FOURTH SERIES. ROUGH SURFACES. 13 



surface ; supposing these grooves to be regularly formed, and capable of numeri- 

 cal estimation, will the effect continue ? 



27. (1.) With respect to the first of these questions, it is satisfactory to be 

 able to answer it affirmatively in a general way. I took two salt plates, of which 

 the surfaces had not been regularly polished for a long time, and which, though 

 bright and clear, were by no means particularly even and true. Of heat from 

 LOCATELLI'S lamp previously sifted by glass, these four surfaces of rock-salt trans- 

 mitted 72 per cent. With dark heat from smoked brass the per-centage was 73, 

 a difference which, in this experiment, could hardly be considered as appreciable. 

 The transmission of these trco very different kinds of heat mas therefore equal. M. 

 MELLONI has shewn that when rock-salt is pure and perfectly polished, .92 of the 

 incident heat is transmitted by a pah* of surfaces, and therefore four surfaces 

 should transmit (.92)* or 84.5 per cent. This estimate I have verified, and am 

 satisfied of its accuracy. The deviation in the present case (which I think it 

 right not to pass over) is due partly, no doubt, to the inequalities of surface 

 but chiefly to some imperfections in the salt itself, which, as the experiment was 

 merely a relative one, were not adverted to. In contrast with this, I used at the 

 same time (December 11. 1839) a piece of salt, which once had been polished 

 on both sides, but which, by being laid aside for some years, had become com- 

 pletely dull and grey on its surface. This specimen, then, was simply depolished ; 

 it contained no furrows, and had been subjected to no mechanical action what- 

 ever. Its per-centage of transmission was, 



Locatelli with Glass. Dark hot Brass. 



Tarnished salt, 66 77 



clearly establishing the general principle. 



28. (2.) With respect to the question, whether roughness of surface has a 

 similar effect in modifying the diathermancy of other substances as well as rock- 

 salt, we are able to give a distinctly affirmative answer. Rock-salt being, so to 

 speak, quite indifferent to the quality arid source of the incident heat, any cause 

 of specific action becomes immediately apparent. Not so with any other sub- 

 stance, which, exercising already a specific action in virtue of its nature, is to 

 have that specific action modified by a modification of surface. At least the 

 question is, whether or not this modification will occur ? An example will best 

 illustrate how this modification may be discovered and expressed. I took a plate 

 of mica with its natural bright surfaces, and so thin as to transmit in considerable 

 abundance heat from different sources. The per-centages in this state were de- 

 termined as follows : 



Locatelli with Glass. Locatelli. Dark heat. 



Mica with bright surfaces, 83.5 74 37 



Both sides of the mica were depolished with emery-paper, and the experiment 

 repeated (27th November 1839), 



VOL. XV. PART. I. D 



