ELECTRIC DISCHARGE WITH THE CHLORIDE OF SILVER BATTERY. 481 
free motion is less than in a confined tube, and it is natural, therefore, that the effect 
of increasing fewness of the molecules should begin to tell at an earlier stage of 
exhaustion : that is, that the pressure of minimum resistance should be greater than 
in a tube.” Later on we shall quote other experiments on the difference of potential in 
different parts of a tube in support of this view. 
Discharge in miniature tubes. 
Experiments on the discharge in miniature tubes gave most unexpected results; for 
example, in a tube of the dimensions represented of the natural size in fig. 4, which 
also represents the arrangement of the apparatus. The terminals in this tube so 
nearly touched, that in order to ascertain that they were not in actual contact, they 
were connected through a sensitive galvanometer with the opposite poles of a battery 
of 10 cells. The battery-power was then increased gradually to 2400 cells, when a 
discharge took place. After this discharge, then a single cell would pass; but if the 
tube was allowed to stand for a short time (5 minutes) then it required 4800 cells 
to reproduce a discharge. From the great heat of the discharge the terminals 
became red-hot and the tube very soon cracked and prevented the repetition of the 
experiments. 
But in another tube of the size shown in fig. 5 fresh experiments were made, the 
terminals in this tube were distant about Truroth of an inch (0’00104) (0’0264 m.m.), 
the points being of the form shown magnified in fig. 6 after a discharge had taken 
place. 
/^> 
After the discharge of 2240 cells had passed, then the number of elements had to 
be increased to 11,240 to produce a discharge, and at last even this number could 
not cause one except by alternating the current first in one and then in the contrary 
direction. Ultimately this failed to produce a discharge, but an induction coil did so; 
