112 Dr. E. Weintraub on the Are in 
anode—or whether the ions, after having reached the positive 
electrode, become positively charged, and move back toward 
the cathode, must be left undecided. It was, however, in- 
teresting to see what results could be obtained if, starting 
with the assumption of unidirectional flow, one attempted 
to solve the second fundamental question in connexion 
with the conductivity of an arc, namely, that referring to the 
amount of matter carrying the current. The mercury are, 
being inclosed, and taking place in a vacuum, presents a 
better opportunity for the study of this question than the ares 
burning in air. Still, the difficulties are such that reliable 
quantitative data could not thus far be obtained. 
Assuming a unidirectional flow of ions, one might think 
that by arranging a vertical tube, having at the bottom a 
mercury cathode, and at the top a graphite anode, providing 
this tube with a well-cooled condensing-chamber around the 
anode, one could get some data by determining the amount 
of vapour collected during a certain time in that chamber, 
the current being kept constant. This procedure would, 
however, be faulty in view of the fact that only a small 
part of that mercury has participated in the actual transport 
of the current. A condensing-chamber at the bottom of the 
Jamp in the way shown in fig. 2 does away with a large part 
of the mercury that simply evaporates from the cathode, and 
if a part of the tube is surrounded with a jacket with circu- 
lating water, large amounts of mercury condense there which 
otherwise would have reached the condensing-chamber at the 
top of the lamp; the current in the lamp is, if anything, 
increased by this condensation taking place. The following 
arrangement is somewhat more satisfactory without, however, 
enabling one to get quantitative results. 
Eaperiment 12.—The lamp has the form shown in fig. 7 
(Pl. ill), and is submerged in a large vessel through which 
cold water constantly circulates. It can be supposed that the 
mercury ions, after having given off their charges at the 
anode, will diffuse and condense in the condensing-chamber 
C, which, being outside of the arc, is the coolest spot in the 
tube. The cooling with water not being sufficient to perfectly 
do away with superfluous vaporization of mercury, experi- 
ments conducted in this way can only give maximum values, 
above which the quantity looked for cannot lie. 
From the experimental data thus obtained, one thing seems 
to be surely established, namely, that the amount of matter 
carrying the current is a very small part of that required by 
Faraday’s law. 
Some experiments have been performed on the loss in 
