578 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
it occurred when the earth connexion was behind the air-spark, reckoning from the 
tube. 
A little consideration suffices to explain how this is caused. Let us suppose that 
the air-spark is in the negative, so that the positive terminal of the machine is 
in metallic connexion with the positive terminal of the tube, while the air-spark is 
situated between the negative terminal of the tube and that of the machine. Then, 
as we have said, the strong shock will occur when the negative terminal of the machine 
is put to earth, and not otherwise. 
Let us consider how the discharge takes place. The negative terminal of the 
machine and all bodies metallically connected with it are maintained at potential zero 
by means of the earth connexion. Hence the negative side of the air-spark interval 
is at potential zero, and would remain destitute of free electricity if free positive 
electricity did not collect upon the surface of the metallic body forming the other side 
of the interval. But, as such is the case, free electricities of opposite signs accumulate 
on the two surfaces until, through their tension and their inductive action upon one 
another, the discharge is brought about. Now, how has this positive electricity 
accumulated upon the positive side of the interval ? It is not due to the influence 
of the negative side, for that, as we have seen, would have been inert. It must, 
therefore, have become charged solely through the accumulation of electricity at the 
positive terminal of the machine which has flowed from thence into the tube and, 
passing through it, has emerged at the negative terminal and charged the positive side 
of the air-spark interval. In this process the tube has become fully charged. When 
a discharge comes, the tube, like a Leyden jar, will empty itself, and thus the amount 
of electricity passing will be augmented by the whole of the charge which the tube 
can hold. The shock, therefore, which is the inductive effect of this sudden discharge 
of the large stock of accumulated electricity, will be proportionately severe. 
Compare this with the case in which (with the same arrangement of air-spark) the 
positive terminal is to-earth. The tube is now kejat at potential zero,—or, more 
correctly, its positive terminal is. The active cause of the discharge is now the 
electricity accumulating on the negative terminal of the machine, and, although it no 
doubt produces a displacement of the electricity in the system on the other side of 
the air-spark interval, it is clear that it cannot cause the tube to become charged, 
connected as it is with earth at the positive terminal. Hence, when the discharge 
takes place, it is only the interchange of the free electricity on the two sides of the 
air-spark interval, and there is no accumulated store in the tube to take part in the 
discharge. The shock is accordingly very much smaller. 
There are two very interesting experiments which serve to confirm the justness of 
these conclusions. The first is an application of an experiment described in Section 
XIII. of our previous paper (Phil. Trans., 1879, p. 220, and Plate 20, fig. 26). In 
order to ascertain the condition of a tube in advance of the discharge, a piece of tinfoil 
was fastened to a glass rod which was then laid along the tube, the piece of tinfoil 
