450 E. Loomis—Barometric Gradient in great storms. 
a center of high pressure gives rise to a centrifugal force, its 
effect must be to diminish the gradient. When we go from | 
a center of low pressure to a center of high pressure, the second 
term of the formula must therefore change its sign on passing 
the line of mean pressure. This must cause an abrupt change 
in the computed values of the gradient, at a point where we 
pass from the low area to the high area. Such a change will 
be noticed in the values of G’ in table L But the observed 
values of the gradient show no such change. The differences 
between the successive values of the gradient given in column 
third of tables I and II, show no abrupt change marking the 
passage over the line of mean-pressure. We therefore conclude 
that by the circulation of the air around a center of high pres- 
sure no appreciable centrifugal force is developed. e also 
my preceding papers, e. g., paper ninth, storm of April 12, 
1874; paper twelth, storm of Jan. 15, 1877, and paper thir. 
a 
Over the Atlantic Ocean for pressures less than 760, the 
average gradient computed by the formula is only 77 per cent 
of the observed gradient; and for pressures greater than 760™" 
the average gradient computed by the formula (omitting the 
last term) is only 78 per cent of the observed gradient. If we 
table I, and if we multiply the gradients computed for a high 
area by 1:36 we obtain the last six numbers in column thir: 
teenth. We perceive that the differences between the numbers 
in columns G and G’” are not very great. 
ver the United States, for pressures less than 30 inches, the 
average gradient computed by the formula is only 61 per cent 
of the observed gradient, and for pressures greater than thirty 
inches, the average gradient computed by the formula (omitting 
the last term), is also about 61 per cent of the observed gra 
