THE FLOWERING PROCESS OF SHORT DAY PLANTS
Just to give one example of this kind of work, I would like to refer to some data obtained with the Michaelmas Daisy which we have studied at Wye College. The cultivar used - an unnamed garden hybrid - was selected as perhaps being somewhat nearer the wild type of Aster novobelgii than other horticulturally more attractive varieties. Originally it was thought that this species would flower when subjected to short days, because it normally flowers as the days shorten in early autumn. It was then found that it would flower neither in continuous light nor in 8-hour short days, and after a more detailed study it was found that the requirements for normal and optimal flowering are much more complicated than had been anticipated. In fact, the Michaelmas Daisy first requires a period of vernalisation or a chilling treatment, next it must be exposed to a period of approximately four weeks of long days, thirdly this must be followed by short day treatment, but finally the short day treatment must neither be too long nor too short or else a form of incomplete dormancy supervenes.
All these requirements are quite definite, at least in the cultivar used. For instance, attempts to give short and long day treatments "simultaneously" by mixtures of cycles were not successful. It is, however, possible to get a very small amount of flowering if the plants are exposed to a daylength which is approximately intermediate in the response between long and short days, e.g. 14 hours, a behaviour not unlike that of Cestrum nocturnum (Sachs, 1956). The long day effect and chilling can be replaced entirely by the application of Gibberellic acid. However, perhaps the most interesting response in the story of the Michaelmas Daisy is the effect of the final length of day; it is this factor which determines whether normal flowers are developed or whether the terminal and lateral apices cease to grow and die after forming only a few floret primordia. This critical control by daylength is itself further