162 Proceedings of the Royal Society of Edinburgh. [Sess. 
the damping of the vibrators — namely, that it is due to electromagnetic 
radiation — is inapplicable to the bands of the anilin colouring matters or 
other broad bands, because it gives much too small a rate of damping. This 
is a view with which I am in thorough agreement. He divides absorption 
bands into two kinds : the first kind characterised by great intensity and 
breadth, and due to an ion or electron in every molecule ; and the second 
kind including bands, such as those of didymium, which are narrower and 
much less intense, and are supposed to be caused only by exceptional 
molecules in the body. Thus by making p a small fraction he explains 
the very small values of pejm obtained. He believes that absorption 
measurements are much more valuable than dispersion measurements from 
the point of view of calculating pe/m, and with this opinion I am in 
emphatic agreement. His views on the absorption spectra of gases are 
specially interesting. Continuously selective absorption is due to the 
normal state of gas molecules ; the band spectrum absorption and emission 
are due only to an intermediary non-stationary condition, and are connected 
with the rupture of the bonds to which the continuously selective absorption 
is due. 
Having stated my own views, I shall now criticise some ideas current 
in the field. 
First of all, take the question of the colour of the ions. According to 
the theory of electrolytic dissociation, in an aqueous solution a certain 
proportion of the salt molecules splits into ions. The absorption spectrum 
of a salt solution should consequently be the sum of the separate spectra 
due to the neutral molecules, anions, and kations respectively. If the 
solution is very dilute it is entirely ionised, and the colour should be 
entirely due to the ions. This conclusion was first drawn by Ostwald, 
and he sought to prove it in one case by showing that the absorption 
spectra of thirteen permanganates were identical at great dilution. He 
did not investigate concentrated solutions, but it was shown later that 
the absorption spectra of concentrated permanganates were also identical, 
so they cannot be adduced as a proof of electrolytic dissociation. They are, 
nevertheless, still cited for that purpose in books on physical chemistry. 
As a result of my studies, both in the literature and at first hand, my 
opinion can be put in the following sentence : There is no spectroscopic 
evidence in favour of the theory of electrolytic dissociation; there is 
some evidence against it, which can, however, be got over by the assump- 
tions of hydrolysis, complex ions, hydration, etc. Few of the researches 
on the subject have been carried out with the necessary thoroughness. 
The photographic method is applicable only to the salts of the rare earths, 
