286 IN STARR* REALMS. 



We have already seen that when hydrogen gas is heated, 

 it radiates forth light which the spectroscope resolves into 

 a certain number of bright lines in various parts of the spec- 

 trum, each with one particular wave length. We have also 

 seen that when a cold atmosphere of hydrogen surrounds 

 a star, the obstruction to the passage of the light is not 

 indiscriminate, but only operates on rays of certain par- 

 ticular wave lengths. 



And now for the extraordinary fact which has provided 

 us with a key to the interpretation of spectroscopic 

 phenomena. It has been discovered that the particular 

 wave lengths which cold hydrogen is capable of stopping 

 are precisely those wave lengths which characterize the 

 light that heated hydrogen is capable of emitting. This 

 is a very singular fact, and to understand it fully we must 

 go a little further into the mechanism of the subject. 

 For this problem, like many another problem in nature, 

 resolves itself as a matter of last analysis into a question 

 of dynamics. 



Hydrogen is an invisible gas, the lightest body in 

 nature, and we know that a gas consists of a number 

 of molecules darting about in all directions ; these mole- 

 cules are so small and so close together, that many 

 billions of them would fit into a lady's thimble. When 

 the gas is heated the activities of the molecules are in- 

 creased, each molecule in fact consists of a number of 

 portions which are themselves moving with almost in- 

 conceivable rapidity, and with a complexity which we can 

 only very partially understand. Fixing our attention for 

 the moment on one particular wave length, corresponding 

 to the ray F emitted by hydrogen, let us suppose that 



