428 M. F. Strenger on the Conductivity 



were not true ellipses, but appeared distorted in the direction 

 from the heated point towards the analogous pole. 



Such observations gave for the ratio between the semiaxes 

 of the isothermal curves, along the principal axis, the mean 

 value of 100 : 129. 



In a second series of experiments, a tourmaline slice, cut at 

 right angles to the crystallographic axis, was fixed between 

 two glass tubes ; into the upper of these a weighed quantity 

 of mercury was poured, while into the lower steam was blown. 

 The time was observed which was required to raise the tem- 

 perature of the mercury through a given number of degrees. 



A difference in the conductivity, according to the direction 

 in which the heat flowed through the tourmaline, was also 

 observed by this method of experimenting. The mean values 

 of the times required to produce an equal elevation of tempe- 

 rature bore to one another the ratio of 100 : 119. 



The few measurements made by Thompson and Lodge 

 differ so widely from one another, that the existence of a 

 unilateral conductivity cannot be inferred with certainty from 

 these experiments. Above all, it should first have been accu- 

 rately ascertained that the slices used were perfectly homo- 

 geneous; that is, that the crystal was not made up of a number 

 of crystalline individuals so grown together as to have the 

 analogous and antilogous poles, of different crystals, lying 

 beside one another in the same surface. It was therefore 

 desirable to find a better method of experimenting, to decide 

 whether tourmaline possesses this abnormal conductivity for 

 heat or not. 



The experiments made by me for this purpose were per- 

 formed upon two slices cut at right angles to the principal 

 axis. The one, about 5*8 millim. thick, was cut from a crystal 

 from Brazil, belonging to the Mineralogical Institute of this 

 University, and was placed at my disposal by Prof. Bucking; 

 the other from a detrital peeble from Ceylon. Neither of them 

 showed any trace of a fibrous structure ; and a careful pyro- 

 electrical examination, according to Kundt's* method (which 

 is founded on the principle of Lichtenberg's figures), showed 

 that the two slices were in fact absolutely homogeneous; for the 

 crystal having been heated, and, while cooling, dusted over 

 with the proper mixture, showed one of the two surfaces per- 

 fectly yellow, the other perfectly red. 



The method employed in experimenting on the conductivity 

 for heat has been adapted from one used by F. Weberf for 

 liquids. A thermoelectric junction, fixed in a copper cylinder, 



* Wied. Ann. xx. p. 592 (1883). 



t F. Weber, Wied. Ann. x. p. 103 (1880). 



