142 PHILOSOPHY OF STORMS. 



the storm towards the south, and these forces not being ex- 

 actly known in quantity, we must wait till a patient induc- 

 tion from accumulated facts shall solve this most interesting 

 problem. 



Another highly interesting question can only be answered 

 by very numerous observations with the barometer. How 

 far is the snow and rain carried by the out-moving current 

 above, beyond the up-moving current in the middle of the 

 storm ? 



This distance will no doubt vary with the violence of the 

 storm. In a case of great violence, if the storm is quite nar- 

 row, the upward current in the middle may be so great that 

 the snow or rain may not be permitted to fall in the centre 

 of the up-moving current at all but be compelled to pass 

 outward above in all directions, and fall down in an annulus, 

 where the barometer may even be above the mean, and rise 

 during the fall. Something of this kind seems to have taken 

 place in the present storm, in the northern part of Pennsyl- 

 vania, extending from Sunbury and Silver Lake, even as far 

 down into the centre of the state as Bellefonte. For, at the 

 two former places the barometer did not fall at all, and at 

 the latter its fall was hardly sensible. At these places, there- 

 fore, it is highly probable, there was no upward current of 

 air, and consequently the snow which fell there, must have 

 been generated at a distance. How far this fall of snow 

 may have been, not only the cause of the irregularities at 

 Silver Lake and Meadville, which were mentioned before, 

 and of the very gentle winds about this region, but also of 

 the general tendency of the winds to move on the east and 

 west side of the storm towards a point south of the centre, 

 it is not necessary for me now to determine ; at present it is 

 sufficient to have pointed out this source of irregularity, 

 and leave it to future investigation to determine its exact 

 amount. 

 Another highly important question is suggested here how 



