AT HIGH PRESSURES BY OPTICAL METHODS. 
141 
these determinations the' crystallization at ordinary temperature must 1;)e briefly, 
described. 
When the acid is fused between two glass plates and allowed to cool super-cooling 
invariably takes place. When crystallization sets in a few round crystal germs of 
the modification melting at 51° C. are formed, and these then grow at moderate 
speed in all directions, forming hexagons remarkably regular in shape. These are, 
however, not homogeneous crystals, but are built up of crystal fibres radiating from 
a central crystal germ. If a small particle of the stable modification is brought into 
contact with these crystals, a transition begins to occur and the stable modification 
grows at the expense of the unstable, very much in the same way as this grows in 
the “ melt.” The growth of the stable modification melting at 64° C. is fan-like, the 
crystal threads being very similar to those of the modification melting at 51° C. 
Both modifications show approximately the same interference colours, and this adds 
to the similarity of their appearance when crystallized between two glass plates. At 
ordinary temperatures the velocity of crystallization of the unstable modification is 
greater than the velocity of transformation to the stable form. 
Considerable difficulty was met with in trying to get a sufficiently pure preparation 
for the melting-point determinations. The substance is very soluble in most solvents, 
and the degree of purity is not much improved by repeated crystallization. It was 
at length found that a pure product may be obtained by treating the acid with a 
large quantity of petrolether at ordinary temperature, and evaporating part of the 
petrolether by an air current. Crystals of the unstable modification about 1 mm. in 
diameter are formed, but these clear, transparent crystals change, while still in the 
evaporating solution, into the stable form whereby they become porcelain-like, white 
and non-transparent. 
Unfortunately a great many measurements on unstable melting-points were made 
on a specimen that was not quite pure, and these had then to be repeated on a pure 
one. But as some interesting observations were made during these measurements 
with the less pure material they are given here in Table YU., beside the measurements 
carried out with the pure substance, Table YI. 
(l) Determination of the Melting-jmint Ourve of the Stable Modificatioyi .—Under 
pressure the molten acid behaves very much in the same way as at ordinary pressure 
in this respect, that it almost invariably crystallizes to the unstable modification. It 
is therefore not possible to work at constant pressure, and a curve connecting the 
points where this modification is spontaneously formed has therefore not been deter¬ 
mined. In order to determine the melting-point curve of the stable modification it 
is necessary to melt the acid between glass plates, and when it has crystallized to 
the unstable form, to inoculate with a crystal of the stable form, whereby the 
transition takes place, and then to bring the preparation into the pressure apparatus. 
Care must be taken during the whole series of measurements not to melt the stable 
form totally, as we then nearly always pass over to the unstable curve, when the 
