2 C. Bar us — Change of Heat Conductivity on 



longed indefinitely below the melting point in the former case, 

 and indefinitely above fusion in the latter case (solid) ; and I 

 can estimate the amount of error involved in such a supposi- 

 tion by actually comparing any given class of properties of 

 the liquid above and below the melting point, by direct ex- 

 periment." 



I may add finally that the application of pressure would 

 enormously complicate the method of measurement, which 

 even in its simplest form, is not without grave difficulties. 

 Indeed the task of keeping thymol liquid between 4° and 15° 

 in a copper vessel, proved to be most wearisome, and out of 

 very many experiments only relatively few were obtained 

 under thoroughly trustworthy conditions. 



2. Method employed. — Seeing that the available substances 

 are usually of low absolute conductivity (below k = 500/1 6 ), 

 and keeping in mind the desideratum of a method which could 

 eventually be used for bodies under pressure, I found none 

 better suited to the present problem than the one which H. F. 

 Weberf has so brilliantly and elaborately worked out. Weber 

 places a thin, wide, plane-parallel plate or layer of the sub- 

 stance to be examined between and in close contact with two 

 thick plates of copper, and it is proved that these are identical 

 as to temperature, with the upper and lower isothermal of the 

 layer. §9. The system is first heated so as to be at a given 

 temperature throughout. It is then suddenly and permanently 

 cooled at the lower surface (copper plate), and the time rate at 

 which heat travels from the top plate to the bottom plate, 

 through the intervening layer, is measured by a thermo-couple. 

 From these data the absolute thermal conductivity of the layer 

 may be computed when the constants of the system are known. 

 To obviate convection in the case of liquids, all plane surfaces 

 are placed horizontal, and cold surfaces are lowermost. 



To cool the bottom plate Weber either bedded the system 

 on a plane of ice, surrounding it by an ice environment, or 

 lowered its temperature by a shower bath of hydrant water 

 with a corresponding environment. It is not feasible to apply 

 this method directly to thymol : clearly, in the first place all 

 manner of disturbance or handling must tend to freeze the 

 under-cooled liquid. In the second place freshly distilled 

 thymol absorbs water or aqueous vapor at an initial rate of 

 •00032 g per cm 2 of free surface per hour, and the interval of 

 undercooling is thereby decreased. At least in contact with 

 copper in air, thymol soon becomes colored and unsuitable for 

 the experiments. Failing therefore utterly in the attempt to 



* The continuation of the present research must for the above reasons await 

 the warm weather. 



•j-H. F. Weber: Wied. Ann., x, pp. 103, 304, 472, 1880. 



