1895.] J. C. Bose — Polarisation of Electric Rays. 293 



Among the possible causes of unsteadiness may be mentioned the 

 following — 



1st. The current actuating the coil may vary after a time. To 

 overcome this difficulty a fairly constant battery was made to charge a 

 small storage cell, and a derived circuit from this cell was led to the 

 Primary coil. 



2nd. The interrupter may have its rate of vibration changed by 

 heating, wearing out of contact points, and other causes. Any change 

 in the periodicity of the vibrator is at once made evident by the corres- 

 ponding change in the pitch of the note given out by the vibrator. 



3rd. The sparking balls may have their surfaces roughened by the 

 disintegrating action of the spark. To avoid this difficulty, the balls 

 were thickly coated with deposit of gold, and were turned round at 

 intervals to expose fresh surfaces. 



The coil with a storage cell is enclosed, with the exception of a 

 horizontal tubular opening, inside a metallic box, not dissimilar in ap- 

 pearance to an Optical Lantern. The interrupter is actuated by turn- 

 ing a key from outside. The sparking balls are at one end of a brass 

 tube 25 cm. long and 5 cm. in diameter. At the further end of the 

 tube is the Polariser. Inside the tube is placed a convex lens with the 

 spark gap at its principal focus. With the help of the lens and suitable 

 diaphragms, the electrical beam is made approximately parallel. By 

 means of an Iris diaphragm, the amount of radiation may be varied. 



Polariser. — The success of the experiment depends greatly on the 

 care with which the Polariser and Analyser are constructed. Fine copper 

 wire "2 m. m. in diameter is carefully wound in parallel lines, round 

 two thin sheets of mica. There are about 25 lines for every cen- 

 timetre. The mica pieces are then immersed in melted paraffin, and the 

 wires thus fixed in situ. By cutting round, two circular pieces, con- 

 taining the gratings are obtained. The mica pieces are too thin to pro- 

 duce any disturbing effect. The gratings are fixed with wires parallel, 

 at the ends of a tube 5 cm. long. This Polariser tube rotates inside the 

 outer end of the tube which sends out the parallel electric beam. 



Analyser. — The Analyser is similar in construction to the Polariser. 

 It rotates inside the Receiving tube, which contains the sensitive sur- 

 face for detecting radiation. 



Receiver. — The Receiving apparatus consists of a ' Coherer ' with a 

 Voltaic cell and Galvanometer in series. The Coherer is modified from 

 its usual tubular form. The filings, a single layer thick, are spread over 

 a large surface. This arrangement secures great sensitiveness. A pair 

 of insulated wires from the ends of the Coherer, are led out to a distant 

 dead-beat Galvanometer of D'Arsonval type in series with a constant 



