THE 



LONDON, EDINBURGH, and DUBLIN 



PHILOSOPHICAL MAGAZINE 



AND 



JOURNAL OF SCIENCE. 



SUPPLEMENT to VOL. XLIV. FOURTH SERIES. 



LIX. On the Heat-conducting Power oj Iron and German Silver. 

 By H. Weber of Brunswick^. 

 [With a Plate.] 

 ^1^7 ITH the exception of the insufficient experiments of Peclet, 

 ▼ ▼ a determination of the heat-conducting power of differ-, 

 ent metals was first accomplished not long since, namely by F. 

 Neumann, Angstrom, and Forbes. The values found by them 

 exhibit not inconsiderable divergencies from one another, which 

 cannot be referred only to the diflference of the sorts of metal 

 employed. It appears that they are much rather to be in great 

 part attributed to the fact that ihe phenomena, from the obser- 

 vation of which the conducting-power was deduced, were very 

 complicated, whence great difficulties arose. Besides, it can be 

 proved that the theory made use of for the calculation of the 

 observations, strictly taken, has only an approximative value. 

 Indeed neither Fourier's nor Poisson's theory takes account of 

 the heat expended for the expansion, and of the alterations of 

 the specific heat and density with the temperature. Moreover 

 the assumption which forms the basis of Fourier's whole theory 

 has been called in question — viz. that the quantity of heat passing 

 in the unit of time through a plate which is in a stationary con- 

 dition is proportional 5o/e/?/ to the difference of temperatures oi (he 

 boundary surfaces ; probably it is dependent also upon the abso- 

 lute temperature of the plate. Lastly, almost greater objections 

 can be raised against the employment of Newton's law of cooling. 

 Nevertheless the results derived from accurately conducted ob- 

 servations might still possess an approximative value for many 

 important purposes, in particular for many problems most inti- 

 mately connected with the mechanical theory of heat; for, to 

 obtain only a nearer insight into the phenomena of heat, such 

 approximative results are often quite indispensable. 



* Translated from a separate copy, communicated by the Author, from 

 Poggendorff's Annalen, vol. cxlvi. p. 257. 



Phil. Mag. S. 4. No. 396. Suppl. Vol. 44. 2 I 



