In view of recent developments attention is directed again at two aspects of the wellknown 'knee' in the cosmic ray energy spectrum at 3PeV: the mass of the predominant particles at this energy and their source,. It is inevitable in a subject such as this that ideas -and conclusions -evolve. Earlier, we had used a particular acceleration model [1] and the nature of the ISM in the local ISM to infer that the particles are mainly oxygen nuclei [2]; direct measurements, when extrapolated ( by at least a decade in energy ) [3], gave a similar result. Initially, [2] no specific source was identified. More recently, however, [4] we have specified the Monogem Ring supernova remnant [5] as the likely source; this is at just the right distance and age and the energies are reasonable. Concerning the mass composition at the knee, a quantity more difficult to determine, recent direct measurements, [6], which extend to higher energies than hitherto, show a likely flattening in the spectrum above ~10 4 GeV/nucleon for He-nuclei, a flattening which, if extrapolated to higher energies, would meet the measured spectrum in the knee region. The other nuclei do not show this feature. He-nuclei in the knee region would also be marginally more consistent with KASCADE extensive air shower data, although there are serious problems with EAS mass estimates in that experiment [7]. Concerning the acceleration, recent models (eg [8]) applied to the Monogem Ring SNR allow a satisfactory explanation in terms of either oxygen or helium, but with the latter being a distinct possibility and perhaps more likely.