SUPPLEMENTARY REPORT ON METEOROLOGY. 145 
(1). To furnish standards of comparison ; 
(2). To establish the laws of phenomena ; 
(3). To fix secular, or normal data. 
The second of these determinations may be made without the 
first and third, and conversely ; just as in magnetism one set of 
instruments serve to measure variations of elements, which yet 
are incapable of establishing the fundamental values of the ele- 
ments themselves. 
322. In respect to the First point, the instruments must not 
only be originally good, but they must be preserved in constant 
repair,—a matter requiring perpetual revision in the case of 
meteorology. Access, under due regulation, should be permit- 
ted to instrument makers and observers to have their instruments 
compared with the standards*. That the instruments be abso- 
lutely as well as relatively correct, is evidently essential to the 
determination of the mean pressure at the level of the seat, the 
actual state of climate with respect to temperature, &c. 
323. Secondly.—Thelaws of phenomena can usually be only 
made out by extended and minute observation, incompatible, 
generally speaking, with private research. Laws of a certain de- 
gree of generality have commonly a pretty wide domination, even 
in a science apparently so capricious as meteorology. The in- 
stance we have mentioned of the “‘ homonymous hours” (37.), 
representing the mean temperature of the day, is one of the kind: 
the constancy of the interval between the hours of mean tempe- 
rature is another: the mean of diurnal extremes being nearly 
the mean temperature of the day is a third : the nearly coincident 
hours of diurnal barometric variation is another :—these, and 
such laws deduced from sufficient data, clear the way for indi- 
vidual exertion, and indeed form the only basis for really useful 
efforts of the kind. A very small number of observatories 
(comparatively), perhaps three or four in the extent of the Bri- 
tish islands, would be sufficient to supply these important data. 
The observations must be made every two hours at least, for 
with less than this the diurnal curve could not be properly drawn. 
But such laborious observations would not require to be indefi- 
nitely continued. Once place the meteorological observations on 
a proper footing, by ten, or perhaps twenty years’ observation of 
this kind, and the great difficulty is overcome: there is not a 
chance of secular changes sensibly affecting these laws; once 
established, they are like the laws of the solar system. 
* As is done, for instance, at Paris. 
+ The specific gravity of the mercury should be actually ascertained, and the 
pressure might be stated in the corresponding height of a column of distilled 
water at a fixed temperature. 
1840. L 
