140 ME. C. E. S. PHILLIPS ON THE ACTION OF MAGNETISED ELECTEODES 
occiuTed Ijetweeii the magnetisation of the electrodes, and the aj^pearance of a 
luminous ring. But in such a case positive electrification of the outer surface of the 
hidl) immediately caused the ring to appear within. Under these circumstances the 
evidence that the luminosity Avas due to the radial streams of positUe ions Avas also 
supjxnded hy tlie appearance of the ring as it l)ecame gradually built up of luminous 
arcs iu tlie manuer shoAvn in tig. 4. The positiA^ely electrified body touched the hulh 
Fie;. 4. 
at tlie points A, I>, and ft Tliese huniuous curves Avere in a tew rare ca.ses clearl}" 
Ausible; they reA’olved round the magnetic axis veiysloAA'ly at first, then more rapidh" 
for a feAv reAmlutions, blending together into a continuous ring as they moved, AAdiich 
Avidened out and soon faded completely away. On a few occasions Avhen the 
electrodes Avere suddenly magnetised the Avhole of the residual gas Avithin the bulb 
appeared luminous—a hluish-AAdiite gloAv—Avhile at the central portion as before, the 
ring, slightly more luminous than the rest, could he seen rapidly reA’olving. An 
electroscope attached to the Insulated magnetic electrodes shoAA-ed that when a 
positiA-ely electrified liody aa^is brought into contact with the outer surface of the glass 
bulb a certain amount of positive electrification aaxis produced upon the iron electrodes 
themselA’'es. In the case of one particular hulh, on a A'ery dry day and under correct 
conditions as to jiressure and stimulation, the sudden magnetisation of the electrodes 
caused a spark to leap across a 6-nnn. air gap betAA'een tAvo Avires, one connected 
to a tinfoil patch at the ecjuator of the Imlh, and tlie other in contact Avith either 
of the iron electrodes. This remarkable efiect took place after all external residual 
electrification ujjcm tlie ap)|)aratus had been removed h}' carefully connecting to earth 
each part of the bulb as AAnll as the electrodes. By means of a tinfoil jiatch, fastened to 
the surface of the hulh at Axrrious jilaces, some useful results AAnre also obtained Avith a 
hulh containing only one magnetic electrode. Fig. 5 shoAvs the distribution of charges 
upon the liidh at the moment the discharge through it aa’us sto^iped. At a pressure 
scarcely Ioav enough to gUn a luminous ring the action of the magnet AAns to 
increase the positive electrification around the etjuator, and also slightly that upon 
the tinfoil ring T at the end. But in this case no luminous ring appeared, nor Avas 
there evidence of the AAuthdraAval of positiA’e ions from the interior glass surface. 
At a loAA’-er pressure, hoAAnA’er, the ring became Aosihle, and then the sign of the 
