November ii, 1883.] 



SCIENCE. 



(ill 



or forty feet to one side, the wind may not be 

 strong enough to brush along tiie sand, and a 

 few hundred feet away it may not be percepti- 

 ble ; but at the centre it makes a distinct rush- 

 ing or roaring sound, and carries liglit objects 

 upwards, sometimes to a height of several tliou- 

 sand feet. This increase of velocity of the sur- 

 face indraught toward the point of its upward 

 escape is a general feature of the motion of a 

 mass of free particles along a path of varying 

 width : the narrower the jiath, the faster the 

 motion. The same increase is seen in the 

 growing velocity of a stream running out of 

 a lake, so beautifully shown where the Rhone 



those of the tornado, as will be sliown far- 

 ther on. 



The second characteristic feature of the 

 wind's motion gives name to the storm. A 

 whirl must necessarily be formed when the air 

 moves inwards from all sides towards a centre, 

 for tiie indraughts will surely fail to follow pre- 

 cisely radial lines. Their aim will be a little 

 ine.Kact; and, as thev pass to one side or the 

 other of the centre, a turning must begin in a 

 direction determined by the strongest current. 

 This, once begun, is maintained by the centrifu- 

 gal force that arises from it ; and the size of 

 the central whirl will then depend on the bal- 



KlG. 3. {Taken IV 



Hows from Lake Geneva, or, more simplj' and 

 prosaically, in the running of water from a tub 

 1)V the escape-pipe. In the case of a desert- 

 whirl, the central wind is held by friction with 

 the surface sands much below the velocity it 

 might attain ; for it must be remembered that 

 these whirls are supplied by a comparativelj' 

 thin layer of superheated air next to tlie 

 ground, often not more than four or five feet 

 thick. Tiic restraint of friction on such a 

 layer will be verj' considerable, and its motion 

 can seldom reach a disastrous strength. It 

 is probable that in the desert sand-storms, 

 which are descril>ed as overwhelming caravans, 

 there is a much thicker mass of air in ac- 

 tivity, and the conditions of motion approach 



ance between the centripetal and centrifugal 

 forces. In ascending at the centre, the wind 

 follows an upward spiral course, like the thread 

 (if a screw of steep pitch, with a diameter of 

 five to twenty feet. The direction of turning 

 is indilferently one w.ay or the other, accord- 

 ing to the side on w'hich the indraught Iiappens 

 to pass the centre. The height to which the 

 whirling column rises will be determined by its 

 mi.xture with the adjoining air, and consequent 

 cooling until its temperature is that of indiffer- 

 ent oquililirium ; and at this elevation the cur- 

 rent will turn and spread laterally to make 

 room for that which follows. Such a whirl will 

 continue as long as its cause lasts ; that Is, as 

 long as it is supplied with warm air at the base. 



