A Lunar Period in the Bates of Evajporation and Rainfall ? 107 
minimum at each end. For, after a wet day, as the air dries, the evapora- 
tion will frequently be slow at first and will increase more or less gradually 
afterwards. Also the approach of a rainy day will often enough be preceded 
by a moister air which will diminish the evaporation, unless an increased 
wind velocity counteracts the humidity. 
The mean hourly, unsmoothed, evaporation for the 530 rainy days omitted 
in making Table IV is given in Table V. 
Table V. 
Hourly Mean Evaporation on Rainy Bays, arrayiged according to the 
Lunar Day. 
Hour. 
Inch. 
Hour. 
Inch. 
Hour. 
Inch. 
I 
•0066 
X 
. -0052 
XIX 
. -0057 
II 
68 
XI 
49 
XX 
59 
III 
62 
XII 
51 
XXI 
59 
IV 
60 
XIII 
48 
XXII 
59 
V 
61 
XIV 
47 
XXIII 
60 
VI 
59 
XV 
49 
XXIV 
59 
VII 
59 
XVI 
49 
XXV 
59 
VIII 
56 
XVII 
52 
IX 
54 
XVIII 
53 
The values shown in Table V are surprising. There would be no appar- 
ent a priori reason why, since rain is to be expected at any hour of the lunar 
day, there should be more evaporation in any one hour than in another. 
Still less that an arrangement of the observed evaporation in any series of 
hours other than twenty-four, should give a fairly symmetrical curve with a 
definite minimum near the central ordinate. Moreover, with a range 
equivalent to 34 per cent, of the mean value. 
Table VI. 
Total Hourly Rainfall in 120 Months arranged according to the 
Lunar Day. 
Hour. 
Inch. 
Frequency. 
Hour. 
Inch. 
Frequency. 
Hour. 
Inch. 
Frequency 
I 
6-876 
123 ' 
X 
8-347 
138 
XIX 
. 7-638 
122 
II 
6-448 
124 
XI 
5-841 
142 
XX 
. 5-758 
122 
III 
9065 
131 
XII 
8-798 
. 143 
XXI 
. 6-288 
. 118 
IV 
8-960 
. 135 
XIII 
7-606 
143 
XXII 
. 4-188 
. 121 
V 
6-765 
141 
XIV 
8-796 
. 137 
XXIII 
. 6-305 
. 121 
VI 
9-366 
. 143 
XV 
5-770 
131 
XXIV 
. 7-055 
. 127 
VII 
8-919 
147 
XVI 
6-180 
. 131 
XXV 
. 6-342 
123 
VIII 
7-446 
143 
XVII 
5-370 
. 128 
IX 
5-721 
142 
XVIII 
7-481 
. 128 
Totals 
177-329 
3304 
Let us inquire, therefore, whether there has been an unequal distribution 
of rainfall in the whole 120 lunar months. Table VI gives the total rainfall, 
and the total number of hours of rain, deduced from the anxanometer 
