( 622 ) 
which is closed on one side, does not present any special particul- 
arities, and starting from the open side a bitangent plane may be 
rolled regularly over the binodal curve as far as the plaitpoint. 
There are, however, also cases where we meet with complications, 
and already in my “Théorie moléculaire’ 1 have allowed in my 
description of the y-surface, for the possibility of the existence of a 
branch plait by the side of the main plait. If two plaits exist 
simultaneously over a very great range of temperature, we may 
properly speak of a transverse plait and a longitudinal plait, and 
the non-miscibility in the liquid state may be ascribed to the long- 
itudinal plait. But if these two plaits occur only over a small range 
of temperature, it is better to speak of a main plait and a branch 
plait; I have chosen these names, because really in such cases 
one of the plaits may be considered as main plait, and the other 
only as branch plait. But, what has not been observed as yet, 
the circumstance may occur, that at a certain temperature these 
two plaits reverse their parts. What was a branch plait, becomes 
a main plait, and the main plait is reduced to a branch plait. 
In saying this I have chiefly in view the description of the 
modifications to which the »-surface is subjected with change of 
the value of 7, to account for the observations of KueNeN on the 
critical phenomena of mixtures of ethane and some alcohols. 
These mixtures have, for a value of 7’ only little greater than 
T;. of ethane, a plait on the y-surface with a continuous course 
without any complication. But with rise of 7’, besides the plaitpoint 
on the ethane side, a new plaitpoint appears lying more to the side of 
the alcohol. So from this temperature 7, we may speak of a three- 
phase-pressure. With further rise of 7’ the new plait extends, and 
at a certain higher value of 7’ = 7,, the first plaitpoint disappears. 
Then the three-phase-pressure vanishes, and from that moment the 
plait has resumed its simple form. Between 7, and 7, we have, 
therefore, a plait with two plaitpoints. If referring to a plait we 
speak of a base and a top, we have between 7, and 7, a plait 
with one base and two tops. Beyond the limits of 7’ equal to 7, 
and 7’,, the plait has only one base and one top. But whereas just 
above 7, the top which has newly appeared, extends but little 
beyond the binodal curve of the original plait, at a higher value of 
T this top will extend further; the top on the ethane side contracts, 
and disappears altogether at 7, and as we shall show, disappears 
as a branch plait. 
As therefore the plait appearing at 7’, is originally a branch 
plait, a transformation must take place with increasing value of 7’ 
