No CrossRef data available.
Published online by Cambridge University Press: 14 August 2015
Arguments are given, suggesting that the most probable type of model is one in which ejection continues after maximum light in the visual region, but with rapid time variations.