248 Prof. J. Thomsen on the Mechanical Equivalent of Light, 



the needle was immediately exhibited as soon as the flame was 

 made luminous, for instance, by chloride of sodium, or other vo- 

 latile compounds, or by a platinum wire. 



It was further seen that the rays which had passed from a 

 non-luminous flame through the layer of water, were absorbed 

 in very various degrees by coloured glasses, while this was only 

 to a small extent the case with the rays before they had been 

 separated by the water. 



From these various investigations the conclusion may be 

 drawn that a layer of water of 20 centims. completely absorbs the 

 heat-rays and only transmits light-rays, or that the absorption of 

 thermal rays is so complete that the error is imperceptible. By 

 experiment, the loss in luminous intensity experienced by radia- 

 tion through such a layer in water which was enclosed between 

 parallel glass sides, was determined, and the result reduced to 

 the luminous intensity of the light which had so traversed the 

 water. The absorption was found to be 0*13. 



Hence between the flame and the thermo-pile was placed a 

 box with parallel sides of plate glass enclosing a layer of water 

 0*2 metre in thickness, and the flame was then brought to the 

 same distance as in the first experiment. All the circumstances 

 were the same, except that now the radiation had to traverse the 

 water, so that the deflection of the multiplier would only depend 

 on the heating- effect which the light-rays produced. The 

 results were as follows : — 



Nature of the flame. 



Intensity of 

 light. 



Thermal and 

 luminous radia- 

 tion per minute 

 for the unit of 

 light. 



Luminous radia- 

 tion per minute 

 for the unit of 

 light. 



Spermaceti candle 



Moderator lamp 



1 



625 



8-6 



7-7 



1-2 



210 



199 



199 

 20 



c 



44 

 3-9 

 41 

 4-2 

 37 



Moderator lamp 





Gas-flame 





The mean of these experiments is 4*lc in a minute; or the result 

 may be expressed : — 



A flame whose luminous intensity is equal to that of a candle 

 which burns 8*2 grms. in an hour, radiates in the form of light, 

 in each minute, a quantity of heat which would raise the tempera- 

 ture of4i'l grms. water one degree Centigrade. 



But the entire radiation of the flame, inclusive of the heat- 

 rays, amounts to about 200c, or is 50 times as great as the 

 luminous rays ; and the entire development of heat in the che- 

 mical process of the flame is, in a candle and in a lamp, about 



