644 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
usual rule) negative relief-effects. These, however, were not very clearly manifested, 
but the remarkable peculiarity was observed that when the wire from earth to the 
latter tinfoil was not allowed actually to touch it, but was held at short sparking 
distance from it, these negative relief-effects were very marked and clear. It is difficult 
to interpret this in any way other than by supposing that the negative discharge in 
the large tube went on accumulating for a certain time instead of instantaneously 
flashing into its full intensity as in the case of the positive discharge. This rapidly 
rising negative charge in the tube of high exhaust would of course drive off negative 
into the other tube in the ordinary way, but this (if the above law be true) would be 
done with less sharpness than in an ordinary negative discharge in such a tube. We 
should thus have a discharge in the latter tube which, though intermittent in its 
nature, yet partook of a continuous character, and the relief-effects would therefore be 
poorly defined. If, however, the relief were only permitted to be given at the exact 
moment when the accumulation of negative reached its height, and was then given 
impulsively (as would be done by allowing positive electricity to spark into the tinfoil 
from the earth wire), the relief would be given with the requisite sharpness, and would 
produce definitely marked relief-effects. 
It will be seen that this explanation requires that the negative should accumulate 
in the original tube. In other words, it shows that it is a difficult (but not necessarily 
a slow) process for negative to get out of the tube as well as a matter of time to get 
in. This will no doubt be found to be the case, and when the law is perfectly 
formulated it will probably include some such statement. The present form is taken 
only as an approximation. 
The second set of observations bearing upon this matter is intimately connected with 
the difference of sharpness of effect which we have already noticed as existing between 
positive and negative intermittence, even in low tension tubes. In high tension tubes 
this difference becomes in some respects immensely exaggerated. If the intermittence 
be positive and the negative terminal be put to earth and the hand placed on the 
tube, strong shocks will be felt. If, however, the intermittence be negative and 
the positive terminal be put to earth the shocks are so much more feeble as hardly 
to be sensible. And the same difference may be made evident in another way. If 
an earth wire be held near to a piece of tinfoil placed upon a high vacuum tube it 
will be found that the sparks will stream between them when the intermittence 
is positive even though they are a considerable distance apart. But if the 
intermittence be negative it is difficult to get any but the very shortest sparks to 
pass between the tinfoil and the wire. And if contact be made, the contrast (if the 
air-spark be considerable) is equally striking. While in the case of the negative 
intermission no special phenomena are observed, in the case of the positive inter¬ 
mission the electricity streams from all sides of the tinfoil and evidences the most 
violent alternations of tension. It is difficult to see any explanation of these 
peculiarities other than that the impulsive change of tension is very much greater 
