1900-1901.] Prof. Knott on Solar Radiation. 
301 
in the same time. We know, moreover, that all the incident solar 
energy cannot be absorbed by the water, but that a considerable 
fraction is reflected or escapes in other ways. It therefore seems 
impossible to explain the afternoon temperature excess down to 
these depths in the Mediterranean as a result of direct solar 
radiation. The only other way out of the difficulty is to suppose 
that there is some considerable error in one or other of the sets of 
experimentally ascertained facts on which the present discussion is 
based. To make the facts compatible we should have either to 
diminish by at least one half the temperature differences observed 
by the officers and crew of the Pola , or greatly to increase the 
value of the solar constant. I do not think that the broad re- 
sults obtained by Langley can be seriously called in question, 
or that there is any ground for believing that the true value of 
the solar constant can be much greater than the value estimated 
by him. 
A careful study of Langley’s measurements and reductions leaves 
on the mind little doubt as to the main accuracy of his conclusions, 
which differ from the conclusions of previous investigators by 
assigning a somewhat greater value to the solar constant. A very 
careful scrutiny of the conditions under which the Pola observa- 
tions were obtained and the methods employed, supplemented by 
similar series of observations carried out in wide oceans, might 
determine how far the results were affected by purely local con- 
ditions. At present it seems to be impossible to suggest any 
satisfactory explanation of the extraordinary magnitude of the 
depth to which the daily solar radiation apparently penetrates in 
the Mediterranean Sea. 
It has been long known that the solar radiation penetrates to a 
comparatively small depth in the rocky material of the earth. In 
1837 Professor Forbes began a valuable series of observations of 
temperature at various depths in the rock of the Calton Hill, Edin- 
burgh ; and the main conclusions from these may be found in several 
of our modern text-books ( e.g . Tait’s Heat). Thus the conductivity 
of the rock is easily calculated by methods furnished by Fourier 
in his classical work Theorie de la Chaleur (1822). From this, 
in combination with the observed rate of increase of temperature 
with depth, an estimate may be made as to the amount of heat 
