APRIL 15, 1897 | 
NATURE 569 
9) 
It may further be remarked that the diameter of the lumin- 
escent ring may be increased or diminished, or finally reduced to 
a point, without altering the degree of vacuum, by moving the 
anti-kathode away from or towards or finally into the focus of 
the kathode stream, the appearance of the ring in each of these 
cases being practically similar to those shown in the figures for 
a uniform distance with varying vacuum. Similarly it may be 
shown that the converging cone of rays between the kathode and 
the focus produce hollow rings upon a carbon anti-kathode 
exactly as does the diverging cone of rays. When the anti-kathode 
surface is not at right angles to the line of the discharge, the 
ring, in place of being circular, takes the proper form of a conic 
section. The holding of a magnet near the tube distorts the 
ring from a circular shape and moves its position on the carbon. 
From these experiments it appears that both the diverging and 
converging cones of kathode rays act as though they were not of 
uniform density throughout their sections, but, at any rate, in 
some instances as if they were completely hollow. 
It should, however, be noted that these hollow effects appear 
only to be obtained with fairly short focus kathodes, such as are 
usually employed in X-ray focus tubes, that is to say, with 
kathodes whose diameter is large as compared with their radius 
of curvature, so that the rays converge and diverge rapidly to 
and from the focus. With comparatively flat, long focus kathodes 
the cones do not show any signs of being hollow, and produce a 
uniformly luminescent spot upon the carbon of larger or smaller 
diameter, according to the conditions of vacuum and the position 
of the screen} 
For instance, while kathodes 1°125 inches diameter and 0708 
inch radius of curvature gave in the manner described dis- 
tinctly hollow convergent and divergent cones, a kathode 1 
inch diameter and 1°5 inches radius of curvature gave con- 
vergent and divergent cones that appeared to be uniformly solid 
under all conditions. 
On the other hand, with rays from flat kathodes brought to 
a focus by magnetic means, both convergent and divergent cones 
are found to produce hollow ring effects. 
The Rays cross at the Focus with no Rotation. 
In order to investigate the kathode rays in a focus tube still 
further, and more especially in order to discover whether the 
various. rays from the kathode cross one another at the focus, 
or diverge again without crossing, and also in order to discover 
whether there is any twist or rotation of the rays, similar to 
what has been observed in the case of rays focussed by magnetism, 
a tube was constructed similar to that used in the previous 
experiments, with a carbon anti-kathode which was also the 
anode, fixed at the opposite side of the focus from the kathode, 
with the focus about equally distant between it and the kathode. 
The peculiarity of this tube consisted in the fact that a sector 
of the aluminium kathode, equal to one-eighth of the total 
area of the kathode, had been entirely removed,|as shown at C, 
Fig. 5. It was expected that on using this tube, with the 
proper degree of vacuum to form a well-defined ring on the 
anti-kathode screen, that a portion of the ring, corresponding 
with the amount of the kathode cut away, would be found 
wanting ; and that by the position of this gap in the ring it 
would be possible to ascertain whether the rays crossed at the 
focus, and whether there was any rotation. What actually was 
observed is shown for three different conditions of vacuum in 
Fig. 5, B being for the highest, and B” for the lowest vacuum. 
As will be seen, the expected gap in the ring was obtained, but 
with the unexpected addition that the dimensions of this gap, 
instead of being only one-eighth of the circumference of the 
ring, was seven-eighths of the circumference. In fact, the 
amount of ring shown corresponded not with the seven-eighths 
of the remaining kathode surface, but with the one-eighth of 
the kathode that had been removed. The portion of ring that 
did appear was of a length corresponding exactly to the arc of 
the removed sector of the kathode, according to its greater or 
lesser nearness to the centre with different conditions of vacuum ; 
and as the portion of ring was in each case exactly in line with 
the portion of kathode that had been cut away, it would appear 
that there is no rotation of the kathode beam as a whole, that 
the rays do cross at the focus; and, further, that when the 
hollow convergent cone is, as it were, split in this manner, some 
unexplained action, similar in effect to the existence of a circular 
surface tension. causes the gap to widen out and the remaining 
portion of the ring-shaped section of the cone to contract corre- 
spondingly, without, however, altering its diameter, 
NO. 1433, VOL. 55| 
In order to further investigate the matter another tube was 
made, in which the concave kathode was complete ; but the 
interior of the tube was furnished with a small movable piece 
of aluminium, which by shaking could be moved up and down 
the tube between the kathode and anti-kathode, and which, 
while not quite reaching the centre of the tube, would fill up 
very nearly one quarter of the circular sectional area of the 
latter. 
With this arrangement of tube, with the aluminium obstacle 
placed just at the focus, as shown in Fig. 6, the point of the 
obstacle just missing the kathode rays, a complete ring was 
formed on the carbon anti-kathode. On moving the obstacle 
slightly into the divergent cone, exactly one quarter of the ring 
on the anti-kathode failed to appear, as shown in Fig. 7, and on 
the obstacle being further moved in the same direction, the 
result was not altered. 
Fic. 6, 7- 
As in each of the latter two cases there was no displacement of 
the gap in the ring, the above showed that there is no rotation 
of the divergent kathode cone. 
Experiments were next tried with the aluminium obstacle, 
moved so that its point just entered the converging cone of 
kathode rays, when a small portion of the ring was cut out ; but 
on the opposite side, as shown in Fig. 8, this confirming the 
previous experiments, which showed that the rays cross one 
another’s paths at the focus without rotation. Upon moving 
the aluminium obstacle a little nearer to the kathode, so that 
