Conductivity in Tourmaline Crystals. .19 



resistance to the passage of electricity (or heat) is greater or 

 less when the flow is in the sense AB than it is when in 

 the sense BA, such a substance possesses a unilateral con- 

 ductivity for electricity (or heat, as the case may be) in the 

 given direction. Some apparent cases of unilateral conducti- 

 vity for electricity had been described by Dr. A. Schuster 

 (vide Brit. Assoc. Bep. 1874). It was to be expected that a 

 phenomenon of unequal heating, the analogue of the unequal 

 electrification of the tourmaline when warmed, would be found. 

 It was also imagined, by a reversal of the known pheno- 

 mena of pyroelectricity, that a pyroelectric crystal when elec- 

 trified from without might have its ends unequally warmed. 

 If the development of opposite electrical states at the two 

 ends, and the establishment of a difference of potential between 

 them, were a result of a unilateral conductivity, all the ana- 

 logies of the conduction of heat and electricity pointed to the 

 probability that the tourmaline would be found to possess a 

 unilateral conductivity for heat also. 



The first named of the authors, therefore, proposed the fol- 

 lowing experiment. Let a slice be cut from a tourmaline 

 crystal having its two faces principal planes of section of 

 the crystal, and therefore containing the crystallographic, 

 optic, and pyroelectric axis. Let the slice be covered with 

 wax, and let it be warmed by a hot wire inserted in a central 

 hole after the method of De Senarmont. The tourmaline we 

 know to be a negative uniaxial crystal ; and the isothermal 

 line marked out by the melting of the wax will be an ellipse 

 having its minor axis along the crystallographic (and optic 

 and pyroelectric) axis. If the crystal, however, possess uni- 

 lateral conductivity for heat, the isothermal lines will be no 

 longer symmetrical about the point of application of heat, but 

 will be displaced along the crystallographic (and optic and 

 pyroelectric) axis toward one extremity. We therefore, as a 

 preliminary trial, procured such a slice of tourmaline (which 

 we will call tourmaline "A") from Mr. Ahrens. It was 

 roughly circular, of about 2 millims. thickness, and measured 

 along the axis 25 '3 millims., across 25*6 millims. Experi- 

 ment proved that when the point of a hot silver wire was in- 

 troduced into the central hole, the isothermal line bounding 

 the melted area possessed the form of a distorted ellipse, 

 always displaced toward the analogous pole of the crystal. 

 Two series of measurements were made : — one, of the areas 

 marked out by the melting of the wax; the other, using Meu- 

 sel's double iodide of copper and mercury, which changes at 

 94° (circa) to a black tint. This method gave isothermals 

 of a higher temperature than the wax. The extremities of 



C2 



