206 Br. L. T. More on the Supposed Elongation of 



that there is always a fall in the tube which is entirely inde- 

 pendent of the kind of liquid in the thermometer. Quincke 

 observed an effect analogous to Govi's, but when the bulb 

 alone was filled with a different liquid and the capillary 

 always with pure water, the variations were eliminated. He 

 concludes that the variations were due to differences of sur- 

 face-tension, capillarity, viscosity, and electric conductivity. 

 But it should be noted that his original variations were in the 

 reverse order to Govi's, as the change with water is here 

 greater than with alcohol. Mercury often showed no change, 

 and turpentine never. In the table*, inserted here for 

 reference, he gives this effect for water and mercury, but 

 unfortunately impairs the value of his deductions by not 

 using the same thermometer in the two cases. 



6 Leyden Jars. 



Therm. ^ * N 



No. 11. 40 20 



Water 5-68 1 2*10 



Mercury... 3*41 1-25 



3 Jars. 



Therm. 

 No. 16. 



Mercury... 



2-787 

 2-728 



0-755 

 0-5 L0 



20 

 2-669 

 2-357 



10 



0-789 

 0-615 



The first table is for the case when liquid fills both the bulb 

 and the capillary ; the second, when water alone is nsed in 

 the capillary. It will be seen that the variations are not 

 accounted for, and indeed Quincke states that this remaining 

 variation must be due to the difference of the conductivity 

 for heat of the two liquids. But this is hardly an adequate 

 explanation, as the variations in the second table are greater 

 for the smaller charges of electricity. It also seems to cast 

 a doubt on the cause of the expansion, for if the relative con- 

 ductivities for heat of mercury and water can change an 

 observation from 051 to 0'75, the entire heating effect of the 

 charge might well cause serious changes in the amount of 

 deflexion observed. 



Experimenting with changes of volume in liquid dielectrics, 

 Quincke found that most of them expand when charged, but 

 that others, as the fatty oils, contract. Roentgen repeated 

 these experiments, using two different methods, and found in 

 no case a contraction. He finally concludes that all cases 

 may be accounted for by an increase of the temperature, and 

 the attraction of the charged surfaces. 



* Quincke, Wied. Ann. x. p. 177. . ..--.•■ 



f Expansion is in inillionths of the original volume. The numbers 

 40, 20, &c, are the relative charges from a unit jar. 



