Prof. G. Magnus on Surface Radiation of Heat. 445 



character of its surface, and indirectly on its chemical durability; 

 but 



(2) It depends also on the more or less non-conducting cha- 

 racter of its internal mass, and in that respect is inversely pro- 

 portional to the amount of soda in the glass. 



(3) In the case of kinds of glass whose surfaces are not of a 

 peculiarly unfavourable nature, the latter circumstance has a far 

 greater influence than the former on the serviceableness of the 

 glass as an electrical insulator. 



It is certainly an unexpected result, that of two oxides so si- 

 milar to each other in their properties as soda and potash, the 

 one should, in combination with silica and lime, produce an 

 electrically conducting glass, the other an almost non-conduct- 

 ing *. But as long as we have no certain knowledge in what 

 proportions the silica in the glass combines with the lime and 

 alkali respectively, and of the conducting-power of the combina- 

 tions thus arising, it appears to me that, from the analyses and 

 experiments here described, we cannot avoid coming to the con- 

 clusion just enounced. I do not, however, mean to deny that 

 the more complete saturation of the bases with silica may con- 

 tribute to the production of the non-conducting character of the 

 substance. 



It would be very interesting, in the first place, to have speci- 

 mens prepared of pure potash-lime-glass and soda-lime-glass of 

 different, but corresponding, equivalent formula?, and then to 

 subject them to an accurate test of their insulating-power and 

 their other physical properties. iVs, however, my occupations 

 will probably not allow me time to continue this research, I do 

 not wish to reserve to myself the completion of the work. 



Stockholm, May 19, 1870. 



LX V. On the Change in the Radiation of Heat by Roughness of 

 the Surface. By the late Prof. G-. Magnus f. 



LESLIEJ, who was the first to observe that a body with a 

 rough surface radiates more heat than one with a smooth 

 one, also expressed the opinion that this was conditional on the 

 density of the surface. Yet he urges against this view the cir- 

 cumstance that the limit between hard and soft bodies cannot 

 be fixed. 



Melloni§ subsequently also maintained that the altered radia- 



* That potassium as a metal is inferior in conducting-power to sodium has 

 been proved by Lame\ Pogg. Ann. vol. c. p. 167. 



t Translated from the Berliner Monatsbericht, October 1869. 



% An Inquiry into the Nature of Heat. 



§ ComptesRendus, vol. vii. p. 238. Poggendorff's;4rmate«,vol. xlv. p. 57. 



