Electricity in Rarefied Air. 129 
Pressure of air. 
millim 
0-011. The spark between the electrodes a and b becomes 
smaller and less extended. A magnificent induction- 
light is obtained by connecting the wires with the 
tinfoil coating. 
0:002. The spark between a and b becomes smaller, and 
passes with difficulty, but the light between ¢ and d 
continues. 
0:0003. The spark leaps over the distance between a and 
b still from time to time, but the light between c and d 
continues. 
Third Experiment.—I used for this experiment a glass tube 
18 millim. wide, bent into the form of a ring of 155 millim. 
diameter, into which two platinum electrodes were melted 
nearly at the ends of a diameter. An open glass tube from 
the outer side of the ring placed it in communication with 
the mercury air-pump. As in the previous experiments the 
tube was surrounded by two tinfoil coatings, which were not 
connected in any way with the electrodes. When the air had 
been exhausted to a certain degree, the current passed between 
the electrodes through both halves of the tube, so that the 
whole tube was illuminated. If, on the other hand, the con- 
necting wires of the apparatus are attached to the tinfoil 
coatings, no induction-light shows itself. If the exhaustion 
is pushed to the highest limit attainable with the air-pump, it 
is impossible for the current to pass between the electrodes, 
whilst when the conducting wires were removed from the tin- 
foil coatings the tube becomes luminous immediately. This 
experiment therefore confirms the former ones. One of the 
Geissler tubes belonging to the collection of physical appa- 
ratus of the Academy of Sciences, a tube 300 millim. long 
and 20 millim. in diameter, having platinum electrodes melted 
into the ends, was so far vacuous that the current of the Ruhm- 
korff’s induction-coil would not pass between the electrodes 
a distance of 200 millim. On surrounding the tube with ie 
foil coating and connecting them with the coil, the tube 
commenced at once to diffuse a brilliant light, due to the in- 
duced currents which traversed it. Consequently we have 
the same result as before. 
Fourth Eaperiment.—The glass tube used for the first two 
experiments was employed as follows :—The air was exhausted 
from it until the current from the induction-coil could no 
longer leap over the interval between the electrodes. The 
tube having been closed by means of a hermetically closing 
