474 
DR. S. P. THOMPSON ON CATHODE RAYS 
When B and C acted jointly as cathodes the shadow of A occupied a mean position ; 
but under no circumstances were two shadows of A seen, or any appearance of 
overlapping shadows. These effects were repeated, using an influence machine to 
electrify the wires casting the shadows. A dry pile, ordinarily used to charge gold- 
leaf electroscopes, proved inoperative. It acted merely as a high resistance. 
Tube [No. G 8], (fig. 7), has three small lateral disk electrodes. A, B, and C, and a 
central electrode D of aluminium wire, about 2 millims. thick. Electrodes A and B 
being connected as cathodes, and C as anode, the shadows of D were about of the 
geometrical size. The exhaustion was at this stage such that a spark would just 
pass in an alternative path between blunt points about 7 millims. apart. 
When the wire D was made anodic its shadows at once became narrower. When 
is was made cathodic its shadow widened enormously, being about 16 mill im s. wide, 
with an ovate end. 
When A and C were made cathodes, and B and the wire D anodes, there were 
produced two narrow shadows at a and c respectively, opposite the two cathodes. 
As the pump was worked, and the exhaustion increased, these shadows grew 
narrower, becoming mere lines, with a brighter spot of luminescence above the tij) of 
each. Then the upper parts of these linear shadows closed up entirely, so that 
instead of presenting each, as at first, a dark line about 1 millim. broad, emarginate 
with a luminous edge, each now presented the appearance, except at the base, of a 
narrow bright luminescent line. At this stage the exhaustion was such that a spark 
would just not pass at a gap of 18 or 19 millims. in the alternative circuit. As 
exhaustion proceeded further, the luminescence on each edge began to overlap ; and 
finally the shadows became two hriglit linear strips each about I'o millim. broad. 
The spark-gap at this stage, when the tube was sealed off, was 30 oi' 32 millims. 
In this tube, then, the effect of making anodic the object which casts the shadow 
was to produce a deflexion of the cathode rays into the geometric shadow, even to 
overlapping; the result being to produce negative shadoivs, that is to say, shadows 
which appear bright upon a less bright background. It also appeared that the 
enlargement of a shadoiv ivhen the ohject is made cathodic, and the diminution of the 
shadoiv ivhen the ohject is made anodic, both depend upon the degree of exhaustion of 
the tube; and both are augmented by raising the degree of exhaustion. 
These two effects are, however, unequal. The enlargement when cathodic exceeds 
by many times the diminution when anodic, under identical conditions of exhaustion 
and excitation of the tube. 
These observations furnished an explanation of the luminous spot observed in tube 
(fig. 3) ad the ends of the shadows of the lateral electrodes, and of the tapering form 
of those shadows when the electrodes were anodic. The surface when anodic deflects 
the cathode rays into the geometrical shadow ; and, having been deflected, they 
continue in a new direction. As the surface of the bulb where the shadow falls is 
curved, the rays that cast the tip of the shadow have to travel a longer distance 
