4 g REPORT ON PHYSICAL OPTICS. 



nating from other sources. He thus arrived at the interesting 

 d sov g ery, that the system of bands in the "h"**^ 

 light wb ich he examined varied with the source ; while it was 

 cottaltlv the same in the number of the bands, and the* relate 

 o the coloured spaces, in the light of the same source however 

 Codified. In the light of Sirius there are three broad band 

 which have no resemblance to those of solar light. The light o 

 the electric spark, on the other hand, when analyzed by the prism, 

 is found to have several bright lines, of which that in the green is 

 remarkably brilliant. Similar phenomena were observed in the 

 light of artificial flames-the flame of an oil lamp, for example, 

 exhibiting a well-defined bright band between the red and yel- 

 low and another not so distinct in the green * This, however, is 

 not universally the case. In the red flame of strontia, as was 

 observed by Dr. Faraday and Mr. Talbot, there are a number oi 

 red rays separated from each other by dark bands ; and in the 

 flame of cyanogen, when similarly analyzed, the violet is found to 

 be divided into three distinct portions, with broad dark intervals.f 

 It is easy to account for the general fact of the deficiency of 

 certain classes of rays in certain lights. "When a body violently 

 heated begins to shine, the phenomenon is simply accounted for, 

 in the wave-theory, by an increase in the frequency of its vibra- 

 tions. In the same manner it seems natural to suppose, generally, 

 that the mechanical agencies at work during combustion accelerate 

 or retard, in various ways, the rate of vibration, and so alter the 

 character of the emitted lights. The light emitted in weak or 

 incipient combustion is generally blue. Sir John Herschel ob- 

 served that when sulphur burns with a feeble flame, its light con- 

 tains all the rays of the spectrum, and particularly the blue and 

 violet ; while, in vivid combustion, these disappear entirely, and 

 the light is a yellow of almost perfect homogeneity.* The various 

 shades of colour in the flame of a common candle from the deep 

 blue of the lower part, which is found by prismatic analysis to 

 consist of five distinct portions, to the yellowish white in the 

 centre, and thence to the dusky red at the apex of the flame 

 seem to be referrible to the same principle, Fraunhofer and Sir 



Munich Memoirs. 



t Phil. Mag., Third Series, vol. iv. p. 114. 



J " On Absorption of Light in coloured Media." Edin. Tram., vol. ix. 



