SUPPLEMENTARY REPORT ON METEOROLOGY. 49 
diminishes with the Jength of the conducting wire. Some ad- 
vantage is undoubtedly gained by using two or three pairs, 
acting simultaneously, instead of one. Wherever distant com- 
parisons are intended, it is of great consequence to have a 
standard thermo-electric combination, by which any change in 
the sensibility of the galvanometer may be tested. 
32. A method of determining the temperature and pressure 
of the atmosphere depending on very different principles, has 
lately been proposed by M. Arago*. It depends on the known 
optical principle, that if a pencil of light radiating from a small 
and distant source be divided into two parts, and one of these 
be in the least degree retarded more than the other, coloured 
fringes will appear when the two pencils are re-united and suf- 
fered to fall upon a screen or received on an eye-piece. If by any 
arrangement, whether of reflexion or refraction, these bands 
have been already formed, their breadth and aspect will be 
changed by such a retardation of either component pencil. 
Now air rarefied, whether by heat, diminished pressure, or the 
intermixture of vapour, acts less energetically in retarding light ; 
consequently a tube with glass ends, filled with such modified 
air made to transmit one pencil, whilst the other passes through 
a precisely similar tube of standard air, will exhibit by the pro- 
duction of movement of the fringes very minute changes in its 
physical condition. It remains to be shown, however, by what 
mechanical adaptations M. Arago proposes to make this delicate 
experiment susceptible of general application. This he pro- 
poses to point out in a future memoirf. 
33. The comparison of ordinary thermometers with standards 
is often a matter of great importance, and too little attended to. 
The increasing demands of science require a proportionable in- 
crease in the consistency of instrumental indications; not so 
much indeed for ascertaining the temperature, as in many other 
experiments connected with the temperature of the ground, 
rivers and hot springs. Travellers should seize every oppor- 
tunity to verify the freezing points of their instruments, and this 
is an easy matter ; but to compare thermometers at temperatures 
above 100° F., is a practical problem of far greater difficulty than 
is commonly imagined. Plunging them together in laboratory 
vessels of hot water is a most unsatisfactory process, even if the 
sensibilities of the instruments be pretty equal. If they are 
very unlike, it is all but impossible, although the correction for 
the gradual cooling of the medium becomes then a pretty mathe- 
matical problem{. Where the investigation is an important one, 
* Comptes Rendus (1840). + Ibid. 
t Fourier, Théorie de la Chaleur, p.357, and Kelland’s Theory of Heat, p. 84, 
1840, E 
