S56 Dr Brewster on a New Specks of Double Refraction, 
The most common form of the Analcime is the icositetraJie-^ 
drouy a solid contained by twenty fmir equal and similar trape- 
ziums', and formed by three truncations on the eight solid angles 
of the circumscribing cube. If we suppose this cube to be dis- 
sected, by planes passing through all the twelve diagonals of its 
six faces, it will be reduced into twenty-four irregular tetrahe- 
drons. The same planes divide the icositetraliedron into twenty- 
four similar pentahedrons. 
If we transmit polarised light through the mineral in a direc- 
tion perpendicular to any of the faces of the cube, we shall find 
that all the dividing planes now mentioned, are planes of no 
double refraction and polarisation, that is, that they consist of 
an infinite number of axes parallel to the four axes of the solid. 
When any of the axes of the cube are placed in the plane of 
primitive polarisation, the tints will disappear, and continue in- 
visible while the ci^,stal is made to revolve round that axis ; but 
when the axis k inclined 45° to the plane of primitive polarisa- 
tion, or when the diagonals of any of the cubical faces are in 
that plane, we observe a black cross AB, CD, Plate VII. Fig. 9. 
separating four luminous sectors covered with the tints of polar- 
ised light. 
In order to determine the character of these tints, we have 
only to cross them with the axis of a plate of any crystal, the 
character of whose action is determined; When the polarised 
tints are crossed by a plate of sulphate of lime, having its axis 
inclined 45° to the arms of the black cross, the tints all descend 
in the scale, and consequently the polarising action of the crystal 
is Negative in relation to each of the four axes of the icositetra- 
hedron. 
In every part of the crystal, the polarised tints are exactly 
those of Newton’s scale, and have all the properties of the tints 
of moveable polarisation. 
From an attentive consideration of the experiments, as detailed 
in the original memoir, it is obvious that the phenomena of the 
tints exhibited in any individual sector, Plate VII. Fig. 9., have 
no relation to the axis of the icositetrahedron passing through O, 
considered as an axis of double refraction. The axis of polarisa- 
tion of every portion in each sector, as COB, is, on the contrary, 
perpendicular to the line GB, or parallel to one of the rectangu- 
