191 



after heating and subsequent cooling has been reproduced (Mallard), 

 which shows the lamellar structure quite clearly. At 265 C. the bire- 

 fringence disappears suddenly, and reappears without retardation 

 if the crystal be cooled down below that temperature. The optical 

 behaviour of plates cut parallel to 110} and (100} is schematically 



{110} II {110} 



Fig. 148. 

 Boracite. Plates parallel to {100} and {HO}. 



II {100} 



shown in fig. 14.$, while in fig. 14.9 a pseudo-rhombicdodecahedron 

 of boracite is reproduced, and the arrangement of the component 

 rhombic individuals is indicated by the direction of their axial plane. 

 Every face of the rhombicdodecahedron is the base of a rhombic 

 pyramid, with its top lying in the centre of the crystal; the biaxial 

 individuals have their optical axial plane 

 parallel to the longer diagonal of each rhom- 

 boid. By means of Ron t gen-rays, patterns 

 for plates parallel to 100], $110}, and 

 111} were obtained by us 1 ) at room-tem- 

 perature, which were in accordance with 

 the symmetry of a cubic space-lattice, but 

 also others which, when parallel to 100}, 

 only manifested a binary axis with two per- 

 pendicular planes of symmetry; the last 

 plate, however, when heated to 300 C., and 

 then passed by a pencil of R on t gen-rays, gave a pattern, the 

 symmetry of which was that of a true cubic crystal-plate. 



Although definite conclusions cannot yet be drawn from these 

 results, the last mentioned experiment nevertheless seems to support 

 the explanation given by Mallard. In the case of leucite we were 

 not able to obtain Ron t gen-pat terns at all, whose symmetry was 



Fig. 149. 

 Boracite. 



1 ) H. Haga and F. M. Jaeger, Proceed. Kon. Akad. van Wet. Amsterdam, 

 Vol. 16, 792, (1914). 



