1642 
place after the oxygen had been consumed and the eolour had 
almost completely faded. 
After some time, beside the residual terephthalophenone another 
non-coloured substance began to deposit; this went on until the 
original terephthalophenone had entirely disappeared and the liquid 
again was almost quite decolorized. 
To judge from the molecular weight, the colour reaction with 
potassium hydroxide (to which we will refer later) and the decom- 
position on heating the new product consisted of the terephthalo- 
phenopinacone. 
When we place the yellow solution in the dark it keeps unchanged 
for months; if, however, air is admitted it is also very rapidly 
decolorized. 
Finally it should be remarked that if an inert solvent but no 
alcohol is present, no yellow coloration sets in, so that the aleohol 
is a necessary condition for this colour reaction. Taking into con- 
sideration our previous results we may conclude from the foregoing: 
1st. The ketone absorbs a part of the violet rays. 
2nd, The following photo-reaction takes place: 
terephthalophenone + alcohol + little light = yeilow sub- 
stance in which the two hydrogen atoms forming part of the 
alcohol molecule are rendered active. 
This substance is permanent in the dark but is decomposed by 
oxygen as follows: 
yellow substance + oxygen = terephthalophenone + aldehyde 
(ketone) + H,O. 
3. If the yellow substance is further illuminated it is converted 
according to the reaction: 
yellow substance + much light = terephthalophenopinacone + 
+ aldehyde (ketone). 
After this elucidation it was only natural to ascertain whether in 
the case of the other ketones, where the yellow colour is not so 
pronounced or could not even be observed, the progressive change 
was the same. 
Indeed it could be ascertained that if reduction to pinacone did 
occur, there could always be noticed absorption of oxygen should 
this gas be present. 
To demonstrate this, a little apparatus was employed consisting 
of a vessel of 20 em. capacity in which was placed a hollow stirrer 
with T-piece and opening. By the rotation of the stirrer the gas 
from the gaseous space was forced through the. liquid and kept this 
saturated with air. The stirrer was also connected air-tight with the 
