346 



SCIENCE. 



[Vol. XVI. No. 411 



miles in all directions, instead of 4° below the three-years' nor- 

 mal, no conclusive argument could be drawn from it. But it is 

 well known that the temperature departures from the normals 

 are often very great and of long continuance. The observed 

 temperatures, therefore, on Oct. 1, may have been 4° C. 

 below the normal, and yet 5* or more above the surrounding tem- 

 peratures at a great distance; and unless these surrounding tem- 

 peratures are observed all around at distances of several hundred 

 miles, and at almost all altitudes, or at least up three or four 

 miles, so that there can be a comparison of the interior and exte- 

 rior temperatures all around at the same levels, no argument can 

 be deduced against the condensation theory of cyclones. In fact, 

 it is readily seen from this that the theory can neither be proved 

 nor disproved in this way, nor by a comparison of the interior 

 temperatures with normals. 



It is well known that surface temperatures, on the average, in 

 cyclones, are generally below the normal temperatures, especially 

 in the summer season. This is due, as Dr. Hann explained sev- 

 eral years ago, to the products of condensation falling from high 

 and cold altitudes. In Dr. Hann"s memoir it is stated that on 

 the day and evening preceding Oct. 1 rain fell in the valleys, and 

 snow on the mountains. Now, it must be noted here that all the 

 observations in the Alps on Oct. 1, from which the interior tem- 

 perature of the cyclone, up to an altitude of 3,100 metres, has been 

 estimated, were surface temperatures, and consequently they were 

 considerably lower than they otherwise would have been, from the 

 ■effect of the recently fallen snow. They cannot, therefore, be as- 

 sumed to be the same as open-air temperatures, even at a little dis- 

 tance on the same levels, and, much less, can they represent the 

 general average of the great mass of air in the interior of the cy- 

 clone, of perhaps five or six hundred miles in diameter and up to 

 a considerable altitude. Since this has been explained bj' Dr. Hann 

 to be a mere surface effect, why now attempt to deduce an argu- 

 ment from it against the condensation theory? 



Another false assumption in the preceding argument is, that 

 cyclones occur in a normal state of the atmosphere ; for, unless 

 they do, it is not logical to compare the observed temperatures in 

 a cyclone, in the average of many observations, with the normal, 

 and, if found to be less, to infer that the temperatures in cy- 

 clones are less than in the surroundings generally. The normal 

 state of the atmosphere is one of stability, whereas cyclones oc- 

 cur in an unstable state of the atmosphere, when the vertical 

 temperattu'e gradient is abnormally large, and so when the parts 

 of the atmosphere on a level with the upper part of the cyclone has 

 a lower temperature than usual in reference to the temperature of the 

 lower part; or, in other words, the average temperatures of the air at 

 considerable altitudes, taken when the air is in an unstable state, 

 and so when the conditions are favorable for cyclones, must be 

 less than the general average of all times. The observed tempera- 

 tures, therefore, in a cyclone at high-level stations, may be lower 

 on the average of many observations, and yet higher than the 

 average surrounding temperatures at the times of the cyclones, 

 for these are below the normal on the average at these times. 

 The observed negative temperature departures from the normal on 

 Oct. 1, 1889, at the high stations in the Alps, may have been due to 

 the fact that the air at the high levels at the time, both in and around 

 the cyclone, had a temperature considerably below the normal on 

 account of the abnormal and unstable state of the atmosphere at 

 the lime: and so on this account the observed temperatures may 

 have been lower than the normal, and yet above the temperature 

 of the surroundings; and so the necessary conditions of a cyclone 

 would still have been fulfilled. The negative temperature depar- 

 tures on Oct. 1 were mostly at the higher stations. 



Another of the high-level tests is the observed high temperature 

 at elevated stations at the times of long- continued high baromet- 

 ricjpressures, and especially that of the high-pressure area of No- 

 vember, 1889, over the Alps, which continued fourteen days. It 

 is well known that in such cases there is a body of abnormally 

 heated air at some distance above the earth's surface, of a foehn- 

 like character, arising from the downward current which must 

 necessarily exist in the high-pressure area. "When the high 

 pressure occurs over a mountainous region, such as that of the 

 A\[j3, with high-level^ stations _^of observation, such abnormally 



high temperatures are frequently observed. Because these tem- 

 peratures are frequently above the normal temperature of the 

 month, or season of the year, and also sometimes found to be 

 higher than the temperatures observed in cyclones at correspond- 

 ing seasons, it is attempted to base an argument upon this against 

 the condensation theory of cyclones. But what connection there 

 is between the observed premise and the conclusion the writer is 

 entirely unable to see. Whatever may have been the peculiar 

 circumstances under which the long-continued high pressure ex- 

 isted, even if the temperature within had been raised 20° above the 

 normal, he cannot see how this would interfere with the existence 

 of the necessary conditions of a cyclone by the condensation the- 

 ory, say in America, at the same time; and especially, it could 

 have nothing to do with their existence at other times; and these 

 long-continued high-pressure areas are not of frequent occurrence. 

 These conditions, as is well known, are simply that the vertical 

 temperature gradient at the time of the cyclone shall be greater 

 than usual, so as to induce the unstable state in which the tem- 

 perature of the air in the ascending current in the interior of the 

 cyclone shall be kept, by the latent heat given out in the conden- 

 sation of the aqueous vapor, a little above that of the surround- 

 ings, and so its specific gravity a little less. There is no reason 

 why such conditions could not exist in America, or even anywhere 

 at a considerable distance, during the time even of the existence 

 of this peculiar state of pressure and temperature conditions over 

 the Alps. It has never been claimed that the conditions of a cy- 

 clone exist in these high-pressure areas, and it is well known that 

 the tendency is for cyclones to pass around such areas. Will 

 Professor Davis be so good as to throw some light upon this dark 

 part of the argument, so that there may be a clear understanding 

 of it, and a thorough discussion of it at some other time ? 



Since by the new hypothesis the energy of C} clones is in the 

 upper poleward-moving current of high latitudes, where the 

 pressure gradients between the equator and the pole are steep, 

 Professor Davis seems to realize the difficulty in applying this 

 energy to the cyclones which originate below the tropics near the 

 equator. He therefore thinks that a little of Espy's ' ' steam- 

 power " may be necessary at first until they get a start. During 

 this time the energy is in the latent heat of the aqueous vapor, by 

 which, set free in condensation, the ascending air is kept warmer 

 and lighter than the surrounding air, and the gyration depends 

 upon the deflecting tendency of the earth's rotation. But being 

 once under wa}', this is changed, and the ascending air in the 

 cyclone is colder and heavier than the surrounding air. At first it is 

 compared with a train of cars, driven by its own store of energy ; 

 but after a time the engine becomes simply a dummy, and the 

 train is driven by an external motor. But this is not strictly a 

 happy comparison; for, instead of the engine becoming a dummy, 

 it becomes a reversed engine. Before the change the ascending 

 air was lighter than the surrounding air, and so the tendency was 

 for it to rise, and for the cyclone to be continued ; but after the 

 change, when the ascending air was heavier, the tendency was 

 just the reverse. Nevertheless the cyclone machine, after the 

 reversal of the engine, seems to run on, all the same, and even 

 with increased energy. Davis says the external motor is the gen- 

 eral circulation of the winds. But why not say oleclricity? This 

 would be just as satisfactory. It must be remembered here that 

 the question is not with regard to the progressive motion of the 

 cyclone, for there is no difficulty here, but with regard to the 

 force which causes the heavier air to rise, and which maintains 

 the gyratory motion. The mere assertion that these arise from 

 the general cnculation cannot be accepted in a scientific argu- 

 ment. Let it be proved, from true physical and mechanical 

 principles, that there is a force arising from the general circula- 

 tion which acts on all sides of the cyclone so as to force the hea- 

 vier air up, and also acts as a couple in keeping up the gyration, 

 or at least make it appear that this is probable; for unfortunately 

 there are many things in science which cannot be absolutely 

 proved, but only be made to appear reasonable and probable. 



From what has been stated, it seems, that, of two rival theo 

 ries, the one is applicable to the cyclone in the first part of its 

 course, and the other in the latter part. But how is it with re- 

 gard to tornadoes ? Does the powerful ascent of air in these arise 



