Fourth Series. — Powdered Surfaces. 115 



proper diathermancy. I therefore included the powders be- 

 tween two polished plates of rock-salt, closed at the edges 

 with wax. The preliminary experiment (27.), to show that the 

 salt surfaces, in the state in which I commonly employed 

 them, exercise no perceptible influence on the quality of the 

 transmitted heat, was evidently a very important one for the 

 conclusions I meant to draw. It was, as I have stated, quite 

 satisfactory. 



36. The first experiments which I made with powders 

 (December 6, 1839), were with chalk and alum, finely dusted 

 between two plates of salt. I selected the chalk on account 

 of its absolutely uncrystalline and opake character; and alum, 

 because its power of stopping rays of heat of low temperature 

 was so very great, that I judged that if the influence as a me- 

 chanical modifier of surface should prove predominant, and 

 allow as much, or more, heat of low than of high temperature 

 to pass, the mechanical influence of a substance in fine powder 

 would be clearly established. 



37. Now, the result at which I arrived, and which was en- 

 tirely conformable to my anticipation, may serve to show the 

 caution requisite in drawing conclusions from limited data, 

 however apparently conclusive. The surfaces powdered with 

 chalk suffered rather more heat of low than of high tempera- 

 ture to pass (viz. 34<*5 per cent, dark heat, and only 30*5 of 

 heat from Locatelli lamp, transmitted through a thick glass- 

 lens), whilst the salt strewed with alum appeared quite indif- 

 ferent to the kind of heat incident* (transmitting only 1 7 per 

 cent, of both, thus showing that the powder was in consider- 

 able quantity). I concluded, therefore, with apparent reason, 

 that the chalk having no specific action, or being (most pro- 

 bably) opake or athermanous, the powder of it acting me- 

 chanically, allowed low temperature heat to pass in excess, 

 whilst in the case of alum, the specific action was entirely 

 counteracted by the mechanical action of the powder. I sim- 

 ply stated the fact amongst others detailed in the preceding 

 pages, in a memorandum presented to the Royal Society of 

 Edinburgh on the 16th December 1339f, and a few days 

 after, in a slightly different form, communicated to M. Arago, 

 and printed in the Comptes Rendus de VAcademie des Sciences, 

 6th January 1840. On the 28th of December, I obtained a 

 similar result for charcoal powder (whose affinity with smoke 



• Yet an alum plate of a certain thickness transmits no less than 27 per 

 cent, of the one kind of heat, and no sensible portion of the other (Mel- 

 loni.) 



t See our July Number, p. 69. note *. 



12 



