461 Mr. J. Evershed's Experiments on 



under certain circumstances as wreaths and streaky clouds 

 moving with the convection-currents in the tube *. 



Being satisfied by repeated trials that the glow was really 

 due to incandescent chlorine, an endeavour was made to 

 determine the spectroscopic character of the light. It 

 seemed almost hopeless to observe the spectrum directly, al- 

 though with a spectroscope of very low dispersion and wide 

 slit the spectrum appeared to be continuous ; but it could 

 only be traced between the positions of D and F of the solar 

 spectrum, probably owing to the greater sensitiveness of the 

 eye to that region. An indirect method was, however, de- 

 vised. The depth of the layer of cool chlorine traversed by 

 the light from the hot gas was, in the first instance, about 

 20 cm., that being the distance between the prism and the 

 hot part of the tube. The absorbing effect of this thickness 

 of gas was tested by varying the distance, heating the 

 tube successively at 15 and 30 cm. from the prism. If the 

 spectrum of the light emitted consisted of lines or bands cor- 

 responding to the absorption-lines of chlorine, there should 

 be a marked increase of absorption after traversing 30 cm. as 

 compared with 1 5 cm. But no difference was perceptible in 

 the intensity of the glow whatever the distance traversed, 

 showing that the cool gas exercises very little absorption on 

 the light coming from the hot. 



The inference, therefore, is that chlorine, like iodine and 

 bromine, emits other rays than those absorbed, and probably 

 shines with continuous light ; the selective absorption of the 

 cool gas merely giving the glow a greenish tint. 



There seems no reason to doubt that the luminosity of 

 chlorine, as of its allied elements, is directly clue to the 

 heating, and that chemical changes are not concerned in the 

 phenomenon. 



The Sulphur Group. — Similar results have been obtained 

 with sulphur and selenium. The glowing may be observed 

 either " end on " in a red-hot porcelain tube filled with hy- 

 drogen or nitrogen, or the element may be simply sealed in 

 a hard glass tube in air. If a small portion of the tube be 

 then heated strongly, a faint glow can be seen while distil- 

 lation is going on, and the heated space is filled with vapour. 



Phosphorus. — Experiments with this substance have so far 

 led to negative results. After all the phosphorescence due to 

 traces of oxygen has disappeared from the tube, the vapour 

 appears to give no light. 



* The glowing of heated chlorine may be more conveniently observed 

 in a porcelain tube connected with a T-tube having a glass plate ce- 

 mented in one end of the T. If the Bunsen is concentrated on the tube 

 by means of a fire-clay arch placed over it, the glow uniformly illuminates 

 the bore of the tube, 



