The Production of Female-sterile Flowers by Hermaphrodite
Plants of Fuchsia procumbens

Text-fig. 1.—Hermaphiodite (left) and male flowers of Fuchsia procumbens from a hermaphrodite plant.

The plants were kept on trolleys which stood from 8 a.m to 4 p.m. in the open greenhouse. At 4, the short-day trolleys were moved into the dark, into a light-proof room adjoining the greenhouse. The long-day trolleys remained in the greenhouse and were there illuminated until 10.30 p.m. by electric light: about 45 f.c at plant height. The greenhouse and darkroom were frost-protected by a thermostat set at 10°C and air was circulated between the two by fans in order to equalise temperature.

Day-length treatment was begun on 25 October, 1951 and at this time, the plants were in leaf but none visibly budded.

As flowers appeared they were pickled in 70% alcohol, either on the day they opened or, if abcissed, as soon as they fell off. Subsequently the diameters of the stigmas were measured with a micrometer eyepiece and a note made of whether the style was longer or less than or equalled the stamens.

Results

The observations cover one summer and part of the next. Unfortunately the experiments had then to be discontinued. In the first season flowering in the greenhouse began at the beginning of December, was at its height in February and continued sporadically until the end of April. The original pair of plants out of doors had a much briefer season. A few buds appeared towards the end of December but all abcissed (and were rejected on that account). Good flowers were collected only from the beginning of February to the end of March.

(in the greenhouse); but in the second season, by which time all the plants had grown and there were more flowers to compare, it would appear that abscission was rather greater in long days—and certainly continued for longer—whereas the production of short-styled flowers was, if anything, the other way.

For abscission, on the other hand, the outstanding feature of the tables is the seasonal trend, the percentage of abscission tailing off from the beginning of the flowering period onwards.

Discussion

but some evidence against it is, as a matter of fact, available. In the second season some experiments had been begun with Fuchsia plants, both seedlings and cuttings, grown in frames at still higher minimum temperatures; with thermostats set at 15 and 25°C respectively. Some of these plants, both seedlings and cuttings, came into flower, and quite abundantly, at both temperatures before the investigation was closed down.

Of recent years interest has quickened in day-length effects in perennial plants, on vegetative growth, rather than on flowering; especially for the particular case of dormancy and bud-break in deciduous trees. These are, as a matter of fact, not unrelated to their seasonal flowering, for in most of them flowers and foliage come out together with bud-break and at a time remote from that of their initiation, which occurs during dormancy or even before. The usual sequence is that dormancy follows spontaneously on leaf senescence, but that bud-break is induced by low temperatures, or by day-length, or a combination of the two (Olmsted, 1951; Wareing, 1956). Now Fuchsia procumbens is deciduous, and out of doors the leafy season includes, but is not much longer than, the flowering season. It was observed in the greenhouse in short days, on the other hand, that at irregular intervals loss of leaves occurred for short periods during the summer season itself and, moreover, that the bursts of flowering came each time with renewal of growth.

This rather suggests then an induction of dormancy by short days (perhaps in combination with falling temperatures in the case of winter dormancy) controlling both leaf and flower growth. It would be necessary to suppose that the leaves alone perceive day-length, so that growth could begin again spontaneously even in continuous short days, once the leaves fall.

2. Flower abscission.

It can be seen in Table II that if buds which abscissed are eliminated from the comparison of flower production in long and short days, the superiority of long days largely disappears. In the data of Table I on the other hand, abscission is hardly evident at all except out of doors. Even in the second season, the percentage of abscissed flowers had fallen to zero in the greenhouse by the time the experiment was concluded, though at that time a few buds were being abscissed from the plants out of doors. This suggests two possibilities: either that when flowering begins the first buds to be initiated are defective, or, that some enviromental factor causing abscission is operative at the beginning of each flowering season.

It will be remembered that flowering began in the greenhouse over two months earlier the second season than the first. It then seems likely, that the difference in the incidence of abscission in the two seasons, may well have been an environment effect. As the day-lengths were fixed in both seasons, some other factor than day-length must be considered responsible. The occurrence of abscission out of doors after it had ceased in the greenhouse the second season strongly suggests that the responsible factor may have been temperature. Indeed it so happens that the flowering season began the first season just about the time, early December, that abscission ceased in the second season in the greenhouse; and it may then be guessed that low temperatures are the cause of abscission and that by mid-summer they were no longer operative in the greenhouse but still effective out of doors.

That day-length modifies the effect of temperature is, however, apparent in the data of Table II, for it is evident that abscission fell to zero some five weeks earlier in short than in long days, and was less in short days in every corresponding week throughout the season. It seems likely that long days may combine with low temperatures to produce abscission out of doors; and as the natural flowering season does not begin until after mid-summer and extends well into the autumn, abscission at the beginning of the season must be mainly due to the day-length; and at the end, to the falling temperatures,

3. Short-styled flowers

In Table III, measurements of stigma diameter are shown. Godley (1955) gives sketches of flowers in which the stigma was shrunken but the style as long as in a hermaphrodite flower. Critical examination of short-styled flowers in the present investigation indicates that abortion affects the stigma primarily, and extends down a varying length of the style below, upon occasion. Stigma measurements should therefore provide a more sensitive detection of female sterility than comparison of the style and filament lengths. The data in Table III do not, however, present a very different picture of the incidence of male flowers than Tables I and II.

These experiments were undertaken initially, on the supposition that production of short-styled flowers by hermaphrodite plants is seasonal. There is, however, little confirmation of this in the data collected: the plants grown out of doors, those subject to natural environment, showed indeed a variable incidence of male flowers from week to week, but no recognisable trend as the season advanced. On the other hand, there is an outstanding treatment difference in Table I, which does suggest that female sterility may have an environmental cause: the general difference between the plants outside and those in the greenhouse. Once again this suggests a temperature effect: that female sterility is likely among flowers initiated at relatively low temperatures. In confirmation, attention may be drawn also to the relative abundance of male flowers in the greenhouse the second season compared with the first; that is, in greenhouse temperatures of the spring and early summer compared with late summer. However, if low temperatures are responsible for both bud abscission and style abortion, the effect must be different in each case. To begin with, the two imperfections are not associated in the same flower—the proportion of short