58 M. MELLONI ON THE IMMEDIATE TRANSMISSION 



The only difference observable between the transmission through a 

 continuous medium and the transmission through a series of detached 

 screens is in the amount of the losses, which, for a given thickness, are 

 found to be greater in the latter, because of the reflexions produced by 

 each separate surface. 



These facts cannot surprise us after the idea we have formed to our- 

 selves of the influence exercised by diaphanous substances on radiant 

 heat. For the calorific sources always emit a certain portion of rays 

 heterogeneous (if we may use the expression) to the calorific tint of the 

 glass, Avhich, through the absorbent action of the matter constituting 

 the continuous medium or the detached screens, are successively extin- 

 guished until no rays remain but those that are homogeneous to this tint. 

 Now these homogeneous rays must suffer a loss greater or less in its 

 amount, but constant in respect to layers of equal thickness, as is the case, 

 in the transmission of light, with red rays passing through a medium of 

 the same colour, and with white rays passing through a medium diapha- 

 nous and colourless. What we have said of glass is equally true of 

 every other partially diathermanous substance. 



The calorific transmission through a series of homogeneous screens is 

 then absolutely of the same nature as that which is effected through the 



of a millimetre of each layer, when referred to the quantities of incident heat, 

 will have the values 



,3-286 0-535 0-160 0-050 0-020 0-010 0-007 

 100 77 54 46 41 37 35 



that is, 0-0328 0-0070 0-0030 0-0011 0-0005 0-0003 0-0002. 



By similar calculations the successive losses sustained by the radiations from 

 the incandescent platina and the copper heated to 390° will be found to be 



0-0614 0-0081 0-0032 0-0019 0-0010 0-0005 0-0003 



0-0943 0-0155 0-0050 0-0022 00014 0-0010 0-0008. 



Now the differences between every two terms of these series are for the 



1st, 0-0258 0-0040 0-0019 0-0006 0-0002 00001 ; 



2nd, 0-0523 0-0049 0-0013 0-0009 0-0005 0-0002; 



3rd, 0-0780 0-0105 0-0028 0-0008 0-0004 0-0002. 



As to the fourth source it is useless to speak of it, as its rays are completely 

 extinguished at the distance of one millimetre. 



Thus, notwithstanding the inequalities of the increase of the distance from 

 the second and the third layer to the surface of entrance, we observe in the 

 three series the two principles we have laid down, namely, 1st, the decrease of 

 the losses; 2nd, the tendency of this decrease towards a limit at which the loss 

 becomes constant : but for each particular case the points of the medium at 

 which the rays begin to suffer tliis constant action are evidently placed at a fixed 

 distance from the origin. Therefore, if the glass be divided into equal layers, 

 the limit of the decrease of the losses will be attained more slowly in proportion 

 as the^ayers are more numerous, that is to say, thinner. It is for this reason 

 that in each series the limit at which the losses become constant depends, as 

 wc have already said, on the thickness of the elementary layers. 



