( 264 ) 
The three-phase line on the ammonia side runs at low temperatures 
far below the liquid-vapour line of .ammonia. 
It ascends very rapidly in the neighbourhood of 110°, intersects 
the ammonia line in the P-T-projection, and runs almost vertical at 
117°. I have been able to follow this line towards higher pressure 
to 225 atmosphel-es without any indication of a critical phenomenon. 
So the point q lies at a pressure which is higher than 225 atmospheres. 
In the point N, which lies at about 50°, contact takes place of 
the minimum pressure line and the three-phase line. Hence the P-x- 
sections have a minimum pressure in the L-fPlines at temperatures 
above 50\ Of the sublimation line of the compound the situation 
in the neighbourhood of the maximum sublimation point was accurately 
determined; this latter point lies at 88.4° and 19.3 atmospheres; 
the melting-point could not be observed so accurately as a slight 
action of the mercury on the mixture in consequence of the generated 
hydrogen can give rise to a great error in the pressure; it lies at 
116 a 117°, and at a pressure which is higher than 150 atmospheres. 
While in the few cases in which the two points mentioned were 
accurately determined, the differences in pressure and temperature 
were slight, a difference occurs here of 28° and more than 130 
atmospheres; this system is the first and only case in which such a 
situation was met with. 
By the aid of closed tubes on one side, and Cailletet tubes on 
the other I have succeeded in determining the 7 -a-- projection of the 
part of the critical line on the hydrogen sulphide side (fig. 7), and 
of the three-phase line on both sides (figs. 8 and 9). 
Fig. 7. 
The critical line runs pretty, much straight from the critical point 
of pure hydrogen sulphide (99.6 ) to the critical end-point p (103.0 , 
8.4 % ammonia). This line has been reproduced in fig. 8 with the 
liquid and vapour branch of the two-phase region coexisting with 
the solid compound, the former of which could be accurately deter¬ 
mined, the latter is indicated schematically. The T-#-projection of 
the same region on the ammonia side (fig. 9j presents the melting- 
