8 E. Loomis— United States Weather Maps. 
the years 1872 and 1873. On the other hand, an increase in 
the velocity of the wind in the eastern quadrant tends to pro- 
duce a greater precipitation on the eastern side of the storm's 
center; that is, tends to push the storm’s center eastward, or 
increase the velocity of its progress. 
Does the velocity of a storm's progress depend upon the rate at 
which the barometer falls when the storm is approaching, or upon 
the rate at which the barometer rises when the storm has passed ? 
In order to answer these questions, I compared the height of 
the barometer at the center of each storm with its height at the 
same place 24 hours before, and also 24 hours afterward. I 
thus obtained the fall of the barometer for 24 hours before the 
middle of the storm, and its rise for 24 hours after the middle. 
These numbers were incorporated in the same table which 
showed the velocity of the storm’s progress for each month. 
The observations were then divided into two classes, one con- 
taining those cases in which the velocity of the storm’s pro- 
cases in which the velocity was less than the mean. It was 
found that the average fall of the barometer in front of the 
storm was nearly the same for both cases; but the average rise 
of the barometer for 24 hours in the rear of the storm was sen- 
sibly greatest in those cases when the velocity of progress was 
eatest. The averages for the two years indicated that when, 
after the center of the storm has passed, the barometer rises 20 
per cent more rapidly than usual, the storm center advances 
seven miles per hour more rapidly than the mean; but when, 
after the storm, the barometer rises 20 per cent less rapidly 
than usual, the storm center advances seven miles per hour less 
rapidly than the mean. 
By the same method of comparison it was ascertained that 
the velocity with which a storm advances is independent of the 
amount of the barometric depression at the center of the storm ; 
and it is not sensibly affected by the circumstance whether the 
ir the center of the storm is increasing or diminishing. 
y the aid of the preceding principles, when we have given 
a weather map showing the position of a storm center for a cer- 
tain hour, it seems possible to predict with considerable confi- 
dence where the storm center will be at the end of 24 hours. 
hold true for the average of a e number of examples, 
numerous and striking exceptions will be found when we at- 
tempt to ap rules to particular cases. It is evident 
"pote variety of circumstances, and all of these circumstances 
ve not been considered in the preceding deductions. 
Sere ene "a 
