352 PROFESSOR FORBES ON THE CLIMATE OF EDINBURGH. 
daily range,” deduced from the average of all the daily extremes for twenty-nine 
years (1822-1850), during which self-registering thermometers were in use.* I 
shall make a few observations under each of these heads. 
(1.) On the inflections, or “periodic anomalies” of the annual curve, com- 
pared with the standard or computed values. 
65. Of these the most marked, indicated by the Curve and the Table last 
referred to, is an excess of temperature above the calculated or normal amount 
in the latter part of January and earlier part of February. This is a well-known 
and long recognised anomaly. In most European instances it affects materially 
the mean temperature of February, which is very commonly far too high, when 
contrasted with the general sweep of the annual curve. At Brussels, for example, 
M. Quetelet finds an excess in the temperature of February above the general 
curve of at least 1°5 Fahr. The same peculiarity may be noticed in the annual 
curves of London, Prague, St Petersburg, Vienna, and many other places, including 
even the Great St Bernard. If it is nearly insensible in the monthly means of 
Table XI. p. 346, this apparently arises from the anomaly occurring rather sooner 
at Edinburgh than in most other places, and therefore affecting the temperature 
of January fully as much as that of February. 
66. In connection with this anomaly, I may observe that it is apparently con- 
nected with an anomalous depression of the thermometer in the early part of 
January, of which it may be said to be the reaction. Although we have seen 
(Art. 57), that in the equalized curve of temperature the minimum is attained on 
the 17th January (and if two terms of the equation alone were used, it would be 
on the 22d), the average coldest day is very decidedly earlier in the month. 
This is well shown in both the detailed curve and the curve of five-day means. 
The 11th January is the average coldest day (Table [X.), and it is also the central 
point of an abnormal depression of temperature. The same thing is well marked 
in M. Quetelet’s curves, which coincide almost to a day with the preceding results.+ 
67. Of the other periodic inequalities of the annual curve we cannot speak 
with much confidence. Even after forty years’ observations, the casual fluctua- 
tions of daily temperature are far from being eliminated ; and for a period of ten, 
or even twenty years, very great uncertainty still remains, as may easily be con- 
cluded from a comparison of the numbers under each day in Table XII. for the four 
decennial periods. Perhaps the general depression of temperature in the month — 
of November, which is also traceable in the curves of Greenwich and Brussels, 
may be considered as a true periodic anomaly, at least in this part of the world. 
* The reductions of the daily range have been less scrupulously verified than most of the other 
computations contained in this paper, but any residual errors are hardly likely to affect sensibly the 
mean results. 
+ Since writing the above, I notice that M. Quetelet, at page 39 of his Memoir, expresses him- 
self as to this anomaly in almost the same terms that I have used. 
