( 251 V 
care, and whs entirely rmmerged in a thermostat. From 50 J to 140° 
a slightly different method was followed; the mercury manometer 
in the thermostat was used as an indicator, as is usually done, while 
the pressure was read on an open manometer, which was connected 
with -i the small manometer by means of a T-joint with cock, and 
was placed outside the bath, the pressure being kept equal to the 
three-phase pressure (by means of a cycle-pump). 
The investigation from + 50° to +140° gave the following result. 
(See table p. 250). 
These observations, which harmonize very well with those of 
Tammann up to 100°, show that this three-phase curve does not 
present any irregularity up to 140°. 
As pressures above 3 atmospheres were dangerous for the apparatus 
which was used up to 140°, we now adopted the method with the 
Cailletet tube, which was up to 217° of glass of a high melting point, 
and above this temperature of quartz, because glass is rapidly attacked 
by water above 200°. 
Unfortunately, however, quartz was not the right material for the 
observation at high temperatures either; for though quartz was not 
changed at all by water, even at the critical temperature, it was 
attacked by a solution of Na 2 S0 4 under for the rest similar circum¬ 
stances [probably in consequence of hydrolysis of the salt], which 
caused the quartz tubes to burst repeatedly, and which rendered the 
continuation of the vapour tension determinations, which require a 
comparatively long time, impossible above 260°. 
From 140° to 280° the investigation gave the following result. 
(See table p. 252). 
If we compare these pressures with those of pure water at the 
same temperature, we notice in the first place that as was to be 
expected in connection with the slight solubility of the natrium- 
sulphate, the differences are small, and in the second place it strikes 
us that, these differences reach a maximum value at ± 234°, which 
points to a point of transition, which has been really found by 
Huttner and Tammann 1 ) at 235°, and by Boeke 2 ) at 239°. Shortly 
after Nacken s ) showed in an accurate investigation that the results 
of Huttner and Tammann are nearest to the truth, as he found 234° 
for the point of transition. His experiments revealed further that the 
modification of Na 2 S0 4 is rhombic below 234°, and monoclinic 
above it. 
l ) Zeitschrift. An. Ghem. 43, 215. 
*) Thesis for the doctorate, Amsterdam 1906. 
) Neues Jahrb. fur Mineralogie, Beilageband 24, 27 (1907). 
