impossible according to our above theoretical considerations, as for 
T = 0 always a pressure of coexistence exists (positive or negative). 
The grounds on which Tammann thinks such a course possible, are 
therefore theoretically not to be justified — as far as the points A* 
and M’ are concerned. 
Therefore Bakhuis Roozeboom has substituted the “half” egg for 
the “whole” egg of Tammann (see loc. cit. p. 93, fig. JO); but such 
a course (see fig. 13 of the plate) too is incompatible with theory. 
Bakhuis Roozeboom tried to explain with this diagram why in 
many cases liquids do not crystallise, so that with lowering of the 
temperature a continuous transition would take place from the 
liquid into the amorphous glassy state, without this latter being 
metastable. 
But this phenomenon is also explained by our theory. For with 
sufficiently low value of q 0 , p, becomes smaller and smaller, and 
even with comparatively very low pressures the solid region (see 
fig. 9, upper arrow) will no longer be reached. Finally the solid 
region will quite descend below the line OK and afterwards below 
the axis p = 0 (see fig. 10 and 11), and we have a continuous liquid 
region — viz. from the moment that the line NM touches the line 
OS'SMK (fig. JO), in which then S' and S will coincide- 
Apart from the not accurately indicated course-of the line SM 
past the point M, Tammann’s figure is therefore closer to the probable 
truth (indicated by our fig. 9) than Bakhuis Roozeboom’s figure with 
two pressures of coexistence in B and C at T = 0. In the latter’s 
diagram (fig. 13) the solid state might sometimes be reached by 
increase of pressure; with us, however (fig. 9), only by decrease of 
pressure. 
The absence of the solid state can now be ascribed to four causes. 
1. The liquid mass is already so viscous before the melting point 
(lying on the line SM) is reached, that it passes into the amorphous 
glassy state. In consequence of this the velocity of crystallisation is 
so slight, when the melting point is reached, that no crystallisation 
takes place, at least not immediately. Prof. Jaeger writes me that 
in such cases sometimes after a very long time devitrification takes 
place. So in this case we can do nothing but wait. 
2. The melting point is reached in liquid state, but the pressure 
(e.g. that of one atmosphere), is too high (fig. 9, upper arrow). In 
this case we might try slowly to cool the liquid in a closed glass 
