TIPOX ELECTRICAL DISCHARGE PHENO^IENA IN RAREFIED GASES. 
141 
electrification upon the glass at the equator reversed to negative while the tinfoil 
became more strongly positive, as before. The range of pressures over which the 
best results were obtained with hidbs nearly splierical and 6 centims. In diameter 
extended from 0'0120 to 0'0004 millim. of mercury, as measured by means of a 
MacLeod gauge. Tlie sign of the electrification upon the luminous ring itself was 
then found to 1)e invariably negative by means of tlie following experiment :— 
A ring of platinum wire stood within the bulb (fig. I). It was sup]5orted by a tliin 
metallic stalk, which terminated in a small exterior loop, and was jfiaced so as to 
occupy the position in wlilch under suitable conditions a luminous ring might have 
been expected to appear. The diameter of the platinum ring was approximately that 
of the luminous ring when at its best. A discharge was then started and stopped 
and the electrodes magnetised, with the result that two luminous rings appeared 
within the bulb, one on each side of the platinum circle and very close to it. On 
slightly electrifying the platinum circle negatively both these rings flew apart, each 
threading itself on one of the electrodes. On the other hand, when the circle Avas 
positively electrified both rings moved towards each other, and filled the space 
bounded by the wire with a brightly luminous whitish-blue glow. The luminous 
ring has, in this rvay, been shown to invarialfiy be negatively electrified. Kecourse 
was tlien had to a set of experimeiits dealing in more detail with the existence of the 
radial incoming streams of positively electrified gas particles previously referred to. 
Diselectrificafion Effects. 
In the first place a new apparatus was constructed, in Avhich tlie soft iron rods 
E^, were sheathed in a glass tube (fig. 6), which AA^as continuous tliroughout and 
melted at the ends on to the short necks projecting from the bulb. The glass tube 
carried, AAU’apped round its central portion, a piece of aluminium foil, 2'5 cm. Avide, and 
connected by means of a fine Avire to an external terminal loop To. Concentrically 
Avith this thin aluminium cylinder stood a larger, though much narroAA^er band A of 
stouter aluminium foil, supported upon a metallic stalk Avhich terminated in an 
external platinum loop T^. 
