178 J.D. Everett on Reducing Observations of Temperature. 
The influence of the term ¢,, in modifying the symmetrical 
curve y=t,+¢, for Greenwich, may be thus described:—It re- 
tards the vernal intersection, and the maximum; hastens the 
autumnal intersection, and the minimum; intensifies the maxi- 
mum, and moderates the minimum. Hence the temperature 
rises higher above the mean than it falls below; but the number 
of days below the mean is greater than the number above. 
These features of climate will Sigs | be found to prevail gen- 
erally in extra-tropical latitudes; and they constitute the most 
marked departures of the curve of temperature from symme- 
try, the terms after ¢, being comparatively insignificant. In so 
far as they are common to all places, they are unimportant in 
the comparison of climates, which was the object proposed in 
my former paper; and, at all events, the first elements that claim 
attention are those in which the actual curve agrees with the 
symmetrical curve, that is to say, the elements of mean tempera- 
ture, range, and general earliness, as represented by the three 
cons 
Of course, three numbers cannot express every feature of the 
curve of temperature at a place, for it is in the nature of things 
impossible that three numbers should express more than three 
independent elements. If the constants A, A, , are cor- 
rect as measures of the three leading elements, it is all that we 
can expect of them; and if a closer approximation to the curve 
of temperature is desired, it can readily be obtained by com- 
puting the constants in one or more of the succeeding terms. 
The subjoined diagram will give a clearer impression of the 
relation between the symmetrical and the actual curve than 
The = 
above either of the curves are together equal to those which lie 
below the same. 
