176 PHILOSOPHY OF STORMS. 



in Upper Canada, north west of the boundary to which our 

 observations extend. It would be highly desirable to ob- 

 tain records of the weather from Upper Canada, which 

 alone can answer this highly important question. 



136. The uncertainty, however, as to the exact direction 

 in which this great storm moved, most fortunately for the 

 cause of science, does not apply to the course of the wind. 

 On this point there was a most wonderful and beautiful 

 uniformity. 



The documents abundantly show, that, for many hours 

 during the 26th, while a very great rain and snow were 

 falling in the region between the eastern coast of the United 

 States and the western parts of New York and Pennsylva- 

 nia, the wind on the coast, from Maryland to Maine, was 

 almost uniformly blowing'from the east or south east, and 

 at the same time, in the western parts of New York, Penn- 

 sylvania, and the middle of Virginia and Maryland, it was 

 from some western or north western direction. Whilst 

 these winds were thus blowing in on both sides towards a 

 central line, the barometer was greatly below the mean in 

 the middle space between the two winds ; and the more the 

 barometer fell, the more violently did the winds blow on 

 each side. 



In the middle, however, the winds were very irregular 

 along the whole length of the line from north to south ; in 

 some places they were north (64, 66), and in others (69, 72) 

 south, violent ; at others (43, &c.) calm, with the barometer 

 at or near the minimum. In one of these places of great 

 irregularity, there fell a very uncommon quantity of rain 

 6 inches, 4 inches, and 3.55 inches ; at Renssetaer, Hamil- 

 ton, and Oxford, in a region forty or fifty miles in extent, 

 and towards that region round about, the winds seemed to 

 incline. In a storm of such great length as the present one, 

 it is quite natural to expect that there would be some places 

 in the central line where much more rain would fall than 



