152 



NA TURE 



{June 14, 1888 



order to point out that, by causing rotation at the surface, 

 M. Weyher found himself unable to produce a gyratory 

 system extending downwards into the liquid from the 

 area of rotation. On the other hand, he always found 

 rotation, whether above or below, produce aspiration (ac- 

 companied by gyration) towards the area initially set in 

 motion. According to these results, therefore, river eddies 

 produced by inequalities in the horizontal flow cannot 

 propagate themselves below the area of flow disturbance. 



Now it is precisely this very form of river eddy which M. 

 Faye takes as his analogue to the aerial tornado, and it 

 is here that his argument fails ; for, while he draws atten- 

 tion to the system of downward motion and gyration in 

 an eddy caused by an outflow through an orifice in the 

 bottom of a vessel containing liquid, where such motion 

 and gyration is evidently caused by the outflow, he is 

 obliged to avoid all reference to outflow at the surface as 

 a cause in the supposed downward atmospheric gyrations. 

 At the same time he imagines that an entirely similar 

 system takes place, in the river, and the atmospheric 

 eddies, as in that produced by efflux, which propagates 

 itself downwards simply through initial rotations taking 

 place in the upper portions, of the liquid in the one case, 

 and of the atmosphere in the other. We have no hesita- 

 tion in saying that even if such an action were possible, 

 which we strongly doubt, it is in direct opposition to all 

 that we know of tornadoes, either deductively from physical 

 theory, or inductively from the facts which have been 

 recorded up to date. 



It would be a laborious, though at the same time dis- 

 tinctly easy, task, to point out the numerous physical facts 

 which accord with the upward aspiration and downward 

 propagation of conditions only, and which are utterly 

 opposed to M. Faye's theory of downward motion of the 

 air. It would be equally easy to quote numerous obser- 

 vations showing the objective reality, which M. Faye 

 questions, of upward motion in a tornado. Prof. Loomis, 

 for example, who is noted for his caution, relates the fol- 

 lowing pregnant incident in his own life, in his preface 

 to the revised "Contributions": — " In February 1842 a 

 tornado of unusual violence passed within 20 miles of 

 Hudson. As soon as I received the news, I started out 

 with chain and compass to make a thorough survey of 

 the track, and succeeded to my entire satisfaction. As 

 the tornado passed over a forest of heavy timber, I had 

 the best opportunity to learn the direction of the wind 

 from the prostrate trees ; and, by measuring the direction 

 of the trees as they lay piled one upon another, I deter- 

 mined the successive changes in the direction of the wind. 

 The facts demonstrated incontestable that the movement of 

 the wind was spirally inward and upward, circulating 

 from right to left about the centre of the tornado. This 

 tornado was but an incident in a great storm which swept 

 over the United States . . . " ; and he goes on to say that 

 the results of his subsequent investigation of the latter 

 showed that neither the purely circular theory of Redfield 

 nor the purely inward theory of Espy was correct. -The 

 truth, as usual, lay between these two extremes, and the 

 wind, like that shown in the diagram of the Manilla cyclone, 

 really blew in a spiral, curving in towards the centre. Any of" 

 the accounts published by the United States Signal Service 

 afford equally strong evidence in favour of both aspiration 

 towards the centres and motion up the axes of the tor- 

 nadoes. Thus, in the Report furnished by Rev. Charles 

 Brooke, of the West Cambridge tornado of August 22, 

 1851, the following remark occurs: "No one saw any 

 object driven downward by it, but all testify to its taking 

 things up" (the italics are in the original); and then 

 follows a list of articles taken up and carried, such as 

 boards and slates, to a distance of 3 miles, a large barn 

 1 5 feet, a freight-car 60 feet, &c. 



Again, in the Official Report of the Iowa and Illinois 

 tornado of May 22, 1873, different witnesses say: " Saw 

 boards whirling round in the funnel." " While the whirl- 



wind was on the river, the water ceased to flow over the 

 dam, although the river at the time was high." " Saw rails 

 flying out from the summit [of the column] ; an aver- 

 age rail weighs about forty pounds." And we may close 

 the list with one quoted by Ferrel as a well-authenticated 

 case, in the tornado at Mount Carmel, Illinois, June 4, 

 1877, in which "the spire, vane, and gilded ball of the 

 Methodist church were carried 15 miles to the north-east- 

 ward." The whole evidence, in fact, both in tornadoes, 

 and in their milder form of water and sand spouts, is 

 overwhelmingly against M. Faye's views, and in favour 

 of upward motion and aspiration to their very summits. 



In his endeavour to bolster up a theory weak at all 

 points, M. Faye seizes upon the well-known phenomenon 

 of the central calm in cyclones, and cites one which 

 occurred in the typhoon at Manilla on October 20, 1882, as 

 proving the general existence of a downward current. In 

 this case, while the thermometer during the first half of 

 the storm marked 24 C, it rose during the passage of the 

 central calm to 31 C, after which it fell again to 24° C. 

 The relative humidity followed analogously inverse 

 changes, falling from 98 to 53 — an extraordinary degree 

 of dryness for such a climate. With reference to this 

 circumstance, M. Faye quotes with considerable triumph 

 a remark of Dr. Sprung to the effect that " such a 

 characteristic phenomenon can only be explained by 

 admitting the existence of a descending current at the 

 centre of this cyclone." Locally, and for a short space 

 upwards, there might have been ; but these particular 

 features, accompanied by a clearing of the sky, and 

 known as the " eye of the storm," are the exception and 

 not the rule, even in tropical cyclones. It is, moreover, 

 readily seen that if there were a descending current of 

 any extent or velocity in cyclones it would necessitate an 

 outflow along the surface for some distance round their 

 centres — a condition utterly opposed to all observation 

 and experience. M. Faye makes one more attempt to 

 support the differential-current-motion hypothesis of 

 tornado and cyclone generation, by referring to certain 

 empirical laws of the relation of the former to the latter 

 disturbances, deduced by Mr. Finley, of the United States 

 Signal Service. For example, (1) the fact that tornadoes 

 are usually found in the south-south-east or dangerous 

 octant of a cyclone ; and (2) a law formulated by M. 

 Faye himself, according to which their trajectories, as 

 traced by the areas of destruction, are parallel to those 

 of the cyclones in which they are generated. 



The first of these facts has been known for some time, 

 and applies equally to thunderstorms. M. Faye considers 

 it to arise from the air shot down from the upper currents 

 reaching its maximum velocity " where the velocity of 

 translation is added to that of rotation," an idea which 

 concentrates in a truly tornadic manner two fundamental 

 errors which pervade his work. Modern investigation 

 has shown that the velocities of rotation and translation 

 in cyclones are quite independent, and is in this matter 

 as far ahead of M. Faye's view as his knowledge of 

 cyclonic systems is superior to that of Franklin, who had 

 no isobaric charts to help him. 



Again, the south-east portion of a cyclone is precisely 

 where, according to the corrected theory of aspiration, 

 the conditions are most favourable to the production of 

 local and parasitical disturbances of equilibrium, and 

 since such disturbances take their birth in or just below 

 the cloud-strata, their trajectories will naturally tend to 

 follow the course of these higher strata, which in this 

 part of the cyclone generally coincides with that of its 

 translation. The violent motion, moreover, which M. 

 Faye considers to be such an essential primary condition 

 in the generation of tornadoes is by no means necessary, 

 as Prof. James Thomson, among others, has pointed out 

 in a recent paper before the British Association (British 

 Association Reports, 1884, p. 641.) 



Besides the objections we have all along pointed out 



