PRESIDENTS ADDRESS—SECTION A. ae 
may regard an apparently continuous phenomenon such as a 
steady current. as being really discontinuous if the discontinuous 
changes take place so rapidly that we are unable to follow them 
in detail, For instance, imagine a charged sphere supported by 
a fused quartz stand—fused quartz is by all odds the best insu- 
lator in practice known to us. Let the sphere be placed close to 
a plate connected to earth and let it be tested from time to time 
by an infinitesimal proof plane. We know chiefly from German 
researches that the dust particles in the air will under such 
circumstances become the chief agents in discharging the sphere ; 
they_are first, attracted to its surface, then charged, then repelled 
to the earth plate, again attracted, Ans. so on. 
However fast in practice we could work our proof plane, the fall 
of potential of the sphere would appear regular—if it were possible, 
however, to work the plane in a period less than the time elapsing 
between the arrival of successive dust particles, we should find 
that the phenomenon of discharge was really discontinuous. Ina 
similar way our continuous steady current may, if it is convenient 
to us and sanctioned by our general knowledge of what is likely 
to happen, be regarded as the time average effect of the alternate 
establishment and breaking down of the electric tield. On this 
view our tubes conveying the electric and magnetic energy would 
. also move by jumps, and if we had fine enough instruments we 
could detect this intermittence. Let us imagine a Leyden jar 
charged, and having terminals attached to its armatures. Let 
the terminals be immersed in some substance whose electric 
strength is infinite, so that we get no spark on joining the said 
terminals. Then, possibly, an instrumental eye affected by inter- 
mittency of electric state placed between the jar and the wire 
might see a flash of light during the discharge of the jar, though 
there would be no spark. Our eyes, however, do. not lend them- 
selves to such experiments, because they are affected by states 
differing as to sign, we must haye electric and magnetic forces 
following each other alternately in opposite directions to produce 
any effect. J do not know of any fundamental physiological 
reason for this, though it is doubtless an advantage, so that we 
might possibly find some animal whose sense organs would respond, 
T he experiment is not possible, however, ee we are in practice 
sure to get.a spark discharge, and then we may have changes of 
electric state differing as_to sign. If we did not, it would 
probably place the conductivity too low, the most promising sort 
of conductivity would be that of a cubic centimeter of mercury. 
I have however been assuming more information than has yet 
been given, so that I will now return at once to the electro- 
magnetic theory of light. I intend to adopt Poynting’s method 
of calculating the velocity of electromagnetic disturbances. This 
method is analogous to Rankin’s solution in the case of sound, and 
proceeds from another consequence of the theory—that the energy 
