18 
HARRINGTON. 
published permit us to study its details only at San Jose, 
which is in this region and is, moreover, quite free from non¬ 
periodic disturbances of the barometer, and therefore from 
cyclonic action. The hourly observations available were 
for the three years 1889, 1890, and 1891. The combination 
of these observations by hours and months is given in Table 
VIII and represented graphically on Plate 4, where the 
hours (two by two) are represented by the horizontal lines, 
and the months by the vertical lines, the annual rainfall for 
the hour in that month in inches being placed at the inter¬ 
section of these lines. The culmination of the rainfall at 
from 4 to 6 in September and October is very manifest. In 
October more than half the rain falls in the three hours 
from 3 to 6 p. m., and two-thirds in the five hours from 1 to 
6 p. m. A similar statement may be made for the months, 
viz., at 4 p. m. more than half the rain falls in August, Sep¬ 
tember, and October. In January the little rainfall is in 
the morning hours, but generally speaking the rain does 
not begin here until noon, nor fall after 10 p. m., so that 
practically all the rainfall is confined to the afternoon hours 
and nearly all to the five hours from 1 p. m. to 6 p. m. 
The same order of facts is shown in the number of hours 
of rainfall for each month of the year, for the calculation of 
which observations are available at San Jose for ten years. 
The results are: 
January.. 
per year. 
February. 
U i t 
March... 
U C6 
April. 
. 16 
<< 
CC U 
May. 
. 74 
(C 
t c u 
June. 
. 66 
u 
11 u 
July. 
.. 65 
u 
u u 
August. 
. 89 
C i 
u cc 
September. 
. 77 
i c 
cc cc 
October. 
. 94 
u 
cc cc 
November. 
. 34 
u 
cc cc 
December. 
. 18 
C l 
cc cc 
Total. 
. 550 
< c 
ti u 
