199 
acid, will no longer have mirror images that cover each other, 
which can easily be demonstrated by means of the well-known 
carbon models. 
Only two positions of the anti tartaric acid are symmetrical, viz. that in which 
the equal groups lie on the same side of the axis of the central C-atoms as close 
‘as possible to each other and that in which they lie on both sides as far from 
each other as possible. It would be quite accidental, that one of these positions 
should be the most stable state of equilibrium under all circumstances. 
This seems in conflict with experience. We should, however, 
bear in mind the following considerations. In view of the limited 
number of isomers it has already long been assumed that the molecule 
halves can move round the single bonds as axes in opposite direction 
or with different velocities. 
When we supplement this necessary condition with the hypothesis 
that these movements also continually take place in this sense that 
the most stable position of all the positions possible at the moment 
will occur most frequently, and that, therefore, the molecule executes 
irregular rotations or oscillations round this position, there is no 
longer any contradiction with experience. 
When we consider these rotations round the bond of the central 
carbon atoms of the anti-tartaric acid, we see that in one full 
rotation of one part with regard to the other, the molecule will 
twice pass a symmetric position. | 
At these moments the molecule is optically inactive, and since 
now the chance has become equally great that the stable asymmetric- 
position will be reached by a movement in the direction of the 
hands of a clock or in opposite direction, a continual racemisation 
will take place. 
Though the state of equilibrium is asymmetric, we shall probably 
never observe the optical activity in liquid or dissolved state. 
This dynamic representation, which is forced upon us by the 
inactivity of the anti-tartaric acid, applies, of course, to all saturated 
molecules. Our observations of the behaviour of the @ glycols towards 
boric. acid and acetone can be explained most simply in this way. 
The saturated non-cyclic a-glycols do not increase the conductivity 
of the boric acid; we have drawn the conclusion from this that the 
hydroxyl-groups are not favourably situated, and have accounted 
for this by the assumption that the OH-groups repel each other, 
through which they can get as far as possible from each other 
owing to the mobility of the molecule. 
According to the dynamic representation this position will be the 
