M. MELLONI ON THE POLARIZATION OF HEAT. 



341 



malines enables us to advance a step further ; for this equality shows 

 that not only the quantity but the quality of the heat transmitted by 

 pure water is the same as the quality of that transmitted by water 

 saturated with salt or alum. In sliort, if we receive on the pile the 

 calorific streams which issue from distilled water, and the solutions of 

 rock salt and alum, we shall find the deviations in the galvanometer 

 very nearly equal in the three cases. It is to be recollected, however, 

 that this will happen only when the three layers from which the rays 

 issue are of the same thickness. Now this invariability of action through 

 la3^ers of equal thickness takes place also if we interpose the same plate 

 of alum or any other substance behind each of the liquid layers in suc- 

 cession ; for the common deviation is always diminished by a constant 

 quantity, even when, by concentrating the calorific radiation with lenses, 

 we have raised to 35° or 40° the arc of impulsion described solely under 

 the action of the stream transmitted by each of the liquid layers. 



In order to conclude the experimental study of the calorific polariza- 

 tion of the tourmalines, we have only to compare with one another the 

 polarizing action of these crystallized substances on the radiations of 

 different sources of heat. For this purpose I select the four systems 

 marked Nos. 1, 5, 8, and 9 in the first table, and these being exposed, 

 in the two principal directions of the axes, to the calorific radiations of 

 an Argand lamp, a Locatelli lamp, a spiral of incandescent platina, and 

 a plate of copper heated to 400°, give the indices of polarization con- 

 tained in the following table. 



Table IV. 



N.B. In this table the arcs of impulsion are omitted in order to avoid exces- 

 sive complexity. We think it necessary, however, to observe tliat these were 

 often more extensive than those of the preceding tables ; and this was indeed 

 absolutely necessary in order to perceive those indices which in many instances 

 are extremely feeble. No. 1, for example, being exposed to the radiation of the 



Vol. I — Part II. 2 a 



