Vol. 4] Thelen. — Thermal Conductivities of Certain Schists. 209 



means of two fine platinum wires. He took readings on a very 

 large number of minerals. Liebisch* says of Jannetaz' attempt 

 to show a dependence of the thermal conductivity on the cleav- 

 age of the crystals, that the conductivity is symmetrical with 

 respect to other properties than that of cleavage, — "class die 

 Warmeleitungsfahigkeit andere Symmetrieeigenschaften besitzt 

 als die Cohasionseigenschaften. " This matter will be brought up 

 again. 



Roentgen! obtained his ellipses by an entirely unique method. 

 He blew his breath upon the polished surface of the crystal, set 

 a warm, pointed rod perpendicularly upon it, removed the rod 

 and hastily spread lycopodium powder upon the surface. 

 Where the moisture from his breath had not evaporated, the 

 powder stuck; and on turning the crystal plate over, the loose 

 powder could be knocked off from the center, thus leaving a 

 negative ellipse with sharp borders. 



S. Thompson and 0. J. Lodge:|: used Senarmont's method to 

 verify the statement that linear conductivity depends upon the 

 direction of flow along the line. They obtained the results they 

 were looking for. The ellipses were elongated toward the 

 analogous pole of the tourmaline (section cut parallel to c) 

 with which they were working. The individual values for the 

 ratios of this elongation varied from 1.17 :1 to 1.42 :1, — results 

 varying so much as to be considered valueless by the writer. 

 More accurate methods of other observers gave determinations 

 of these values in which the maximum difference was under 

 one per cent. ; and we may yet be able to devise means sufficiently 

 accurate to verify experimentally that heat flows with equal 

 facility in either direction along a given line, regardless of 

 pyroelectric or other properties. The only reason this matter 

 is mentioned here§ is to show what the personal equation can 



* Liebisch. Physikalisehe Crystallographie. Veit & Co. Leipzig, 1891. 



t W. C. Roentgen : ttber eine Variation der Senarmont. 'schen Methode 

 zur Bemessung der isothermen Flaeheu in Krystallen. Ann. Pogg. 151, 

 613, 1874. 



W. C. Roentgen: "fiber eine Methode zur Erzeugung von Isothermen 

 auf Krystallen. Zeit. f. Kryst. 3, p. 17, 1879. 



X On Unilateral Conductivity in Tourmaline Crystals. Phil. Mag. (5) 

 S, p. 18, 1879. 



§The matter is discussed in some detail in Liebisch, loe. cit. 



