—_—_—- 
PROFESSOR FORBES ON THE CLIMATE OF EDINBURGH. 339 
determinations, and where any discrepancy appeared, the cause was carefully 
sought out. 
26. Farther, the accuracy of the copy from the original registers, and of each 
part of the calculations, was tested in a great many instances (taken at random) 
by Mr Batrour Stewart (now of Kew Observatory), in whose exactness I have 
the utmost confidence. Although some errors were thus detected, they were not 
sufficiently numerous or important to shake my confidence in the care with 
which the reductions had been made, but rather indeed confirmed it. 
27. I ought to add, that the observations were all made with instruments 
manufactured by Mr Avie himself, and certainly equal in accuracy to any then 
constructed. I have not been able to ascertain when or how often the thermo- 
meters were renewed. We must therefore rely on their general exactness, with- 
out attempting a correction for index error. 
28. The thermometers were commonly read to whole degrees merely. Conse- 
quently, in taking the mean temperature of the day from two observations, 
we have not usually to do with any fractions differing from 0°°5. Tosave room, 
I have adopted, in the MS. calculations, a plan which may be usefully employed 
in all such cases. I have written such a temperature as 52°5 thus, 52°, and have 
allowed for these half degrees in making the summations. Where quarter degrees 
occurred, the nearest whole degree was written, and this little error is evidently 
-self-compensatory in the long-run. 
Srct. 3. On the Annual and Monthly Mean Temperatures of Edinburgh for Fifty-six years. 
29. A peculiar interest attaches to long-continued series of thermometrical 
observations in one locality. When made by one observer, and with comparable 
instruments, the interest is greatly heightened. We then see clearly the great 
extent of what M. Dove calls the “ Non-periodic fluctuations” of temperature, 
and the great length of time required to attain any certainty, even as to the 
precise mean temperature of a place, still more the peculiarities of the inflection 
of the diurnal and annual curves of temperature. 
30. A slight inspection of M. Dove’s large collection of monthly average 
temperatures,* will show in how few instances anything like fifty years of con- 
tinuous thermometrical observations can be depended on. The interesting de- 
ductions by M. QuetEeLeT from his own perfectly comparable observations at 
Brussels for only twenty years, show how much may be done to reduce the appa- 
rently lawless changes of climate to order, by the judicious combination of long 
-averages.t I have kept M. QuETELET’s reductions in view, in the course of those 
which follow. 
* See his Memoirs in the Berlin Transactions. In the Index to these which I have made 
and printed in the Edinburgh Transactions, vol. xxii. p. 75, I have, indicated by an asterisk the 
stations where the observations have been most continuous. 
t Mem. de l’Acad. de Bruxelles, tom. xxviii. Paper read June 1853. 
