SPECTRAL LINES 441 



lines in the spectrum could be separated into four well-defined 

 groups : the displacements of the components of each group 

 varied as the cube of the wave length, whilst the displacements 

 in the four groups are in the ratio 1 : 1*5 : 3*4 : 6'6. For titanium, 

 on the contrary, the experiments indicate a dependence upon the 

 square, and not upon the cube of the wave length. 



The same experimenters obtained some interesting results in 

 connection with the enhanced lines of titanium. In general it 

 was found that there was but little difference between the dis- 

 placements of corresponding arc and spark lines at the same 

 pressure : the enhanced lines provided the exception, for with 

 them the displacements of the spark lines are much larger than 

 the displacements of the arc lines. In general also, the dis- 

 placements of the spectral lines are practically identical, whether 

 the arc is surrounded by hydrogen or carbon dioxide, but with 

 the enhanced lines, which are known to be strengthened in a 

 hydrogen atmosphere, there was an increase in the displace- 

 ment amounting to about 25 per cent, at a pressure of 4 

 atmospheres. 



These results are of importance in the study of solar pheno- 

 mena, and as Gale and Adams remark : " The fact that the 

 enhanced lines show materially larger displacements both at 

 the sun's limb and also under pressure than do the other lines, 

 strengthens greatly the view that pressure is the effective agent 

 in producing the solar displacements." In the sun, of course, 

 there are necessarily a number of complicating factors, such as 

 scattering and absorption and varying displacements due to 

 different levels ; but a more complete study of the whole 

 phenomenon in all its aspects should be of considerable value 

 in any discussion of the relative merits of the various solar 

 theories. In determinations also of the radial motions of stars, 

 based upon the measurements of the displacements of spectral 

 lines due to the Doppler effect, the possibility of there being a dis- 

 placement due to pressure must be considered : if the reversing 

 layer of a star is under heavy pressure, the displacements 

 resulting from this cause will be quite appreciable. 



To explain these phenomena several theories have been 

 advanced. The first attempted explanation was based upon 

 Lommel's l theory of absorption and fluorescence, and attributed 

 the phenomena to the damping of the vibrations to which the 



1 IVied. Ann. 3, p. 251, 1878. 



