646 
MESSRS. W. SPOTTISWOODE AHD J. FLETCHER MOULTON 
The fundamental experiment in this matter was made by us in our experiments 
with the standard-tube. We had found in working with positive intermittence that 
there is but little difference between the variations of electric action undergone by a 
piece of tinfoil upon the tube, an intermediate terminal projecting within the tube, 
and the non-air-spark terminal, so that it was comparatively a matter of indifference 
to which the tinfoil on the standard-tube was joined. It then occurred to us that we 
would try the case of a unipolar discharge. Accordingly a tube was taken in which 
there was an intermediate terminal situated very near to one of the ends of the tube. 
The other terminals were in the usual position at the two ends of the tube. The two 
terminals that were very near together were joined to the terminals of the machine, 
so that the current only traversed the very short portion of the tube lying between 
them. A positive air-spark was now introduced into the current. As has been 
described in the previous paper, this caused a pointed tongue of positive luminosity 
gradually to advance into the unoccupied portion of the tube, forming a positive 
unipolar discharge. The terminal at the other end was then joined to the tinfoil 
on the standard-tube. No effect was, however, produced until the air-spark was 
lengthened (and with it the unipolar column) so much that the unipolar luminosity 
came up to that terminal and formed the usual blank-space around it. Instantly the 
most perfect positive effects appeared in the standard-tube. The air-spark was then 
decreased so as to make the unipolar luminosity retreat from the terminal, and at 
once the effect ceased. This was repeated several times, and the result was always 
the same, showing that so local is the effect of the free electricity discharged into the 
tube that it is only across a true negative blank-space that it exercises upon another 
terminal any decided influence of the type required to produce interference. The 
experiment was repeated, a ring of tinfoil round the tube being substituted for the 
distant terminal. The result was the same. So soon as the tongue of luminosity got 
within a like distance of this ring so that the blank space was formed either completely 
or partiallv the effect appeared in the standard tube, but not otherwise. 
We need not point out how strongly this supports the theory suggested in the 
previous section as to the signification of the blank-space. We wish rather to point 
out its bearing upon the question of the significance of the positive column. The 
luminosity in the unipolar discharge is in reality a positive column which doubles back 
upon itself when it does not find any negative terminal to which it can discharge. It 
thus may be taken for our present purposes as a representative of an ordinary positive 
column, and it becomes strictly a positive column when it reaches the further terminal 
and forms a blank space round it. Regarding it as such, we see that it is unable to 
produce interference in the standard-tube through the medium of the further ter¬ 
minal, unless it extends to within the breadth of the ordinary blank-space from that 
terminal. The effect produced upon the standard-tube could only be caused by the 
presence of free electricity in the immediate neighbourhood of the terminal. This 
seems to show that the luminosity marks much more clearly the local situation of 
