AND SOME ANALOGOUS EAYS. 
473 
be varied by sliding the wire along the wood. That patch expanded toward which 
the cathode wire was moved, while the other patch contracted. 
The first conclusion to be drawn from these observations was that the size of the 
cathodic shadoiv of an object depends upon its own electric state* If it is j^ositively 
electrifed the shadoiv contracts, if negatively the sliadoiv expands. The same result 
was found to occur when the electrification of the object was produced independently 
by use of an influence machine. 
Fig. 4. 
Tube [No. G 4], (fig. 5), was constructed to verify the above results. It also had 
three electrodes, one a slightly convex disk. A, at one end, the other two, B and C, 
short cylindrical wires inserted transversely to the tube, near together, and parallel 
to one another. The degree of exhaustion of this tube was made rather higher than 
that of the preceding. When electrode A was made cathode, shadows of B and C 
were cast on the broad end of the bulb. If B and C were both made anodes their 
shadows were shghtly narrower than the geometric shadow, and became extremely 
w'ell defined. When B was left as anode and C connected to the cathode through a 
rod of wmod or through another vacuum-tube of high resistance, the shadow of C at 
once swelled out to an oval shape (fig. 6), while that of B was shifted a little 
sideways, as if repelled from B. If A was made anode and C cathode, a shadow 
of B was thrown upon the side wall of the tube. If then B was made anodic, 
by connecting it to A through a rod of wood, its shadow thinned down and became 
more brightly marginate. If B was made cathodic, by similarly joining it to C, its 
shadow widened out to an oval patch, while at the same time an oval shadow of C 
was cast on the opposite wall. By varying the resistance in the connexions, in the 
w^ay described above, for the first-mentioned tube, the sizes of these two oval 
shadows could be varied, one enlarging when the other diminished. While the 
disk A was thus serving as anode, it also cast a shadow of itself upon the smaller end 
of the tube. This shadow shifted slightly sideways according as B or C was cathode. 
* Ckookes, in ‘ Pliil. Trans.,’ 1879, Part II., p. 648, Las alluded to tLe widening of a cathode shadow, 
and to the production of a penumhi’a under an unsteady electrification of the object casting the shadow, 
VOL, CXC.-A. 3 P 
