496 CONTINUOUS SPECTRA. 



volatilised, and the image of the lithium flame will have disappeared 

 before it can be seen. 



The separation of the single images is the greater, the greater the 

 distance between the flame and the prism ; it should not be less 

 than l m , or the images will not appear quite separated, and if the 

 distance can be made greater it is much better. Shortsighted 

 persons should use spectacles or an opera-glass to see the small 

 flames at that distance. 



If more substances, capable of colouring the flame, 

 should be introduced into it, the number of distinct 

 images would obviously increase; they would appear so 

 close to one another that each single image could no 

 longer be clearly distinguished ; that is, a continuous 

 spectrum would be produced, similar to the spectrum 

 which is observed when a beam of sunlight or daylight 

 passing through a slit is seen through a prism. The 

 flame of a candle or a lamp does not give out only one 

 colour, but an infinite number of colours, and hence of 

 images; it becomes thus impossible to distinguish be- 

 tween the individual images, and they merge into one 

 another like the colours of the spectrum produced by 

 direct or diffused sunlight. 



The platinum wire itself, when held in a vertical posi- 

 tion in the hydrogen flame until it becomes white-hot, 

 produces a continuous spectrum, and likewise every solid 

 or liquid body when at a white heat; only gases orvapours 

 emit light of one colour when heated until they become 

 luminous. The light emitted by the flame of a candle, or 

 a lamp, or a gas flame, is not caused by the glowing va- 

 pours, but is due to the infinitely small particles of 

 white-hot carbon which are set free by the decomposition 

 that takes place at a high temperature of the stearine, tal- 

 low, oil, coal-gas, &c. During combustion these particles 

 usually disappear, but their existence can be shown by 



