150 ME. C. E. S. PHILLIPS OX THE ACTION OF MAOXETISED ELECTRODES, ETC. 
becomes luminious, and is then, as it were, blown upon by streams of negative ions 
which are attracted to the outer surface of the luminous ring, and thus cause the 
wliole to revolve at a higli velocity. 
The expansion of the luminous ring is accounted for by the outward Hying streams 
of negative ions which go forth to meet the incoming positive ones. Moreover, this 
view receives strong support from the experiment in which a mica vane, when partly 
immersed in the haze sometimes visible about the ling itself, rapidly rotated in the 
opposite direction to that in which it turned when the brighter portion enveloped 
the wings. It is also seen why the size and shape of the bulb itself exerts an 
influence upon the eftects and why the distribution of charged gas in the interior 
plays so important a part. The other minor phenomena, such as the lateral move¬ 
ment of the ring as a whole, from one part of the I)ulb to another, as well as the 
formation of luminous rings in planes other than at right angles to the magnetic 
axis, may all be explained in accordance with the theory here put forth. 
I desire to acknowledge thankfully the help which I have received from my 
assistant, Mr. CTiarles Coton, during the progress of the work, and to express my 
indebtedness to the managers of the lloyal Institution for having kindly placed for a 
considerable time the exceptional resources of the Davy-Faraday Laboratory at my 
disposal. 
