In contrast to earlier expectations, it is estimated that generation of runaway electrons from close collisions of existing runaways with cold plasma electrons can be significant even for small electric fields, whenever runaways can gain energies of about 20 MeV or more. In that case, the runaway population will grow exponentially with the energy spectrum showing an exponential decrease towards higher energies. Energy gains of the required magnitude may occur in large Tokamak devices as well as in cosmic-ray generation.

12 The spectral line shapes are temperature independent up to 20 K, showing the predominance of rigid-lattice spin-spin relaxation mechanisms. Thus field-dependent phonon processes [see R. Orbach, Proc. Roy 0 Soc. London, Ser. A 264, 485 (1961)] are not importantThe use of relativistic electron or ion rings trapped in magnetic mirror fields to produce absolute-minimum-2? configurations for confining thermonuclear plasmas was originally proposed by Christofilos, 1 ' 2 and McNally. 3 As presently envisioned, rings of several hundred MeV ions will be needed for a practical fusion reactor. As recently proposed 4 such ion rings may be produced by adiabatic compression of low-energy ion rings produced either by slow pulse injection of H 2 + ions 5 or by injection of intense ion beams which have recently become available. 6 In an effort to investigate the stability of such proposed ion rings, model experiments with easier-toproduce electron rings were performed by the Astron group at Lawrence Livermore Laboratory and more recently by the relativistic-electroncoil-experiment (RECE) group at Cornell.At Lawrence Livermore Laboratory attempts to produce field-reversing electron rings by multiple-pulse injection of 6-MeV, 650-A linac pulses proved unsuccessful. 7 However, single-pulseinjection experiments using an additional toroidal field, JB e , approximately equal to the basic mirror field, B z0 , and a tank filling of 100 mTorr here. In addition, exchange coupling in similar structures has been found to be negligible [E. Bucher, H, J. . hydrogen gas led to the generation of electron rings exhibiting axial field changes, bB/B ZQ^ 40%, and overall lifetimes of several milliseconds. On the other hand, field-reversing rings (6B/B Z0 ^ 190%) were first produced at Cornell University by injection of intense 400-keV, 20-kA, 80-nsec electron beams in an Astr on-type magnetic-mirror geometry, 8 * 9 and with an annular beam injected through a magnetic cusp field. 10 The rings appeared to decay stably, with the exception of the well-known precessional mode 11 which was found to be stabilized by the image currents in the conducting walls or by adding small toroidal or quadrupole field components to the basic mirror field. 12 ' 13 Also, distortions of the applied magnetic field of up to 1% did not appear to trigger instabilities. 14 The overall ring lifetimes, however, were limited to less than 30 /xsec due to collisional diffusion 15 of the fast electrons in the background gas of several hundred millitorr of hydrogen which was needed to obtain sufficient ring trapping.The present paper reports on new experiments in which higher-energy electrons (^2.1 MeV) are injected into a new facility, RECE-Christa. In Field-reversing electron rings, exhibiting initial axial field changes 6B/B e0^ 170%, and having overall lifetimes of greater than 1300 jusec, have been generated by injection of 2.1-MeV, 30-kA electron-beam pulses into the relativistic-electron-coil-experiment (RECE)-Christa mirror field, B Z Q(Z) . These lifetimes, which are co...

At this Conference organized by the Ecole Polytechnique Fédérale de Lausanne and the European Physical Society, and attended by more than 400 participants, approximately 200 talks were presented, including 21 invited papers. These papers were approximately equally divided into experimental and theoretical contributions although the invited papers mostly concerned experimental work. As was to be expected from the present trends in funding and interest, the Conference proceedings were dominated by reports on tokamak research, a topic which concerned more than half of the invited talks and approximately 30% of the contributed papers. In contrast, the representation of other, normally strong areas of research, as laser fusion, high-β plasmas, and stellarators was significantly reduced owing to topical conferences in these fields which were held at about the same time, and because some major machines are still in the construction stage. The Conference included new results on TFR which tend to explain the confinement-time limitations observed in that machine, extensive new work on internal and macroscopic disruptions in tokamaks and an order-of-magnitude advance in plasma confinement time observed in the 2XIIB mirror machine at Livermore.