192 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
striated the hollow cone will end in a blunt rounded head of considerable brightness 
(Plate 16, fig. 9, and Plate 17, fig. 11). Each of these hollow cones will have its hazy 
irregular internal surface turned towards the bright end of the sectional positive 
column immediately behind it, i.e., towards the positive terminal of the tube. If 
now we bring the finger or a conductor to the side of the tube these columns will 
all display sensibility, but each willjnove independently of the others and of the 
remainder of the positive column and behave as though it started from the tinfoil 
ring at its base, still preserving, however, its position relative to the sectional columns 
on each side of it. If a magnet be used, it deflects all these columns, but each column 
still moves independently and as though it had its tinfoil ring as base. The magnet 
also shows, as might be expected, that the discharge through each of the columns is 
in the same direction, viz. : the positive passes from the hollow cone towards the 
bright termination of the column of which it forms the base or commencement. 
Now, taking a tube in which the above phase Is clearly marked, let us gradually 
increase the air-spark interval, so as to render the electrical pulsations more violent. 
The sharply-defined head of the terminated column will be found to flatten and widen 
out, growing less bright in the centre till it becomes almost an annular ring of light. 
As the air-spark interval is still further increased, this annular ring widens and gets 
less definite, until at last the column on the positive side of the tinfoil assumes 
somewhat the form of a hollow cone pointing towards the positive terminal of the 
tube. The base of this cone is however not cpiite close to the tinfoil, 44 and is formed, 
as above stated, of the last phase of the annular luminous ring formed out of the head 
of the terminated positive column (Plate 17, fig. 12). 
.If the air-spark interval be too small these phenomena will not be obtained, but 
only a more or less faint foreshadowing of them mixed with diffused luminosity. 
This is not to be wondered at, as in such cases the electrical pulsations to the 
outside of the glass are probably insufficient to produce within it discharges of 
sufficient intensity to give complete satisfaction to the discharge from the positive 
terminal, and hence we have an effect produced by the superposition of an ordinary 
luminous discharge and one interrupted in the way above described. But careful 
examination of various tubes has convinced the authors of this paper that the forms 
above described are the truly typical forms of the special or non-relief-effect when the 
positive is the air-spark terminal. 
VIII.— Examination and interpretation of special and relief effects in general in 
the case of interrupted discharges. 
We have hitherto confined our examination of these effects to the single case in 
which the air-spark or other cause of the interruptedness of the current is situated in 
* The authors of this paper believe that such is always the case, but the phenomenon is complicated by 
the presence of a reverse current which accompanies the use of a large air-spark in the positive circuit. 
