154 MR JOHN AITKEN ON 



way in which the air currents in the atmosphere could produce eddies hundreds of miles 

 in diameter. 



Are there not, however, differences in these two forms of cyclonic movement by 

 means of which we can distinguish the one from the other, and say whether the cyclones 

 in our atmosphere are convectionally or dynamically driven ? I think there are. One 

 way in which we may distinguish between the two kinds is in the direction of the 

 circulation. In a convectionally driven cyclone the circulation is spirally inwards, 

 whereas in a dynamically driven one we would expect it to be spirally outwards. In 

 a dynamically driven cyclone in our atmosphere we would be entitled to expect an 

 inward current near the surface of the earth, owing to the velocity there being retarded 

 by friction ; the walls of the cyclone would therefore be weak at that part, and air would 

 be drawn in at the surface of the ground, and it would also probably be drawn in at the 

 top. Now, is there any evidence of a general outward circulation in cyclones with in- 

 draught at top and bottom ? 



For information on this point we cannot do better than turn to the valuable and 

 very important results obtained by observations on clouds made by means of the 

 nephoscope and theodolite, and published in the Report of the International Cloud 

 Observations, prepared under the direction of Willis L. Moore, Chief of the Weather 

 Bureau, by Frank H. Begelow, M.A., and published last year by the Weather Bureau, 

 Washington, U.S.A. If we turn to the series of diagrams in the above Report 

 showing the direction and velocity of the circulation in cyclones, we shall see that at only 

 one cloud level does the circulation show any tendency to be outwards. The following 

 is a summary of the results taken from the Report, pp. 435-6 : " Beginning with the sur- 

 face, it is seen (1) that all the vectors have an inward component up to the cumulus, the 

 inner increasing in strength ; (2) that they continue inward in the levels from the strato- 

 cumulus to the cirro-cumulus, except that there is a slight tendency to turn outwards on 

 the exterior circle, the alto-stratus again showing average conformity with the others ; 

 (3) the tendency is still inward in the cirro-stratus and cirrus level in the interior, but 

 apparently lawless over the outer parts." From the above it will be seen that at all 

 levels except the strato-cumulus the circulation is inwards. This tendency to outward 

 circulation at the one level is evidently due to the cyclone having at the strato-cumulus 

 level its greatest rate of rotation, as shown by the diagrams, and this high velocity 

 of rotation at this level would appear to be due to the air in the anticyclone also 

 having its maximum velocity at the same level. Though the evidence here shows the 

 direction of almost the whole of the circulation to be towards the centre, and thus 

 points to the circulation being convectionally driven, yet as there may be some hesitation 

 in accepting this evidence as final on the point, as all the circulation is not inwards to 

 the top, I shall therefore now proceed to point out another and more definite way in 

 which we can distinguish between dynamically and convectionally driven cyclones. 



In a convectionally driven cyclone the velocity of movement of the air, both absolute 

 and angular, increases from the outside towards the centre, whereas in a dynamically 



