854 



SCIENCE. 



LN. S. Vol. XV. No. 387. 



use of a miniature anemometer similar to 

 those I have used to show the oscillatory 

 motion of the air in a stationary sound 

 wave.* 



The anemometer possesses the convenient 

 property of rotating in one direction, Avhat- 

 ever may be the direction of the particles 

 acting upon it, so that an oscillatory motion 

 of the ions should produce rotation pro- 

 vided their amplitudes of oscillation are at 

 least as great as the radius of the cups. 



The arrangement of the apparatus was 

 as follows : Pour large twenty-liter Leyden 

 jars were joined two in parallel, the inner 

 coatings of each pair being connected with 

 the spark gap and the outer coatings vnth 

 the coil B. Within this coil, which con- 

 sisted of eighteen turns of coarse wire, was 

 placed a cylindrical glass vessel g, 5^ cm. 

 in diameter. The vessel was in permanent 

 connection with a mercury air pump. A 

 miniature anemometer A, consisting en- 

 tirely of glass, was mounted on a needle 

 point so as to turn with great freedom. 

 The anemometer was 3^ em. in diameter 

 and the cups, which were half cylinders, 

 were each 2 cm. long and 6 mm. in diam- 

 eter. Between the coil and the vessel was 



* Amer. Jour. So., February, 1902. 



placed a Faraday cage, which was made by 

 attaching narrow strips of tin-foil on a 

 glass cylinder so that the strips were per- 

 pendicular to the plane of the coil. This 

 was found to shield the vessel very well 

 from external electrostatic effects. The 

 jars were charged by a large induction 

 coil excited by an alternating current of 

 forty complete periods per second. The 

 length of the spark gap was 11 mm. When 

 the proper degree of exhaustion was ob- 

 tained, upon the passage of the sparks, the 

 white ring discharge was produced and the 

 anemometer rotated in the direction of the 

 convex side of the cups. 



The experiments are given below: For- 

 ward rotation signifies rotation in the direc- 

 tion of the convex side of the cups. 



1. Pressure 3.6 cm. mercury. A faint 

 red light in vessel. Anemometer does not 

 rotate. Vessel cool. 



2. Pressure 1.1 mm. The red light deeper 

 and stronger. The anemometer does not 

 rotate. Vessel cool. 



3. Pressure .64 mm. The white ring dis- 

 charge is obtained. Anemometer rotates 

 forward about two revolutions per second. 

 The vessel becomes very hot. 



4. Pressiire .17 mm. The anemometer 

 rotates forward but not so rapidly as in 

 experiment 3. After the interruption of 

 the Tesla current the anemometer rotates 

 backward. If the current is kept on for 

 some time, especially when the vessel is 

 wrapped closely in paper to confine the 

 heat, the walls of the vessel become nearly 

 as hot as the vanes of the anemometer and 

 the anemometer rotates but very little 

 backward. 



5. Pressure .058 mm. Immediately on 

 the appearance of the white electrodeless 

 discharge the anemometer rotated back- 

 ward. Upon vsrrapping the vessel with felt 

 or paper to confine the heat the following 

 is observed : When the current is turned 

 on the anemometer rotates at first back- 



