TRANSACTIONS OF SECTION A. 641 



During- the ensuing year -with improved apparatus the author hopes to make a 

 more complete series of observations by wlrch the variations in the velocity of the 

 air at different heights may be discussed with reference to the action of the several 

 factors, time of day, season, temperature, humidity, pressure, &c. The conclusions 

 at present tentatively arrived at may be classified thus : 



1. The velocity of the air invariably increases up to a height of 1,100 feet 

 above sea-level. 



2. The increase is rapid near the surface, and diminishes sensibly at heights of 

 200 feet or more above the surface. 



3. The velocity at any moderate height may be very neaily found from an 



empirical formula of the form — = ( - ) vh m+A; where V, v, II, h, are the velocities 



and heights at the upper and lower altitudes, and o a constant to be determined 

 by comparison of observations. 



4. That Koppen's theory of the diurnal period in the velocity of the wind is 

 strongly supported by the behaviour of the kites. 



5. That there is an ascending current in the front and a descending current in 

 the rear of travelling cumuli and cumulo-strati. 



14. On the recent Sun-glows and Halo in connection with the Eruption of 

 Krakatoa. By Professor E. Douglas Archibald, M.A. 



15. On Whirlwinds and Waterspouts. 

 By Professor James Thomson, LL.D., F.E.S. 



Whirlwinds, whether on sea or on land, have their characters in great part 

 alike. For simplicity it will be convenient to begin by taking up only the case of 

 whirlwinds on sea, as thus the necessity for alternative expressions to suit both 

 cases, that of sea and that of land, will be avoided. 



It may be accepted as a fact sufficiently established both by dynamic theory and by 

 barometric observations that, at the sea level, the pressure of the air is less in the 

 neighbourhood of the axis of whirl, than it is at places further out from the axis, 

 though within the region of the whirl. The apocentric force (centrifugal force) of 

 the rapidly revolving air resists the inward pulsive tendency of the greater outer 

 than inner pressure. But close over the surface of the sea there exists necessarily 

 a lamina of air greatly deadened as to the whirling motion by fluid friction, or 

 resistance, against the surface of the sea ; and, all the more so, because of that 

 surface being ruffled into waves, and often broken up into spray. This frictionally 

 deadened lamina exerts, because of its diminished whirl speed, less apocentric 

 force than the quicker revolving air above it, and so is incapable of resisting the 

 inward pulsive tendency of the greater outer than inner pressure already mentioned. 

 Hence, while rushing round in its whirl, the air of that lamina must also be 

 flowing in centreward. 



Tie influx of air so arriving at the central region cannot remain there con- 

 tinually accumulating: it is not annihilated, and it certainly does not escape 

 downwards through the sea. There is no outlet for it except upwards, and, as a 

 rising central core, it departs from that place. This is one way of thinking out 

 some of the conditions of the complex set of actions under contemplation, but 

 there is much more yet to be considered. 



Hitherto, in the present paper, nothing has been said as to the cause or mode 

 of origin of the diminished barometric pressure which, during the existence of the 

 whirlwind, does actually exist in the central region. Often in writings on this 

 subject the notion has been set forth that the diminished pressure is caused by the 

 rapid gyratory motion of the whirling air ; but were we to accept that view, we 

 would have still to ask : ' How does the remarkably rapid whirling motion receive 

 its own origin ? ' The reply must be that the view so ottered is erroneous, and that 

 in general a diminished pressure existing at some particular region is the causft 



18b4. x t 



