1645 
axis of a mercury seal that could be rotated with a little motor 
whilst the gaseous space was connected with a mercury manometer 
by means of a capillary. 
In this manner, not only the oxidation could be ascertained but 
also its velocity could be measured. 
A small vertical plate placed perpendicularly to the wall and 
projecting just above the surface of the liquid served to keep the 
surface tranquil; a similar little plate was also found in the small 
mercury reservoir of the seal causing the mercury surface to keep 
in the horizontal position. 
The reduction to pinacone and the oxidation of the alcohol to 
aldehyde and water seem to be inseparately connected with each 
other; a priori this is not necessarily so. We may conceive tbat 
the H is rendered active in such a manner that it combines with oxygen 
without being yet capable of attacking the ketone, and conversely. 
As these two reactions probably take place together, the action 
of the CO-group on the H of the alcohols, which leads to an activity 
towards the O of the air as well as to that of the ketone, must be 
attributed to a same cause. 
Only there is a difference and that is that the oxidation reaction 
can proceed further without the assistance of light, seemingly because 
this is a catalytic reaction, whereas the pinacone formation requires 
the continuous assistance of the light energy. 
From our previous communication it appears that there are a number 
of ketones, which, although they powerfully absorb the light, are 
not at all, or at least not perceptibly, capable of activating the 
alcoholic hydrogen. *) 
Many of these ketones may, however, be conveniently reduced 
by other means, the non-occurrence of the photo-reduction must, 
therefore, be looked for in the particularly unfavourable condition 
of the CO-group in connexion with the influence of other groups 
in regard to the light. 
The condition of the CO-group in the ketone molecule as deter- 
mined by all surrounding molecules will regulate the power of 
rendering active alcoholic hydrogen atoms. Moreover, a reciprocal 
reaction between fwo ketone molecules and one alcohol molecule 
must be able to take place. 
The catalytic aleohol-oxidation with the co-operation of the CO- 
1) It is presumably a quantitative difference; o-methyl-benzophenone does it 
slowly so that we have just been able to show it by means of the oxidation 
reaction; fluorenone acts much slower still 
