34 H. C. DANNEVIG. 



number of miles that the atmosphere has travelled in the 

 different directions. These automatically taken records 

 have been totalled up each month, and to that extent I 

 found the material conveniently prepared. It seems 

 customary in most meteorological work to attach greatest 

 importance to the duration of the "blow" (prevailing 

 winds) and the total mileage, which is proportionate to the 

 expended energy is frequently disregarded. That point of 

 view might be quite good in some instances, and the pre- 

 vailing wind is often gentle and harmless, but the domin- 

 ating wind which travels a greater number of miles in a 

 shorter time is also of consequence, on account of its greater 

 violence and powers of destruction. 



The method adopted in my treatment of the winds, and 

 also its justification may be seen from the following illus- 

 trations : If in the centre of an imaginary borderless sheet 

 of perfectly smooth ice were placed a 'feather-light' object 

 and all friction could be disregarded, the successive winds 

 from different points would carry this object about from 

 place to place. The distance traversed in each direction 

 would be equal to the mileage of each blow, and with a 

 complete record of this mileage, it is easy at any moment 

 to determine the whereabouts of the object. The final 

 resting place of the latter at the end of a certain 

 period would, in its relation to the starting point, be a 

 means of discovering the general drift of the atmosphere 

 during that time. The distance and direction would be 

 expressed respectively in miles and degrees, and it would 

 be easy to compare one period with another. 



The feather-light object represents the atmosphere itself 

 (or a particle), and the 'borderless' ice a sufficient section 

 of the surface of the globe; the starting point or centre is 

 in this case identical with the wind recorder, i.e., Sydney 

 Observatory. 



