SUPPLEMENTARY REPORT ON METEOROLOGY. 47 
Magnus* revived by Walferdin, and his instruments are usually 
employed in Parist. The overflowing principle has been used 
(in the same way as in Wollaston’s thermo-metrical barometer) 
for constructing a thermometer sensible to small change of ani- 
mal heat by Dr. Marshall Hallt. Jurgenson, the chronometer- 
maker of Copenhagen, has proposed a clock for estimating the 
mean temperature of limited periods, in which the compensation 
shall be inverted, and the natural effect of temperature on the 
rate of an uncompensated clock exaggerated§. 
27. Several other methods or new applications of known 
principles have been proposed for the measure of temperature 
besides that of expansion commonly employed. The extensive 
introduction of the thermo-electric pile of Nobiliand Melloni, in 
delicate researches on radiant heat, has suggested the use of the 
same principle in other cases. Indeed it seems peculiarly 
adapted for ascertaining the condition of inaccessible points in 
respect of temperature. M. Peltier, of Paris, I believe, first pro- 
posed the application of long wires of heterogeneous metals to 
ascertain at a distance the temperature of their point of junction, 
however distant. The method is briefly this: suppose two 
wires of copper and iron respectively, and of equal lengths, laid 
side by side and soldered at each end. Let one of the wires be 
cut, and a galvanometer introduced into the open circuit thus 
made. It is well known that the galvanometer needle will not 
stand at zero under these circumstances, unless both solderings 
have a common temperature. As the temperature of one ex- 
ceeds or falls below that of the other, the index will move in one 
or other direction, and in a well-constructed instrument the de- 
viations will be almost exactly as the variations of temperature ||. 
Thus, one junction of the wires being plunged in water of known 
temperature, and the positive or negative deviation of the galva- 
nometer being known and converted into thermometric degrees, 
the temperature of the other soldering (which may be inaccess- 
ible and removed to a considerable distance) becomes known. 
This very ingenious arrangement was shown to me in June, 
1835, by M. Peltier, whose acquaintance I owe to the attention 
of M. Elie de Beaumont. 
28. The application of the thermo-electro-magnetic princi- 
ple becomes easy in very many cases. It has been proposed for 
pyrometers by M. Pouillet, who has compared the march of such 
* First Report, p. 209. + Comptes Rendus (Paris), ii. 505, 619. 
{ British Association Reports. 
§ Comptes Rendus, iii. 142. Compare First Report, p. 213. 
|| I am aware that this position has been controverted. I am satisfied, how- 
ever, of its general truth from careful experiments. 
