence of the False Point cyclone of September 1885, 

 says * : — 



"It is therefore clear, from these autographic records, 

 that there was a very strong indraught towards the 

 storm-centre, and that for a considerable portion of the 

 time, even when the storm-centre was comparatively close 

 to Hazaribagh, the winds were part of a well-defined 

 spiral system. In fact, for a large part of the time they 

 subtended an angle of less than 45^ with the radius of 

 the storm. . . . The records of five anemographs within 

 the influence of the storm . . . show that the theory of 

 the circular movement of winds in a cyclone, which was 

 advanced by Reid and Piddington, and has been sup- 

 ported by some later writers, is utterly untenable. At 

 considerable distances from the storm-centre the winds 

 approach more to the radial direction of indraught to- 

 wards the centre, as advocated by Espy, than to any 

 circular movement. As the centre of the storm is 

 approached, the circulation appears to become more 

 defined; but even just outside the storm-centre there is 

 no evidence to show that the direction is tangential." 



The reports here quoted and many others, all leading 

 to the same conclusions, have been communicated 

 officially to a large number of scientific bodies in Europe 

 and elsewhere, and taken together they probably furnish 

 the most copious and complete body of existing evidence 

 relative to the cyclones of a tropical sea. Not long since 

 I examined the whole of the charts given in these reports, 

 in order to verify Mr. Wilson's rule (quoted above) for 

 ascertaining the bearing of the storm-centre when the 



' "Report on the Midnapore and Burdwan Cyclone of October 15 smd 

 i6. 1874, ' P- 86. The italics are as in the original Report. 



" " Hand-book of Cyclonic Storms in the Bay of Bengal," p. 14 (1890). 



3 ''Indian Meteorological Memoirs," vol. iv.. Part 2, p. 127. The baro- 

 metric reading recorded when the centre of this sloim was p.-issing False 

 Point Lighthouse is the lowest that has ever been observed at the sea-level. 



