During the last few years, ideas regarding heating solid D-T mixtures up to thermonuclear temperatures'v' by high-power electron beams have been further developed.v" It is important that the requirements for beam focusing and beam duration have essentially decreased, as compared with results of earlier works. I ,2 In order to realize recent proposals, an electron beam with a current density of about 10 8_109 afcm 2 , a pulse duration of (5-10) ns, and a beam energy of 3 MJ has to be generated. We have no such beam today.5-8 Therefore, the main experimental investigations of high-current beams being carried out at I. V. Kurchatov Institute are concerned with the problems of focusing and decreasing pulse duration.It appears that solution of the beam-focusing problem may have two stages. In the first stage it is necessary to find a way to transmit the energy of high-power electromagnetic waves to diodes with sufficiently small dimensions.The main obstacle in the transmission problem is the low breakdown strength of well-known dielectrics and, in particular: the vacuum insulation; however, recent studies of the improvement of vacuum insulation by means of magnetic fields 5 ,9, 10 bring us hope of success.The beam-focusing problem is more complicated. Some experimental and theoretical investigations [11][12][13] show that the desirable beam focusing will be obtained in a drift chamber with longitudinal or azimuthal magnetic fields if the beam produced in the diode gap is very cold. Hence, progress now being made in the problem of beam focusing directly in the diode region is of considerable interest. Good results were achieved when plasma filled up all or part of the diode region. In principle, there are two possibilities for plasma effects on beam generation. First, under certain conditions, intensive beams can be generated in a currentcarrying plasma. This effect was discovered in Ref. 14 and apparently was closely connected with the plasma-current instabilities. Further studies with the powerforming lines as a current source l 5 -17 have shown the theoretical possibility of high-current beam generation in a plasma discharge, but have not resulted in an understanding of this phenomenon.By contrast with these results, the experiments performed and explained by Yonas and others6,18 are more clear. They used a plasma channel created along the diode axis. The electron beam emitted by the edge of flat cathode was pinched 668