20 Prof. Thompson and Dr. Lodge on Unilateral 



the minor axis were scratched with a needle upon the surface, 

 and afterwards measured. The following is the series of mea- 

 surements made : — 



Crystal "A." 

 Measurements of Semiaxes minor of Isothermal Curves. 

 Series 1. (Wax.) 

 Seniiaxis — . Semiaxis -f. 



Series 2. 



2-5 3-1 or 100 



124 



3-1 4-4 100 



142 



4-3 5-9 100 



137 



5-6 7-2 100 



129 



6-7 8-4 100 



124 



10 12-8 100 



128 



(Double Iodide of Cu and Hg.) 



2-0 2-5 or 100 : 125 



3-1 4-2 100 : 135 



3-9 5-1 100 : 130 



5-9 6-9 100 : 117 



Mean 100 



: 129-1 



(a) 



(•) 



(d) 



W 



(/) 



(A) 



(0 



(k) 



These results differ somewhat widely from the mean figure ; 

 yet it is significant that they all tend in the same direction. 

 It is to be remarked that in all cases these experiments were 

 made quickly, and that in no one case was a condition of 

 thermal equilibrium established in the crystal between the 

 central gain of heat and the loss by radiation &c. on the outer 

 edges. The result then appeared to be that in a crystal ivlrich 

 was getting warmer the heat flowed more easily towards the 

 analogous pole than away from it. 



It is clear that the isothermal curve produced thus will 

 be one possessing geometrical discontinuity ; for the resolved 

 parts of the radii from the centre of heating parallel to the 

 axis will on one side of the centre be lengthened, and on the 

 other side reduced, in a ratio dependent on the unilateral co- 

 efficient of conductivity. Hence the curve will consist of tAvo 

 semiellipses having a common major axis, but having different 

 semi- minor axes in the ratio given above, viz. on the average 

 1'3. These results were obtained in July 1877. They w^ere 

 not published at the time, because it w^as desired to obtain 

 further confirmation with other specimens of tourmaline. No 

 suitable specimens, however, could be met with ; and the mat- 

 ter rested at this point for some months*. 



* Another tourmaline, a transparent green crystal, obtained from Lau- 

 rent of Paris, has since afforded equally significant results — the isother- 



