78 



The author has sought a variation of viscosity by observing tlie rate of de- 

 scent of small drops of mercury through castor oil, which served as the dielectric 

 n a plate condenser. The condenser consisted of a tall, glass tank, the middle 

 part of which served as a condenser, tin-foil being glued to the middle half of the 

 outsides of the glass plates. To eliminate temperature effects the ratio of the time 

 of descent through the condenser part of the tank to the time of descent through 

 the non-condenser part, the condenser being uncharged, was compared with the 

 ratio similarly obtained when the condenser was charged. In this way any change 

 of temperature affecting the whole tank may be eliminated. To further eliminate 

 anv variation affecting different parts of the tank unequally, a long series of 

 readings was taken with the condenser alternately charged and uncharged ; and 

 each ratio obtained with the condenser charged was compared with the mean of 

 the adjacent ratios obtained with the condenser uncharged. The experiment was 

 performed in a cellar of fairly constant temperature, temperative effects being 

 thus almost perfectly eliminated; long series of readings made as described showed 

 invariable increases of viscosity on the application of electro-static stress. The 

 variation of viscosity seemed to be dependent rather on a non-uniform or varying 

 electro-static field than on a steady field. Castor oil and glycerine showed an in- 

 crease of viscosity. 



As the above method could only be applied to very viscous liquids, the au- 

 thor also constructed an analogous apparatus on the capillary tube principle 

 suitable lor mobile liquids. It consisted of a capillary tube placed vertically be- 

 tween condenser plates, and connected above to a large tube with four constric- 

 tions in it dividing it oft' into three compartments. The times of emptying of the 

 compartments by fiow through the capillary tube were observed, the condenser 

 being first uncharged during the emptying of the middle chamber and then 

 charged. If the ratio of the time of emptying of the middle chamber to the sum 

 of the times of emptying of the other two be taken, the condenser being uncharged, 

 and compared wi h the ratio similarly obtained, the condenser being charged, a 

 method free from temperature effects is again obtained for detecting a viscosity 

 variation. In this way it was found that under a varying electro-static field, car- 

 bon di-sulphide showed an increase of viscosity and paraffine oil a decrease. 



The above methods are being applied to other liquids, and a determination of 

 the law of variation of the effect discovered under varying strengths of electro- 

 static field will be made later. 



