( 919 ) 
the meniscus vanishes between the two other phases in the middle 
of the tube under critical phenomena. So it is this temperature and 
pressure which are indicated by fig. f. These appeared to be 37°.3 
and 83.15 Kg. per eM’. The x of the point with horizontal tangent 
is, therefore, smaller than 0.0245. 
As mixture III taught us, however, this 2 is greater than 0.0100. 
For this mixture of the concentration mentioned yielded the three- 
phase-pressures mentioned in the table up to 37°.1, but at 37°.2 
three-phase-pressure was no more observed. The dense liquid (3), 
(which distinguished itself so greatly from the other liquid by its 
viscosity, colour, and refractivity that they could not possibly be 
confounded) which formed already at very low pressure, increased 
to a maximum, then decreased, and had entirely. disappeared at 
83.00. Then the mixture remained homogeneous up to the highest 
attainable pressures. So at 37°.2 and «= 0.0100 we are clearly in 
the case of fig. ¢, on the left side of 1. At 57°.1, however, three- 
phase-pressure appeared again, so the point 1 moves from left to 
right between 37°.1 and 37°.2 past the concentration «= 0.01. At 
37°.2 no double retrograde condensation can, of course, be observed, 
because then the line for the equilibrium 1, 2 no longer inclines to 
the left, which is by no means astonishing, as it has already entirely 
disappeared O°.1 higher. The point with horizontal tangent of fig. 7 
lies, therefore, evidently between « = 0.0100 and «= 0.0245, and 
that probably much closer to the first-mentioned than to the last- 
mentioned value, as already at 37°.1 with increase of the pressure 
the meniscus disappears in the top of the tube, i.e. the plaitpoint lies 
already on the left side of «= 0.0100. 
As a fourth mixture (w = 0.00765) did not reveal anything new, 
we immediately passed on to a fifth of «= 0.00540 and then to a 
sixth of w= 0.003875. In both we could very clearly ascertain the 
retrograde condensation of the dense liquid observed already for the third 
mixture, but we did not succeed in finding double retrogade con- 
densation. For in order not to get on the line of the three-phase- 
pressure, the temperature had even to be raised above 36°.15 for 
the mixture of «= 0.00375. Then, however, after the dense liquid 
had disappeared everything remained homogeneous up to the highest 
pressures. At 36°.10 the three-phase-pressure reappeared. The point 
1 of fignre d, where the tangent to the branch plait is vertical for 
the three-phase-pressure is, therefore, still on the left of 0.00375. 
We have resolved, at least for the present, not to continue our 
investigations at still smaller «2. Not because the preparation of 
such mixtures would give rise to difficulties, in the way described 
