Journal of Agricultural Research 18(11): 553-606 (March 1, 1920)
Effect of the relative length of day and night and other factors of the environment on growth and reproduction in plants.
W. W. Garner, Physiologist in Charge, and
H. A. Allard, Physiologist, Tobacco and Plant Nutrition Investigations,
Bureau of Plant Industry, United States Department of Agriculture
1

[579]

RELATIONSHIPS BETWEEN ANNUALS, BIENNIALS, AND PERENNIALS

It is well known that there are no hard and fast lines of distinction separating annuals, biennials, and perennials; for plants may change from one of these types to another under influences of environment, although in the past the particular factors of the environment involved have not for the most part been understood. The experiments recorded in this paper make it clear that in any particular region the relative lengths of the days and nights running through the year constitute one of the controlling factors in determining the behavior of plants in this particular. The soybean is commonly regarded as a typical annual in that its entire life cycle is completed in a single season, and coincident with or soon following the maturation of the seed the plant as a whole perishes. As recorded on page 567, however, a suitable change in the length of the daily exposure to light revived the vegetative life of the matured plants. After the first crop of seed had ripened and the foliage had yellowed just as usual immediately before the plant died, new shoots developed on the old stems, vegetative activity was resumed, and, finally, with the approach of the shorter days of autumn, the plants blossomed and fruited a second time. Thus, under controlled conditions the plant simulated the behavior of a flowering perennial except that the two cycles of alternate vegetative and reproductive activity have been crowded into a single season. Ragweed behaved in essentially the same manner. To make the analogy more convincing, attention is directed to the fact that aster, a flowering perennial, under the same treatment gave exactly the same results as the soybeans and ragweed. Thus, the aster readily completed two complete annual cycles within a period of about four months, except that, in the absence of low temperature, the original growth above ground, of course, was not killed. Moreover, in the second period of vegetative activity new shoots were sent up from the roots in addition to the new axillary shoots appearing on the original stems. The first flowering and fruiting of the soybeans, ragweed, and aster were forced by artificially shortening the length of the day. When the plants were restored to the full exposure of the normal summer day, vegetative activity was resumed, and, finally, the natural shortening of the days in August and September resulted in the second flowering and fruiting periods. The factor of the environment which makes the cycle of alternating vegetative and reproductive activities an annual event would thus seem to be the annual periodicity in the length of day. If temperature differences are assumed to be the primary factor, annual periodicity in tropical regions (not including the immediate vicinity of the equator) is not readily explainable.

As has already been pointed out, the Mammoth or giant type of tobacco behaves as a typical flowering annual, like the ordinary tobaccos, when grown under the influence of days not exceeding 12 hours in length. During the winter months the plant blossoms readily and, in fact, becomes practically an ever-blooming type. It is an interesting fact, however, [580] that as the seed capsules mature the seed-bearing stem dies back only to the first node which may have sent up a new branch. This holds true even though the new branch be but a few inches below the seed head. The portion of the stem below the new branch and the root system henceforth function as parts of a new plant. In winter the new branch blossoms and fruits promptly, perishes, and is succeeded by new branches. As spring advances the new branches coming out assume the giant or nonflowering type of growth which continues till fall brings a return of the short days, when blossoms promptly appear. It would seem that the new branch acts as a rejuvenating or a protective agent against the death of the older organs to which it is attached. Obviously the Mammoth tobacco resembles both the annual and the perennial types of plant life. The sharpness with which the new branch controls the extent of the dying-back of the mother stem is shown in Plate 76, A.

In the latitude of Washington the radish is an annual unless planted very late in the season. It has already been shown that under a shortened light exposure, on the other hand, while vegetative development may continue, flowering does not occur. It would appear from this that the radish might not flower in regions where the maximum length of days is relatively short; and, in fact, according to Dr. Walter Van Fleet, of the Bureau of Plant Industry, the radish as a rule does not blossom when grown in the equatorial region. Similarly, the radish blossoms only occasionally as far north as Porto Rico, where the principal growing season is during the winter months (13). This behavior of the radish, again, is obviously an approach toward the nonflowering type of perennial. Similarly, Dr. Van Fleet states that a lima bean coming under his observation in the Tropics had continued to grow as a perennial for a number of years, having attained giant proportions, while there was only occasional and sparse fruiting. Conversely, the beet ordinarily is a biennial in the latitude of Washington, but when grown in Alaska where the summer days are very long, it is likely to develop seed and thus complete its life cycle in a single season. The intimate relationship existing between the length of day and the attainment of the reproductive stage is strikingly shown by the behavior of the radish under special conditions. In the box of plants used as controls in the experiment described on page 565 and discussed above, the great majority of the individuals developed normal flowering stalks and seed pods in due season (see Pl. 75, B). A few individuals, however, developed considerably later, because of delayed germination or some other reason; and these delayed plants began the formation of flowering stalks. The length of the day having decreased to the critical length, the growth of the seed stalk was arrested after a height of a few inches was attained; and instead of the normal flower head, a crown of foliage leaves developed, as shown in Plate 69, B, thus indicating the resumption of vegetative activity. What is believed to be another example of the directing [581] action of relative day length is the behavior of certain northern varieties of pepper (Capsicum) when planted in Porto Rico in the spring months (13). Under these conditions the peppers imported from the higher latitudes of the United States were able to form only a very few fruits before they began to yellow and shed their foliage, after which the plants soon perished. Also, it is stated that the radish when grown in Porto Rico during the winter months behaves as it does when grown nearer the equator. The above-mentioned experimental results and observations seem to justify the conclusion that the relative length of the day through the year is a factor of the first importance in determining whether many plants behave as annuals, biennials, or perennials, and whether reproduction in such plants is vegetative or sexual or both in any particular region.

The forcing of two flowering periods in a single season under controlled conditions naturally directs attention to another phase of periodicity in plant activity—namely, the appearance of the blossoming period in both spring and fall, or only in one of these seasons in regions outside the Tropics. This question is of special interest with respect to perennials. It is apparent that plants blooming only in the spring or fall or in both seasons are to be regarded as requiring relatively short days for attaining this stage. In annuals, ordinarily a period of vegetative development must necessarily precede flowering, so that the latter stage is likely to be deferred till autumn; but when propagation is by means of bulbs or other reproductive storage organs, blossoming may well occur in the spring. In hardy shrubs arid trees a typical condition is that in which the formation of flowers or flower buds is inaugurated in the autumn under the influence of the shortening days, while the flowering process is interrupted before completion through the intervention of cold weather. The result is that actual blossoming usually takes place in the spring; but if the fall or early winter temperatures are abnormally high, the flowering process may be completed before cold weather intervenes. This phenomenon is occasionally observed in the apple. In the spring, temperature would be the chief factor in determining the date of blossoming for this class of plants. It is suggested that the seasonal distribution of flower-bud formation in the lemon which is considered in a recent interesting article by Reed (22) may be due to these light and temperature relations. The process is most active during the late fall and again in very early spring, with a winter period of low activity. Throughout the summer period of long days, also, activity is at a minimum.

LENGTH OF DAY CONTRASTED WITH LIGHT INTENSITY

As early as 1735 Reaumer (21) undertook to make accurate comparisons of the total quantities of heat required to bring plants to given stages of maturity. At intervals since that time this idea has been [582] revived, and serious efforts have been made to establish some form of quantitative relationship between plant development and the quantity of heat received from the sun. The work of Linsser (15, 16) and of Hoffman (11, 12) in this field is worthy of special mention. In this connection, also, Abbe's critical review of investigations having to do with the relations between climates and crops is of interest (1). It is believed that the results of the present investigation have an important bearing on the subject. Since the quantity of solar radiation received directly by the plant is the product of the intensity and the length of the exposure, it might be expected that any relationship existing between plant processes and the total quantity of radiation received would be disturbed by changes in either the intensity of the light or the duration of the exposure to its action. It has been shown that the relative length of the day is a factor of the greatest importance in relation to reproductive processes in the plant, and it will be of interest to consider whether the intensity of the solar radiation is also of special significance. At the outset it may be observed that it hardly seems likely that light intensity could exert a controlling influence on reproduction in plants, in view of the extent to which the response of plants to differences in light intensity has been studied by investigators without discovery of any very significant relationships so far as concerns reproduction. In the experiments discussed in preceding paragraphs it was found that where daily exposures of 7 hours and 12 hours, respectively, were equally effective in shortening the vegetative period, a total daily illumination aggregating on an average 9 to 10 hours but consisting of two separate exposures, with a 4-hour period of darkness intervening, was vastly less effective in this respect. This shows at once that the total quantity of radiation received can not be responsible for the shortening of the vegetative period produced by shortening the single daily exposure to light. Furthermore, since in the double daily exposures the intervening period of darkness to which the plants were subjected, 10 a.m. to 2 p.m., was at the time of day when the intensity of the solar radiation reaching the earth's surface is at its maximum, the average intensity of the radiation received by these plants is less than that received by those plants which were exposed continuously from 9 a.m. to 4 p.m.

The Stewart Cuban Mammoth tobacco which requires a day length of 12 hours or less to attain the blossoming stage has been grown commercially to some extent under an artificial shade of coarse cheesecloth estimated to reduce the intensity of the sunlight by approximately one-third. It has been observed that this shade has had no noticable effect on the date of blossoming of the tobacco. Again, the aster used in the present investigation grows in the wild state under a variety of situations, some of which are very shaded, but observation during the past [583] season showed that there was no appreciable difference in dates of flowering under these varied exposures.

Further evidence on this subject is furnished by the following experiments in which soybeans were subjected to different degrees of shading, primarily for determining the effect on oil formation in the seed. Different types of shade were employed, and in some instances shading was combined with regulated differences in the water supply of the soil. In all these experiments the aim has been to use a type of shade which would reduce to a minimum secondary effects, such as modifying the air temperature and the temperature and moisture content of the soil. The object, in short, was to measure, as far as practicable, only the direct action of different light intensities on the plant itself, though, of course, this goal can not be fully attained. With this aim in mind the triangular type of shade, shown in Plate 76, B, was used in a series of tests made in 1916. For this shade the standard cheesecloth of best grade, extensively used for surgical dressings, was employed (see Pl. 77, E). The opening extending around the shade near the top, with loose overhanging flap, is for the purpose of facilitating ventilation. The arrangement is such that the frame of the shade can be raised from time to time to accommodate the growth of the plants. The width of the frame was 4 inches at the base and 18 inches at the top, and it was 30 inches high. In these as in the later tests the Peking variety of soybean was used. It will be recalled that this variety is quite sensitive to changes in the length of the day.

The simplest and perhaps the most satisfactory type of shade was that employed in 1917 and 1918. A frame of iron pipe, 30 inches high, 40 inches wide, and of the desired length, was used to support the cloth. The shades in all cases extended almost due east and west. The beans in each instance were planted in a row 6 inches to the north of the center line of the shade to allow for the southerly swing of the sun's course through the sky. Comparatively open, loosely woven cloth, of the type used for the commercial culture of cigar-wrapper tobacco in New England and Florida, was used for this shade. Four different weaves of cloth were used—6 by 6, 8 by 10, 12 by 12, and 12 by 20 mesh, these figures indicating the average number of threads to the linear inch. These cloths are shown in natural size in Plate 77, A-D.

In 1918 tests were extended to include differences in water supply in combination with three different degrees of shading (see Pl. 78, A). This was accomplished by planting the beans in wooden boxes 24. feet long, 12 inches wide, and 14 inches deep, each box being divided by partitions into three 8-foot sections. These boxes were set in the soil so as to extend about 2 inches above the surface and were filled with soil up to 2 inches of the top. Under each degree of shading, three different soil-moisture contents were maintained, designated as wet, [584] medium, and dry. Rainfall was largely excluded by laying boards over the boxes on each side of the plants, the boards having sufficient pitch outward to turn the flow of the water. In addition, control plantings were made in the field, a portion without shade and the remainder covered with the shade cloths; and these received no water except the rainfall. Only those features of the test which relate to shading will be considered here, details of the differences in water supply and their effects pertaining more properly to the next section of the paper. To ascertain whether the simplified form of shade exerted any decided indirect effect through the soil, soil thermographs were installed in the soil at a depth of 3 inches under the 12 by 20 cloth shade, in a position near the plants and in a similar position on the field row receiving no special treatment. No significant differences in the temperature records were obtained.

A matter of special importance, of course, is the degree of shading produced by the different types of shade and different weaves of cloth used. For several reasons only approximations can be had as to the intensity of the light received by the plants under the shades. The positions of different plants and different parts of the same plant with respect to the light necessarily vary, and the shape of the shade involves a constantly changing transmission rate by the shade cloth. The normal daily range in light intensity is magnified by the shade, since the coefficient of transmission of the cloth is greatest at midday and decreases toward sunrise and sunset. [n the 1916 type of shade there is a relatively small coefficient of light transmission furnished by the sloping side walls covered with cheesecloth. In the simplified type of shade only the transmission through the top comes into consideration, since there are no side walls. The southward extension of the top is such, however, that only diffuse light reaches the plants from the side, with the exception of their extreme lower portions, which are exposed to the direct sunlight in the early morning and late afternoon. In the open type of shade, diffuse light naturally becomes a larger factor. Observations made by Prof. H. H. Kimball, of the United States Weather Bureau, by means of the pyrheliometer gave transmission coefficients of 0.441, 0.292, 0.452, 0.613, and 0.727, respectively, for cheesecloth 12 by 20, 12 by 12, and 8 by 10, and for 6 by 6 mesh netting when exposed normally to the sun's rays. Formulas also were developed by Prof. Kimball which make it possible to compute the shading effect at any hour of the day and for any date. Since the sun's rays never strike the shade cloth at normal incidence, the maximum intensity of the transmitted light, which is attained at midday, is slightly less than indicated by the above values. The computed shading effect produced by each type of netting at various hours of the day on June 1, July 1, and August 1 is shown in Table V. It is seen that for horizontal exposures the shading effect is almost constant from 10 a.m. to 2 p.m. but increases considerably from [585] 10 a.m. to 8 a.m. and from 2 p.m. to 4 p.m. and increases very rapidly from 8 a.m. to 6 a.m. and from 4 p.m. to 6 p.m. For vertical exposures the reverse relations, of course, obtain.

TABLE V.—Computed shading effect of netting of various weaves and of cheesecloth at different hours of the day during the summer months, with horizontal exposure of the setting and cheesecloth and also with vertical exposure of the cheesecloth
[Complete shading represented by unity]

To obtain further information as to the shading effect of the nettings used, a section of the simplified type of shade, without side covering, was set up and covered with the 12 by 12 netting. Under this shade (about 6 inches below the netting) Livingstone standardized black and white spherical atmometer cups were installed, and corresponding control cups were placed in full sunlight in the open air. In general, it was found that satisfactory results could not be secured when the wind was blowing; but when there was no appreciable breeze, readings were obtained which seemed to indicate a coefficient of light transmission reasonably close to that determined by Prof. Kimball. Typical readings obtained on clear, calm days are given in Table VI.

TABLE VI.—Readings of black and white spherical atmometer cups under 12 by 12 netting and in direct sunlight, and the indicated coefficient of light transmission, 1919

In the 1916 experiments the soybeans were planted June 21, and the shade was placed in position July 5. Detailed observations were made on the growth and development of the shaded plants and of the unshaded controls. There were 93 individuals under the shade and 67 in the [586] control row. The summarized data in Table VII will bring out the comparative behavior of the shaded and unshaded plants.

TABLE VII.—Effect of shading soybeans with cheesecloth, 1916

The shaded plants show the typical effects of reduced light intensity so often observed-increased elongation of stem, slender growth, enlarged area of leaves, reduced production of dry matter. Besides these effects the yield of seed was considerably reduced. For present purposes the important fact is that although the maximum intensity of the direct light reaching these plants was only about 43 per cent of the normal, the date of blossoming was not affected in the slightest degree. This is a striking contrast with the fact that by reducing the length of the daily light exposure from an average of approximately 14 hours to 12 hours, or about 15 per cent, the length of the period from germination till blossoming was reduced from x to 21 days. It was observed, however, that the seeds of the shaded plants were about a week later than those of the control plants in reaching final maturity.

In the 1917 tests the beans were planted June 27 and the shades placed in position a few days after germination for the first series, while in a second series the shades were set up at the time of blossoming. Two grades of netting were used, the 6 by 6 and the 8 by 10 mesh. The general behavior of the plants is shown in Table VIII. Here, again, it is seen that reducing the intensity of the direct sunlight to maxima of about 70 and 59 per cent, respectively, of the normal has shown no effect on the date of flowering.

TABLE VIII.—Effect of shading soybeans with 6 by 6 and 8 by 10 mesh cotton netting, 1917

[587]

TABLE IX.—Effect of various degrees of shading in combination with differences in water supply on the growth and development of soybeans, 1918
a This planting differed from the control immediately preceding only in that the plants were spaced s to 6 inches apart in the row while in all other cases they were spaced 5 to 6 inches apart.

In the 1918 experiments the plantings were made from June 4 to 6. Two different degrees of shading were used in combination with three different rates of water supply in each of two series, one covering the period from germination to maturation and the other extending only from blossoming fill maturation. In addition, two corresponding series were run, in each of which three different degrees of shading were employed without variation in the water supply, the plants in this case being grown in open field rows without use of boxes, so that the actual rainfall of the season was received by the soil. As controls, a series was arranged without shade but with the three rates of water supply, which extended only from the blossoming period till maturation, the plantings being in buried boxes as in the other experiments having to do with water supply. An additional control consisted of a planting in the field without any special treatment as to either shade or water supply; and, incidentally, a similar planting was made which differed only in that the plants were spaced 5 to 6 inches apart instead of the standard distance of 2 to 3 inches used in all other cases. The shades for the two periods of shading were placed in position, and the special water treatments were begun on June 12 and 13 and August 9, respectively. The results of the tests are summarized in Table IX. It appears that the effect of the shade on the size, weight, and relative proportions of the plant parts is [588] dependent to a considerable extent on the relative water supply. In general, however, reduction in light intensity during the period from germination till maturity gives results similar to those obtained in the preceding tests; and there is a tendency toward a reduced yield of seed, as previously noted. Reducing the intensity of light during the period between blossoming and final maturity, on the other hand, appears to increase somewhat the yield of seed. Without exception, the plants began blossoming on August 7 under all treatments as to shade and differences in water supply, applied either singly or in combination. In these tests it is estimated that under the heaviest shading the maximum intensity of the direct sunlight reaching the plant was only 32 per cent of the normal, and the average for the day could scarcely have exceeded 25 per cent of the normal.

RELATION OF OTHER FACTORS OF THE ENVIRONMENT TO REPRODUCTION

Having seen that under the conditions of the experiments described in the previous section differences in light intensity were without effect on the length of the vegetative period which precedes flowering in soybeans, it is worth while considering whether other factors of the environment, especially water supply and temperature, are of significance. In studying the relation of the water supply to the formation of oil in the seed, a number of tests have been made with soybeans, beginning with 1912; but it will suffice to consider here only the results obtained for the years 1916 and 1918 with the Peking variety. In 1916 plantings were made in a series of four boxes set in the soil and provided with board covers, just as has been described in the preceding section (see p. 583). Each of the boxes was 12 feet long, 12 inches wide, and 12 inches deep. In one of these boxes the soil was maintained in a relatively moist condition from germination to final maturity, and in a second one the soil was kept comparatively dry during this period. In the third box the soil moisture was kept the same as that in the first box till the most active flowering stage was past, after which the moisture content was reduced to that of the second box. In the fourth box the soil was kept relatively dry till the flowering stage was past and thereafter in a relatively moist condition. A control planting receiving the actual rainfall was also made in the field. The beans were planted June 2 I, and the addition of water to the boxes began July 17. The transition in the moisture relations of the third and fourth boxes was begun August 19. The appearance of the plants in the boxes in the late summer is shown in Plate 79,A. The quantities of water supplied to the boxes each week, together with the rainfall for the period of the tests, are given in Table X. Determinations of the moisture content of the soil in the boxes were made at intervals through the month of August. Experience has shown that in the field the soil used in these tests contains 16 to 18 per cent moisture when in best condition [589] for most crop plants. The results of the moisture determinations in the boxes are shown in Table XI.

TABLE X.—Quantities of water added to boxes and the rainfall during the period of the tests dealing with effect of differences in soil moisture on the development of soybeans, 1916

 

TABLE XI.—Moisture content of soil in boxes used for growing soybeans 1916

The comparative growth and development of the plants under the different treatments are indicated by the data presented in Table XII. It appears that the control plants in the field were somewhat larger and considerably more productive than the best plants in the boxes, which were those receiving the larger water supply from germination to maturity. These differences were possibly due to the larger volume of soil available to the plants in the field. There are large reductions in the size and productiveness of the plants in the boxes resulting from a deficiency in the water supply. It appears also that a more favorable water supply during the period preceding the flowering stage resulted in greater vegetative development, while a more favorable water supply after the flowering stage gave a larger yield of seed. In spite of the well-defined [590] differences in the size of the plants and their fruitfulness brought about by differences in the water supply, the date of blossoming was not affected at all, the first blossoms in all boxes and in the field appearing August 7. No important differences were observed in the time of final maturation under the different treatments.

TABLE XII.—Effect of differences in the moisture content of the soil on the growth and development of soybeans, 1916

It may be observed in passing that the wooden boxes placed in the soil as indicated above have been found to be very satisfactory for conducting field tests dealing with the effects of the water supply on plants. By the arrangement of sloping covers on either side of the plants rainfall can be very largely excluded, and losses of soil moisture from causes other than transpiration are reduced to a minimum. Boxes of any convenient size and length may be used, and placing the boxes in the soil insures a close approach to general field conditions.

The general plan as well as a summary of the results of the 1918 tests on the effects of differences in water supply in combination with different degrees of shading have been given in the preceding section on light intensity. It is appropriate to give here further details of the water treatments. The surface area of the soil in each 8-foot section of the boxes was 8 square feet, so that nearly gallons of water would be required to supply the equivalent of a rainfall of 1 inch. The water was applied in measured quantities by means of a garden hose. Although nearly all water lost by the soil was through transpiration, it was found necessary to water heavily each day in periods of hot and dry weather. The quantities of water added weekly and the rainfall during the period of the tests are shown in Table XIII.

Soil moisture determinations were made at intervals during the season from samples taken from the field and from the boxes. The samples were taken to a depth of 12 inches—that is, to the bottom of the soil in the boxes. Composite samples were made up from boxes receiving the same quantities of water. The samples were taken in all cases just before [591] adding water to the soil, so that the average moisture contents would be somewhat higher than these figures. Results of the moisture determinations are shown in Table XIV.

TABLE XIII.—Quantities of water added weekly to soybeans, and the rainfall during the period of the tests dealing with effect of differences in soil moisture, 1918
a None b Inches

TABLE XIV.—Water content of soil in boxes and in the field during the period of the test3 with soybeans, 1918

[592]

The rainfall was relatively light but extraordinarily uniform in distribution during the season; and, though the field soil was comparatively dry most of the time, the plants did not show wilting at any stage. In the boxes the plants in the "dry" sections were kept in a condition in which wilting frequently occurred in the middle of the day through the season. The boxes were 12 inches in width while the distance between field rows was 3.3 feet. It is interesting to note, therefore, that the plants in the boxes, which had slightly less than a third as much lateral soil area from which to draw their moisture as was available to the plants in the field, required an addition of water equal to three times the rainfall in order to attain the same development as was reached by the field plants. When this condition was attained—that is, in the "wet" boxes—the size of the plants was almost exactly the same as that of the plants in the field.

It is seen at once (Table X) that in the boxes the water supply is the chief limiting factor, for the height of the plants, the size of the stalks, the production of seed, etc., are greatly affected by the quantity of the water supplied. On the other hand, reduction of the water supply, even to the point where almost daily wilting of the plants occurred, did not change the time of flowering by a single day. Changing the water supply after the flowering stage likewise produced decided effects on the further development of the plants, and in the same general direction as noted above, although naturally the changes are not so great as when the differences in water supply are maintained throughout the active period of the plant's life. As regards maturation, it was observed that the plants in the wetter soil of the boxes were perhaps a week later in shedding their leaves and ripening their seed pods than those in the field and in the drier soil of the boxes.

While water supply is the chief factor influencing plant development in the boxes, these tests furnish a clear case of the simultaneous action of two limiting factors, for the different degrees of shading likewise affected the development of the plants. Quantitatively, these two limiting factors are of decidedly unequal significance. Within the limits covered by the tests, the effects of the differences in water supply could be demonstrated in nearly all cases even if the light intensity were an uncontrolled variable. On the other hand, the effects of the differences in light intensity would be completely masked in most instances if the water supply were not rigidly controlled. This experiment illustrates the problems of soil productiveness and crop yields which confront the agronomist and clearly points to the futility of attempting to deal with limiting factors of relatively small significance, such as comparatively narrow distinctions in fertilizer requirements of given soils or crops or in the crop-yielding powers of different strains or varieties of plants. [593]

That temperature is a factor of first importance in influencing and controlling plant activities is well understood, and it needs to be considered here only in its relations to the length of day as a factor playing a dominant rôle in the reproductive processes of the plant. It is well known that in accordance with Vant Hoff's law the speed of chemical reactions is doubled for each increase of 5 to 10 in temperature; and similarly plant activities and processes such as respiration and growth are accelerated by increase in temperature, provided the optimum is not exceeded. Conversely, decrease in temperature may moderately retard the plant's activities, or this effect may increase to more or less complete inhibition. Extremes in either direction, of course, may result in killing the flowering buds or fruits, the vegetative shoots, or the entire plant. It is a matter of common knowledge that low temperature retards the development and the unfolding of flowers. An interesting interrelationship of this action of temperature and that of the seasonal decrease in length of day is seen in the behavior of such trees as the apple, previously referred to. Under the influence of the relatively short days fruit trees of this type might be expected to unfold their flowers regularly in the fall instead of in the spring were it not for the interference of low temperatures. The low temperature of winter would seem to have the effect of changing what would otherwise be among the latest flowering plants of the fall into the early flowering ones of spring.

For the temperate and frigid zones the results of the present investigation have made it clear that in some species, at least, distinctions in the time of flowering and fruiting of different varieties, which may be classed as early maturing, late maturing, nonmaturing, and sterile or nonflowering, are due primarily to responses to different day lengths which come into play as the season advances. Here, again, low temperature becomes a factor of increasing importance as the season advances, and, so far as concerns "short-day" plants, it controls the situation with respect to the conditions of nonripening of fruit and of nonflowering. With increasing latitude this relationship between the opposed action of the length of day and the falling temperature becomes more critical for the later maturing varieties. With decreasing latitude a condition is reached in the subtropics which is much more favorable to late maturing or short-day varieties, for the length of day may fall below the critical maximum for flowering without the inhibiting or destructive action of low temperatures coming into play to prevent successful fruiting. At the equator, annual periodicity of both temperature and length of day cease to play an important rôle in plant processes.

LENGTH OF DAY AS A FACTOR IN THE NATURAL DISTRIBUTION OF PLANTS

1 In this connection the tables showing the time of sunrise and sunset at 10-day intervals through the year for various latitudes in North America, as given in SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE. v. 21 (1876), p. 114-119, will be found very convenient for reference.

In an intelligent understanding of the natural distribution of plants over a particular area those factors which are favorable or unfavorable to growth and successful reproduction for each species must be given [594] consideration. Heretofore temperature, water, and light intensity relations have been considered the chief external limiting factors governing the distribution or range of plants. In the light of the observations and experimental results presented in this paper it seems probable that an additional factor, the relative length of the days and nights during the growing period, must also be recognized as among those causes underlying the northward or southward distribution of plants.1 It is evident that the equatorial regions of the earth alone enjoy equal days and nights throughout the entire year. Provided the water relations are favorable, the warm temperatures in these regions favor a continuous growing season for plants. Passing northward from the equatorial regions into higher latitudes, temperatures promoting active vegetative growth and development are restricted to a summer period which, other conditions being equal, becomes progressively shorter as the polar regions are approached. Coincident with these changes from lower to higher latitudes, the summers are characterized by lengthening periods of daylight and the winters by decreasing periods of daylight. We may now consider how these different day and night relations operating during the summer growing period will exercise more or less control upon the northward or southward distribution of certain plants.

It is evident that a plant can not persist in a given region or extend its range in any direction unless it finds conditions not only favorable for vegetative activity but also for some form of successful reproduction. For present purposes only sexual or seed reproduction need be considered. The experiments above described have indicated that for certain plants—for example, ragweed and the aster—the reproductive or flowering phase of development in some way depends upon a stimulus afforded by the shortening of the days and the consequent lengthening of the nights as the summer solstice is passed. It remains to consider more specifically the bearing these facts may have when plants characterized by this type of behavior are subjected to the daylight relations of different latitudes. In the vicinity of Washington, D. C., the ragweeds regularly shed their first pollen about the middle of August. It may be considered that the earliest flowering plants bloom about this date each season because they react to a length of day somewhat less than that of the longest day, which is about x hours in this latitude. In other words, as soon as the decreasing length of day falls somewhat below 15 hours, a condition which obtains about July 1, the period of purely vegetative activity is checked, and the flowering phase of development is initiated. Should the seeds of such plants now be carried as far as northern Maine into a latitude of 46° to 47°, these plants would not experience a length of day falling below 15 hours in length, for which it is assumed they are best [595] suited, until about August 1. In this latitude, then, provided other conditions did not intrude, flowering would be delayed until about August 1, and the chances of successfully maturing seed before killing frosts intervened would be greatly lessened. If the seed were carried still farther north, the plants might not blossom at all, owing to the fact that even the shortest days of the summer growing period would exceed those to which they were best suited in their normal habitat. Although in such instances these failures naturally would have been explained in the past on the basis of unfavorable temperature relations alone, it is obvious that length of day, primarily, is the limiting factor which has retarded the reproductive period so that unfavorable temperature relations have intervened to prevent the ripening of seed.

Although the Arctic summers are very short, plants have become successfully established under such conditions, largely by the development of specialized perennial types, which find the extremely long days favorable both to vegetative growth and to flower production. Although it has been usually considered that the purely Arctic forms are confined to Arctic conditions because of certain temperature requirements, etc., it is possible that length of day, hitherto overlooked as a factor in plant distribution, may have much to do with their restricted range apart from other factors of the environment.

In tropical regions it is probable that the success of many native plants is more or less closely dependent upon the conditions of equal or nearly equal days and nights which prevail there during the entire year. The varieties of bean coming from Peru and Bolivia appear to be of this type. It is evident that such plants, whose flowering conditions depend more or less closely upon a length of day little if at all exceeding 12 hours, can not attain the flowering stage attended by successful seed production in higher latitudes, at least during the summer season, which would necessarily be characterized by days in excess of 12 hours. It is indicated by the beans in question, however, that some plants of this class may grow and attain successful seed production under day lengths less than 12 hours. This being the case, such plants could at least extend their range beyond the Tropics in so far as the temperature conditions of the winter months in these latitudes were favorable to growth and reproduction.

In any study of the phenological aspects of different species of plants the fact stands out that certain plants bloom at definite seasons of the year. This is quite as marked in subtropical regions as in more northern regions having a definite summer growing season. In this connection it is probable that the relative lengths of the days and nights are of particular significance in many instances. The behavior of the composite Mikania scandens, as observed under specially controlled conditions and under winter conditions in the greenhouse, may be more critically considered in relation to its normal blooming season throughout its range. [596] This plant normally blooms from late July to middle or late September, indicating that blossoming becomes more or less inhibited as the autumnal equinox is passed in late September and the length of the day falls below 12 hours. In the greenhouse at Washington the short days of the winter, ranging around 9 to 10 hours in length, have completely inhibited the flowering phase of development of this plant. The shorter 7-hour daily exposures to light under controlled conditions have produced identical results. Thus it appears that the normal flowering period of Mikania scandens even in the warmer portions of its range should not occur much later in the season than the period when the days are not less than 12 hours in length. This seems to be the case in Florida, where the blooming season of Mikania is confined to August and September, as it is in much more northern portions of its range. Plants of this type, attaining their best development under daylight lengths of approximately 12 hours, should also find a more or less congenial environment under truly tropical conditions where the days are never much less than 12 hours in length. It is probable that in the Tropics, however, many plants of this type would not only become perennial in their aerial portions, but would also have a more or less continuous flowering period.

Since it has been shown that the stature of some plants increases in proportion to the length of the day to which the plants are exposed under experimental conditions, this factor should be expected to have some influence upon the stature of such plants in their normal habitat. In general, exceptional stature would be attained in those regions in which a long day period allowed the plants to attain their maximum vegetative expression before the shorter days intervened to initiate the reproductive period. This condition should hold true not only for different latitudes where a plant has an extensive northward and southward range but for different sowings in the saine locality at successively later dates during the season. It is a matter of common observation that the rankest growing individuals among such weeds as the ragweed, pigweed (Amaranthus), lamb's quarters (Chenopodium), cocklebur (Xanthium), beggarticks (Bidens), other conditions being equal, are those which germinated earliest in the season, and consequently were afforded the longest favorable period of vegetative activity preceding the final flowering period. It is also a matter of common observation that all these weeds, when germinating very late in the summer and coming at once under the influence of the stimulus of the shortening days, blossom when very small, often at a height of only a few inches.

Many species of plants have an extensive northward and southward distribution. In these instances it may be that such species are capable of reacting successfully to a wide range of different lengths of day, or it is possible that the apparent adjustment to such a wide range of conditions may depend upon slightly different physiological requirements of different types which have been developed as a result of natural selections. It yet [597] remains to be seen whether those individuals of a given species which grow successfully in high latitudes have the same physiological requirements with respect to length of day as those growing quite as successfully near the equator. In any study of the behavior of plants introduced from other regions, with a view to determining certain economic qualities, it is evident that the factor of length of day must be taken into consideration as a matter likely to have great significance.

LENGTH OF DAY AS A FACTOR IN CROP YIELDS

From the facts which have been developed in this paper it would seem that the seasonal change in length of day is a hitherto unrecognized factor of the environment which must be taken into account when dealing with the problems in crop production. So far as is now known, the length of the day is the most potent factor in determining the relative proportions between the vegetative and the fruiting parts of many crop plants; and, in fact, as already pointed out, fruiting may be completely suppressed by a length of day either too long or too short. In some crop plants the vegetative parts alone are chiefly sought, while in others the fruit or seed only are wanted, and in still others maximum yields of both vegetative and reproductive parts are desired. It is apparent that the merits of different varieties or strains may depend largely on the relative length of day in which they are grown, arid, therefore, the date of planting may easily become the decisive factor. These are matters of vital importance to the plant breeder and the agronomist. Obviously, a delay of even two or three weeks in seeding certain crops because of inclement weather conditions or other considerations may bring about misleading results. It is to be remembered, furthermore, that planting too early may be equally inadvisable, for crops requiring relatively short days for blossoming may thus come under the influence of short days in early spring, resulting in "premature" flowering and a restricted amount of growth. An impressive lesson as to the influence of length of day on the size attained by the plant before blossoming is seen in the relative heights of consecutive plantings of the Biloxi soybean, as shown in figure 3. For maximum yields of many crops it is essential that the date of planting be so regulated as to insure exposure of the plant to the proper length of day, due regard being had for the specific light requirements of each crop as well as for the relative values of the vegetative and fruiting portions of the plant.

RESULTS OBTAINED WITH ARTIFICIAL LIGHT USED TO INCREASE THE LENGTH OF THE
DAILY ILLUMINATION PERIOD DURING THE SHORT DAYS OF WINTER

These results are of particular significance, since increasing the duration of the illumination period of the short winter day by the use of electric light of comparatively low intensity has consistently resulted in [598] initiating or inhibiting the reproductive or the vegetative phases of development, depending upon whether the plants employed normally require long or short days for these forms of expression. In these experiments a greenhouse 50 feet long, 20 feet wide, and 12 feet high to the ridge, with side walls of concrete 5 feet high to the eaves, was provided with 34 tungsten filament incandescent lights, each rated at 32 candlepower, evenly distributed beneath the glass roof. As a control, a similar greenhouse without artificial light was used. The long axis of these houses was on a north and south line. The temperature was approximately the same in the two greenhouses, ranging at night around 60° to 65° F. and 75° to 80° during the day. The unlighted greenhouse, however, tended to run two to three degrees higher than the illuminated house. Beginning on November 1 the electric lights were switched on at 4.30 p.m. and turned off at 12.30 a.m., this procedure being followed throughout the course of the experiments. Supplementing the natural length of the winter days with this 8-hour period of artificial illumination has given about 18 hours of continuous daily illumination, approaching in length the summer days of southern Alaska. Under these conditions the following results have been obtained:

A large dump of Iris florentina L., with all earth intact, was transplanted October 20, 1919, to each of the two greenhouses. The plants exposed to the long daily period of illumination began growing vigorously at once, soon attaining the normal size for this species, and produced blossoms on December 24 and December 30. The controls remained practically dormant and showed no tendency to blossom as late as February 12, 1920.

1 Horticultural variety.

Seed of spinach (Spinacea oleracea L.), Bloomsdale Curley Savoy,1 was sowed November 1, 1919, and came up in both houses on November 6. The plants in the control house, 20 to 25 in number, grew very slowly, producing low, compact, leafy growths or rosettes, and gave no evidence of blossoming as late as February 12. The plants in the lighted house elongated very rapidly, soon developing flower stalks, and all blossomed in the period between the dates December 8 and December 23. These have continued to elongate more or less, blossoming and shedding pollen continuously, thus becoming in effect "everblooming" plants.

Seed of cosmos (Cosmos bipinnata Cav.) was sowed November 1, 1919, and germinated in both houses November 5. In each greenhouse 40 to 45 plants were grown. The plants in the control house quickly flowered, and all blossomed in the period from December 22 to January 2. The plants in the lighted house grew well but remained in the strictly vegetative stage and were showing no indications of blossoming on February 12. On this date the control plants averaged 30 inches in height and the plants in the lighted house 60 inches. [599]

1 Horticultural variety.

Seed of radish (Raphanus sativus L.), Scarlet Globe,1 was sowed November 1, 1919, coming up in both lighted and unlighted houses November 5. On February 12 the control plants, although more stocky and having larger roots, showed no indications of developing flower stems. In the lighted house, however, the plants had developed smaller roots, and flower buds were plainly in evidence, showing that the plants would soon blossom, as is their normal behavior in response to the long summer days out of doors.

Seedlings of the Maryland Mammoth variety of tobacco were transplanted to 12-quart iron pails on November 10, on which date they were placed in the control and the lighted houses. The control plants, six in number, exhibited the typical behavior of winter-grown Maryland Mammoth plants, all blossoming during the period from December 31 to January 8. The plants in the lighted house, six in number, behaved as typical summer-grown mammoths, becoming very compact, stout and leafy, with no indications of blossoming on February 12. On this date these plants had already produced many more leaves than the control plants.

Bulbs of Freesia refracta Klatt were placed in soil in 5-inch pots on July 11, 1919. Pour pots of these plants were kept in the large dark house previously described from p.m. to 9 a.m. daily from July 23 till November 15, when they were transferred to the greenhouses, two pots being placed in the control house and two in the lighted house. None of these plants when taken from the dark house on November 15 showed any indications of blossoming. Both lots began blossoming about December 27. In the control house, however, the plants produced many flower stalks and continued to blossom profusely for a long period. The plants in the lighted house, on the contrary, produced but few flower stalks and few blossoms and soon ceased blooming entirely.

Large, robust dumps of wild violets, of the species Viola papilionacea Pursh, were transplanted to pots and boxes and placed in the control and lighted houses on October 35, 1919. At the time these plants were removed from the field the abnormally warm autumn weather had forced them into bloom, and many purple, petaliferous blossoms were in evidence. As the winter days continued to shorten naturally in the control house, blossoming was suppressed and no new leaves were produced. These control plants appeared to be almost dormant, except for the production of numerous short, thickened stems which were crowded close to the ground among the old leaves. In the lighted house the production of the purple, petaliferous blossoms also ceased, but vegetative growth was initiated and new leaves appeared in great abundance. Coincident with this marked vegetative activity, the plants continuously produced fertile, cleistogamous flowers in great abundance. This furnishes another example of ever-blooming in response to a favorable length of day. In all respects this behavior of the violets in the lighted house [600] simulates the normal behavior of these plants out of doors under the influence of the long summer days.

Through the kindness of Dr. D. N. Shoemaker, several varieties of Lima beans (Phaseolus lunatus L.), F. S. P. 1. No. 46133, from Chincha, Peru, and F. S. P. 1. No. 46339, from Guayaquil, Ecuador, were transplanted from the field to the greenhouse. Several plants of each of these varieties were transplanted into each house on October 18, 1919. Up to that time the plants had not flowered but gave evidence in the field of being extremely late varieties in this latitude. The control plants grew rather slowly but soon became markedly floriferous, setting pods freely. On the other hand, the plants in the artificially illuminated house produced an exceptionally rank growth of vines but did not flower.

Biloxi, Tokyo, Peking, and Mandarin varieties of soybeans were sowed November 1, 1919, and came up November 10 in both houses. In the control house the first blossoms appeared on the Biloxi and Mandarin about December 24, and on the Tokyo and Peking about December 18. In the artificially lighted house the early variety, Mandarin, blossomed on about the same date as in the control house; but under the longer light exposure the plants continued to grow vigorously, and only a very few blossoms appeared, suggesting a tendency toward gigantism. The few blossoms which formed, however, were normal and developed normal pods, while those on the control plants were cleistogamous and sterile. As late as February 12 the other three varieties showed no indications of blossoming. On that date all varieties were much taller in the electrically lighted house than in the control house.

Seed of Beggar-ticks (Bidens frondosa L.) were sowed in both houses on November 19, 1919, and came up in each on December 1. When the plants were very small they were transferred from the flats to 5-inch pots. The transplanting took place on December 19, and on January 12 all the control plants in the unlighted house, 8 or 10 in number, were showing tiny flower heads, although these had attained a height of only 1 to 2 inches. These flower heads came into expression as soon as the plants had developed the second pair of foliage leaves above the cotyledons. The plants in the lighted house continued to produce vegetative growth and gave no evidence of producing flower heads as late as February 12. Two of these plants which had attained a height of & to 9 inches in response to the lengthened period of illumination in the artificially lighted house were transferred to the control house on December ig. On January 12 both plants had produced flower heads in response to the naturally short winter days prevailing in this house and gave promise of blossoming in a short time. The sister plants remaining in the illuminated house continued to produce vegetative growth, with no evidence of blossoming.

Buckwheat (Fagopyrum vulgare Hill) was sowed November 1, 1919, and came up in both houses on November 7. In the control house 28 [601] plants were grown, and 32 plants were grown in the illuminated house. The dates on which the first blossoms appeared on the control plants exposed to the short winter days extended over a range of only about a week-from December 4. to December 10, inclusive. On the other hand, the dates on which first blossoms appeared on the plants exposed to the artificially lengthened day extended over a period of about four weeks-from December 6 to January 2, inclusive. On February 12 the control plants averaged uniformly only 24 inches in height and had practically ceased growing and blooming. The plants in the artificially illuminated house, on the contrary, continued to grow vigorously and to flower freely, having attained an average height of 8 inches, some of the taller being more than 9 feet in height on February 12. These taller plants blossomed much later than the others and produced very few blossoms, thus showing a tendency to become giant forms in response to the artificially produced longer day. The ever-blooming tendency of the plants as a whole, however, was much more marked under the influence of the lengthened illumination period than in the control greenhouse. Again, although the control plants showed very uniform behavior in the range of their earliest blossoming, it is evident that the artificially lengthened period of illumination has in some manner led to a greatly extended range in the time of blossoming. Whether this really represents an unequal response of several more or less distinct, intermingled races to the artificially increased length of day or may be due i part to a more profound physiological variability which has been induced can not be determined until systematic selection and breeding studies have been carried on.

It will be evident that these data dealing with an artificially lengthened illumination period obtained by means of the electric light greatly strengthen the results of the experiments secured during the previous summer by artificially shortening the natural period of illumination through the use of dark houses. The results with the Maryland Mammoth variety of tobacco, the several soybean varieties in question, and the radish are of special significance since they were obtained by methods the direct converse of those used during the summer. Although the intensity of the electric light was undoubtedly far below that of normal sunlight, it was sufficient to initiate or to suppress the reproductive and vegetative activities of these three species as did the long days of the summer time. With respect to the ever-blooming behavior of certain of the plants under study, the results obtained indicate that this behavior is likely to follow when an approximately constant daily illumination period of a duration favorable to both growth and reproduction is maintained for a sufficient length of time. It thus seems possible that the comparatively uniform length of day prevailing in the Tropics accounts for the particular abundance of ever-bloomers in that region. [602]

IS THE RESPONSE TO DIFFERENCES IN THE LENGTH OF DAY A PRINCIPLE OF GENERAL APPLICABILITY IN BIOLOGY?

Experience has abundantly demonstrated the fact that the biologist who attempts to draw sweeping generalizations regarding responses of plants or animals as a whole to conditions of the environment is in serious danger of going astray, even though his observations he based on the behavior of relatively large numbers of species. With this fact clearly in mind, the following suggestions are put forward tentatively but as possibly being of sufficient interest to justify careful consideration on the part of biologists especially concerned in the fields touched upon. It has been dearly brought out in this paper that for a number of plant species the appropriate length of day acts, not merely as an accelerative, but rather as an initiative influence in bringing into expression the plant's potential capacity for sexual reproduction. Perhaps, as an equally satisfactory way of expressing the fact, it may be said that the length of the day exercises a truly determinative influence on plant growth as between the purely vegetative and the (sexually) reproductive foams of development. The response to length of day may be expected to hold for other species, although it would be premature at present to assert that all higher plants will be found to respond to this factor.

One is naturally inclined to inquire whether, also, the length of day is a controlling factor in sexual reproduction among the lower forms of plant life. The observed behavior of some of these lower forms certainly suggests that they come under the influence of the seasonal range in length of day. A single instance will suffice to illustrate the parallelism existing between the vegetative and the reproductive periods of activity, on the one hand, and the periodical change in the length of the day, on the other. Reference is made to the work of Lewis (14), in which it is shown that in certain species of red Algae there is a definite seasonal periodicity in the appearance of sexual and asexual forms. In brief, the July growth of these species consists primarily of tetrasporic or asexual individuals, while through August the growth is characterized by a predominance of sexual plants produced from the tetraspores of the July crop of plants. The carpospores of autumn become sporelings which persist through the winter and give rise to the tetrasporic plants of the early summer period. Should it be true that lower plants respond to differences in length of day as do some of the higher species it may be expected that various relationships between annual and perennial forms, differences in sensibility to relatively long and short days, and other facts which have been shown to apply to these higher species would likewise hold true for lower organisms. It is possible, even, that the seasonal activities of some of the parasitic microorganisms are the result of response to changes in day length.

As to animal life nothing definite can be said, but it may be found eventually that the animal organism is capable of responding to the [603] stimulus of certain day lengths. It has occurred to the writers that possibly the migration of birds furnishes an interesting illustration of this response. Direct response to a stimulus of this character would seem to be more nearly in line with modern teachings of biology than are theories which make it necessary to assume the operation of instinct or volition in some form as explaining the phenomena in question.

CONCLUSION

The results of the experiments which have been presented in this paper seem to make it plain that of the various factors of the environment which affect plant life the length of the day is unique in its action on sexual reproduction. Except under such extreme ranges as would be totally destructive or at least highly injurious to the general well-being of the plant, the result of differences in temperature, water supply, and light intensity, so far as concerns sexual reproduction, appears to be, at most, merely an accelerating or a retarding effect, as the case may be, while the seasonal length of day may induce definite expression, initiating the reproductive processes or inhibiting them, depending on whether this length of day happens to be favorable or unfavorable to the particular species. In broad terms, this action of the length of day may be tentatively formulated in the following principle: Sexual reproduction can be attained by the plant only when it is exposed to a specifically favorable length of day (the requirements in this particular varying widely with the species and variety), and exposure to a length of day unfavorable to reproduction but favorable to growth tends to produce gigantism or indefinite continuation of vegetative development, while exposure to a length of day favorable alike to sexual reproduction and to vegetative development extends the period of sexual reproduction and tends to induce the "ever-bearing" type of fruiting.

The term photoperiod is suggested to designate the favorable length of day for each organism, and photoperiodism is suggested to designate the response of organism to the relative length of day and night.

SUMMARY

(1) The relative length of the day is a factor of the first importance in the growth and development of plants, particularly with respect to sexual reproduction.

(2) In a number of species studied it has been found that normally the plant can attain the flowering and fruiting stages only when the length of day falls within certain limits, and, consequently, these stages of development ordinarily are reached only during certain seasons of the year. In this particular, some species and varieties respond to relatively long days, while others respond to short days, and still others are capable of responding to all lengths of the day which prevail in the latitude of Washington where the tests were made. [604]

(3) In the absence of the favorable length of day for bringing into expression the reproductive processes in certain species, vegetative development may continue more or less indefinitely, thus leading to the phenomenon of gigantism. On the other hand, under the influence of a suitable length of day, precocious flowering and fruiting may be induced. Thus, certain varieties or species may act as early- or late maturing, depending simply on the length of day to which they happen to be exposed.

(4) Several species, when exposed to a length of day distinctly favorable to both growth and sexual reproduction, have shown a tendency to assume the "ever-blooming" or "ever-bearing" type of development—that is, the two processes of growth and reproduction have tended to proceed hand in hand for an indefinite period.

(5) The relationships existing between annuals, biennials, and perennials, as such, are dependent in large measure on responses to the prevailing seasonal range in length of day. In many species the annual cycle of events is governed primarily by the seasonal change in length of day, and the retarding or more or less injurious and destructive effects of winter temperatures are largely incidental rather than fundamental. Hence, by artificial regulation of the length of the daily exposure to light it has been found that in certain species the normal yearly cycle of the plant's activities can be greatly shortened in point of time, or, on the other hand, it may be lengthened almost indefinitely. In certain cases, annuals may complete two cycles of alternate vegetative and reproductive activity in a single season under the influence of a suitable length of the daily exposure to light. Similarly, under certain light exposures some annuals behave like nonflowering perennials.

(6) In all species thus far studied the rate of growth is directly proportional to the length of the daily exposure to light.

(7) Although the length of the daily exposure to light may exert a controlling influence on the attainment of the reproductive stage, experiments reported in this paper indicate that light intensity, within the range from full normal sunlight to a third or a fourth of the normal, and even much less, is not a factor of importance. It follows that the total quantity of solar radiation received by the plant daily during the summer season, within the range above indicated, is of little importance directly so far as concerns the attainment of the flowering stage.

(8) In extensive tests with soybeans, variations in the water supply ranging from optimum to a condition of drought sufficient to induce temporary wilting daily and to cause severe stunting of the plants were entirely without effect on the date of flowering, although in some cases drought seemed to hasten somewhat the final maturation of the seed. Similarly, differences in light intensity, in combination with differences in water supply, failed to change the date of flowering in soybeans.

(9) The seasonal range in the length of the day is an important factor in the natural distribution of plants. [605]

(10) The interrelationships between the length of day and the prevailing temperatures of the winter season largely control successful reproduction in many species and their ability to survive in given regions.

(11) The relation between the length of the day and the time of flowering becomes of great importance in crop yields in many instances and in such cases brings to the forefront the necessity for seeding at the proper time.

LITERATURE CITED

(1) ABBE, Cleveland.
1905. A FIRST REPORT ON THE RELATIONS BETWEEN CLIMATES AND CROPS. U. S. Dept. Agr. Weather But. Bul. 36, 386 p. Catalogue of periodicals and authors referred to ... p. 364-375.

(2) ALLARD, H. A.
1919. GIGANTISM IN NICOTIANA TABACUM AND ITS ALTERNATIVE INHERITANCE. In Amer. Nat., v. , no. 626, P. 218-233. Literature cited, p. 233.

(3) BAILEY, L. H.
1891. SOME PRELIMINARY STUDIES OF THE INFLUENCE OF THE ELECTRIC ARC LIGHT UPON GREENHOUSE PLANTS. N. Y. Cornell Agr. Exp. Sta. Bul. 30, p. 83-122,13 fig., a pl.

(4) ————
1892. SECOND REPORT UPON ELECTROH0RTICULTURE. N. Y. Cornell Agr. Exp. Sta. Bul. 42, p. 133-146, illus.

(5) ————
1893. GREENHOUSE NOTES FOR 1892-93. I. THIRD REPORT UPON ELECTROHORTICULTURE. N. Y. Cornell Agr. Exp. Sta. Bul. 55, p. 147-157, 2 pl.

(6) BONNIER, Gaston.
1895. INFLUENCE DE LA LUMIERE ÈLECTRIQUE CONTINUE SUR LA FORME ET LA STRUCTURE DES PLANTES. In Rev. Gén. Bot., t. 7, no. 8, p. 241-257; no. 79, p. 289-306; no. 80, p. 332-342; no. 82, p. 409-419. pl. 6-15.

(7) CORBETT, L. C.
1899. A STUDY OF THE EFFECT OF INCANDESCENT GAS LIGHT ON PLANT GROWTH. W. Va. Agr. Exp. Sta. Bul. 62, p. 77-110, pl. 1-9.

(8) FLAMMARION, Camille.
1898(?). PHYSICAL AND METEOROLOGICAL RESEARCHES, PRINCIPALLY ON SOLAR RAYS, MADE AT THE STATION OF AGRICULTURAL CLIMATOLOGY AT THE OBSERVATORY OF JUVISY. In Exp. Sta. Rec., V. 10, 110. 2, p. 103-114, 4 fig., 2 Col. Pi.

(9) GARNER, W. W., ALLARD, H. A., and FOUBERT, C. L.
1914. OIL CONTENT OF SEEDS AS AFFECTED BY THE NUTRITION OF THE PLANT. In Jour. Agr. Research, V. 3, no. 3, p. 227-249. Literature cited. p. 249.

(10) GIROU DE BUZAREINGUES, Ch.
1831. SUITE DES EXPERIENCES SUR LA GENERATION DES PLANTES. In Ann. Sci. Nat,, t. 24, p. 138-147.

(11) HOFFMAN, H.
1882. THERMISCHE VEGHTATIONSCONSTANTEN; SONNEN- UND SCHATTENTEM-PERATUREN. In Ztschr. Österr. Gesell. Met., Bd., 17, p. 121-131.

(12) ————
1886. PHAN0LOGISCHE STUDIEN. In Met. Ztschr., Jahrg. 3 (Ztschr. Österr. Cesell. Met., Bd. 21), p. 113-120, pl. 6.

[606]

(13) KINMAN, C. F., and MCCLELLAND, T. B.
1916. EXPERIMENTS ON THE SUPPOSED DETERIORATION OF VARIETIES OF VEGE-TABLES IN PORTO RICO, WITH SUGGESTIONS FOR SEED PRESERVATION. Porto Rico Agr. Exp. Sta. Bul. 20, 30 P., tab., diagr.

(14) LEWIS, I. F.
1914. THE SEASONAL LIFE CYCLE OF SOME RED ALGAE AT WOODS HOLE. In Plant World, v. 17, no. 2, p. 31-35.

(15) LINSSER, Carl.
1867. DIE PERIODISCHEN ERSCHEINUNGEN DES PFLANZENLEBENS IN IHREM VERHÄLTNISS ZU DEN WÄRMEERSCHEINUNGEN. In Mém. Acad. Imp. Sci. St. Petersb., s. , t. n, no- j, «p.

(16) ————
1869. UNTERSUCHUNGEN UBER DIE PERIODISCHEN LEBENSERSCHEINUNGEN DER PFLANZEN. ZWEITE ADHANDLUNG. In Mém. Acad. Imp. Sci. St. Petersb., s. 7, t. 13, no. 8, 87 p.

(17) LODEWIJKS, J. A., Jr. 1911. ERBLICHKEITSVERSUCHE MIT TAHAK. In Ztschr. Induk. Abstam. u. Vererb., Bd. , Heft 2/3, p. 139-172; Heft 4/5, p. 285-323, illus. Literathrverzeichnis, p. 171-172, 322-323.

(18) MACDOUGAL, Daniel Trembly.
1903. THE INFLUENCE OF LIGHT AND DARKNESS UPON GROWTH AND DEVELOPMENT. Mem. N. Y. Bot. Gard,, v. 2, 319 p., illus.

(19) MOOERS, Charles A.
1908. THE SOY BEAN. A COMPARISON WITH THE COWPEA. Tenn. Agr. Exp. Sta. Bul. 82, p. 75-104, illus.

(20) RANE, F, William.
1894. ELECTROHORTICULTURE WITH THE INCANDESCENT LAMP. W. Va. Agr. Exp. Sta. Bul. 37, 27 p., illus.

(21) REAUMUR, R. A. F. DE
1738. OBSERVATIONS DL' THERMOMETRE, FAITES Â PARIS PENDANT L'ANNÉE M.DCCXXXV. COMPARÉES AVEC CELLES QUI ONT ÉTÉ FAITES SOUS LA LIGNE, À L'ISLE DE FRANCE, À ALGER, ET EN QUELQUES-UNES DE NOS ISLES DE L'AMERIQUE. In Hist. Acad. Roy. Sci. [Paris], Mém. Math. & Phys., ann. 1735, p. 545-376.

(22) REED, Howard S.
1919. CERTAIN RELATIONSHIPS BETWEEN THE FLOWERS AND FRUITS OF THE LEMON. In Jour. Agr. Research, V. 17, no. 4, p. 153-165, I fig.

(23) SCHÜBELER.
1879. THE EFFECTS OF UNINTERRUPTED SUNLIGHT ON PLANTS. (Abstract.) In Nature, v. 21, p. 311-312. 1880. Original article (Studier over Klimatets Indflydelse pan Plantelivet) in Naturen, Aarg. j, No. 6, p. 81-89; No. 8, f. 2X3-123, illus. 1879.

(24) TOURNOIS, J.
1911. ANOMALIES FLORALES DU HOUBLON JAPONAIS ET CHANVRE DÉTERMINÉES PAR DES SEMIS HÂTIFS. In Compt. Rend. Acad. Sd. [Paris], t. 153, no. 21, p. 1017-1020.

(25) ————
1912. INFLUENCE DE LA LUMIÈRE SUR LA FLORAISON DU HOUBLON JAPONAIS ET DU CHANVRE. In Compt. Rend. Acad. Sel. [Paris], t. 155, no. 4, 2. 297-300.

(26) WIESNER, J.
1907. DER LICHTGENUSS DER PFLANZEN . . . vii, 322 p., illus. Leipzig.