USDA Yearbook, 1920, pp. 377-400

Flowering and Fruiting of Plants as Controlled by the Length of Day

By W. W. GARNER, Physiologist in Charge, and H. A. ALLARD, Physiologist, Tobacco and Plant-Nutrition Investigations, Bureau of Plant Industry

ONE of the most characteristic features of plant growth outside the Tropics is the marked tendency shown by various species to flower and fruit only at certain periods of the year. This behavior is so constant that certain plants come to be closely identified with each of the seasons, in the same way as the coming and going of migratory birds in spring and fall. In midwinter the blossoms of cyclamen, freesia, the brilliant color of poinsettia, and the fruits or berries of ardisia, all are reminders of the season; in spring we expect to see the unfolded blossoms of forsythia, wild violet, crocus, redbud, dogwood, and other typical plants; as summer approaches, poppy, rhododendron, iris, and columbine begin flowering; in the autumn salvia, aster, cosmos, dahlia, and chrysanthemum herald the approaching end of the open growing season.

The thought at once suggests itself that the underlying cause or causes of flowering or fruiting occurring only at a particular season must be purely internal, else the vagaries of the weather and other variable external conditions would seriously upset the regular cycle. It is true, of course, that plants can flower and fruit successfully only within certain limits of temperature and moisture supply, and it has long been known, also, that light is indispensable. Thus, plant development may be retarded in the spring by cool weather, and at times drought or excessive rainfall may interfere, but, in general, flower and fruit are produced regularly in their seasons in spite of these temporary disturbances. The ripening of seeds as a sequel to flowering is obviously of great importance to many plants, in that it affords the only means of avoiding extermination. We might easily conclude from this that the plant's entire activities are directed toward this means of propagation, all preliminary growth and development of root, stem, and leaf being incidental. This view, however, is not correct. The plant merely inherits the capacity to flower and fruit in response to certain favorable external conditions. It is both interesting and practically important. therefore, to determine these conditions.

While marked regularity in the time of flowering and fruiting is the rule it plants so long as they are grown in any particular locality in temperate regions, transferring plants from one region to another may greatly change their habits. A species which flowers and fruits readily in one region may become sterile in another, or, in some instances, the time of flowering may be changed from spring to fall, or vice versa. Again, plants behaving as annuals in one region may become biennials in another. These changes in the behavior of plants when grown outside their native regions furnish strong evidence that external conditions control the processes of flowering and fruiting and also suggest the possibility of artificial control,

Does Change in Temperature Account for Seasonal Flowering and Fruiting?

We instinctively think of temperature as the outstanding external factor causing one season to differ from another in its effects on plants. In particular, we associate the opening of spring flowers with moderate temperatures, following the chill of winter. Likewise, as the characteristic flowers of autumn make their appearance we have been inclined to assign decrease in temperature as the cause, mainly perhaps for the reason that there has seemed to be no other obvious cause for the flowering of these plants. Temperature unquestionably is a very important factor in plant development, and plants differ widely in their temperature requirements. Nevertheless, change in temperature fails to explain why plants flower and fruit at certain periods; that is to say, even though the appropriate temperatures are provided out of the regular flowering and fruiting season, as a rule the flower and fruit fail to appear except in their usual seasons. For example, common iris, which flowers in May and June, will not blossom under ordinary conditions when grown in the greenhouse in winter, even under the same temperature conditions that prevail in early summer. Again, one variety of soy beans will regularly begin to flower in June of each year, a second variety in July, and a third in August, when all are planted on the same date. There are no temperature differences during the summer months which could explain these differences in time of flowering; and, since "internal causes" alone can not be accepted as furnishing a satisfactory explanation, some external factor other than temperature must be responsible.

The ordinary varieties of cosmos regularly flower in the fall in northern latitudes if they are planted in the spring or summer. If grown in a warm greenhouse during the winter months the plants also flower readily, so that the cooler weather of fall is not a necessary condition. If successive plantings of cosmos are made in the greenhouse during the late winter and early spring months, maintaining a uniform temperature throughout, the plantings made after a certain date will fail to blossom promptly, but, on the contrary, will continue to grow till the following fall, thins flowering at the usual season for this species. This curious reversal of behavior with advance of the season can not be attributed to change in temperature. Some other factor is responsible for the failure of cosmos to blossom during the summer months. In this respect the behavior of cosmos is just the opposite of that observed in iris.

Certain varieties of soy beans change their behavior in a peculiar manner with advance of the summer season. The variety known as Biloxi, fore example, when planted early in the spring in the latitude of Washington, D. C., continues to grow throughout the summer, flowering in September. The plants maintain growth without flowering for 15 to 18 weeks, attaining a height of 5 feet or more. As the dates of successive plantings are moved forward through the months of June and July, however, there is a marked tendency for the plants to cut short the period of growth which precedes flowering. This means, of course, that there is a tendency to flower at approximately the same time of year regardless of the date of planting. As a necessary consequence, the size of the plants at the time of flowering is reduced in proportion to the delay in planting. This behavior is well shown in figure 1, for all plantings had flowered when photographed. Like cosmos, the Biloxi soy beans show a marked tendency to flower at a definite season of the year, and if planted early they wait, as it were, till this season arrives. It is easy to see the advantage which a plant has in being able to shorten the growing period which must precede flowering if, for any reason, the plant gets a late start. In such a case the chances of successfully maturing seed before frost and thus avoiding extermination in a given region are greatly increased, and the production of seed constitutes the plant's method of perpetuating itself in the face of the destructive action of cold. It is important, however, to make a distinction between advantage and cause with respect to time of flowering. The Biloxi soy beans by curtailing the period of vegetative activity when beginning growth late in the season are actually able to forestall the arrival of cold weather; hence, low temperature can not be considered as a cause of this behavior. The response of the soy beans to the advance of the season begins before there is any decrease in temperature.


Soy Beans Planted at Regular Intervals during the Summer.
FIG. 1.—From left to right Plantings were made at intervals of three to five days, beginning July 14. All plantings had flowered and growth had almost ceased when photographed September 8, The progressive decrease in vegetative development as the dates of planting become later and later is very striking.

Effect of Shortening the Duration of Daylight.

It is perfectly clear that the time of flowering and fruiting of many plants is inseparably linked in some way with the advance of the season, and necessarily there must be some external factor which maintains this relationship.

With temperature eliminated, there remains one change from season to season which proceeds with great regularity, namely, the change in length of day and night. At Washington. D. C., the time between sunrise and sunset ranges from nearly 15 hours in late June to about 9 1/2 hours at Christmas. To determine whether this change in the length of day is a cause of regularity in the time of flowering and fruiting, a series of experiments was made in which a number of plants were darkened for a portion of the day during the long days of summer. The results obtained were remarkable. The plants no longer persisted in their usual habit of deferring the flowering period till a particular time of the year had been reached. The normal seasonal periodicity was completely broken up. The experiments included a large variety of plants both wild and cultivated, and it was found that the reaction to differences in the length of the day is of very wide occurrence.

The method followed in these tests is very simple. A "dark house" was so constructed as to admit air freely at the bottom and allow its escape at the top, without the admission of daylight. For convenience a series of small steel tracks leading into the dark house was provided, and on these tracks were mounted a number of trucks with steel wheels capable of supporting the containers in which the plants were grown. With this equipment it was a simple matter to transfer the plants into and out of the dark house at regular intervals each day. For example, if it were desired to give a particular lot of plants eight hours of light each day the truck bearing these plants would be rolled into the dark house at, say, 4 o'clock in the afternoon each day and rolled out into the open air again at 8 o'clock the following morning. The outfit used in the experiments is well shown in figure 2. For comparison, in each test a second lot of plants, known as "control" plants, was grown under exactly the same conditions as those to which the darkened plants were exposed, except that the control plants were exposed to light throughout the day.


Dark House Used to Shorten the Daylight Period.
FIG. 2.—The receptacles in which the plants are grown are placed on trucks fitted with steel wheels. The trucks are run into and out of the dark chamber over steel tracks. In this way the plants receive only the desired number of hours of light each day.

The response of the plants to this artificial shortening of the daylight period was prompt and clean cut. Biloxi soy beans which germinated May 17 were allowed to receive seven hours of light daily, beginning May 20. These plants were in blossom in 26 days, whereas a similar lot of plants exposed to light throughout the day required 110 days to flower. This variety of soy beans, which ordinarily flowers in September, even though planted in May, was forced into blossom in June, simply by shortening the daylight period. In further tests it was found that a daylight period of 12 hours was as effective as the 7-hour period in forcing the flowering of the soy beans. It is easily seen, therefore, why this variety of soy beans ordinarily does not flower till September, for it is at that time that the length of the day is reduced to 12 hours.

An experiment was made with another variety of soy beans known as Peking at the same time and in the same way as with the Biloxi. In this case the plants receiving 7 hours of light daily flowered in 21 days, while those exposed to light for the entire day required 62 days to reach the blossoming stage. This is fully in accord with the fact that the Peking regularly blossoms in the field in July, two months in advance of the Biloxi. The Peking, therefore, is capable of flowering under a considerably longer day than the maximum day length which will cause the Biloxi to blossom.

A common wild aster which ordinarily flowers in September was found to behave in the same manner as the Biloxi soy beans when exposed to, a shortened daylight period. 'When exposed to 7 hours of light daily the aster was in bloom in 36 days, as against 122 days when exposed to light for the entire day. A variety of Lima bean imported from Peru which ordinarily does not flower till late in the fall at Washington, D. C., was caused to blossom in 28 days by reducing the daily light period to 7 hours. The common ragweed behaved in a similar manner.

Some Effects of Short Daylight Periods.
FIG. 3.—A, Cosmos requires comparatively short days in order to flower. These plants were grown in the greenhouse during the short days of winter. The plants on the left received electric light after sunset, so as to prolong the daily light period, and thus were prevented from blossoming. This explains why cosmos does not flower during the summer months, the days being too long.
FIG. 4.—B, Forcing flowering and fruiting in soy beans by shortening the daylight period. The plants on the left were exposed to the full day length of summer, while those on the right received only 10 hours of light each day, all other conditions being the same. Many plants will not flower and fruit when the days are long.
FIG. 5.—C, Chrysanthemums are made to flower in summer by shortening the daylight period. The plant in blossom on the left was allowed to receive only 10 hours of light daily, beginning May 12, and the first blossoms opened July 17. The plant on the right, receiving light during the whole day, did not flower till fall.

One scarcely expects to see chrysanthemums in bloom in midsummer but, as is indicated in figure 5, these typical fall-flowering plants are readily made to flower in summer by shortening the length of the daily light period. Late-flowering varieties of dahlia are readily forced into blossom during the summer by reducing the length of the daily light exposure to 10 hours or less. A highly colored specimen of poinsettia, the plant so typical of the Christmas season, was developed in August by reducing the daily light period to 10 hours.

In the light of these experiments there is no longer any element of mystery concerning the fact that when plantings of cosmos are made at successive dates in early spring a point is reached at which the plantings suddenly swing over from flowering in the spring to flowering in the fall. Cosmos begins to flower in the fall when the length of day has decreased to about 12 hours (sunrise to sunset) and, in the same way, it is no longer able to flower in the spring after the days become much in excess of 12 hours in length.

There is, then, a large group of plants, including most of the so-called summer annuals, which regularly flower after midsummer as a result of decrease in the length of the day. While relatively short days favor flowering and fruiting in these plants, long days are more favorable to rapid and extensive vegetative development. Some of these plants, therefore, if they receive the full benefit of the long days of summer, may reach giant proportions before being brought into the flowering condition. Thus, we can understand why it is that when the farmer plants some crops too early, there is a tendency toward excessive development of leaf and stem with little flowering or fruiting. Late planting, on the other hand, may lead to dwarfing in growth but abundant flowering and fruiting. Again, it is easily seen why carrying some plants into northern latitudes causes very rank growth, with a tendency toward barrenness, since the length of the day in summer increases as we go northward. Plants in this group differ widely as to the extent to which the longest summer days must be shortened to induce flowering, with the result that some flower in July while others may not flower till November. Even the latest of these are readily forced into flowering and fruiting during the hottest part of the summer simply by shortening the daylight period, so that there is no reason for considering the cooler weather of fall as a factor of importance.

Effect of Darkening Plants in the Middle of the Day.
FIG. 6.—A, The Biloxi soy beans in box on the right were exposed to light from daylight to 10 a.m. and from 2 p.m. to dark, in all 9 to 10 hours daily. The plants in the box on the left were exposed to light from 6 a.m. to 6 p.m., 12 hours daily. The 4-hour period of darkness in the middle of the day was not effective in hastening flowering and the ripening of seed, although the plants thereby received less than 12 hours of light daily.

Red Clover Flowers under the Influence of Long Days.
FIG. 7.—B, The plants in the can on the left were exposed to the light for only 10 hours daily, while those in the can on the right were exposed throughout the day during the spring and early summer. Long days favor flowering in this type of plant. The prostrate habit of growth during the short days of winter is characteristic of this group of plants.

A modification of the method of shortening the daily light period used in the above-mentioned experiments gave somewhat surprising results. Instead of giving Biloxi soy beans a single exposure to light each day, they were transferred into the dark house at 10 o'clock in the morning and returned to the light at 2 o'clock in the afternoon. As is shown in figure 6, the midday period of darkening was almost without effect in hastening flowering, although the two daily light periods aggregated considerably less than 12 hours in duration.

Another important feature of the effect of shortening the daylight period should be mentioned. Just as many plants may be forced into flowering by artificially shortening the daylight period, so also is the ripening of the fruit or seed greatly hastened. Thus, in a test with Peking soy beans, two similar lots of plants were grown tinder natural summer conditions of daylight till flowering had taken place and very small seed pods could be seen. At this stage one lot of plants was darkened for a portion of the day, so that they received only 7 1/2 hours of light daily, while the second lot continued to receive light during the entire day. The result of the test is indicated in figure 8. Six weeks after flowering, the leaves were falling from the plants which received the shortened light exposure, and sonic of the seed pods were fully ripe. The plants under the natural length of day did not mature their seed till several weeks later. Several other plants have responded in a similar manner to artificial shortening of the daylight period.

Some Plants Require Long Days for Successful Flowering and Fruiting.

In striking contrast with the group of plants already discussed is a second group regularly flowering in late spring and early summer. It is obvious that these plants do not require short days to reach the flowering stage. On the contrary, it has been found that short days prevent, or at least greatly delay, flowering and fruiting. To this class of plants belong the so-called winter annuals: also many of our common vegetables. The radish has given some interesting results which are fairly typical for the group. The ordinary varieties of radish when planted in the spring first produce a thickened edible root and somewhat later develop a flowering stem, which in due season matures seed. Thus the Scarlet Globe variety, planted May 15, began to blossom June 21 when exposed to the natural length of day. A similar planting, made at the same time, but allowed to receive only 7 hours of light daily, grew slowly and formed no flowering stem. Under the shortened daylight period the roots of the radishes continued to enlarge slowly throughout the summer, with a corresponding increase in size of the rosettes of leaves surmounting the roots. One of the plants which was transferred to the greenhouse in the fall continued its slow growth through the winter months. Finally, as the days lengthened in early spring this plant was able to send up a flowering stem and perished after seed formation was completed. Thus the radish, which ordinarily is a typical annual, was made to behave as a biennial. The radish furnishes a case in which flowering may be prevented for a more or less indefinite period by shortening the daily period of illumination, in contrast to the group of plants previously considered, which are prevented from flowering by long days and are forced into flowering by shortening the daylight period.

The Length of the Day is a Controlling Factor in the Ripening of Seeds and Fruits.
FIG. 8.—A, Peking soy beans which were exposed to light during the entire day in summer till flowering had taken place, but thereafter were allowed to receive only 7 1/2 hours of light daily. B, Peking soy beans exposed to light during the whole day throughout the test. The two lots of plants are of the same age and were treated exactly alike except as to the length of the daily light exposure after flowering had taken place.

The behavior of the radish is in no sense exceptional. Failure to send up a flowering stem during the short days of winter and early spring is a characteristic feature of many hardy plants which maintain more or less vegetative activity at those seasons of the year. The tendency is toward a prostrate type of growth, with free stooling or a rosette form of leaf development. As the longer days of spring come on, the character of growth changes, and upright-growing stems appear, in preparation for flowering and fruiting. Our small grains belong to this class of plants. Red clover furnishes a good illustration of this behavior, as may be seen by referring to figure 7. By allowing the test plants to receive only 10 hours of light daily, the prostrate nonflowering type of development was continued long after a corresponding lot of plants which were exposed to light all day had developed upright stems and had successfully flowered and fruited. Likewise, the common evening primrose transplanted from the field in early spring continued the prostrate rosette type of development for several weeks when allowed a daylight period of only 10 hours, whereas similar plants exposed to light throughout the day quickly developed tall, erect flowering stems.

Under ordinary conditions spinach can not be grown successfully for table use during the summer months, because it quickly goes to seed instead of forming the desired rosette of large leaves. This behavior has been generally attributed to high temperature. It is quite true that within suitable limits an increase in temperature, as a rule, speeds up plant development. Nevertheless, experiments have shown that spinach will produce an excellent rosette in summer if the light period is reduced to 8 or 10 hours. Under these conditions the flowering stems are unable to form, or, at least, their appearance is greatly delayed.

Tubers of the groundnut (Apios) planted on March 11 sent up shoots which appeared above the ground on April 6. By April 20 flower birds were showing on all these plants. On one lot which was exposed to light all day, the first open blossoms appeared June 1, and flowering continued till late in August. On a second lot which received only 10 hours of light each day, beginning May 20, only one or two blossoms were able to open, the other flower buds dropping off. Thus, in spite of the fact that the flower buds had been laid down before the daylight period was shortened, these buds were unable to unfold under the new conditions.

The above examples illustrate the fact that there is a large group of plants which are brought into the flowering and fruiting stages of development because of the increase in length of day as spring advances into summer. A5 a matter of convenience in discussing flowering and fruiting activities. this group may be spoken of as "long-day plants," in contrast with the group previously discussed, which are forced into flowering and fruiting by the shortening of the days in fall and therefore may be called "short-day plants." While as a whole there are sharp contrasts between the two groups, there are many plants which perhaps may be regarded as occupying an intermediate position. There is, in fact, no hard and fast line between these two classes of plants. There are some plants, indeed, for which it is possible to provide a daylight period too long, on the one hand, and too short, on the other, to induce flowering and fruiting.

It has already been pointed out that while the short-day plants are diverted toward the flowering and fruiting, or reproductive, stage of development by shortening the daylight period, the rate and amount of vegetative growth, on, the other hand, are increased in proportion to the lengthening of the daylight period. In the case of the long-day plants the reproductive stage is induced by a lengthening of the daily period of illumination, so that vegetative growth is necessarily restricted more or less through the influence of long days. This refers more particularly, however, to the final size attained by the plant rather than to the rate of growth. For example, as already has been detailed, long-continued exposure to a short day length eventually produced a radish of exceptionally large size, but it required nearly nine months to accomplish this result. The rate of growth was less than when the radish is exposed to the light for the whole day in summer. It is true, however, that there are plants whose rate of growth is less during the longest days of summer than during the days of spring and fall, which are of intermediate length.

How Length of Day Controls Everflowering and Everbearing.

In temperate regions most plants have a comparatively short period of flowering and fruiting each year, though plants differ in the length of this period. In some cases, however, this period of reproductive activity continues through several months, and plants behaving in this manner are known as everbloomers or everbearers. In the preceding discussion the fact is brought out that most plants tend to continue the purely vegetative form of development as long as the days are of a certain length, while under another length of day vegetative development quickly gives way to flowering and fruiting. Not all plants are equally sensitive, however, to changes in the length of day. With these two fundamental facts in mind it is easy to understand the relation of the length of day to the condition in plants known as everblooming or everbearing. If Biloxi soy beans or cosmos plants are subjected to an artificially shortened period of daylight of 9 or 10 hours in midsummer the purely vegetative form of activity is promptly checked and flowering and fruiting quickly follows. Subjecting Biloxi soy beans to a somewhat longer daylight period of 12 hours in midsummer has resulted in a considerably larger stature for the plants, and blossoming has been considerably delayed. Furthermore, lengthening the daylight period from 10 hours to 12 hours has markedly slowed down the rate of development of the pods, and consequently the ripening of the seed. In other words, we have been working in the direction of vegetative activity and to a greater or less degree away from the condition of free and rapid flowering, ripening of seed, and final death of the plants. This suggests the possibility of a nice balance or adjustment between the vegetative and the reproductive phases of development which would express itself in more or less prolonged everblooming and everbearing tendencies. From this viewpoint the everflowering tendency simply means the ability to continue both vegetative and reproductive activities more or less successfully together.

Two features of the relationship between length of day and everblooming are of special importance, namely, (1) the occurrence in different latitudes of the proper range in length of day continuing over a sufficiently long season and (2) differences among plants in their sensibility to changes in length of day. In the case of those plants which are readily changed from the vegetative to the reproductive form of activity by a change in the length of the day, the proper intermediate length of day favorable to both forms of activity must persist over a sufficiently long period if we may expect the everblooming habit to appear. As one advances from the poles toward the equator both the seasonal and the daily changes in length of day decrease till at the equator a fixed day length of 12 hours prevails the year round. In extreme northerly or southerly latitudes. On the other hand, there is a constant and relatively rapid change in length of day. It is clear that under these latter conditions the tendency would be for plants to be swept rather rapidly through the particular range in day length which would permit the vegetative and reproductive activities to proceed simultaneously. Therefore, there would be little opportunity for the everblooming habit to develop in far northerly or southerly regions, even during the open growing season. In these regions everflowering would be confined mostly to those plants which happen not to be particularly sensitive to changes in the length of day. For plants having a daylight requirement for both growth and flowering ranging around 12 hours, conditions at the equator would be ideal for the development of the everflowering habit. As a matter of fact, everflowering is a characteristic feature of plant life in the Tropics, and this form of reproductive activity steadily becomes less prominent as we advance toward the poles. In temperate regions comparatively few plants can be regarded as typical everbloomers.

By suitable control of the daylight period the explanation of everflowering offered above can be directly tested. With a daily light period intermediate between that required to induce free flowering and that which favors vegetative development exclusively a given plant should continue to flower for a more or less indefinite length of time so long as the light period is held constant. For example, one of our common wild violets (Viola papilionacea) after a brief period of winter dormancy renews its activity in early spring by unfolding new leaves. A little later the familiar blue spring, blossoms make their appearance. As the longer days of May and June come on vegetative activity is increased, there is greater development of foliage leaves, and the characteristic blue blossoms disappear. Obviously, these plants are approaching a strictly vegetative form of activity. In reality, however, flowering in the botanical sense does not cease, for in place of the shiny spring blossom a peculiar type of flower is produced beneath the leaves which does not open, though it produces seeds. This appears to be a case of fine adjustment to day length, for evidently the peculiar summer type of flowers represents a stage nearer the purely vegetative condition than does the richly colored spring blossom. Now, when these plants were allowed to receive only about 8 hours of light daily they continued to produce only the blue spring type of blossom and made but little vegetative growth. Surprising as it may seem, by this method the plants were kept in bloom constantly from March till November, with a minimum growth. Flowering finally ceased only because the daylight in December fell below the minimum requirement, so that the plants were forced into dormancy.

But, by keeping the plant under a daylight exposure in excess of 12 hours, it is possible, also, to maintain this violet for an indefinite period in the more nearly vegetative condition of midsummer, in which the inconspicuous, nonopening type of flower is formed. As will be explained later, this may be done by the use of artificial light after sunset to prolong the daily light period. Thus, in the broadest sense, this plant is in blossom from early spring till late fall under the natural range in length of day in our latitude. Considering either of its two alternative forms of blossoming separately, however, the violet behaves as a true everbloomer only when, by artificial means, the appropriate length of the daylight period is held approximately constant. Thus, two distinct types of everblooming are possible in this violet, involving the formation of different sorts of blossoms, and both types of everblooming can be produced at will by artificially regulating the daily light period. This plant furnishes a striking example of the marvelously fine balance between vegetative and reproductive activities which the length of the day controls.

Other plants have shown similar tendencies toward everflowering when exposed to a suitable, fixed illumination period. In fact, under these conditions there is a tendency in plants generally to become everbloomers. Under natural conditions, however, the seasonal change in day length in our latitude is such that only a few of our plants show a pronounced type of everblooming. A number of our common weeds, including the ubiquitous chickweed and the dead nettle (Lamium), are of this class. These plants continue to grow and to flower more or less persistently throughout the winter in the warm greenhouse, and likewise in the field throughout the summer. Such plants stand out conspicuously as essentially different in this behavior from the majority of our plants, which have their definite floral seasons.

Electric Light to Prolong the Daily Light Period.

In summer the daily light period is readily shortened by use of dark chambers, into which the plants are placed for a portion of the day. In this way various plants may be forced into flowering and fruiting out of their natural season, or plants normally flowering and fruiting in summer may be prevented from doing so. On the other hand, to initiate flowering out of season in long-day plants during the short days of winter, or to prevent its occurrence in short-day plants, it would be necessary to lengthen the daily period of illumination With this in view, a greenhouse was fitted with a series of 40-watt electric lights, evenly distributed overhead, so that an average intensity of about 3 to 5 candlepower was obtained immediately above the soil surface. The electric light was used from sunset till about midnight each day. The intensity of the light used seems insignificant in comparison with daylight, which on clear days in winter may reach as high as 5,000 foot candles or more. Yet some striking results were obtained. For comparison, plants were grown in a similar greenhouse without the use of electric light.

As a general proposition, the long-day plants, so called, should tend to remain in the purely vegetative condition in the "control" house without electric light and hasten toward reproductive activity in the electrically lighted house. Short-day plants, on the contrary, should flower readily in the control house and assume a purely vegetative form in the illuminated house. In the control house cosmos has invariably flowered, showing reproductive tendencies when very small. Flowering actually took place within 50 to 60 days from germination. In the illuminated house the plants grew vigorously, greatly exceeding the control plants in stature, and showed no indications of flowering, months after the controls had flowered. These plants were removed from the illuminated greenhouse in June and placed out of doors, where they received only the normal daylight of the long summer days. Under these conditions the plants remained in the actively growing, sterile, vegetative stage and did not flower till they had reached a height of 15 feet in October, when they were finally forced into the reproductive stage by the natural decrease in day length.

Various species of beggar-ticks (Bidens), comprising some of our best known and most persistent weeds in moist, rich bottom lands, have shown a behavior similar to that of cosmos. In response to the short winter days, these have quickly flowered in the control house when only a few inches high, and flowering in turn has been promptly followed by the decline and death of the plants. This is just the way these plants behave when subjected to an artificially shortened daylight duration of 9 to 10 hours in midsummer. In the greenhouse where the daily duration of light had been artificially lengthened by electric illumination the plants behaved just as they have done during the midsummer period of longest days—i.e., grew to great stature, with no indications of flowering. To make these results even more, striking, plants of various ages and statures were from time to time transferred from the illuminated house to the control house, where they at once came under the influence of the relatively very short daylight duration of the winter time. Within a few weeks flowering was initiated simultaneously on all the branches, and decline and death of the plant ultimately followed. This is just what happens in summer time when out-of-door plants are suddenly subjected to artificially shortened daylight periods of 9 to 10 hours' duration.

In the control house, where no electric light was used, the Peking and Biloxi varieties of soy beans, although producing only a dwarfed growth, flowered in the characteristic winter manner, i.e., with the production of poorly developed blossoms. This is also the behavior of these plants when grown under the influence of artificially shortened daylight in summer time. In the illuminated house, on the other hand, vegetative growth was favored and the plants reached an unusually large stature without flowering, thus showing a general similarity to their summer behavior when the days are long.

In the above plants the purely vegetative development is favored by long days, and flowering is initiated when the days have been sufficiently shortened. We will now consider the behavior of iris, which flowers during the long days of May and June. Plants taken from the field in autumn started into growth at once and flowered within 55 to 60 days in the house where electric illumination was used from sunset till midnight to supplement the short daylight period of the winter season. In the control house, the plants remained practically dormant vegetatively until March or April, since they showed practically no growth, and flowers did not appear till June. In spite of the warm temperatures in the control greenhouse, this plant was unable to flower in winter because the days were too short. In the same way the common goldenrod, which regularly begins flowering in June, was readily forced into flowering in winter by the use of the electric light, whereas without electric light no flowering stem was formed, even after an exposure of several months to short-day conditions. Spinach planted in the house provided with electric light on November 1 was in bloom in six weeks, while in the control house the plants remained in the rosette stage throughout the winter.

The above examples are enough to show that artificial light of low intensity used to prolong the daily illumination period during the short days of winter effectively prevents many short-day plants from flowering and is equally effective in forcing long-day plants into flowering and fruiting. In other words, comparatively weak artificial light used as a supplement to daylight of short duration during the winter will produce much the same effects as the daylight of long duration in summer.

In the above-mentioned tests the electric light was supplied by 34 tungsten filament lights of 40 watts each evenly distributed beneath the glass roof of a greenhouse 50 feet long, 20 feet wide, and 12 feet high to the ridge. While the average intensity of 3 to 5 foot candles thus obtained was sufficient for many plants, it was found that others require higher intensities. The number of hours of artificial light needed after sunset, of course, depends on the particular plant concerned, since each variety and species has its own requirements as to duration of the light period. Naturally, the best indication of this requirement is the prevailing length of day at the regular season of flowering for the plant under consideration.

Practical Uses of the Discovery.

The experiments briefly discussed in this paper have opened a wide field for experimentation and study. The full significance of the discovery that the activities of plants are profoundly influenced by seasonal change in the length of day can not be understood until the field has been more fully explored. At present it is possible only to indicate broadly some of the directions in which it seems most likely that practical application of the principles involved can be made.

A correct interpretation of the effects of length of day upon the plants will be a great aid in reaching a better understanding of the causes which limit the natural habitat of most plants, a problem which has been a difficult one to solve. To the farmer. the facts which have been established will strongly emphasize the importance of accurately knowing the correct season for planting each of his crops in order to secure the highest returns. Under some conditions a difference of no more than 10 days in time of planting would definitely direct the plant's activities toward either the purely vegetative or the reproductive form of development, as the case may be. Now, in one case the farmer may be chiefly concerned with extensive vegetative growth, while in another he may be interested primarily in flower, fruit, or seed development. Of course, much has already been learned empirically as to the proper time of planting various crops, but recognition of the importance of the relative lengths of day and night as a factor in a measure reopens the question.

The plant breeder should be able to gain a better insight into some of his problems, such as securing for any particular region earlier or later varieties, more fruitful or larger growing forms, and improved everbloomers and everbearers. In the same way the problem of extending the northern or southern ranges of crop plants may be more clearly defined. In many cases breeding work can be hastened through artificial control of light duration, which will make it possible to work more or less independently of natural conditions of day length, both as to time of year and as to geographical location of the worker. It often happens that plant breeders are unable to make crosses between certain plants because of differences in time of flowering of the two parental types. In such instances artificial control of the daily light period should be of great value, for in this way the date of flowering can be accurately controlled. The plant introducer will have at his command a more adequate basis for analyzing the factors which determine whether any particular plant is adapted to a new region. Moreover, in special cases it may be possible to introduce successfully new plants through artificial control of light conditions or by taking fuller advantage of seasonal differences in length of day.

Solution of the Problem of seed Production in the Maryland Mammoth Tobacco.
FIG. 9.—A, The plant at the left, which was grown in the greenhouse in winter, shows the characteristic behavior of the Mammoth tobacco when propagated under a short day length. The plant at the right was grown under exactly the same conditions except that the daylight period was lengthened by the use of the electric light, and flowering thus prevented. The plant does not flower in the field in Maryland because the days are too long.

—B, A crop of Maryland Mammoth tobacco estimated to yield 2,000 pounds or more per acre. Under the influence of the long summer days exceptionally large yields may be obtained with this variety in southern Maryland, but the plants normally fail to mature seed. The seed may be readily obtained by growing the plants in southern Florida in winter, thus exposing them to short days.

Within suitable limits of temperature and other important factors in plant growth, there would seem to be no reason why almost any plant may not be made to flower and fruit at any season of the year and in any region. By shortening the daily light period through the use of dark chambers or lengthening it by means of artificial light, reproductive activities may be induced almost at will. With proper knowledge of the specific requirements of each plant, therefore, the florist should be able to force flowering at any desired time of the year. It has been possible to secure excellent flowering specimens of iris in midwinter and chrysanthemum, poinsettia, and other plants in summer by utilizing these principles. In the same way wild violets have been kept in the everblooming stage as long as 9 months. The principles involved are so simple that anyone interested in plants can easily obtain instructive and convincing results.

In conclusion, it may be of interest to cite a specific instance in which time day-length effect has been applied to the solution of a practical problem in tobacco culture. Several years ago a new type of tobacco was discovered in southern Maryland. Under suitable conditions this type grows to an unusually large size, the plant in some cases producing more than 100 leaves; hence the name Maryland Mammoth by which this variety is known. Because of its high yielding capacity this variety has been grown with great success in southern Maryland. An excellent crop of Mammoth tobacco is shown in figure 9. A peculiarity of this tobacco is that either it does not flower at all in the field in Maryland or flowering occurs so late in the season that the seed does not mature. Farmers, therefore, can not obtain seed by the usual methods. It was found, however, that Mammoth tobacco flowers very readily in the greenhouse under the natural day length of winter, whereas artificial lengthening of the daily light period of winter prevents flowering, as shown in figure 9. The plant does not flower in the field in Maryland, because the summer days are too long. The problem of securing seed is easily met by growing the plants in southern Florida during the winter, for under these conditions the Mammoth flowers and fruits much the same as the ordinary varieties of tobacco.