Bartlett—The Cold-Waves of SoutJihCentral Wisconsin . 297 
whence the cold air to form the cold-wave is usually drawn, the 
resulting temperature drop at Madison, the wind direction 
prevailing during the cold wave; and the highest 
wind velocity for 5 minutes. An examination of this table 
reveals that, preceding a cold wave at Madison, a storm center 
usually overlies the Great Lakes, or the adjacent states, while 
an area of high barometric pressure is as a rule central in the 
Dakotas, Montana or western Canada (see Plate XX). The 
strength of the storm, as shown by the central isobar, may 
vary from 29.15 to 30.05 inches, with an average strength of 
29.65 inches. The central pressure of the High usually ranges 
from 30.2 to 30.9 inches, with an average of 30.6. The average 
difference in pressure between the High and Low centers thus 
amounts to 0.95 inches. The greatest difference was 1.5 inches 
on February 2, 1907, while the least was 0.55 inch on Hovember 
13, 1905. 
Preceding a cold-wave at Madison the temperature is gener¬ 
ally near or below zero in the Dakotas, though in all except the 
winter months it may be higher. During the past two years 
most of the cold-waves have been attended by northwest winds; 
although occasionally west or north winds prevail and on one 
date, Xovember 13, 1905, the temperature fall was accompanied 
by a northeast wind. This last wind direction is quite unusual 
during a cold wave as usually it brings a rise rather than a fall 
in temperature, or signifies a breaking up of any threatening 
cold-wave. It is to be noted, however, that this was in late 
fall rather than in winter and that the temperature did not 
need to fall extremely low to verify a cold-wave forecast. The 
table also shows that high to gale winds often accompany the 
cold-wave. 
It occasionally happens that, with the 7 a. m. weather map 
conditions favorable for a cold wave, no decided drop in tem¬ 
perature occurs. Such a case, with the resulting thermograph 
trace at Madison, is shown in Plate XXI. One would nat¬ 
urally expect much colder weather in southern Wisconsin fol¬ 
lowing such a map. Some disturbing influence, possibly the 
storm in Utah, appeared to force the Saskatchewan High north 
of Lake Superior on the following day so that the wind at Madi- 
