638 Scientific Proceedings, Royal Dublin Society. 
Over against these conclusions we may set the facts regarding — 
leucite.? gy 
The pseudo-holohedral-cubic erystals of this substance are — 
marked with numerous striz, which are due to the presence of — 
twin lamelle, whose direction is that of the rhombic-dodecahedral — 
faces, whether such are actually present on the crystal or not; — 
there is also optical evidence of the presence of thin lamelle which — 
have apparently served as gliding planes in the transition to the 
lower symmetry.° 
The following is evidence that the crystals have undergone 
such a transition. 
At a red-heat they become isotropic, all the surface markings 
which are attributable to the presence of twin lamelle: vanish, and 
every face becomes as to form and position uniform throughout 
and free from deformation ; internal irregularities also disappear. 
When reversal takes place local peculiarities are to some extent, 
but not entirely, found reappearing at the same spots. In the 
case of thin plates especially the alterations which present them- 
selves on the reversion to the former symmetry are considerable. 
Three principal interpenetrant individual portions are in many 
cages traceable by their optical properties, but they are commonly 
of various magnitudes; they approximate in the more perfect 
examples to square right pyramids meeting at their vertices, and 
of which the opposite ones, two and two, have their general struc- 
ture similarly orientated, and so appear to form a single individual. 
At ordinary temperatures the angles differ somewhat from the 
values proper to the cubic symmetry. 
Like leucite, one of the uranyl double acetates shows twinning 
lamellee when in the less symmetrical of its two dimorphous 
forms, and loses them as the temperature is raised. And the 
behaviour of re-entrant angles but slightly removed from 180°, 
which are found on the rhombohedron faces is, as a certain critical 
point is reached, particularly instructive, the indented edges moving 
outward and forming straight lines as isotropy sets in, and the 
indents in the rhombohedron faces at the same time vanishing. 
The process is reversible. 
1 Brauns, loc. cit., p. 106. 
2 This is Rosenbusch’s suggestion. Comp. Jdid., pp. 51 and 109. 
