The intensi ty of the photoluminescent (PL) emission in ZnS is known to be controlled by the electric field. I t follows that oscillating fields produce a modulated emission (brightness waves). When this phenomenon takes place in electroluminescent (EL) ZnS, i t is usually referred to as photoelectroluminescence (PEL) (1). Investigations on this effect were recently performed, concerning principally the dependence of the brightness on the oscillating voltage amplitude. Evidence was obtained about the connection between the EL and PEL processes, which turned out to be due to the same lattice defects (2). In the present paper, some experiments on the initial behaviour of the PEL emission are reported. The need of these experiments arises from the fact that investigations on the phenomenon in a t ransient phase are more suited to obtain information on the kinetics than measurements in steady conditions. Condenser-shaped cells about 0.1 mm thick were used, filled with ZnS microcrystals embedded in araldite and doped with about 10 -a Cu and C1 atoms per Zn atom. The cell electrodes were buil t with conducting quartz, in order to allow the UV light to reach the ZnS crystals. The cell was irradiated with parallel light coming from a quartz lens in front of a 250 W d.c. low-pressure Hg lamp. A filter (Jena Glaswerk UG1) and a collimator were used to shield the cell from visible light. Light from the cell, suitably filtered (Jena Glaswerk GG 495) to remove scattered UV light, was detected by a 6810 A RCA phototube which fed directly the a.c. input of the scope (a). While the cell was i r radiated for several minutes with the UV light, a sinusoidal voltage (100 V ...... amplitude, variable frequency) was applied, with an ini t ia l zero phase, to excite the PEL emission. Under these conditions the brightness waves were detected by the scope. In Fig. 1 some typical pictures are shown as recorded on the same t ime scale; a) is the 103 Hz exciting voltage, b) is the customary building-up of the EL emission alone (in this case, of course, UV light was tu rned off and the scope d.c. input was used), c) and d) correspond to the ini t ia l behaviour of the PEL brightness waves. Specifically, while in picture c) during the first half-cycle of the applied voltage the electric field was opposite to the UV flux direction, in picture d) the electric field and the UV flux were, on the contrary, directed in the same way. I t is to be noticed that the