1900-1901.] Prof. Knott on Solar Radiation. 
The greatest positive value of this is when 
305 
27 rt 7T 
-Y+I=l 
and the least positive value or greatest negative value is when 
2t rt 5tt 3tt 
— +q = -£ or . 
The times corresponding to these values are - 0-0307 and + 0’4693 
expressed in fractions of a year and reckoning from the middle 
of September, that is, about the beginning of September and the 
beginning of March. 
Hence there is more heat accumulated within the Calton Hill 
rock in the month of September than in the month of March by 
an amount equal to 
J_ 2c Y (,/ + p) 
J2 p' 2 +P 2 
cYJ2 
P 
approximately 
— 2000 nearly (Fahr. degree). 
= 1111 (Cent, degree). 
A better estimate may, however, be made from the temperature 
observations themselves if we first of all calculate the values at the 
surface. This requires us to work out the successive harmonics in 
the same way in which the first has been treated. The results 
for the second harmonic are as follows. The aim being to express 
the four harmonic terms in the form 
Ve 
q' x cos 
/ 47 rt 
\T~ 
r -qx + e 
the three values obtained for q were 0-00659, 0-00592, 0-00497, 
and the values of q and e worked out from the four-phase rela- 
tions by the method of least squares were 0-00515 and 1*84. 
These give 1’656 as the mean value of the amplitude of the tem- 
perature variation at the surface. 
The comparative smallness of the amplitudes of the third and 
fourth harmonics, and the shortness of the period of the fifth 
harmonic, render it quite unnecessary for these to be taken into 
account. The two harmonic expressions for the surface varia- 
VOL. XXIII. 
U 
