114 Dr. E. Weintraub on the Arc in 
which would have similar properties, and the action of the 
magnetic field on the arc presents, therefore, features which 
are at first sight paradoxical. If the arc is vertical and the 
field horizontal, the deflexion of the are and that of the spot 
are in opposite directions. The best way to carry out this 
experiment is to have a bulb about 24 in. in diameter, pro- 
vided with two mercury cups at the bottom and a graphite 
anode suspended from the top. The vertical arc is started 
between the graphite anode and one of the cups. The spot 
on the cathode is usually wandering about the surface. To 
give it a fixed position it is best to have an iron wire pro- 
truding a little above the surface of the mercury. Ihe 
bright cathode-spot is then centred on the mercury surface 
around the iron wire. Under these conditions the arc is 
first deflected one way, and after a while the bright spot is 
seen to move away from the iron wire in the opposite 
direction. 
The action of the field on the mercury are is complicated 
by the fact that the conductivity of the mercury-vapour 
changes with its temperature, so that the observed effects 
change in their appearance when the amount of mercury- 
vapour present in the arc is changed. 
It is interesting to note that the action of the field on the 
cathode-spot can be in most cases formally accounted for by 
assuming that positive current elements leave the cathode 
surface in a direction perpendicular to that surface. 
§ 4. 
Arcs in Vapours of Alkali Metals. Behaviour of Ama/gams. 
The most volatile metals, after mercury, are those of the 
alkali group, the boiling-points of the most common of them, 
sodium and potassium, lying between 600° and 700°. On 
the subject of the arcs of these metals in an exhausted space, 
there are in the literature only a few preliminary experiments, 
published by Arons in the article mentioned. The difficulties 
one encounters in this work are many, and it was only after 
many failures that I succeeded in making the simple form of 
a tube in which the arc could be started and maintained for a 
time long enough to make observations. The tube had the 
form shown in the figure. 
Heperiment 13 (Pl. III. fig. 8).—CD is a wire connected to 
the negative pole of the source. The part OD of this wire is 
best made of platinum or platiniridium. EH is a piece of iron 
about 7 in. long and } in. diameter. It swings around the 
pivot P, and is to serve as the auxiliary anode; FP is a 
platinum wire ; the two seals F and C are purposely placed 
