Physical Significance of the Mean Diurnal Curve of Temijerature. 191 
January Temperature Variation. 
Hour Ending 
Normal Curve. 
Component Curves. 
A. 
B. 
c. 
t — t. 
t-t'. 
t-t'. 
t-t'. 
o 
0 
0 
T 
1 1 .O 
+ 1-^ 
+ 0-1 
+ 0-9 
+ 1-9 
TT 
1 1 .1 
+ 11 
+ 0-2 
+ 1-2 
+ 2-0 
TTT 
+ 0-4 
4-0-9 
+ 1-8 
1 n.Q 
+ 0-2 
+ 1-2 
( + 2-0) 
\J 
-|-U o 
+ 0-1 
+ 0-9 
+ 1-8 
VT 
— u o 
-0-1 
-10 
-2-0 
VTT 
— 0 u 
-4-0 
-5-0 
-5-9 
VTTT 
— tt -dt 
-3-L 
-4-0 
-5-4 
— o o 
-3-0 
-40 
-4-9 
V 
— 0 u 
? 
-3-0 
-4-1 
— A o 
-2-1 
-2-9 
-3-4 
1 -ft 
— 1 o 
-1-2 
-2-3 
-2-9 

— i ± 
-1-2 
-1-8 
-2-4 
XIV 
-0-4 
-1-1 
-0-3 
+ 0-3 
XV 
+ 0-1 
-0-9 
0-0 
+ 1-0 
XVI 
+ 0-9 
+ 0-1 
+ 0-9 
+ 1-9 
XVII 
+ 1-5 
+ 1-0 
+ 1=8 
XVIII 
+ 2-3 
+ 1-1 
+ 2-0 
+ 2-9 
XIX 
+ 4-0 
+ 3-1 
+ 4-0 
+ 5-0 
XX 
+ 3-1 
+ 2-1 
+ 2-9 
+ 4-0 
XXI 
+ 20 
+ 1-0 
+ 2-1 
+ 30 
XXII 
+ 1-5 
+ 0-1 
+ 0-9 
+ 1-9 
XXIII 
+ 14 
00 
+ 0-9 
+ 2-0 
Midnight ... 
+ 1-3 
0-0 
+ 1-0 
+ 1-9 
that indicate the tenths, and hence readings to whole degrees should be 
the most frequent."" Apart from this, however, it is not unlikely that what 
is called the mean diurnal curve of temperature is, for Kimberley, made up 
of at least three superimposed component curves of the same period — 
curves proper perhaps to various outstanding types of weather. They are 
given below in smoothed values of t -t', but rather in the way of inference 
from the results than of proof; and also, for comparison, the corresponding 
differences of the normal mean diurnal temperature curve. Of these three, 
marked A, B, and C respectively, the B curve resembles the normal 
curve ; the A curve is probably proper to damp weather ; while the C 
* At any rate this is the most important reason. 
