178 ON TWO STORMS EXPERIENCED THROUGHOUT THE UNITED STATES, 
rature and diminished pressure within the United States, and we can only account for 
the long delay of this inward rush, by supposing that a similar state of things prevailed 
far to the north of the United States. To know the full extent of this storm, we need 
observations far more extensive than I have been able to obtain. The storm is now sure 
to be contracted within narrower limits. On the morning of the fourth, the rain was very 
much reduced, and two very distinct centres of diminished pressure are shown, one near 
Boston and the other near St. Louis, Near Boston the pressure is at a minimum, and 
the temperature at a maximum, and there is a tendency of the winds inward, with a dis- 
position to circulate against the sun. Thus in the vicinity of the North river, the winds 
are north; along the Sound they are south-west, and at Gardiner in Maine, they are 
north-east. In the vicinity of St. Louis the same fact is observable. At sunset of the 
fourth, some indications of subordinate centres of action may be noticed near New York 
and in New Hampshire; but throughout almost the rest of the United States, the winds 
are plainly governed by one centre near the middle of Lake Erie. Here the pressure is 
least, and the winds, taken as a whole, plainly circulate around this centre, with conside- 
rable tendency inward. Thus on the north side the winds are north-east ; on the west 
side, generally north-west; on the south side, west and south-west ; and on the east side, 
generally east, except towards New England, where there are to be seen other centres of 
action. Here, in the centre of this ellipse, was formed a violent tornado, exhibiting fea- 
tures very similar to those already described ; namely, an inward motion of the air, and 
a circulation against the sun, The cause of this tornado is now obvious. Although its 
effects were mainly local, its origin was far otherwise. The centre of this vacuum 
travelled north 62° 17’ east, eight hundred and sixty miles in twenty-four hours, equal to 
thirty-six miles per hour. The tornado in question travelled north 33° 30’ east, and if we 
suppose it to have kept pace with the vacuum, the velocity of the tornado must have 
been forty-one statute miles per hour, a result a little greater than I have given in my 
published account of the tornado, derived, however, from a much less number of 
observations. 
On the morning of the fifth, the storm was driven into the north-east corner of the 
United States, but exhibited features very similar to those of the previous evening. 
IV, OSCILLATION OF THE BAROMETER. 
Local changes in the density of the atmosphere seem to be the chief cause of the 
oscillations of the barometer; nevertheless these oscillations are propagated by the laws 
of waves, and are felt much beyond the limits of the original disturbing cause. As this 
is a very important principle, I propose to adduce conclusive evidence of it. This prin- 
ciple, then, is shown by the fact, that the barometric depression, of which the storm is the 
centre, often extends beyond the limits of the storm. Thus February 3 and 4, there wag 
considerable depression of the barometer south of the region of rain, or even cloud, This 
phenomenon must be still more striking in higher latitudes. Thus at Melville island, 
Captain Parry found the sky generally clear. Though it was sometimes overcast or 
obscured by a slight general haziness, there were no separate clouds. Some well de- 
fined clouds appeared on the sixteenth of April, which were nearly the first that had 
