204 
MESSES. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
straight wire, the negative glow forms itself into a cylindrical envelope around it with 
a rounded or hemispherical head. If the negative terminal he in the shape of a ring 
the negative glow envelops it, remaining at a constant distance from the surface of the 
terminal and thus taking the shape of an anchor ring,* or else (especially if the ring 
be a small one) no portion of it lies within the circuit of the ring, but it envelops 
the ring as a whole, the bright surface of the glow having a boss or protuberance 
opposite to the centre of the ring. If the negative terminal be large and flat, it will 
be found that the negative glow does not substantially differ in appearance from a flat 
stria; and by making the negative terminal in the shape of a hollow cone, which is 
the natural shape of the negative terminal of what we have considered to be the 
physical unit of a striated discharge, there is little doubt that its resemblance to an 
ordinary stria would be rendered still more striking.! 
If we are right in our conclusion as to the nature of the negative glow, it follows 
almost necessarily that the negative dark space and Crookes’ space are merely the 
representatives of the dark space which occurs as we have seen in every unit of a 
striated discharge. The former belongs to the unit of which the stria at the head of 
the positive column is the positive terminal, and the negative haze is the negative 
terminal, and the latter belongs to the unit of which the negative glow is the positive 
terminal, and which has for its negative terminal that of the tube itself. No doubt 
there is one great difficulty in this interpretation, viz.: the contrast in the lengths of 
the dark spaces in these two units, both when compared with one another and with 
the dark spaces in the other units throughout the tube. But if we remember the 
different circumstances of the discharge in the cases of these terminal segments when 
compared with those that prevail in the other segments of the discharge, we shall see 
no reason to give great weight to these unexplained difficulties as telling against the 
truth of the theory. In the first place we are dealing, in the case of Crookes’ space, 
with the passage of electricity from metal to gas or gas to metal, instead of simply the 
passage of electricity from one portion of gas to another. And further, we have the 
far more important circumstance that both in Crookes’ space and the negative dark 
space we have a case of a discharge in which the negative terminal of the unit is of a 
shape and disposition wholly unlike that of the negative terminal of any other of the 
* These last remarks will explain a difficulty which would otherwise be felt, as to why it is only in 
particular cases that we get any analogues of strife from those special or relief effects which cause negative 
discharges within the tube. Like all other fixed negative terminals, the glass insists on having its dark 
space round it in all directions. This dark space is hounded by a glow which, as we have seen, is the 
analogue of the bright surface of a stria. But its arrangement is such that all we can expect is a stria 
turned inside out, as at an ordinary negative terminal; or more particularly, at a negative terminal in the 
fonn of a flat ring placed close to the interior surface of the tube. And this is precisely what we do 
obtain. 
f Other remarkable instances of the modification of the form of the negative glow, and of the extent 
to which it may he made to assume the appearance of an ordinary stria under suitable local conditions, 
will he found in the Postscript to this paper. 
