“…A short electron bunch [42], a powerful laser pulse [43], a sharp cut in the bunch current profile [40,44], or a relativistic ionization front copropagating within the drive bunch can seed the SMI quite well. Analytical and numerical calculations, however, have shown that bunches with long rise times (longer than or about the plasma wavelength) do not produce stable bunch trains [30,38,39].…”