466 Mr. J. Evershed's Experiments on 



remains at a normal density corresponding to Cl 2 between 

 about 200° and 1200° C. * With regard to the former element, 

 the temperature at which dissociation commences (say 600°) 

 is not much above that at which the glowing is first seen, and 

 as in most of the experiments the iodine or bromine vapour 

 is largely diluted with a neutral gas, so that the partial pres- 

 sure is a good deal less than one atmosphere, it might well be 

 supposed that dissociation was going on even at the lowest 

 temperature at which the glow can be seen. But in the case of 

 chlorine dissociation begins at some 500° above the temperature 

 of my experiments (assumed at about 700°) : moreover, there 

 is no dilution of gas, which is observed at the atmospheric 

 pressure, so there can be no question of dissociation here ; or 

 at any rate, as there is no independent evidence of it, we have 

 no more right to assume it as a cause of the luminosity than 

 we have in the case of glowing solids. 



But, apart from the fact that chlorine can be made incan- 

 descent although it is not dissociating, it appears to me that 

 the general relation mentioned above between radiation and ab- 

 sorption of the visible rays, and the fact that the intensity of 

 the glowing of the more absorptive vapours (the others being 

 too difficult to observe) appears to closely follow that of a 

 solid raised simultaneously through the same range of tempe- 

 rature t? gives strong support to the view that there is no 

 essential difference between gases and solids in the manner 

 in which they radiate, at any rate under the conditions of the 

 foregoing experiments. If dissociation were concerned, say, 

 in the case of glowing iodine, one would expect the intensity 

 of the light to rapidly increase when the temperature is made 

 to approach the actual temperature of dissociation, where the 

 maximum interaction of the atoms occurs. It should in fact 

 increase in a much greater ratio than in the case of a glowing- 

 solid. But I have failed to detect any evidence of such rela- 

 tive increase on the part of either iodine or any other glowing- 

 gas. Further, a decrease of density (by exhaustion or dilu- 

 tion) will facilitate dissociation, and thus should tend to 

 counteract the reduction of luminosity due to a smaller 

 number of molecules concerned. But no such effect is in 

 fact to be seen under these conditions. 



* J. M. Crafts, ibid. xvi. ; also Jahn, ibid. xv. 



t The radiation from iodine may be easily compared with that of a 

 solid at the same temperature, by placing a small piece of carbon inside 

 the heated portion of the glass tube described in exp. I. Also when the 

 glass contains opaque particles, these are seen to glow with the same 

 intensity as the iodine, whatever the temperature, when the vapour is of 

 sufficient density to give the maximum luminosity. 



