On the Conductivity of Tourmaline for Heat. 427 



And hence on dividing by | a l b. 2 c^d 4i | | aib 2 c 3 d i e 5 (, we have 



1 a&d&fz 1 _ 1 a x c 2 d z e± 1 1 a-J) 2 c z d A f^ \ \ aiC 2 d 3 f 4 \ _q 

 I « Ac 3 ^ 6 1 | aA<¥*4 1 1 a^c^d^ ) | aj) 2 c z d± ( 



Similarly, 



_ 1 (hCidzfi 1 1 giCgja 1 1 aAcsA 1 _ 1 gigg^i 1 



| ajwrf* I I ^ A^3 1 1 a^cA | | aAcg | 

 and 



= 0; 



and therefore by addition, 



1 ^1^2^3^4/5 1 _ I a^d^ 1 [ aAwhfs 1 _ 1 «iC2^3 1 1 « 1^2^3/4 1 

 I a-Jj^d^ I I aj^c^ | | a^b 2 c t d A e h \ \aj) 2 c z \ \ aj) 2 c z d± \ 



I ^ic 2 1 1 a A/3 1 1 ^i,/*2 1 



In the quotient on the left of this identity the determinants 

 may evidently be of any higher order, the right-hand member 

 having then a correspondingly greater number of terms. In 

 Schweins' final result the said determinants are of the 

 order oc. 



[To be continued.] 



XL VII. On the Conductivity of Tourmaline for Heat. 

 By Franz Strenger*. 



THEORETICAL considerations on the causes of pyroelec- 

 trical phenomena led S. P. Thompson and 0. J. Lodge f 

 to the supposition that tourmaline possesses a unilateral con- 

 ductivity for heat in the direction of the principal axis, and 

 that therefore the conductivity in the direction from the ana- 

 logous pole to the antilogous pole is different from the conduc- 

 tivity in the reverse sense. To verify their assumption they 

 made two different series of experiments. In one of these 

 they used De Senarmont's method: — Slices of the crystal, cut 

 parallel to the axis, were covered with thin layers of wax, and 

 the curves were then observed to which the wax melted when 

 a point of the plate was warmed by applying a hot wire. They 

 then found, as they had supposed, that the isothermal curves 



* Translated from the Annalen der Physik und Chemie, vol. xxii. p. 522 

 (1884), by Fred. II. Hatch. 



t S. P. Thompson and O. J. Lodge, Phil. Mag. [5] viii. p. 18 (1879). 



2 F2 



