Mr. L. Fletcher's Crystalhgrapkic Notes, 479 



which is not concealed by the matrix : no fewer than ten faces 

 of this semiform it (3 1 2) can be seen, while not a single face 

 of the complementary semiform ir (3 2 1) is to be found. 



Gustav Rose has remarked that, in cases of diplohedral 

 hemisyinnietry, the poles of all the faces present belong in 

 general to one set of systematic triangles: this crystal of Skut- 

 terudite forms another of the rare exceptions to this rule, 

 which would require the association of it (3 2 1) instead of 

 7T (3 1 2) with the semiform it (3 1 0). 



This semiform tt (3 1 2) is again to be observed on crystal 

 No. 13, though it is there represented by only a single face. 

 The angle with the adjacent face of the form (2 1 1) was in 

 this case measured to be 10° 52^, a result according well with 

 the calculated angle 10° 54/ : on the same crystal eight faces 

 of the semiform it (3 1 0) are developed. 



The last crystal, No. 14, differs from the rest in showing 

 three faces, which, if the crystal were simple, would un- 

 doubtedly be attributed to the complementary semiform 

 7r (1 3 0) ; the crystal, however, still presents a hemihedral 

 habit, since the three faces of this complementary semiform 

 7r (1 3 0) only appear at quoins where the faces of tt (3 1 0) 

 are missing. It is very possible indeed that the crystal may be 

 twinned about the normal to a dodecahedron-face, as is at 

 times the case in iron pyrites — which theory would likewise 

 account for the presence of reentrant angles, otherwise to be 

 attributed to parallel growth. 



The subordinate forms (3 3 2), (6 4 3), described by vom 

 Rath, do not seem to be present on any of the above crystals, 

 and must be very rare. In some cases indeed the edges of 

 intersection of the octahedron with the form it (3 1 0) are 

 " rounded off" by very small faces not susceptible of mea- 

 surement; while the edges of the octahedron itself are bevelled 

 by narrow planes, which, if crystal-faces at all, can only be 

 approximately determined. On one crystal the angles made 

 with the adjacent octahedral faces by the four tautozonal 

 rudimentary planes of the triakisoctahedron were measured 

 by the method of maximum illumination as 18^°, 17^°, 13°, 

 16|° respectively — thus suggesting the form (2 2 1), which 

 requires an angle of 15° 48', and, as we have seen, was given 

 by Miller, no doubt in mistake, as an observed form. On 

 another crystal a series of images could be obtained from each 

 of two faces of the triakisoctahedron ; and the limiting-values 

 thus determined for the angle corresponding to the one just 

 mentioned were, in one case 13^°-16£°, and in the other 

 15f°-16^° ; on a third crystal two similar angles were mea- 

 sured at 15£° and 19°, thus again indicating the form (2 2 1). 



