458 E. Loomis—Barometric Gradient in great storms. 
stant, but diminishes in approaching the center of the low 
area. This change in the angle of inclination is not simply an 
occasional occurrence, but is an invariable characteristic of 
great and violent storms. 
In my 2d and 8th meteorological papers I gave a table show- 
ing the mean distance between the isobars in the neighborhood 
of a storm’s center (the isobars being drawn at intervals of one- 
tenth inch) for all velocities of the wind from zero to 50 mules 
per hour, according to the United States Signal Service maps. 
If for these velocities we compute the corresponding gradients 
we shall find that the results do not accord well with the pre- 
n 
gradient amounts to 0-075 inch for a distance of 100 miles, and 
for a velocity of 50 miles per hour the gradient is only three 
times as great as for a velocity of zero. These anomalies re- 
sult partly from the mode in which the tables were constructed. 
A sheet of paper was ruled with 40 vertical columns and at 
the head of the columns were placed the numbers 0, 1, 2, 3, 
te. i 
within an area of low barometer the distance between the two 
adjacent isobars was measured, and the distance recorded under 
the corresponding velocity found at the top of one of the Nes 
tical colamns. Now it frequently happens that the velocity a 
the wind at some station is zero; but according to the ae 
Service maps the gradient is never zero. Indeed it frequently 
happens that at some station the velocity of the wind is zeT0 
when the gradient is very steep. The following examples” 
taken from the volume of observations for September, 1877, 
will show the truth of this statement. In each of these cases 
se ae 8 as 
