( 1116 ) 
time the bearers of the conden- 
satorplates. The wall of the con- 
densatorroom, ground at the out- 
side, forms the cone of a ground- 
in joint, the neck of which is 
the end of a eylindrical tube 
forming the principal part of the 
apparatus. On this neck is made 
a stop-coek ; in the wall of the 
condensatorroom there isa little 
hole, which can be placed oppo- 
site the stop-cock. In this way 
we can exhaust the room be- 
tween the condensatorplates and 
fill it with the different gases 
under research. 
In the other, straight end of 
TRH KRAAN. : eke 
the above mentioned cylindrical 
tube the discharge tube is cement- 
in ne ed in. rhe open space between 
Yt eZ the discharge tube and the ioni- 
zationroom forms the absorption- 
room. By means of a stop-cock 
this room too can be exhausted 
and filled with other gases. 
The discharge tube was made 
in the way deseribed by Lyman’), 
with internal capillary and ring- 
shaped aluminium electrodes. li was closed by a fluorite window, 
20 m.M. diameter, 3 m.M. thick, and provided with a glass mantle 
fitting in the eylindrical tube of the apparatus. A pair ,of electrodes 
in the absorptionroom were not fit for use because of photoelectric 
effects presenting themselves. 
The arrangement of the apparatus insures perfect air-tightness 
of both ionization and absorptionroom. The distance between the 
fluorite windows of the discharge tube and the ionizationroom was 
regulated in such a way that no rays from the discharge tube could 
strike the plates of the condensator, in order to preve’t mistakes by 
photoelectric effects on the electrodes in the ionizationroom. 
As during the examination no changes may take place in the 
1) Astrophys. J. 19, 181, 1906. 
