WOOD. — SPECTRA OF SODIUM VAPOR. 249 



I am quite at a loss as to how the series formed by the other mem- 

 bers of the doublets is originated. It appears to coincide with the series 

 excited by helium 5014, as will be seen by reference to the chart. It 

 appeared at first as if the exciting line might lie between two adjacent 

 fluorescent lines, and in that way excite a double series, but cadmium 

 5086 is slightly on the short X side of the magnetic line 5087.3, while 

 the wave-length of the other line of this pair is 5092, i. e. on the long 

 X side. 



This is the only case recorded where a spectrum of doublets is 

 excited by monochromatic stimulation, though I am of the opinion 

 that the copper line 5152 behaves in the same way. 



What is still more remarkable is the fact that if the excitation is at 

 a different point we no longer get doublets. The lithium line at 4971 

 takes hold of one of the more refrangible components of one of the 

 doublets, but only a single series of lines appears in the fluorescence 

 spectrum (see chart). The other series, i. e. the less refrangible com- 

 ponents, can be separately excited by stimulation with the helium line 

 5014 (see chart). If we are dealing with anything in the nature 

 of electron doublets, we should expect both the lithium and helium 

 radiations to excite a fluorescence showing double lines. 



If we try to explain the phenomena by assuming two chains of 

 electrons fastened together at the point 5086, we must account for 

 the fact that the 5086 vibrator excites the other chain when it is 

 acted upon by light of its own frequency, but not when it is vi- 

 brated by the lithium radiation acting at a different point on the 

 chain. I have adopted this hypothesis of electron chains merely to 

 aid in describing the physical phenomena, and not with much hope 

 that it will explain anything. 



It seems much more likely that the different lines represent vibra- 

 tions of different frequencies of the same system. We must not try to 

 make the molecule too much like a piano. The vibrations may be 

 ripples running over its surface or they may be unlike anything with 

 which we are familiar. If we had never seen a bell, it would be diffi- 

 cult to work out the theory of its very complicated vibrations from a 

 study of a set of simple pendulums. Possibly stimulation at some 

 other points might give rise to the double lines. 



I attempted to do this with the monochromatic illuminator, but 

 without success. The band of exciting light cannot well be made 

 much narrower than the distance between the components of the 

 doublets. Even with the instrument set at 5086, I could detect no 

 evidence of the doublets. I am planning to investigate this matter 

 further with a larger monochromatic illuminator designed to furnish 

 more nearly monochromatic light. 



