The origin of cosmic rays and applicable laboratory experiments are discussed. The principle problems of shock acceleration for the production of cosmic rays in the context of astrophysical conditions are found to be: (1) The presumed unique explanation of the power law spectrum is shown instead to be expected as a universal property of all accelerators with high loss; (2) The extraordinary isotropy of cosmic rays and the limited diffusion distances implied by supernova induced shock acceleration requires a more frequent and space-filling source than supernova; (3) Near perfect adiabaticity of reflection by strong hydromagnetic turbulence is required in order to accelerate the particles. In each doubling in energy roughly 10(5-6) scatterings are required with negligible energy loss. This seems most unlikely by strong hydromagnetic waves; (4) The evidence for acceleration due to quasi-parallel heliosphere shocks is weak. There is small evidence for the expected strong hydromagnetic turbulence, and instead, only a small number of particles are observed to be accelerated after only a few shock traversals; (5) The acceleration of electrons in the same collisionless shock that accelerates ions is difficult to reconcile with the theoretical picture of strong hydromagnetic turbulence that reflects the ions. The hydromagnetic turbulence will appear adiabatic to the electrons at their much higher Larmor frequency and so the electrons should not be scattered incoherently as they must be for acceleration. Therefore the electrons must be accelerated by a different mechanism. This is unsatisfactory, because wherever electrons are accelerated these sites, observed in radio emission, may accelerate ions more favorably. Because of these difficulties, an alternate explanation is given where reconnection of twisted magnetic fields (from gravitational condensation, accretion, and conservation of angular momentum) accelerates particles along field lines. The acceleration is coherent provided the reconnection is coherent, in which case the total flux, as for example of collimated radio sources, predicts single charge accelerated energies significantly greater than observed The same acceleration process should occur during the formation of stars associated with the T-Tauri phase and bi-polar outflows. Because of the ubiquitous nature of matter condensations in the universe, and the near universal excess of angular momentum, the acceleration is inherently isotropic and space-filling in nature. Four laboratory experiments are suggested that would form the basis of a substantiated science of plasma processes of acceleration. These include an alpha-omega dynamo using liquid sodium, a one-ended plasma pinch that simulates the formation of collimated radio sources or bi-polar out-flows, reconnection or current interruptions in a tokamak, which simulates acceleration by coherent reconnection, and finally a collisionless shock experiment to simulate the classical shock acceleration process.
