i5o 



NATURE 



[J tine 14, 1888 



that M. Faye insists upon all through is, that these arise 

 solely through inequalities in the upper currents, causing 

 gyration round a vertical axis, which, like a river eddy, is 

 propagated from above downwards, by a descending motion 

 of the air. M.Colladon,referringtoM. Faye's view, describes 

 the supposed action as " un mouvement tourbillonnaire 

 aerien constituant a son interieur une trombe aspiratrice 

 a mouvement descendant." M. Faye, therefore, postulates 

 two points : (1) that the movement commences above ; 

 (2) that it is propagated downwards by a descending 

 motion, accompanied by gyration round a vertical axis. 

 The opposite theory, as presented by M. Faye, is the 

 exact inverse of this, since it makes the action (1) com- 

 mence at the earth's surface ; (2) propagate itself up- 

 wards ; and (3) borrow its gyration from that of the 

 earth. Here, however, we find ourselves distinctly at issue 

 with M. Faye, for we do not believe that the leaders of 

 modern meteorology entertain any such view as the 

 latter. The surface of the earth is the most unlikely 

 birthplace for a tornado, whirlwind, or waterspout. In 

 order to maintain an ascending current, the air must be 

 nearly saturated, and this will generally occur only in 

 and near the lowest cloud stratum. The vertical tem- 

 perature gradient and disturbances which start the action, 

 will likewise operate most effectively at this level, so that 

 all the conditions which unite to cause a tornado will tend 

 to commence at some distance above the earth's surface. 

 On the question of level, therefore, we may invite M. 

 Faye to agree with us. Then comes the question of the 

 downward propagation. 



The entire gist of the question appears to us to lie in 

 this downward propagation. The physical theory de- 

 veloped by Ferrel and Sprung makes the action commence 

 in a slight upward motion in unstable air, due to a tem- 

 perature inequality or some other cause, the only other 

 condition being a gentle gyratory motion relative to some 

 central point, which is never wanting in a cyclonic area. 

 Once the motion is started, and the air which feeds it is 

 nearly or quite saturated, the action will go on and be 

 propagated downwards, not by a descent of the air, but 

 by the transference of the physical conditions which 

 favour the continuance and maximum development of the 

 " courant ascendant." The increasing rapidity of gyration 

 of the air as it approaches the axis, however gentle it may 

 be at starting, only allows it to partially feed the initial 

 and continually reproduced vacuum, which is thus com- 

 pelled to draw its supplies chiefly from the non-gyrating 

 air at the lower end of" the aerial shaft. As this is drawn 

 upwards, the centrally aspired surrounding air is made to 

 gyrate more rapidly (partly by the friction of the super- 

 jacent rotating layer), and thus the gyratory and other 

 conditions are propagated downwards until a balance is 

 struck between supply and demand. 



The theory thus sketched may be termed the modern 

 theory of aspiration as applied to tornadoes, and will, we 

 venture to think, be found to meet all M. Faye's objections 

 to the first crude notions which prevailed in past years 

 from a study of a few isolated surface conditions. 



Before proceeding to notice the objections which M. 

 Faye brings against the existence of either an upward 

 current or any sort of aspiration in tornadoes, we must first 

 touch upon the cognate question of cyclone generation, 

 which he explains on the same principles ; l and here, with- 

 out attempting to give any review of the modern theory, 

 which involves as a primary condition a horizontal 

 temperature anomaly over a considerable area, we 

 may observe that the two main objections brought by M. 

 Faye against the ordinary view of their formation are, 

 (1) that it assumes the existence of centripetal currents, 

 and hence aspiration towards their axes ; (2) that it gives 



• ' There is no real connecting link between the two, i.e. the smaller 

 cyclones do not begin where the larger tornadoes leave off. The average 

 size of 600 tornadoes in the United States was found to be 1085 yards. The 

 average size of the cyclones is as many miles. 



no explanation of their movements over the earth's 

 surface. 



With respect to the first objection, M. Faye draws 

 attention to a principle which he develops on p. 46, 

 according to which the isobars in the temperate zone do 

 not correctly represent the motions of the air in a cyclone, 

 and says we must look at the isobars in a tropical cyclone 

 if we wish to arrive at correct conclusions. 



Here, according to M. Faye (pp. 2 and 46), where "by 

 the ancient theory the direction of the wind ought to cut 

 the isobar at an angle of nearly 90°, the angle is sensibly 

 nothing ; the pretended centripetal component disappears ; 

 and the isobars and the wind arrows display an almost 

 rigorous circularity." Again, on p. 12 he ridicules the 

 idea of a barometric gradient in the tropics, " where the 

 wind blows precisely along the isobars." It is with no 

 desire to indulge in mere polemic that we take up the 

 gauntlet thus thrown down, but the magnificent work of 

 that most careful and renowned inductive meteorologist 

 Prof. Loomis which he has been lately revising, enables 

 us to show most conclusively not only that in the latitude 

 of the Philippines which is nearly the equatorial limit of 

 true cyclones, the direction of the wind in a particularly 

 violent and well observed typhoon cut the isobars right 

 through at the large angle of 62 ; but that an extensive 

 comparison of similar conditions, embraced in a large 

 number of violent storms in different latitudes, shows that 

 the angle between the winds and the isobars increases as 

 it should do according to theory from the poles to the 

 equator. 



The accompanying figure represents the observations 

 accurately, except that the isobars were not as there 

 exactly circular ; while the following table shows at 

 a glance how entirely opposed M. Faye's statement is 

 to the true facts, in the very region where he says, a les 

 isobares elles-memes dessinent sur le sol comme les fleches 

 du vent un edifice cyclonique non encore deforme." We 

 have no hesitation therefore in saying that these obser- 

 vations of Prof. Loomis not only give the death-blow, if 

 one were needed, to the purely circular theory of Reid and 

 Piddington, but constitute a coticlusive argument against 

 M. Faye's theory of downward gyratory currents and non- 

 aspiration in cyclones. 



Inclination of the Wind to the Isobars in Violent Storms. 



It is true both from theory and observation that the 

 inclination is less on sea than on land, and usually less as 

 we approach the centre ; but the above cases suffice to 

 show the danger which might attend an unmodified 

 adherence to the circular theory, or the rough empirical 

 law of Buys Ballot, which is its practical expression. Dr. 

 Meldrum, F.R.S., of Mauritius, as we have pointed out 

 in a previous article (Nature, vol. xxvi. p. 31), has 

 frequently exposed the danger of following the purely 

 circular theory, and in a number of the Journal of the 

 Mauritius Meteorological Society for July 1883, he 

 mentions a case in which the captain of the ship 

 Caledonien on January 24, 1883, deliberately ran it into 

 the centre of a cyclone by following the circular rules. 

 Fortunately he subsequently became aware of his error, 

 and altered his course just in time to escape the centre. 



The second objection brought by M. Faye against the 

 physical theory of cyclones is, that it cannot explain their 

 general motions and course over the earth. We admit 

 that the partial theory, sketched in his opening statement, 

 which he considers to represent the modern meteoro- 



