[ 58 ] 



X. On Thermal Radiation. By Prof. Magnus*. 



IN a previous "Note on the Constitution of the Sun"f I 

 communicated the results of some experiments on the 

 thermal radiating powers of sodium, lithium, potassium, &c. 

 These experiments were made so as to compare the radiation of 

 a platinum plate, heated in a Bunsen's burner, with that of an 

 exactly equal and similar plate covered with fused carbonate of 

 soda, of lithia, &c. The great radiating power of these substances 

 appeared to be due not improbably to the circumstance that, at 

 the high temperature to which they are exposed, small particles 

 continually detach themselves, which particles produce the in- 

 tense and peculiarly coloured light of the flame, At the moment 

 of detachment these particles may be regarded as so many points ; 

 so that it appeared possible that the radiation might be deter- 

 mined by them or by the roughness of the surfaces of the glow- 

 ing substances. It is well known, in fact, that rough metallic 

 surfaces do radiate more heat than smooth ones — either in con- 

 sequence of the points which they present, or, as MelloniJ and 

 Knoblauch § assert, because their density when rough is smaller 

 than when they are smooth. 



In order to determine whether the greater radiation of sodium 

 and of other similar substances does in reality depend upon the 

 detachment of such small particles, their radiating powers at 

 10Q° C. were compared with that of platinum. For this purpose 

 a small apparatus was expressly constructed; it was heated by 

 steam, and to its radiating surface, which had a diameter of 22 

 millimetres, several plates could successively be attached. The 

 result was, that even at 100° C. a platinum plate covered with 

 fused sodium was found to radiate much more heat than a pla- 

 tinum plate not so covered. It is difficult to obtain exact mea- 

 surements of the relative magnitudes of these radiations, since 

 the cohesion or tendency to the formation of drops is so strong 

 in carbonate of soda that the substance readily flows to one place 

 on the plate, and not only refuses to distribute itself uniformly 

 over the latter, but frequently recedes altogether from certain 

 patches. Nevertheless the experiments were sufficiently accu- 

 rate to prove that the radiating powers of platinum and sodium 

 have a similar relation to each other at 100°, to that which they 

 have at the temperature of a Bunsen's flame. 



It follows from this that the great radiating power of sodium, 



* From the Monatsberichte for August 11, 1864. 



t Monatsberichte for 1864, p. 166. Poggendorff's Annalen, vol. cxxi. 

 p. 510. Phil. Mag. S. 4. vol. xxvii. p. 376. 

 X Thermochrose, p. 90, Remark. 

 § Poggendorff's Annalen, vol. lxx. p. 340. 



