6 E. Loomis— United States Weather Maps. 
hours, it is presumed that the correspondence would have been 
still closer. 
The connection here discovered between the progress of storms 
and the extent of the rain area cannot be regarded as accidental, 
and it is not difficult to discover, at least in part, the origin of 
this connection. The fall of rain, that is, the precipitation of 
the vapor of the atmosphere, is generally accompanied by a fall 
of the barometer. According to the theory advocated by the 
la r. Espy, when the vapor of the atmosphere is condensed, 
its latent heat is liberated, which raises the temperature of the 
surrounding air, causing it to expand and flow off laterally in 
all directions in the upper regions of the atmosphere, thus 
causing a diminished pressure over the region of precipitation, 
. 
and an increased pressure on all sides beyond the area of the 
rain. 
The progress of the storm eastward is not due wholly to a 
drifting, resulting from the influence of an upper current of the 
atmosphere from the west, but the storm works its own wa 
eastward in consequence of the greater precipitation on the 
eastern side of the storm. Thus. the barometer is continually 
falling on the east side of the storm and rising on the west side, 
in consequence of the flowing in of colder air upon that side, as 
will be more fully shown on a subsequent page. 
Influence of the wind’s velocity upon the progress of storms. 
In order to determine whether there is any connection be- 
tween the velocity of the storm’s progress and the velocity of 
the wind upon the different sides of a storm, I selected all those 
cases in which a storm center was so situated that the velocity 
of the wind was given ata considerable number of stations both 
upon the east and west sides of the center. I then took two 
knitting needles and soldered them together at right angles so 
as to form a Greek cross. Then placing this cross upon one of 
the weather maps over a storm center, with the wires point- 
ing northeast and southwest, the area surrounding the storm 
center was divided into four quadrants, which I designate as 
the north, south, east and west quadrants. The average veloc- 
ity of the wind for the stations of observation in the different 
quadrants was then determined, including all stations within the 
influence of this storm center. The isobar 29°90 was generally 
taken as the limit of the storm, but sometimes it was necessary 
to reject observations within this distance when they were 
clearly under the influence of another storm center. Only 
79 cases were found suitable for this comparison. In each of 
these cases the average velocity of the wind was determined 
for the east and west quadrants, and generally also for the south 
quadrant; but in a majority of the cases no observations could 
i i il) ae 
