566 



SCIENCE 



[N. S. Vol. XXX. No. 773 



The principle which underlies the whole 

 research is that the absorption spectrum of 

 a solution consists simply of the superposed 

 absorption spectra of all the molecular species 

 present in the solution. In a solution of even 

 a single solute there may be a large number 

 of these molecular species, namely, ions, undis- 

 sociated molecules, aggregates of the ions or 

 of the molecules, and compounds of ions and 

 undissociated molecules with the solvent. It 

 is evidently no simple matter to unravel the 

 spectrum of so complicated a system and to 

 determine the origin of the various bands. 



The method by which the authors have at- 

 tempted to do this has been to keep the num- 

 ber of molecules of some one particular species 

 in the path of the beam of light constant, 

 while varying the amounts of the other spe- 

 cies, and then observing the effect produced 

 upon the absorption spectrum. Unfortu- 

 nately, the only molecular species about which 

 we know enough to make it possible to apply 

 this method are the simple ions and undisso- 

 ciated molecules. The authors therefore only 

 carried out experiments, keeping, first, the 

 total amount of salt, second, the number of 

 undissociated molecules and, third, the num- 

 ber of ions in the path of the beam of light 

 constant. 



Many solutions were studied under the first- 

 named conditions, that is, keeping the total 

 amount of salt in the path of the light con- 

 stant. Only a very few of these showed no 

 change in their absorption spectra with 

 changing concentration. This, of course, was 

 to have been expected from our general knowl- 

 edge of solutions, for the absorption spectra 

 would only remain unchanged when either the 

 relative concentrations of the different ab- 

 sorbers did not change with the concentration, 

 or where the absorption spectra of all the dif- 

 ferent kinds of absorbers were identical. The 

 first alternative is perhaps never fulfilled, but 

 the second is very probably the explanation of 

 the constant band of nickel sulphate solutions 

 in the ultra-violet and of the whole constant 

 spectra of dilute neodymium and praseodym- 

 ium solutions. 



Nearly as many solutions were studied 

 under the second or third of the above condi- 



tions, that is, keeping the number of undisso- 

 ciated molecules or of ions in the path of the 

 light constant. If, when the number of un- 

 dissociated molecules was kept constant, the 

 absorption decreased with the dilution, or if, 

 when keeping the number of ions constant, the 

 absorption increased with the dilution, we 

 should be forced to the conclusion that the 

 change in the absorption spectrum with the 

 dilution could not be explained as being due 

 simply to the differing absorption spectra of 

 the ions and the undissociated molecules, as 

 Ostwald at first proposed. Instead, we should 

 be obliged to assume that other absorbers than 

 the ions or undissociated molecules must have 

 been present and that their formation or de- 

 composition with the changing concentration 

 of the solution was responsible for the observed 

 variations. Just such variations were observed 

 in the ultra-violet bands of copper salts and 

 of cobalt chloride, in the red bands of cobalt 

 salts, and in the whole spectrum of ferric 

 chloride. It follows then that in these solu- 

 tions, at least, other absorbers than simply 

 ions and undissociated molecules must be 

 present. 



Two possibilities are suggested as to the 

 nature of these additional absorbers. They 

 may be either aggregates of the undissociated 

 solute molecules or of the solute ions, or they 

 may be compounds of the solute ions or mole- 

 cules with the solvent. To decide between 

 these two possibilities the authors cite the 

 observations of Hartley, on the change of the 

 absorption spectrum of salt solutions with the 

 temperature. Hartley found that a rise in 

 temperature in general produces the same 

 effect as an increase in concentration. This, 

 the authors consider, is evidence against the 

 assumption of aggregates, for they reason that 

 a rise in temperature would tend to break up 

 the aggregates, while increase in concentration 

 would have just the opposite effect, and hence 

 produce an opposite, instead of the same effect 

 on the absorption spectrum. It is not easy to 

 see that this argument is conclusive, for 

 whether or not the aggregates will break up 

 with rising temperature will depend uiK)n the 

 heat change incident to their formation. If 



