626 
MESSRS. W. SPOTTISWOODE AND J. ELETCHER MOULTON 
part in the effects with a positive air-spark. The negative-special presents precisely 
the same phenomena as the positive-relief. We get perfect virtual shadows and 
bright patches of phosphorescence on the other side of the tube. Indeed, so perfect 
is the identity of effects that it would be just as possible to work with one as with the 
other, were it not that it is so much more convenient to examine rehef effects than 
special effects on account of the simpler character of the manipulation. 
There is, however, in all cases a very superior sharpness of action in the positive 
effects of which we shall have to speak later ; but, allowing for this, it would be 
difficult to overstate the jDerfect correspondence that there is between the positive- 
special and the negative-relief, and between the positive-relief and the negative- 
special. The splendid phosphorescent effect and virtual shadows produced by the 
two last mentioned are so strikingly alike as to need no further comment; but this 
resemblance extends to the less striking phenomena. For instance, if we produce the 
negative-special by the aid of a wire from the negative terminal brought parallel to 
the tube, and near to it, we shall find that the previously diffused haze is driven into 
a thin column on the other side of the tube, closely resembling the pencil-like column 
of the positive air-spark. This can be driven about in like manner, and even attendant 
phosphorescence has at times been faintly visible. If the negative-relief be produced 
by a similarly placed wire connected to earth there will be a concentration of the haze 
on the side of the tube nearest to the wire, as in the case with the positive-special. 
We have little doubt that further investigation will show that there is here also atten¬ 
dant phosphorescence, though it is difficult to detect it. 
The patch of phosphorescence on the glass immediately beneath the tinfoil of which 
we have spoken in treating of positive-relief has also been seen in negative-relief. 
The existence of phosphorescence on the glass opposite the tinfoil and that of the 
correlative virtual shadow are, however, more difficult to determine. We have not 
obtained the former in a satisfactory manner, and though we have often seen feeble 
virtual shadows with the negative-relief it has always been a matter of doubt with us 
as to whether some slight positive intermittence had not crept in and caused it. Our 
own opinion is that these two secondary phenomena are in the case of the negative- 
relief too feeble to be generally visible, though the virtual shadows (which require a 
much less violent action than does the production of phosphorescence on the opposite 
side of the tube) may occasionally do so. 
But although for these reasons we do not find it profitable to examine with special 
care the phenomena of the discharge with negative air-spark, the importance of the 
general identity of these effects with the converse effects in the case of the positive 
air-spark is very great as a proof of the main proposition in the theory which the 
authors of this paper have put forward with regard to the disruptive discharge in its 
intermittent form, viz.: that it takes place from either terminal, and that, in general, 
the electricity passes along the whole of the tube in the form of a discharge of the 
same name as that of the terminal from which it proceeds, and only meets with a 
