198 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
to a point considerably on tire same side of the tinfoil as the positive terminal of the 
tube, so that the bright tongue on that side is separated by it from the blue haze for 
a considerable interval. The other tongue, on the contrary, seems to proceed from a 
point close to the blue haze. The magnet shows that both these tongues are formed 
by currents where the positive electricity is passing in the same direction as that 
in the tube. It is therefore evident that the interior of the tube beneatli the tinfoil 
acts as a negative terminal to the tongue first described, and that this tongue is 
a positive luminous column, which is separated, as in ordinary cases, from the halo or 
glow of its negative terminal by a dark space. The other tongue is a positive 
discharge which proceeds from the interior surface of the tube after the satisfaction of 
the negative discharge therefrom in the manner just described. This positive discharge 
is doubtless due to the positive electricity that was fixed on the interior of the tube by 
the same negative pulse in the tinfoil which caused the negative discharge from 
the interior of the tube. 
This leaves unexplained only the state which precedes the discharge-relief-effect, i.e., 
that in which the luminous column is repelled. This cannot be taken to be analogous 
to the repulsion in tubes which do not give both forms of relief-effect (which repulsion 
we have seen to be due to the formation of a Crookes’ space within the glass), for it is 
very much greater in amount, and moreover, when the air-spark interval is increased, 
it disappears, giving place to the discharge-effect which we have seen is the true 
analogue of the repulsion in tubes which do not give both forms. The interpretation 
of this repulsion has been felt by the authors of the present paper to be a matter of 
great difficulty, but they have come to the conclusion that it represents a discharge of 
the same type as the non-luminous discharges that pass through vacuum tubes under 
certain conditions of exhaustion and current. This identification must be considered 
to a great extent hypothetical, but the authors of the present paper believe that it will 
on examination be found to be free from objections, and to satisfy the conditions of the 
problem more completely than any other. 
Thus we see that the four effects divide themselves into those caused by positive 
pulses and those caused by negative pulses,and that the effects of these are respec¬ 
tively to form a positive and a negative terminal within the tube beneath the tinfoil.t 
* Now that the relief-effects can be divided with certainty into those which are produced by positive 
pulses to the outside of the tube and those which are produced by negative pulses, and these can be 
recognised by the nature of the appearances within the tube, the full force of the experiment described on 
pp. 178, 179, becomes felt. If the air-spark be in the positive the effect of joining the tinfoil to earth will 
be to produce negative pulses to the tube whether the positive or negative terminal be joined to earth, that 
is to say, whether the tinfoil be maintained at a potential higher or lower than any portion of the circuit in 
which the tube is situated. And, of course, a similar remark applies to the case where the air-spark is in 
the negative. 
f These are the primary effects; but in each case there is, as we have seen, a secondary effect of a 
precisely opposite character, viz.: the formation of a negative and a positive terminal in the respective 
cases. 
