1906.] 0)1 Electric Discharge and Spectroscopy. 205 



absorption will depend upon the velocity of the corpuscle, for we can 

 easily show that when a corpuscle passes at a fixed distance from a 

 system of corpuscles having a definite period of vibration, there is one 

 velocity of the corpuscle, depending on this period — fast if the period 

 is short, slow if it is long — for which the energy given by the corpuscle 

 to the system is a maximum. Thus the relation between the amounts 

 of energy absorbed by two systems from the corpuscles depends upon 

 the velocity of the corpuscle. The velocity of the corpuscles in a 

 discharge tube depends upon the pressure of the gas, so that even 

 though the rate at which the electrical forces are doing work may be 

 the same at two different pressures, the relative intensities of the 

 lines of two systems A and B may be different. 



Again, we might expect that the coefficient of the rate of absorp- 

 tion of energy would be different according as the energy is given 

 to the atom by means of the large systems which form the positive 

 ions, or by means of small corpuscles ; and that the relative brightness 

 of lines might be different in the two cases. In the Canal-Strahleu 

 we have positive ions moving through gas and producing luminosity, 

 and the spectrum of this luminosity possesses interesting peculiarities 

 differentiating it from the spectrum of other parts of the tube. 

 Perhaps the most striking difference, however, is when the positive 

 ions strike against a salt like lithium chloride : they make the red 

 lithium line appear with great brilliancy, while if corpuscles strike 

 against the chloride the red line is not visible. It is remarkable 

 that the spectrum of the metal is produced much more readily by the 

 positive ions when they strike against a salt of the metiil than when 

 they strike against the metal itself. This is shown in a striking way, 

 if we take the liquid alloy of sodium and potassium and direct a 

 stream of Canal-Strahlen upon it the clean parts of the alloy appear 

 quite dark, but the specks of oxide scattered over its surface shine 

 with a bright yellow light, giving the sodium spectrum. 



AVhen the internal energy of the atom is increased by means of 

 light, as in Professor Wood's beautiful experiments on the fluorescence 

 of sodium vapour, the coefficient of absorption of a system will depend 

 upon the relations between the periods of that system and the period 

 of the light vibrations incident upon them. Thus, as Professor Wood 

 found to be the case, the numerous lines in the spectrum given out by 

 the vapour alter greatly in character and wave-length when the 

 period of the incident light is changed. 



The same principles which explain the variation in the intensities 

 of the spectra given out by two different systems in the same atom, 

 can be applied to explain the variations in the intensities of the 

 spectra of two gases A and B when these are mixed together. AVe 

 know that under some conditions the lines of only one constituent of 

 the mixture appear, while under others we get the lines of both the 

 gases. Let us suppose that the hues of A appear with a lower rate 

 of work of the electric forces than those of B, and that we send a 



