476 
DR. S. P. THOMPSON ON CATHODE RAYS 
produced when the conducting surface was shielded from direct action by the inter¬ 
position of a non-conducting layer of glass. Attempts to deflect the cathodic shadows 
by holding charged bodies outside the tubes led to no defi.nite result. In some cases 
it was possible by laying the finger on the outside of the tube to produce slight dis¬ 
placements, particularly if the cathode pole of the coil was earthed at the time, so 
that the finger acted cathodically. Tube [No. G 17], (fig. 9), was constructed to test 
this point. It is of a pear shape, with a small disk cathode, h, at one end, and an 
anode, a, in the side tube by which the bulb was connected to the pump. A wire, 1), 
to cast a shadow, was inserted near the broad end, and two dimples or depressions, 
cl, were impressed into the sides, leaving an internal distance of about 5 millirns. 
between their faces. These depressions were covered externally with tinfoil. The 
wire 1) gave a shadow the size of wdrich varied, as in previous experiments, according 
as the state of the wire was neutral, anodic, or cathodic. On each side of this 
shadow appeared a shadow of the dimple cl. When the metal coatings of the two 
dimples wmre made cathodic, the space between their shadows decreased slightly. 
When made anodic, there was no measurable increase in the space between their 
shadows. When one w'as made anodic and the other cathodic, their shadows 
appeared to shift slightly as from the cathode side, and the shadow of the wme h 
was very slightly shifted in the same sense. 
Tube [No. G 16], (fig. 10), was of a vertical j^ear shape,with a small disk cathode 
of about 6 millirns. diameter inserted at the side, and a similar disk anode at the 
bottom. Through the top was inserted a narrow glass tube, closed at the bottom, 
open at the top to receive mercury. This tube cast the usual shadow on the oppo¬ 
site wall, and gave identical shadows when empty and when filled. A wire was then 
inserted into the mercury to enable it to be electrified. At low degrees of exhaustion 
the shadow remained unchanged in size, whether the mercury thread within it was 
neutral, anodic, or cathodic. But, as the exhaustion was increased, a point was 
reached at which, almost suddenly, sensitiveness set in, and the size of the shadow^ 
became variable, contracting very slightly when the mercury was made anodic, 
expanding enormously when made cathodic. It was noticed that this change from 
the non-sensitive to the sensitive state occurred at the stage of exhaustion at which 
the “ splash” phenomenon appeared on the bulb-wall opposite the cathode. It was 
also noticed that the sensitiveness depended upon the conditions of excitation, being- 
greater when the break of the coil was lightly adjusted, so as to operate the coil with 
the spark-gap at 3 millirns., than when tightened up to operate with the spark-gap 
at 15 or 20 millirns. or more. 
Several tubes were made of the general form of [No. C 2], (fig. 11), having a small 
disk cathode at one end, an anode in a side tube, and, as object to cast shadows, a 
tube containing mercury. These mercury tubes were made of several different 
*■ This tube, and the succeeding one, were exliibited at the evening gathering of the Royal Society', 
October, 1896. 
