122 B. A, Gould—Diurnal Variation of Temperature. 
inequality comparatively slight, the first variable term, the 
period of which is twelve hours, acquires an overwhelming 
significance; and the residuals which it leaves are disposed of 
with comparative ease by the other terms of long period. But 
when the sun rises later the reverse must be the case. By 
means of the data above collected, provisional assumptions of 
the form of the diurnal curve may often be so made as to 
afford important assistance in deducing its true form, with great 
approximation, from a small number of observations. 
On the other hand, no argument is needed to show how un- 
justifiable is that frequent misuse of the general formula by 
ich the values deduced for certain terms from observations 
made during the day time or at inadequate intervals, are em- 
curve, give a close approximation to the true fo. 4 
daily variation, so far as this can be expressed without includ- 
