Electricity and Light. 339 



across between the two induction-wires, at the rate of about one 

 per second, without cessation of the principal discharge. The 

 insulation was so good that the dielectric was not in any way 

 damaged. The discharges took place through the air in dense 

 white sparks, from end to end of the block, a distance of 7 or 8 

 inches. In these circumstances, the part of the glass block be- 

 tween the induction-terminals must have been subjected to a 

 strain little short of the utmost it could bear. Electric forces of 

 such intensity were hardly ever applied in the experiments, were 

 not indeed required. 



3. The polariscope consists of two Nicol's prisms. A flat 

 paraffin flame, presented edgeways, is used as a source of light. 

 Next to the lamp and close to it comes the first Nicol; then at a 

 distance of 2 feet or more comes the dielectric of plate glass ; 

 then at a like distance comes the second Nicol. The pieces are 

 so arranged that the observer, looking horizontally through the 

 polariscope, and keeping the first Nicol at the centre of the field of 

 vision, sees the flame through the centre of the dielectric, midway 

 between the induction-terminals. The light crosses the dielec- 

 tric at right angles to the plate-faces, and therefore at right 

 angles to the lines of force. 



4. Neutralizing Plate. — Every very thick plate of glass which 

 I have yet worked with, exerts at most of its points a slight de- 

 polarizing action upon transmitted light. When such a plate 

 is inserted between the two Nicols without compensation, any 

 thing like perfect extinction is generally unattainable, and the 

 sensibility of the polariscope is lost. 



The principal section of the first Nicol being in any desired 

 position, and that of the second Nicol perpendicular to it, the 

 dielectric (still unexcited though connected with the coil) is 

 inserted properly between the Nicols, and the light reappears 

 well in the polariscope. 



The neutralizing plate, a piece of glass about 6 inches 

 square, taken from the same original plate as the dielectric, is 

 then placed upon a stand immediately in front of the second 

 Nicol, and is moved by trial into such a position that the re- 

 stored light is again extinguished perfectly by a very small 

 rotation of the analyzer. The apparatus is now ready for 

 work. 



5. The highest powers applied in the experiments are suffici- 

 ently indicated thus : — battery in the primary circuit, a series of 

 six Grove's or Bunsen's elements; corresponding spark-length 9 

 or 10 inches ; actual spark-length, or thickness of air between 

 the spark-terminals, 6 inches, rarely 7 ; thickness of glass be- 

 tween the induction-terminals T 3 F inch. But powers a good deal 

 lower give effects distinct enough. 



Z2 



