Tropical and Extra-tropical Cyclones. 17 



Fig. 8. 



Tolcano of Abayon. 



a. Before typhoon. 



b. After typhoon, and usually. 



These sketches were shown to the author as proofs of a descending- 

 current in front of a typhoon, but he is not altogether satisfied w r ith 

 the evidence. It is very difficult to decide on such a question with- 

 out personal observation. It has been suggested that the downward 

 curl of the smoke was merely the vortex eddy on the lee side of the 

 volcano, and that a similar phenomenon is obvious in any factory 

 chimney during a high wind. The original sketches, however, did 

 not the least convey that impression ; and there must be some reason 

 why the smoke does not curl that way with other winds. In England 

 there is a well-known saying, that when smoke falls down rain may 

 be expected. This is often true, and is undoubtedly due to some 

 peculiarity in the way of blowing before certain kinds of rain, and not 

 merely to the velocity of the wind. The author believes that the 

 downward curl of the smoke of Abayon is due to some peculiarity in 

 the wind preceding a typhoon, though he is not prepared at present to 

 assume the existence of a regular downward current in the hio-h 

 pressure area that precedes the typhoon. 



Some physicists have apparently doubted the existence of downward 

 currents at all, on the ground that the stream lines of an air current 

 must be tangential to the earth's surface, and that wind cannot slant 

 downwards in the ordinary sense of the word. This would be 

 perfectly true, if air were a perfect fluid, and wind blew like a con- 

 tinuous theoretical current ; but in practice wind blows irregularly, 



VOL. XLIII. - C 



