HOURLY OBSERVATIONS OF AIR TEMPERATURE AND PRESSURE. 
635 
of P and p become identical; the negative sign in the above table signifies that the 
component has a negative maximnm value at midnight; and the positive sign that 
the component has a positive maximum value at the same hour. The values of p for 
the several components will be seen to be as follows :—For the first component 
jx-^ = 180°, or noon ; for the second = 0°, or midnight; for the third jUg = 0° for the 
summer, and p-g = 60° for the winter, the change taking place at the equinox when 
the component becomes zero ; for the fourth, in the lower latitudes, pg = 45° at all 
seasons, in the higher latitudes at the equinox = 45°, passing to 0° both in the 
winter and summer. 
For comparison with the above results, are given below the components of the 
diurnal temperature curve as actually observed at certain selected stations, varying in ' 
latitude, for the three seasons of the year. 
Table G. 
Mid-winter. 
Equinox. 
Mid-summer. 
Stations. 
Components. 
Components. 
Components. 
Pi 
Ps- 
Ps- 
Pi- 
Pi- 
P 2 - 
P.S- 
P.1- 
Pi- 
P 2 - 
P;r 
Pi- 
Singapore. 
Lat. 1° 15' N. 
-5-02 
+ 1-78 
-•28 
-■18 
-6-47 
+ 2-04 
+ •68 
-•48 
-4-62 
+ 1-54 
+ ■32 
-■51 
Hong Kong. 
Lat. 22T8' N. 
-2-39 
+ 1-01 
-•67 
-•81 
-1-98 
+ -70 
-■02 
-•12 
-1-86 
+ -63 
+ •09 
-•13 
L^ons. 
Lat. 45°46'N. 
-2'5l 
+ 1-01 
-•41 
+ •07 
-6-26 
+ 1-49 
+ •11 
-•31 
— 7^55 
+ -76 
+ •59 
+ •16 
Greenwicli. 
Lat. 51'“30' N. 
-l-6o 
4- -76 
-■33 
+ ■09 
■-4-67 
+ 1-34 
+ ■18 
-•23 
-7-71 
+ •56 
+ •61 
+ •21 
Fort Rae. 
Lat. 62°40' N. 
-1-09 
+ -67 
-•31 
+ 29 
-7-71 
+ 1-90 
+ •42 
-•50 
-6-01 
+ -59 
+ •19 
+ •06 
The signs indicate the sign of the components at midnight. 
It will hence be seen that while the intermittent curves representing the heating 
power of the Sun, under varying conditions of meridional altitude and length of day, 
are fairly well reproduced by combining the components of the first four orders having 
the values shown in Table F, these components bear a very close resemblance to 
those shown in Table G, obtained from the observed temperatures of places at the 
latitudes and seasons corresponding to those adopted in the computations from the 
intermittent curves. . 
The conclusion is unavoidable, that although l:»oth in the real and hypothetical cases 
the harmonic components, when combined, are truly representative of the peculiar 
4 M 2 
