160 Transactions of the Amebic an Institute. 



a pure color, we see simply an image of the slit. If, however, we 

 direct it toward a colored light consisting of several simple hues, we 

 see just as many distinct images of the slit as there are separate colors. 

 The reason is obvious. This glass prism bends these differently 

 colored lights to a different extent. It bends the red colors the least, 

 and the violet hues the most ; and therefore these separate tints 

 become spread out and form what we call the spectrum. If our 

 colored light consists of a single tint, we get but a single image ; if it 

 consists of three tints we get three separate images ; if it consists of 

 pure white light, which contains every gradation of color, we get an 

 infinite number of images of the slit, which succeed and overlap each 

 other ; and these blending images form the band which I have already 

 pointed out. 



Understanding now, I hope, what is meant by pure white light, 

 we may return to the principles with which we started ; that an 

 incandescent solid body always emits pure light. What is true of 

 solid bodies is also true of liquids which can be heated to the tempe- 

 rature at which they become incandescent, as for example, the molten 

 metals. They also emit pure white light. Mark now the important 

 conclusion to which this at once leads. If we can reason from analogy, 

 it follows that whenever we see white light, the luminous source is a 

 solid or a liquid body. So far as our experience upon the surface of 

 the earth goes, this is universally true; and analogy would lead us to 

 suppose it to be a general law of nature. Now the light which comes 

 to us from the sun and the light which comes to us from most of the 

 fixed stars, is pure white light. Then the sun and the fixed stars 

 must be incandescent solid or liquid bodies. They cannot be vapors 

 or gases; gases and vapors differ from solids and liquids in two 

 important respects. In the first place, the light which they emit is 

 far less intense than that emanating from a solid or a liquid at the 

 same temperature. There is a great difference between gases and 

 vapors, but, as a general rule, the denser the vapor or the gas the more 

 brilliant the light which they emit. Hence more light is emitted by 

 a dense metallic vapor than by the light gases, like our atmospheric 

 air. In the second place, the light from a vapor or gas differs in that 

 it is colored ; and moreover its color is characteristic of the substance 

 from which the light emanates. These points are very important, 

 especially the last. 



But, first of all, I must explain an apparent contradiction in my 

 general statement. We are here burning a gas. The hall is illumi- 



