Microscope Spectrum Apparatus. By H. G. Sorby. 201 
4. Relation between the Absorption Bands in different Solutions. 
Much still remains to be learned before we can form a satisfac- 
tory opinion as to the full significance of the absorption bands seen 
in the spectrum of any substance. It is quite clear that they 
must indicate that its ultimate particles are in some way or other 
so related to waves of light of particular length that their force is 
expended either in raising the temperature of the solution or in 
causing chemical change, whereas waves of other lengths pass 
through comparatively unimpeded, as though the bands were in 
some way or other connected with the size of the particles. The 
general character of the spectrum — whether having one or more 
bands, which are narrow or broad, dark or faint — must, I think, 
depend on the shape or constitution of the particles, and therefore 
it seeems probable that the spectrum as a whole furnishes us with 
evidence both as to the character and magnitude of the ultimate 
molecules. 
It may perhaps be premature to conclude that it is a law 
subject to no important correction, that when the spectrum of 
a single substance contains a number of well-marked absorption 
bands they are all related to one another in a perfectly definite 
manner, but as far as I am able to judge from my present know- 
ledge, there is a far more uniform connection between the wave- 
lengths of their centres than between any other condition. At all 
events, in many spectra having a series of bands whose centres are 
at wave-lengths a, b, c, and d, there is the same ratio between each 
consecutive two, so that y = - = ^ • Possibly this may be a true 
general law, subject to modifications depending on particular condi- 
tions. Another important point is that, in the case of substances 
giving two or more well-marked bands, though the position of these 
bands may vary very considerably in the spectrum of the solid sub- 
stance and in that of its solutions in different solvents, that is to 
say, though the actual wave-lengths of the centre of the bands 
may vary with the conditions in which the substance occurs, the 
ratio between the wave-lengths of the bands remains almost if not 
quite constant, the discrepancies being no greater than may be 
due to errors of observation. As a good example, I will refer to 
yellow xanthophyll. 
Condition. 
Centre of the 
two Bands. 
Ratio. 
Ia free state and solid 
501 469 
l : 
•936 
Dissolved in carbon bisulphide .. 
498 467 
l : 
■937 
Dissolved in absolute alcohol 
471 442 
l : 
•938 
Combined with Canada balsam . . 
488 457 
l : 
•936 
Q 2 
