286 Proceedings of the Royal Society of Edinburgh. [Sess. 
is especially the case in connection with glucinum, cerium, lanthanum, and 
didymium. Large quantities of the minerals were worked up, pure salts 
prepared, and much work was done, which has no doubt since been con- 
firmed by others, although it may be questioned whether even now all 
Gibson’s results have been re-established. 
About the time when the paper on glucinum was published, Gibson 
started some experiments on the effects of light on such changes as the 
conversion of chlorine water into hydrochloric acid, the resulting observa- 
tions being published in a short paper on “ Photochemical Action ” in the 
Proceedings of the Royal Society of Edinburgh in February 1897. This 
was followed by a short paper, “ A Preliminary Note on a Characteristic of 
Certain Chemical Reactions ” ( Proc . Boy. Soc. Edin., Dec. 1897). The origin 
of these papers was as follows : In studying the action of light on these 
mixtures, Gibson discovered the fact that the amount of change depended 
on whether the final result of the reaction was to increase the electrical 
conductivity of the solution as a whole or to diminish it, there being a 
tendency for any such solution to move in the direction of increased 
electrical conductivity. This led him further to investigate the question as 
to how far other reactions, apart from those caused by light, were influenced 
by these conditions. 
No particular physical value had been, so far, associated with the 
electrical conductivity of a system as a whole, and the whole direction of 
research was proceeding towards experiments on very dilute solutions, 
with a view to the application of the laws laid down by van’t Hoff, 
Arrhenius, Kohlrausch, and Nernst to the problems of electrochemistry. 
It was probably for this reason that more attention has not been directed 
to the very interesting results obtained by Gibson in this direction. 
In the preliminary paper already referred to he gives examples of the 
law that many chemical reactions are governed by the tendency of a 
solution to develop a state of maximum conductivity in the system, these 
examples being : the dehydration by hydrochloric acid of hydrated 
cobaltous chloride ; the dehydration of sugar by sulphuric acid ; the re- 
duction of chromic anhydride by hydrochloric acid ; the oxidation of 
hydrogen iodide by sulphuric acid ; and the oxidation of nitric oxide by 
nitric acid. 
In order to carry these investigations further, he decided to redetermine 
the conductivity curves of some of the best known acids and salts, and 
devoted a great deal of time and labour to these measurements, with the 
result that there can be no doubt that the most exact conductivity curves 
that we have for hydriodic, hydrobromic, and hydrochloric acids, and 
