4 02 SCIENCE PROGRESS 



blende, it is parallel to the longer diagonal of the rhombic 

 dodecahedron. The transformation to the negative form is 

 retarded on the crystal surface, and, where it does occur, the 

 optic axis is often perpendicular to that in the still positive 

 crystals. The retarded transformation generally occurs about 

 the temperature of isotropic fusion. In all the other liquids 

 the optic axis is normal to the crystal surface. 



The same author (ibid. 280-3) nas applied a modified form 

 of the magnetic theory of Langevin and Weiss to anisotropic 

 liquids and concludes that, at the point of contact of these 

 with solid bodies, the anisotropic condition should persist above 

 the temperature of isotropic fusion, and therefore after the 

 remainder of the liquid has become isotropic. This is explained 

 by the existence of local molecular fields produced by the solid 

 round the points of contact and equivalent to the exterior field 

 of the magnetic theory. That this is in agreement with the 

 facts is proved by the experiments described in another paper 

 (ibid. 431-4). There it is shown that, on heating an anisotropic 

 liquid above the temperature of isotropic fusion, and then 

 allowing to cool, the doubly refracting regions appear at the 

 same points, and with the same orientation, as before. On 

 passing the transition temperature during heating, these regions 

 can still be detected in polarised light, and those residual 

 traces, which the author terms " pellicles," have the same 

 orientation as the original anisotropic regions. 



Crystal-growth. — The accuracy of the experiments made by 

 Becker and Day in 1905, to show that a crystal growing under 

 a load developed a linear force in the direction of the load, was 

 questioned by Bruhns and Mecklenburg in 191 3. The former 

 authors have therefore repeated these experiments (Journ. 

 Geol. 24, pp. 313-33, 19 16) and have shown that their results 

 were correct, and that the comparative method of the latter 

 does not test the point at issue. Taber (Proc. Nat. Acad. Sci. 

 3, pp. 297-302, 191 7) has obtained definite evidence that pres- 

 sure phenomena accompany the growth of crystals of am- 

 monium nitrate. The importance of this in connection with 

 the growth of crystals in veins and concretions in rocks, which 

 are difficult to explain otherwise, is indicated. 



Wright and Hostetter (Journ. Wash. Acad. Sci. 7, pp. 

 405-17, 191 7) have made numerous observations on the growth 

 of crystals under strain with reference to the reversibility of the 



