an. Oberhalb 700 V setzten beim Abschalten von U, abbauende Vorentladungen ein.

For direct and absolute determination of wall charges an electrostatic probe method, the field mill and the Faraday cup technique are used. The first scans the charge density distribution, while the two latter methods measure total charge magnitudes. Indirect methods, estimating the \wll charges by integration of discharge current pulses are also used in the arrangement. The dischnrges are generated in (L. cylindrical qurrrtz insulated N r + N?-filled dischiirge gap by individual voltage pulses having small risetimes nnd long duration.Preliminary tests, comparison of the methods and first rcsultu concerning charge density distriblttions will be given in pwt TI of this pnper.

A bs lrac tJleiisurements have been accomplished of wall charge distributions and total charges which were generated in h'e-",-filled discharge cells with insulated electrodes under the impact of ultraviolet irradiation and of bipolar pulse sequences. The non-self-mnintained pulse discharge, generated by means of UV-light, leads to a build-up or destruction of wall charges in dependence of the efficient gap voltage. In the course of the discharge sequences wall charge transients, similar to those in ac plasma displays, are produced. Discussion of wall voltage transfer curves shows the dependency of discharge activity in the gap on the charging conditions of the insulating surfaces, placed in front of the electrodes.

First experiments with the arrangement described in part I of this paper resulted in three spurious effectx: shifting of characteristic values due to ageing effects, dissymmetric behaviour of wall charges and currents with respect to the voltage pulse sign, and backfiring. The comparison between the direct and the indirect methods demonstrated the sensitivity of the indirect ones to iincertainties iind variations in effective cell capncities. The wall charge density distributions arc axially symmetric and can be described in the electrode range by f (2) = aebt'.

No abstract