TRANSACTIONS OF THE SECTIONS. 383 



refrangibility of the heat. From this we derive almost as easy 

 a mean for estimating the refrangibihty of a degree of heat as 

 in the corresponding case of hght. M. Melloni took a clear 

 glass, and coloured glasses of various tints of the spectrum, and 

 adapted each glass to the opening in a screen, through which 

 rays fell from a lamp upon the sensible part of the thermo- 

 multiplier. The distance of the lamp was varied till, with each 

 glass successively, a deviation of 40° was produced in the index- 

 needle of the galvanometer. In each case the rays transmitted 

 by the glass were then caused to traverse a plate of sulphate of 

 lime. The needle always indicated a smaller thermal effect 

 than before, and its new position was the same for the white, 

 violet, indigo, blue, yellow, orange, and red glasses ; only in 

 the green, instead of the angular deviation being reduced 

 from 40° to 18°, as in the former cases, the needle stood at from 

 7° to 10°. When alum was substituted for sulphate of lime, 

 the deviation for green glass was from 1° to 1°'6, whilst for all 

 the others it was 8°. Hence we conclude that red, orange, 

 yellow, blue, indigo, and violet, as well as clear glasses, pro- 

 duced no elective action on heat, since the rays issuing from all 

 these traverse an interposed plate with equal facility, which 

 would not be the case were they differently refrangible. Green 

 glass, on the other hand, transmits rays easier stopped than 

 any of the others ; and therefore, from what has been already 

 said, admits the passage of the least refrangible rays, whilst it 

 stops the more refrangible, which we know to traverse the sul- 

 phate of lime or alum most easily. To confirm this deduction 

 was the object of a second series of experiments, since the 

 transmission of the first and the stoppage of the second ought 

 to be made manifest. 



The author took an Argand lamp, and a spiral of platinum 

 wire placed over a lamp of alcohol, which, it is easy to show, 

 had a much lower temperature, and therefore its heat had a 

 lower degree of refrangibility than that of the Argand burner. 

 The differences of the quantity of heat from these two sources, 

 transmitted by all the kinds of glass except the green, showed 

 that the transmitted rays, in the case of incandescent platinum, 

 were almost exactly half those transmitted when the source was 

 an Argand burner. In the case of green glasses, the difference 

 was little or nothing. Hence the conclusion that, as was antici- 

 pated, the green glass was more transparent for the rays of 

 small refrangibility than for those of greater refrangibility (or 

 temperature). 



Another experiment was performed with the same object, 

 and with similar results. 



