Inductive Capacity of Electrolytes. 197 



3° and 33° C, in a few cases being carried up to 45°. As 

 soon, however, as the temperature rose much above that of the 

 surrounding air, vapour condensed on the mirror and window 

 before it, and thus rendered observations difficult or impos- 

 sible. In every case, time was allowed for the temperature 

 of the liquid (water or alcohol) to become nearly stationary 

 before observations were made ; otherwise convection-cur- 

 rents made the needle unsteady. However, it was not possible 

 to keep the temperature perfectly steady, and the measure- 

 ments, which are plotted out in Plate VI., figs. 6 and 7, are 

 not so accurate and consistent as though made at one tem- 

 perature. In the case of alcohol the needle was very unsteady 

 above about 26°, and several rough measurements (Nos. 3, 4, 

 and 9) made at higher temperatures are omitted from the 

 diagram ; but none are omitted below 26°, although Nos. 5 

 and 6 were somewhat unsteady. Ordinates represent tem- 

 peratures, and abscissae represent the excess of deflexions 

 above 4*4 centim. in case of alcohol, and 8*0 centim. in case 

 of water. The observations were made in groups in the order 

 of the numbers on the diagram, and extended in the case of 

 alcohol through a period of twelve hours. It will be noticed 

 that the alcohol was carried several times through the given 

 range of temperature ; observations Nos. 2, 10, and 16 fall 

 closely together near 25°, but between Nos. 2 and 10 the 

 alcohol was cooled to 7°, and between Nos. 10 and 16 to 3° 

 (see fig. 6). The observations fall upon a straight line as 

 nearly as could be expected from the circumstances of the 

 case. They indicate a rather large temperature-coefficient, 

 the force falling off about 10 per cent, between 3° (5*20) and 

 26° (4*70). The conductivity, on the other hand, rapidly 

 increases with the temperature, although no attempt was made 

 to determine a value for its temperature-coefficient. Accord- 

 ing to Foussereau, the conductivity of alcohol increases nearly 

 30 per cent, between 3° and 26°. 



In the case of water the observations do not lie as clearly 

 upon one straight line, but the later observations are nearer 

 to a second straight line nearly parallel to the first. This is 

 believed to be due to polarization effects, which increased 

 during the course of the day, the observations having extended 

 over a period of 14 hours. However, the difference between 

 the first and last measurements is scarcely more than one per 

 cent., and the temperature-coefficient obtained from either or 

 both combined will differ but little. The force falls off' about 

 10 per cent, between 3° and 31°, a somewhat smaller change 

 than with alcohol. 



Assuming the specific inductive capacity to be proportional 

 Phil. May. S. 5. Vol. 31. No. 190. Mar. 1891. Q 



