106 Transactions of the Royal Society of South Africa. 
Excepting that the values in the second half of the lunar day are some- 
what greater than those in the first, there is nothing in Table III to 
suggest outside influences. In particular there is nothing in the table that 
would indicate a lunar influence. We can, however, as it happens, show 
how such variation as there is in the hourly rate arises : 
First let us compare the evaporation on dry days with that on wet days. 
The evaporation, smoothed in threes, on dry days is given in Table IV : 
Table IY. 
Hourly Mean Evaporation on Dry Days, arranged according to the 
Lunar Day. 
Hour. 
Inch. 
Hour. 
Inch, 
Hour. 
Inch. 
I 
•0074 
X 
. -0075 
XIX 
. -0076 
II 
73 
XI 
75 
XX 
75 
III 
73 
XII 
76 
XXI 
75 
IV 
73 
XIII 
76 
XXII 
75 
V 
74 
XIV 
76 
XXIII 
75 
VI 
74 
XV 
76 
XXIV 
75 
VII 
74 
XVI 
76 
XXV 
74 
VIII 
74 
XVII 
76 
IX 
74 
XVIII 
76 
In Table IV, by " dry days " is meant all those days upon which there 
was no rain at all, or none sufficient to materially disturb the ordinary run 
of the evaporation. In a few instances, however, when there have been 
short passing showers yielding no great quantity of rain, but where the 
weather has made itself felt, a small plus correction has been necessary to 
the observed quantities of evaporation, so as to bring them into line with 
the normal diurnal curves of evaporation for the months in which they 
occur. This correction has in all cases been made with due caution. 
Table IV gives a definite curve with a minimum at the second and third 
hours and a maximum at the fifteenth, with a mean daily range, M — -m, 
of '0003 inch, corresponding to a total range, M — of nearly an inch in 
the whole 2878 days accounted for in the table. If we could regard the 
numbers shown in Table IV as a definite lunar effect, we should have to 
admit an effect so great as to give a range of upwards of 4 per cent, of the 
hourly mean rate of evaporation. Which is to say that the evaporation 
is enlarged or diminished by 2 per cent., according as the moon near the 
meridian is respectively above or below the horizon. This amount of devia- 
tion is undoubtedly greater than could have been reasonably expected. 
Upon reflection, however, it will be obvious that no matter what the 
arrangement (lunar, solar, or what not) of the hourly quantities of evapora- 
tion, the effect of eliminating wet days from the account will be to give a 
curve for the dry days with a maximum near the central ordinate, and a 
