104 Prof. Forbes*s Researches on Heat. 



necessary consequence of the construction, that the heat passes 

 through such piles as I use unaltered, or nearly unaltered, 

 in its character, whilst in passing through bundles of de- 

 tached plates laid together, the thickness of mica to be tra- 

 versed is sufficient to modify the heat by absorption, in such a 

 way that the difference of quality has vanished, whatever he the 

 source, in the very act of transmission. It it hardly likely, con- 

 sidering the size of M. Melloni's mica plates (4 inches long 

 and 2 wide), that they could be less than one fifteen hundredth 

 of an inch thick each ; a pile of ten would then be ten times 

 as thick as my pile of equal energy, and at an incidence of 55° 

 the thickness traversed would not be much shorter than that 

 of the mica plate alluded to in art. 20, which we have there 

 seen to be sufficient to obliterate all distinctive character as 

 to polarizability between an Argand lamp and dark heat. 



Being now fully aware of the importance of the construc- 

 tion of piles of mica which I had adopted, I thought it worth 

 while to examine the proportions of heat from different 

 sources, which these very delicate laminae were capable of 

 transmitting, which, I presumed*, would be found far less 

 variable than when plates of the usual thickness are employed. 

 My expectations were more than realized, as is seen in the 

 following table, the second column of which shows the pro- 

 portion, to the whole incident heat, of that transmitted by the 

 two mica piles I and K placed parallel to each other ; by far 

 the greater proportion of the loss being that due to the obli- 

 quity of reflection and the number of surfaces f . By way of 

 contrast, I have placed in the third column the proportion 

 of the whole incident heat transmitted at a vertical incidence 

 by a plate of mica '016 inch thick. 



Rays out of 100 transmitted by 



Source of Heat. Plates I and K Mica plate '016 



parallel. inch thick. 



Locatelli lamp .... 18*8 51 



Ditto, with plate of glass '06 \ i /^.o 72 



inch thick interposed . . J 



Incandescent platinum . . 17*6 50 



Dark hot brass (700°) . . 15-5 15 



Heat from boiling water .10* 8 



It is very evident that, for the first four sources of heat at 

 least, the transmissive power of the plates I and K varied 



* I do not state this as a new idea ; it has been repeatedly remarked 

 by M. Melloni, that, in proportion as substances are thinner, they possess 

 a more equable diathermancy for heat of different qualities. 



t The part of the effect due to reflection, I had previously established to 

 be nearly the same for different kinds of heat. 



