i8o IRature Studies in Bcrfcsbire. 



a shower make up to the north-east of us, and take 

 the wholly erratic and unusual course of moving 

 south-west, and almost west. In all these cases, the 

 nucleus of the storm was a mass of cumulus which 

 seemed to draw all the neighbouring clouds into the 

 warm vortex of its own ascending vapours, so that 

 the cloud grew by attraction, spread over a continu- 

 ally larger area, and, acquiring intensity and motion 

 with its growth, moved off at last, a full-fledged 

 storm, to fleal dampness and devastation far and 

 wide. And, in its development, the clouds seemed 

 almost to reach backward, to roll up in the rear of 

 their actual direction, and thus create a sort of an 

 eddy which caught whatever smaller clouds were 

 moving near and swept them into the main current. 

 Speaking of currents, I am reminded that nothing 

 about the great cloud drive has been more interest- 

 ing than to watch how it has revealed certain tracks 

 in the atmosphere, through which the clouds seem 

 to drift as naturally as waters between river-banks. 

 Are there channels in the air, in which the clouds 

 run by some atmospheric gravitation ? They cer- 

 tainly follow regular routes, and their course can be 

 predicted as accurately as the flow of the streams on 

 a watershed, or the currents in Long Island Sound. 

 Clouds may form anywhere in the heavens and 

 move in any direction in which the wind lists to 

 bear them. But just as soon as they begin to flow 

 together, like the trickling rills that run to the brooks 

 and the rivers, they seem to seek certain set direc- 



