148 
POPULAR SCIENCE REVIEW. 
smaller the bore of the tube the greater the correction, hence 
one advantage of having tubes of large bore, fths of an inch 
or more. 
When great nicety is required there is a fourth correction, 
namely, that for index error, to he taken into account ; but 
when the instrument is a good one, and is only read to hun- 
dredths of an inch, this correction is, or ought to be, too insig- 
nificant to demand notice. 
The time for taking observations of the barometer is 9 A.M. 
and 3 p.m., but when two readings a-day cannot be managed, it 
is best, as it is also most convenient, that the observer should 
confine himself to the forenoon hour. The height of the metal 
should be taken as from the summit of. the convexity of the 
curved surface due to capillarity, which is discussed above. 
The subject of Weather Prognostications as derivable from 
the study of the barometer is too large a one to be examined in 
a paper like the present; but this, however, is the less to be re- 
gretted, because Admiral Fitzroy’s Barometer and Weather 
Guide , published by the Board of Trade, lays before the reader, 
in a compact compass and cheap form, all, or nearly all, that can 
be said on this topic.* As to the Aneroid Barometer, I suppose 
I ought to say a few words. This instrument is rising in public 
estimation, and a first-class one is capable of doing good service 
to the traveller, to whom portability is an essential object ; but 
the stationary observer will not for one moment set it against a 
mercurial barometer, unless it be for domestic purposes. 
Temperature. — When a sudden extreme of temperature arises, 
be it of heat or be it of cold, everybody becomes an enthusiastic 
thermometrician , if I may be pardoned for coining such a work. 
It is to be regretted that this enthusiasm is usually so epheme- 
ral, because the laws of nature, as regards weather and tempera- 
tures, though no doubt in one sense obscure, are nevertheless 
full of interesting coincidences. 
To describe such a common instrument as a thermometer 
seems superfluous ; rather let us dwell on the way of using it. 
The most comprehensive method of determining the annual 
temperature of a place is by observing daily the maximum and 
minimum temperature of the air during every complete period 
of 24 hours, and taking the arithmetical mean of the two read- 
ings as ‘the temperature of the day. Thus, let us suppose that on 
Monday, April 1, 1867, at 9 a.m., the indices of registering ther- 
mometers showed that during the previous 24 hours the highest 
temperature of the air had been 62*5° and the lowest 40*7° 
From these data we conclude the mean temperature of the 
* This pamphlet does not of course go into the “forecast” system: that 
is an independent subject — at least the Admiral (who was the best judge) 
so treated it. 
