X
PREFACE.
nature being of influence here and there, but so it had to be done
for making the more apparent what is analogous and what is not.
The elaboration is detailed in the introductory part of this
volume; the methods of deducing the means had to be analysed
with special care, the comparison of previous publications with the
numerical details of the original registers now in my hands showing
but too often a quite arbitrary combination of hours. Attention must
be drawn in this regard to the combination of Sunrise and Four o'clock
in the afternoon (SR. and 4h p.m.) by the plain mean of which a very
favourable daily mean is obtained from two hours only; if 10h a.m. is
added the variation also of pressure and wind becomes sufficiently well
defined. For showing the daily range together with the calculation,
or the final correction employed, the observation at sunrise and that
of an hour in the afternoon are given besides the mean resulting;
wherever the observers' registers allowed to do so Four o'clock p.m. was
chosen for the afternoon. The original materials contained for every
station a great deal more; the additional hours, though not used for
the calculation, afforded many a data for the better defining the
character of the climate.
Where the observations were not as detailed as usual, no deci-
mals are added to the resulting means; the decimal 0 • 5 when
dropped was not counted over. I was the more induced to adopt this
rule since experience had showm me that most observers when esti-
mating fractions of a degree have a slight tendency to read rather a
little too high. Brackets are added where either the materials were
incomplete in the registers, or where, as seen in the tabular arrange-
ment of the results, one year only existed for a month, whilst several
years could be combined for the other months.
The hourly variations in the daily and yearly period had to be
given only for some stations. Next to these tables the comparison of
the therrnometrical scales has still to be mentioned; I gave it with so
PREFACE.
nature being of influence here and there, but so it had to be done
for making the more apparent what is analogous and what is not.
The elaboration is detailed in the introductory part of this
volume; the methods of deducing the means had to be analysed
with special care, the comparison of previous publications with the
numerical details of the original registers now in my hands showing
but too often a quite arbitrary combination of hours. Attention must
be drawn in this regard to the combination of Sunrise and Four o'clock
in the afternoon (SR. and 4h p.m.) by the plain mean of which a very
favourable daily mean is obtained from two hours only; if 10h a.m. is
added the variation also of pressure and wind becomes sufficiently well
defined. For showing the daily range together with the calculation,
or the final correction employed, the observation at sunrise and that
of an hour in the afternoon are given besides the mean resulting;
wherever the observers' registers allowed to do so Four o'clock p.m. was
chosen for the afternoon. The original materials contained for every
station a great deal more; the additional hours, though not used for
the calculation, afforded many a data for the better defining the
character of the climate.
Where the observations were not as detailed as usual, no deci-
mals are added to the resulting means; the decimal 0 • 5 when
dropped was not counted over. I was the more induced to adopt this
rule since experience had showm me that most observers when esti-
mating fractions of a degree have a slight tendency to read rather a
little too high. Brackets are added where either the materials were
incomplete in the registers, or where, as seen in the tabular arrange-
ment of the results, one year only existed for a month, whilst several
years could be combined for the other months.
The hourly variations in the daily and yearly period had to be
given only for some stations. Next to these tables the comparison of
the therrnometrical scales has still to be mentioned; I gave it with so