96 MODERN SCIENCE READER 



nous by cooling, introduced chlorine into it to break down 

 the hydrocarbons, and obtained a brilliant light. A porce- 

 lain rod introduced into the lower part of a flame cooled it 

 and decreased its light, but collected no carbon, while, if 

 introduced into the upper part, its under side became coated 

 with soot. Heumann argued that if Frankland was right 

 and the light is reflected from dense hydrocarbon vapors, 

 these should be condensed on all sides of the rod at once in 

 a quiet flame, while, as a matter of fact, soot was deposited 

 only on the under side; and furthermore, soot can also be 

 collected upon a surface too hot to condense hydrocarbons 

 at all. He therefore concluded that the surface merely 

 stops carbon which is formed lower down in the flame. If 

 one luminous flame is allowed to play against another, the 

 carbon is rolled up and can be seen as glowing particles in 

 the outer non-luminous sheath. 



Frankland had said that flames cannot contain solid 

 particles because they are transparent. Heumann pointed 

 out that thick flames are opaque and that thin ones are no 

 more transparent than is an equal layer of soot rising from 

 burning turpentine ; the rapidity of the motion of the par- 

 ticles preventing any obstruction to the view, just as is the 

 case with a rapidly revolving, spoked wheel. 



Heumann next took up the phenomena of shadows and 

 showed that the luminous portion casts a definite shadow 

 when interposed between sunlight and a screen, and that 

 the shadow is continuous for a luminous turpentine flame 

 and the column of soot above it. And further, that a 

 hydrogen flame which ordinarily casts no shadow and gives 

 no light will cast a sharp shadow and emit a fairly bright 

 light if passed through suspended lampblack or if it sweeps 

 any solid matter into the flame. Luminous vapors do not 

 cast shadows, absorption bands being very different from 

 true shadows. 



C. J. Burch found that when sunlight is reflected from 

 a luminous flame it is polarized, while if reflected by glow- 

 ing vapors, however dense, it does not exhibit this phe- 



