United States Weather Maps for 1872-1874. 5 
stations for Sept., Oct., and Nov., 1872, these being the only 
months for which the observations have been fully published, 
and I confined the comparison to those stations which are north 
of lat. 35°. The total rainfall for a month at all the stations 
north of the parallel of 35° was determined for each of the 
three intervals into which the day is divided; these results were 
divided by the number of days in the month, giving the aver- 
age daily fall for each of these intervals; and these last results 
were divided by the number of hours in the corresponding inter- 
val, giving thus the average hourly rainfall for all the stations 
for these three portions of the day. The following is the final 
result : 
Hourly rainfall at the 8. 8. Stations. 
I—2 2—3 3—1 
September  0°244 inch 0.308 inch 0°288 inch 
October "174 "200 192 
November "155 "143 ‘180 
Mean 0°191 0°217 0°220 
_ These numbers indicate that the average rainfall for the sta- 
tions employed is nearly uniform throughout the day. If, 
then, there is a decided diurnal inequality in the rainfall at 
Philadelphia, it seems probable that for other parts o 
country the maximum takes place at a different hour, so that 
in taking the aggregate for a series of stations stretching across 
the continent, the total rainfall is nearly uniform for all hours 
of the day. 
The British Government has published the hourly observa- 
servations, 
Influence of rainfall upon the course of storms. 
In a former paper (this Jour., vol. viii, p. 4) I endeavored to 
show the connexion between the velocity of a storm’s progress 
and the extent of the rain-area on the eastern side of the storm. 
T have made a similar comparison of the observations of 1874, 
and now present the results. The number of cases suited to 
