44 REPORT— 1840. 
termination), such as solar radiation, the proper heat of the earth, 
the proper heat of the atmosphere, the proper heat of space. 
(3) Supposing these preliminaries rightly assigned, next comes 
the arduous mathematical investigation of the way in which the 
unknown constants which represent these various energetic 
sources of modification, enter into the final expression of atmo- 
spheric temperature, or whatever be the particular problem under 
consideration. (4) For the determination of the constants 
(which are supposed to be reduced to the least number of inde- 
pendent elements), the analyst is further bound to mould his 
formule into such a shape as to insulate a certain number of the 
constants in such a way that they may be determined by direct 
physical experiment ; whence the others not so determinable 
may be inferred by the principle of exclusion, which assigns the 
difference between the sum of effects due to causes already esti- 
mated, and the observed effect, to a remaining cause (such as 
the temperature of space) which does not admit of such direct 
estimation. 
20. The duty of the experimentalist is thus clearly defined. 
Perhaps the most important observations which man can make 
for furthering the theory of the universe, are such as no general 
sagacity, no patient attention to mere facts as they are presented 
in the course of nature, could possibly have indicated. And it 
is the most satisfactory and encouraging proof that such a syn- 
thetic mode of treating the problem is not an injurious or an 
illusory one, when we find many concurrent observations set on 
foot in the way which theory has indicated, leading under vary- 
ing circumstances to a common result. Such instances are not 
uncommon, and will fall to be noticed hereafter*. 
21. It is plain, then, that there are three departments of sci- 
ence which must go hand in hand to perfect a mixed science 
like that of meteorology. First, the experimental philoso- 
pher must advise generally on the experimental axioms, and on 
the modifications which they are to receive in order to allow for 
causes which yet do not admit of nice numerical estimation ; 
he must further be fully persuaded that »o energetic cause of 
modification has been left out of account; for in so delicate an 
inquiry, where the principle of exclusion is to be acted upon, 
enormous errors would result from any oversight in this stage 
* Such, for instance, as the geometrical diminution of the range of animal 
temperature beneath the surface of the ground, as we descend in arithmetical 
progression : the wniform retardation of epoch in the same circumstances, and 
perhaps we might have added a few years ago, the agreement of several different 
methods of approximation to the actual temperature of space. See former Re- 
port, p. 203, and Mahlmann’s Translation, p. 14. 
