The American Naturalist 804-815 (Sept. 1895), 904-913 (Oct. 1895)



There is something surprising in the degree of immaturity at which seeds will grow. The usual opinion is, I believe, that seeds not fully ripe will be shrunken and light, and quite worthless for sowing. To some extent there is truth in this, and yet seeds will vegetate when taken from fruit not hail grown, and in which the pulp and even the seeds themselves have the color of fresh, green leaves. Plants from such seeds may flourish, bloom and fruit, and with a certain moderate amount of deviation, show all the usual phases of existence incident to the particular kind of plant life.

This is by no means a recent discovery, but was known to Theophrastus,2 as early as the third century before Christ, who expressed his surprise at the fact, and says that it is wonderful that unripe, imperfect seeds should be able to grow. The fact was established experimentally, however, by several early investigators, notably by Duhamel,3 in 1760, using flowering ash and walnut, by Senebier,4 in 1800, using peas, and by Lefebure,5 in 1801, using radish. In 1822 a successful trial with green seed was made by Seyffer,6 of Stuttgart, which has attracted much attention. The Japanese Sophora, although growing to be a fine tree in Germany, does not often set fruit, and never ripens any, at least in Würtemberg, on account of the cool summers. Despairing of ever securing ripe seed from which to propagate the tree, Seyffer took a branch bearing green fruit, not yet half full size, hung it up until dry, then removed and planted the seed in a cold frame. In this way he obtained 500 young plants, many of which still were to be seen as handsome trees in the grounds of the forestry school at Hohenheim, and in the vicinity, sixteen years afterward, when the paper from which we quoted was read. The economic importance of such a procedure, and its applicability to numerous contingencies, has brought the incident much well merited attention.

  1. Read before the section of botany of the A.A.A.S., Madison meeting, August, 1893.
  2. De causis plantarum, lib. iv., cap. 4.
  3. Duhamel du Monceau, Des semis et plantation des arbres, p. 83.
  4. Senebier, Phys. végétale, iii, p. 877.
  5. Lefebure, Expériences sur le germination des plantes, p. 27.
  6. Seyffer, Isis, 1838, p. 113.

It would be possible to cite many other instances7 of the successful germination of green seed, but it is unnecessary, for all doubt regarding the viability of such seed was set at rest long ago in the very exhaustive treatise upon the subject by Ferdinand Cohn, entitled, "Symbola ad seminis physiologiam," 1847, in which he not only reviewed the previous history, but also himself grew plants of more than a score of widely diverse species from seed in various stages of immaturity.

  1. Waitz, with morning glory (Convolvulus Nil) Bot Zeit, 1835, p.5.
    Kunze, with wheat. Bot. Zeit., 1835, p. 5.
    Kurr, with rye (?), ten-weeks-stock. Bot. Zeit., xviii (1835), p. 4.
    Seyffer, with peas. kidney beans (Phaseolus vulgaris), English beans (Vicia Faba), soja beans, lentils, laburnum, Sophora Japonica. Bot. Zeit., 1836, p. 84; Isis, 1838, p. 5.
    Treviranus, with turnips and peas. Physiologie der Gewachse, ii (1838), p. 576.
    Göppert. with rye. Bot. Zeit., v (1847). p. 886.
    Cohn, with beans (Phaseolus vulgaris), lupines, radish, shepherds purse, corn, sorghum, datum, apple, cucumber, canna, evening primrose, princes' feather (Amarantus caudatus), morning glory, (Ipomoea purpurea), Salvia verbascifolia, pinks, squirting cucumber (Momordica Elaterium), bladder senna Colutea arborescens), marshmallow (Althaea officinalis), castor bean. Symbols ad seminis physiologiam, 1847; Flora, xxxii (1849), p. 481.
    Lucanus, with rye. Landw. Vers.-St., iv (1860), p. 262.
    Siegert, with wheat. Landw. Vers.-St, vi (1863), p. 134.
    Nowacke, with wheat. Untersuchungen uber das Reifen des Getreides, 1869, p. 37.
    Nobbe, with spruce (Picea vulgaris). Tharander forstl. Jahrbuch, xxiv (1874), p. 203; Landw. Vers.-St, xvii (1876), p. 277; Handbuch derSamenkunde, 1876; p. 338.
    Sagot, with wheat (?). Arch. Des. Sci. Phys. et Nat, 1876; Just's Bot. Jahresb, iv, p. 1243.
    Tautphöus, with rye. Ueber die Keimung der Samen, 1876, p. 23
    Wollny, with winter rye. Forsch. Geb. Agrik.-Phys., ii (1886), p. 294.
    Sturtevant, with maize. Rep., N. Y. Exper. Sta, ii (1883), p. 39.
    Goff, with tomatoes, peas, turnips, lettuce. Rep. N. Y. Exper. Sta., ii (1883), p. 205; iii (1884), pp. 199, 211, 224, 232; iv (1885), pp. 130, 182; v (1886), p. 174, 197.
    Atwell, with morning glory (Ipomoea purpurea). Bot. Gaz., xv (1890), p. 46; Bot. Centr., xlvi (1891), p. 162.
    Bailey, with tomato. Bull. Cornell Exper. Station, 1892, p. 207.

At the very beginning of the agitation of the subject, a curious misusage in terminology arose, which at one time led to considerable controversy, but which gradually disappeared with the better elucidation of the subject. The confusion was in regard to the application of the terms viability, or power of germination, and maturity, or ripeness. The implied reasoning of most writers, especially the earlier ones, seems to have been this: The object of maturity is to render the seed capable of becoming an independent plant through germination, therefore a seed must be mature before it can germinate, per contra, the seed that germinates has already reached maturity.

In Gaertner's monumental work on seeds and fruits, published in 1790, is the statement8 that seeds are ripe as soon as they can germinate, although from their color, weight and size, they may not appear so. Senebier,9 in the year 1800, held that seeds must be ripe in order to grow, and yet at the same time says that he has seen green tender peas, taken from equally green pods, germinate. The same confusion of ideas is shown in the defense which Keith made when DeCandolle10 pointed out that it was an error to place maturity of the seed as one of the conditions for germination, as Keith11 had done in his work on vegetable physiology, published in 1816. Keith12 says: "The seed that will germinate is, physiologically speaking, ripe; that is, its fluids have been so elaborated in the process of its maturation, and its solids so vitalized in the assimilation of due aliment as to be now fully and profitably susceptible of the action of the combined stimuli of the soil and atmosphere. Hence I contend, notwithstanding the objection of M. DeCandolle, that the maturity of the seed is rightly and legitimately placed in the list of the conditions of germination." Treviranus13 held essentially the same views, and expressed himself quite as strongly in his work on vegetable physiology, about the same time. Even Cohn, in his clear and scholarly paper, did not quite set the matter straight. He came to the conclusion,14 that although the proper ripening of the seed is dependent upon the parent plant, yet when prematurely separated it will still pass through the ripening stage before germinating; there is thus an after-ripening for green seeds, which fits them for continued growth. Although he seemingly held that seeds cannot germinate until they in some way ripen, yet he asserted (and it is a most important deduction, correctly worded) that viability does not usually coincide with maturity, but precedes it.15

  1. "Semen maturum, ut docet, non ex colore suo saturato, nec ex sua in aqua subsidentia, neque etiam ex duritie sua satis tuto cognoscitur; sed certior maturitatis nota ex ipso trahenda est nucleo; quippe que, si ex gelatinosa sensim factus sit solidiusculus, si testae suie cavitatem repleat exactissime, atque si intra se ipsum nullum prorsus contineat spatium vacuum, indubitatisimum praebit seminis maturi signum quia ita conformatum, germinando aptum an. Quaecunque etiam fuerit reliqua ejus conditio." Gaertner, De fructibus et seminibus plantarum, ii (1790), I, p. cxii.
  2. "Les graines doivent être mures pour germer; pour l'ordinaire elles ne ment pas quand on lei a cuillies avant leur maturite; j'ai pourtant vu germer des pois verts and tendres otes de leurs siliques vertes and molles." Senebier, l. c. iii, p. 377.
  3. Phys. Veg., ii (1832), p 662.
  4. Keith, System of vegetable physiology, ii (1816), p. 3.
  5. Phil. Mag., viii (1836), p. 492.
  6. "Zum keimen gehort, dass der Same reif sei; das heisst, das der Embryo in dem Grade eniwickelt sel, dam er von der Mutterpflanze getrennt, unter Aneig. nung des Vorrathes nährender Materie im Perisperm oder, in den Samenlappen für sich fortleben kann." Treviranus, i. c., ii, p. 574.
  7. Quum maturatio seminis propria non afficiatur a planta, sumendum videtur, ut etiam processura sit, semine soluto a planta; vel, ut postmaturari possint semina. Cohn, 1. e, p. 72.
  8. Facultas germinandi non in idem tempus coincideresolet cum inaturitate; hanc ills pracedit Cohn, l. c., p 73.

Since the time of Cohn the terminology adopted has agreed well with the facts. The present usage is presented in Nobbe's large and excellent treatise upon seeds. He says:16 "The continued life of the embryo is not dependent upon the completion of the storing of reserve material in the seed; the power of germination appears much earlier, even in a stage of development of the seed undoubtedly to be designated as 'unripe.'

  1. Die Lebenfihigkeit des Embryo 1st an die Vollendiingder Reservestoff-Aufspeicherung in Samen nicht gebunden. Die Keimfahigkeit tritt weit fruther, schon in einem unzweifelhaft als 'unir eif" zu bezeichnenden Entwicklungstadium des Samen ein. Nobbe, Samenkunde, p. 339.

Wiesner17 has given a concise definition. "The condition," he says, "in which a seed loosens itself from the plant in order to continue its development independently, is designated as maturity." We are, therefore, to regard maturity as applying to the seed as a whole, and viability as applying to the embryo, the physiological processes associated therewith being quite distinct. After-ripening, which takes place when partly grown seed is separated from the parent plant, only leads to partial maturity.

  1. Der Zustand, in weichem ein Same sich von der Pflanze loslöst, um sich selbständig weiterzuentwickeln, wird ais Reife bezeichnet. Wiesner, Biologie der Pflanzen, 1889, p. 40.

It is an inquiry full of interest as to the minimum development at which a seed will germinate. Goff,18 in 1884, planted tomato seed in March in boxes in the greenhouse, saved the previous season from fruit still thoroughly green, and obtained only 2 per cent of vegetation. But seed from fruit of full size, and which had begun to lose its green color, although not yet showing any tinge of redness, vegetated 84 percent, while from fruit with a faint reddish tinge the percentage of vegetation reached 100. In another experiment he found19 that peas planted in the usual manner in the open ground in April, that had been gathered when in the condition best suited to table use, gave only 3 per cent of vegetation, while those just past this stage of edible maturity gave 9 per cent. But in all probability the conditions of growth at the time were not particularly favorable, as fully ripe seed in the same experiment gave only 54 per cent. of vegetation. In a very carefully conducted experiment with wheat made by Nowacki, selected seed saved from grain when in the milk gave 92 per cent of vegetation, and from grain when turning yellow, as well as when fully ripe, gave 100 per cent., the seed being sown in the open ground (see table III.) Nobbe20 found that seed of Spruce (Picea vulgaris Lk.) gathered on the first and fifteenth of each month from the middle of July to the first of November, and tested in the laboratory in the following January, gave increased percentage of germination according to degree of maturity (see table I). In experiments performed by myself in 1889 tomato seed from green and ripe fruit of the previous season, tested in April in the laboratory, gave 60 per cent germination for the immature seed against 100 per cent for the fully mature. Considerable other data are on record, all going to show that seeds are more certain to germinate the nearer they approach to maturity, or conversely, the more immature the seed, the less number of chances for its germination.

  1. L. c., iii, p. 224.
  2. L. c., iii, p. 232.
  3. L. c.

Experiment conducted by Nobbe.

Spruce seed, gathered July 15,
Spruce seed, gathered Aug. 1,
Spruce seed, gathered Aug. 15,
Spruce seed, gathered Sept. 1,
Spruce seed, gathered Sept. 15,
Spruce seed, gathered Oct. 1,
Spruce seed, gathered Nov. 1
gave 0 per cent germinations.
gave 40.8 per cent germinations.
gave 61.2 per cent germinations.
gave 75.3 per cent germinations.
gave 71.6 per cent germinations.
gave 84 5 per cent germinations.
gave 88.2 per cent germinations.

The internal examination of the seed to determine the actual stage of development, in connection with such studies, has been rarely attempted. Seyffert and Cohn21 agree, however, that with such seeds as peas, beans, lentils, canna and evening primrose, the embryo must be sufficiently formed to be detected with a hand lens, in order that the seed should be capable of growth. If the embryo is watery and unformed, according to these observers, the seed will not germinate.

Probably most of us would at first think, as Cohn21 did, that it is a curious circumstance in this connection, that while in the ripening of the seed innumerable stages are run through, passing one into the other without interruption, in germination, which is as it were a function of maturity, no transition exists. For evidently a seed can only either germinate or not germinate; there can be no third course." But this is very fallacious reasoning, and is founded upon a misunderstanding of the nature of the seed. In the first place germination is not, even constructively, a function of maturity, as it readily occurs both before and after maturity. From our present standpoint, in whatever way the earlier writers may have viewed the matter, a seed is simply a young plant enclosed in a protective covering derived from the parent plant, and accompanied by surplus nutriment. The resting condition of a seed is purely incidental and designed to aid in distribution and in guarding the plant against injury while very young. From the time of the first cell division in the forming embryo until the new individual becomes established as a free growing plant, there need be no check in the continuous growth, except through untoward conditions, or inherent tendency to provide for such conditions. The germination of seeds inside the fruit of oranges, and gourds, and the ready growth of the mangrove, are familiar instances where the resting period has been practically evaded, and development of the plantlet has been nearly or quite continuous

  1. Es zeigt sich hierbei der eigenthümliche Umstand, dam während bei der Reife der Same unzahlige, ohne Unterbrechung in einander ubergehende Stufen durchläuft, bei der Keimfahigkeit. die gleichsam Function der Reife ist; kein Uebergang existirt. Dean offenbar kann ein Same nur entweder keimen, oder nicht; ein drittes giebt es nicht. Cohn, Flora, xxxii (1849), p. 500.

In the growth of green seed we have a case where an attempt is made to give the plantlet the conditions for continued development without passing through the full protective stage. There is nothing in the nature of things, except the want of skill, to prevent the plantlet being removed from the parent plant at any point in its early development, even before its organs can be detected, and by supplying it with the necessary nutriment, heat and moisture, and protecting it against the inroads of destructive organisms (bacteria, molds, etc.), securing to it by these artificial means the conditions for uninterrupted growth, with the entire omission of the usual resting stage.

With this view of the subject it is easy to explain why green seed generally gives fewer germinations as a rule than mature seed; the more exacting conditions for its growth are not well met. And, further, it is evident that Cohn's aphorism that a seed can only germinate or not germinate is saying that a seed can continue to grow or not continue to grow, and is thus robbed of all its mysticism.

To fully understand the problem before us it will be well to inquire into the meaning of maturity. In the course of normal development of the seed the testa becomes more firm and less permeable, the organic constituents of the cells are transformed into solids or semi-solids, there is a loss of water, growth finally ceases, the organic connection with the parent plant is severed, and the seed is ripe. It remains in an inactive, dormant condition a longer or shorter time and then germinates. Maturity is reached in this metamorphosis when the protecting testa, or pericarp, as the case may be, has become sufficiently solid, and the inner parts sufficiently advanced to permit separation from the parent plant without endangering the life of the embryo.

A most curious thing in connection herewith is the fact that the seed, and sometimes the associated parts of the fruit, will continue to develop under circumstances which put a stop to all growth in the vegetative parts of the plant. If a branch is severed from a tree, all growth in its buds and leaves ceases at once, it wilts, and shortly dies. But the fruits and seeds attached to it continue to develop, and will so continue as long as sufficient moisture remains to transport what food material exists, from the leaves and stem into the fruit and seed. This process is known as after-ripening. So far as I know, it has not been intimately investigated, but I am inclined to think that during this process the embryo continues in actual growth, forming new cells, and elaborating its organs, but that little or no growth takes place in the surrounding parts, although great chemical changes and accumulation of substances do occur.

It was observed by Cohn,22 who was the first to note such phenomena, that green seeds entirely removed from the fruit and laid in moist earth or sand passed through the various changes of color of normal ripening. If very young, they did not progress far, but if sufficiently grown, although still perfectly green in color, they underwent the intermediate changes, and finally gave every appearance of full, mature seeds. He experimented with the seeds of apple, pear, beans, lupines, Amarantus caudatus, Polygonum tartaricum, Colutea arborescens, Koelreuteria paniculata, and Canna orientalis. An experiment in after-ripening by Lucanus,23 is very instructive (see table II). He gathered rye in five stages of maturity, ranging from very small kernels, not yet milky, up to fully ripe kernels. Each collection was separated into four lots; in the first the kernels were removed from the heads at once, in the second, they were allowed to remain in the heads, but the heads were removed from the stalks; in the third they remained attached to the plant which was cut near the ground, and in the fourth the plants were pulled, the roots washed, and set in distilled water. A thousand air-dry seeds from each lot were finally weighed. In all cases the grain weighed more when permitted to remain in the head than when removed at once, still more when all the stem and leaves were attached, and very much more when the uprooted plant was supplied with water. After-ripening is thus seen to play a very important part in the handling of immature seed.

  1. Symbola, pp. 67-70; Flora, pp. 508-510.
  2. L. c.

Experiment conducted by Lucanus

  I II  1II IV V
Weight of 1000 air dry
kernels, in grams
Gathered June 28.
Grain very small,
soft and green
Gathered July 3.
Grain becoming
Gathered July 10.
Juices thick
and milky
Gathered July 18.
Grain solid,
straw yellow
Gathered July 26.
Fully ripe
Kernels removed at once 10430 14655 18366 20294 22230
Left in the separated heads 10575 14830 18510 20302 22250
Left on cut plant 11310 14930 18620 20302 22280
Roots in distilled water 13790 15440 20220 21070 22325

There is a state of over-maturity of seeds, which has importance in this connection. It is well known that the life of the seed is limited; some seeds will not grow after a few weeks or months, although most seeds are good for from one to several years. In all cases the seed gradually loses its vitality, and sooner or later ceases to live, unless in the mean time given the means for germination.

In view of these facts we can better appreciate the importance of the discovery made by Cohn24 that there is an optimum for most rapid germination which falls, as a rule, just before obvious maturity, (or possibly at the end of the resting stage, where this is very pronounced, a point not yet investigated), and before and after this optimum the germination is slower.

  1. Ich selbst babe bei Canna, OEnothera, Lupinen und anderen ein mittleres Stadium im Reifungsprocesse beobachtet, in dem die Samen sich am schnellsten entwickelten; von da aufwärts und abwärts die reifen und die weniger ausgebildeten schienen mir langsamer zu keimen. Cohn, Flora, xxxii, p. 504.

We are thus led to consider the seed as accumulating energy up to the approximate time of its maturity, and then gradually losing this energy so long as it remains an inactive seed; and that the measure of this energy is the vigor of its germination. There is a wealth of data to substantiate this theory of the life of a seed, but which would be burdensome to further present at this time.

Turning now to a more detailed consideration of the deviations from normal development in plants from immature seed, the weakness of the seedlings will be one feature to first attract the attention of the investigator. In a number of trials with green seed of tomatoes, made at various times since 1889, I have found25 that the young plants are under size; the stems being shorter and cotyledons smaller. They have less strength, and in consequence many perish in the vain attempt to lift the covering of soil. Some are unable to extricate the cotyledons from the ruptured testa, and often perish from this cause, even after having reached the light. If the seeds are germinated between folds of moist cloth or bibulous paper, such miscarriage will show even more clearly. Similar effects were observed by Cohn, in the use of canna seed. He says:26 "All plants obtained from the youngest seeds were slender and weak, and scarcely progressed beyond the first leaf." Goff,27 who has made experiments with immature tomato and other seeds at intervals from 1884 to the present time, early noted this characteristic of the seedlings.

  1. The data are recorded in the manuscript records of the Indiana Experiment Station, and have not yet been published.
  2. Dagegen waren alle au, den jüngsten Samen gezogenen Pflänchen hinfallig und schwächlich und gediehen kaum über das erste Blatt. Flora, xxxii, p. 501.
  3. L. c., iii, p. 225; iv, p. 182.

The rate of germination is in general slower for immature than for mature seeds. This has been observed by Seyffert, Göppert, Cohn, Toutphöus and others, but this depends upon many internal and external conditions affecting the seed, and it is, therefore, not inconsistent with our theory of the process to find that some observers (Duhamel, Senebier) have noted an increased rate of germination for immature seeds. In an experiment by the writer (manuscript record No. 82) in 1890, tomato plants (24) from the seed of ripe fruit planted in a cold frame, came through the soil in an average of 12 days, plants (5) from seed of half-ripe fruit in 12.2 days, and plants (13) from seed of green fruit in 14.2 days. Other trials with tomato, as well as with peas, wheat, and other kinds, made in the laboratory, using folded cloth, have also given tardy germinations for unripe seeds. Nowacki28 removed seeds from the heads of wheat when in the milk stage, when turning yellow, and when fully ripe, and sowed carefully selected kernels in the garden (see table III). The rate of germination, judging by the time of appearance of the plants above ground, was much slower for the immature seed, the number on the eleventh day after sowing, being respectively 12, 19, 25.

  1. L. c.

Experiment conducted by Nowacki.

    Germinations Stalks
Degree of
On 11th
Total Av. No.
per plant.
Av. height
in cm.
Product of
No. by Ht.
In the milk 50 12 45 4.6 128 589
Turned yellow 50 19 50 5.4 125 675
Fully ripe 50 25 50 5.9 121 714

Owing to their weakened condition the plants from immature seed are less able to withstand unfavorable conditions than those from ripe seed, the difference being more marked the younger the seeds. In my own attempts to grow very green tomato seeds in the green-house, fully eighty-five per cent of the plants that had unfolded the cotyledons, perished before reaching the third leaf. Wollny29 observed a great loss of plants from immature seed of winter rye, taking into account the number of plants growing in the fall and in the following spring, while the plants from ripe seed under the same conditions experienced no loss whatever (see table IV).

  1. L. c.

Experiment conducted by Wollny.

Degrees of
in fall
in spring
per cent.
Very green  100 97 40 41
In the milk 100 96 88 91
Pale yellow 100 100 100 100
Fully ripe 100 100 100 100

(To be continued.)

(Continued from page 815.)

Such deviations as have been mentioned are readily seen, and are more or less to be anticipated. But what shall we say about the final recovery of such plants? Even if plants are feeble while young, will they not eventually become firmly established and outgrow all traces of early weakness? I think we would say a priori, that such would doubtless be the case. It looks reasonable; and yet from both experimental and theoretical data it can be shown that rarely, and only by accident, does the entire restoration of the vigor of the plant under such circumstances take place. I am aware that the majority of observers and writers have held the contrary view, and that Cohn in his admirable treatise came to the conclusion that "in general plants raised from unripe seed are not weaker than those from ripe seed." It is undoubtedly true, that as the plants grow, the differences, which were at first readily detected by the eye, largely or quite disappear. Eventually it is necessary to resort to careful weighing and measuring to bring out the actual facts. This does not mean that the differences are slight and immaterial, but only that the eye cannot detect small variations distributed throughout large objects having irregular surfaces, although in the aggregate they may be considerable.

In the experiment with tomato plants from seed taken from green, half-ripe, and fully ripe fruit, already referred to, (manuscript record No. 82), no essential difference could be detected between the plants after they came into bearing. But weighing exposed the fact that the ripe fruit of the plants from green seed averaged ten per cent lighter than those from ripe seed (see table V).

Experiment conducted by Arthur.

Degree of
of plants
Number of
ripe fruit
Total weight of
fruit in grams
Average weight of
single fruit in grams
Fruit green 13 1044 18304 17.5
Fruit half ripe 5 439 7858 17.9
Fruit fully ripe 24 1889 36622 19.4

The experiment with wheat, conducted by Nowacki, and already referred to (see table III), shows a larger number of stalks from ripe than from green seed; and although not so tall, the total growth of stalks in length is greater for the plants from ripe than from green seed. Without going into further details, the general principle may be stated, that plants from green seed will, as a rule, attain a smaller development in both vegetative and reproductive parts than those from ripe seed.

It is furthermore to be pointed out in this connection, that not only are all parts of the plant smaller and less vigorous, but that the different organs bear a different reciprocal proportion. We may classify plant organs roughly as reproductive (fruit, seed, etc.) and vegetative (leaf, stem and root.) The use of immature seed increases the reproductive parts at the expense of the vegetative, and thus it comes about, that there is more fruit formed in proportion to the amount of foliage than normal. In an experiment, or rather a series of experiments originated by Goff,30 and continued by the originator and the writer, in which the changes due to the use of unripe seed have been made more than ordinarily prominent by the cumulative effect of repetition through several generations, it was found by the writer (see table VI) that a tomato plant, selected as representative of the series grown from unripe seed, bore 3 pounds of fruit to one pound of the vine (leaves, stems and roots taken together), while a plant of the same variety grown each year under the same conditions, but always from ripe seed gave only 1 1/8 pounds of fruit for each pound of the vine. In this case we have an enormous relative increase of fruitage from unripe seed, which in fact was quite apparent to the casual observer upon looking at the plants of the two series as they grew in the garden, although it required the scales to disclose how surprisingly great the difference really was. With this increased fruitfulness is also associated an increase in the number of fruit, although they are individually smaller, as also are the seeds. It is stated that von Mons,31 of Belgium, has applied this method of using green seed to the raising of apples, in order to check too vigorous growth and to increase the fruitfulness.

  1. For history of these experiments, see Bot. Gaz., xii (1887), pp. 41-42; Rep. Wis. Exper. Sta., viii (1891), pp. 152-159.
  2. Williams, E., Rural New-Yorker, 1890, p. 798.

Experiment conducted by Arthur.

Degree of ripeness Weight of vine Weight of fruit Ratio of
vine to fruit
  lb. oz. lb. oz.
Immature series 2 10 9 2 1: 3.475 (3 1/2)
Mature series 5 10 1/2 6 9 1: 1.127 (1 1/8).

In connection with the increase of the number of fruit borne by a plant, there is also a tendency to increased earliness in ripening the fruit. In the cumulative trials with tomatoes by Goff, which have just been referred to, the strain from green seed ripened from ten days to four weeks earlier in different years, than the corresponding series from ripe seed. In another experiment with tomatoes by Goff,32 seed saved from fruit of the same variety, in different stages of maturity, described as very green, pale green, tinged red, light red, deeper red, and fully ripe (see table VII), gave an advantage in earliness of nearly three weeks for the plants from the very green seed compared with those from the fully ripe seed, and of two weeks compared with those from the half ripe seed; and there was also about two-thirds as much gain in the ripening of the first ten fruits upon the same plants respectively. But such marked difference in earliness, or in fact any difference at all, in favor of plants from immature seed does not always occur; and several observers have noted the reverse results.

  1. L. c., iii (1884), p. 224.

Experiment conducted by Goff.

Degree of ripeness Number of seeds Vegetated per cent. First ripe fruit First ten ripe fruit
Very green 50 2 126 days. 137 days.
Pale green 50 84 143 days. 157 days.
Tinged red 50 100 140 days. 151 days.
Light red 50 96 141 days. 147 days.
Deeper red 50 88 141 days. 147 days.
Fully ripe 50 96 146 days. 152 days.

This is not surprising in view of the fact that it is the weaker plants from which the greater earliness in fruiting is expected, and such plants must necessarily be most affected by the conditions of weather, soil and cultivation, and so their uniform development be most interfered with. It was noted by Goodale,33 in 1885, and since by Goff,34 that some early market varieties of vegetables indicate that they may have been originated through the use of green seed.

  1. Physiological Botany, 1885, p. 460.
  2. Bot. Gazette, xii (1887), p. 41.

I have now stated the principal deviations from normal development in plants due to the use of immature seed, which I have myself observed, or for which I find authentic recorded data. They may be grouped and briefly summarized as follows: (1.) There is a loss of vigor, shown in the smaller percentage of germinations, the weakness of the seedlings, and the greater number of plants which die before maturity; (2) the full vigor of the plants is never recovered, although they may and usually do, produce an abundant harvest, and one acceptable to the cultivator, in case of economic plants; (3) the reproductive parts of the plants are increased in proportion to the vegetative parts, resulting in a greater number of fruits and seeds (although individually smaller) and more rapid ripening of them, than in similar plants from mature seed.

In explanation of these changes, and to bring the phenomena into proper relation with other phenomena of plant and animal life, I venture to assert that the deviation in development, which comes from the use of unripe seed, does not differ in kind from that resulting from any other method of weakening the organism, and is to be considered as only a special instance of the effect of checking the uniform normal growth of the individual.

I have in my possession a large amount of data with which to substantiate this proposition, but it would be tiresome to present it here, and I shall content myself with a bare reference to a few facts, and trust to your being able to further convince yourselves of its correctness by recalling facts from your own researches or observations.

Imperfect seed of any kind germinates poorly and produces weak plants. This is true of seed shriveled because of injury to the parent plant from insects, fungi, drouth, etc., of seed infested with fungus, of seed that is too old, or of seed deprived of part of its nutriment or otherwise seriously mutilated. That weak seedlings from any cause, as a rule, are likely to remain weak and produce a poor crop, I think is a statement that will be generally accepted without elaboration. It is in reference to the third general feature of the deviations due to immature seed that the chief interest rests; an interest that has sprung up very largely in consequence of the numerous experiments by Professor Goff, extending over the last ten years, and now very widely known, more especially his long series of experiments with tomatoes, in which notable results have been obtained, suggestive of wide economic application, but to which I have been able to make but brief reference in this paper.35

  1. Goff's work upon unripe tomato seed and resulting strains is recorded as follows:
    Rep. N. Y. Exper. Sta., iii (1884) pp. 224-226; iv (1885), pp. 182-183; v (1886), p. 174. Bot. Gaz. xii (1887), p. 41-42.
    Garden and Forest, iii (1890), p. 427; (see also pages 355 and 392).
    Cited by Hunn, Bull. N. Y. Exper. Sta. No. 30 (1891), p. 478.
    Rep. Wis. Exper. Sta. viii (1891), pp. 152-159.

While the use of immature seed brings about greater activity in reproduction, and a tendency to early maturity, the same is also true of plants from very old seed, as has been recognized for a very long time. It is probably best known in reference to melons,36 which are generally believed to give more and better fruit when the seeds are five to twenty years old,37 although the plants will be weak. Observations have not, however, been confined to melons, but are recorded for pears, beans, lentils, etc.

  1. "Es ist behaupted worden, dass Melonenkerne nach mehrjähriger Aufbewahrung Pflanzen liefern, welehe welt weiniger ♂ Blüthen bringen, al Pflanzen aus frischen Samen; nach 5 Jahren sollten angeblich gar keine ♂ Blüthen gebildet werden. Verf. säete 1878 Melonensamen von 1876 und von 1870. Von den afteren Semen keinite eine geringere Zahl; die daraus hervorgegangenen Pflanzen waren etwaa weniger kräftig." Baillon (Bull. mens. soc. Linn. de Paris, No. 23, 1878) Just’s Bot. Jahresb. vi (1878), p. 828.
  2. Fleischer, l. c., p. 17; Schulz, quoted by Cohn, Symbola, p. 9.

The retardation of the germination due to age is well shown by the tests of tomato seeds made by Lovett,38 in which seeds from 2 to 6 years old showed the first germination in 10 days, 7 years, in 11 days, 8 and 9 years in 12 days 10 and 11 years, in 14 days, and 13 years, in 18 days. It will be observed that the effect of over-maturity is the same as results from immaturity (cf. table III). The similarity of effect is even better shown by a test of red clover seed made by Nobbe39 in 1874, in which mature and immature seed of the crop of that year was compared with that of the crop of 1870, the trial being made in December, 1874. The germination of the immature seed was slower than that of the mature seed which had been kept four years, while the total number of germinations for both immature and over-mature seed was much decreased by four years' keeping (see table VIII).

  1. Rep. N. Y. Exper. Sta., ii (1883), p. 267.
  2. Samenkunde, p. 346.

It is evident, therefore, that aging as well as immaturity of seed leads to weakness of the seedlings, and a general lowered vitality.

Experiment conducted by Nobbe.

Degree of
Per cent of total germination in 2 days Total germination.
Soon after
4 years after
Soon after
4 years after
Immature seed 63 0 48 6
Mature seed 90 24 88 58

Some of the same characteristics which we have seen in the plants from immature seed may also be observed when plants grown on good and on poor soil are compared. It has been noticed by tomato growers that more seed is obtained on poor than on rich soil,40 which accords with the record for immature strains.41 The difference in fertility of soil need not be especially marked to secure the effect, if other conditions are reasonably uniform, even good soil compared with yet richer soil produces the characteristic results. In some experiments on wheat made by Latta,42 the yield on good wheat land was one pound of straw to .55 of a pound of grain, but the same land richly fertilized gave one pound of straw to only .48 of a pound of grain (see table IX); that is, the poorer soil brought about a greater development of the reproductive parts of the plants, as compared with the vegetative parts, than did the richer soil, without regard to the mode of fertilization. This phase of the subject might be extended to great length and many statistics given, but it will suffice for illustration to appeal to common observation of the remarkable size of the flowers and seed pods of depauperate weeds and other plants, and on the other hand, the tendency of plants in rich soil to produce foliage shoots rather than fruit.

  1. Allen, Amer. Gard., xi (1890), p. 858.
  2. Goff, Rep. Wis. Exper. Sta., viii (1891), p. 157.
  3. Bull. Ind. Exper. Sta., No. 41 (1892), p. 94.

Experiment conducted by Latta.

Plat unfertilized
Plat with (bone black, ammonia, potash,)
Plat with (bone black, ammonia, potash,)
Plat unfertilized
Plat with horse manure
Plat with horse manure
Plat unfertilized

produced 1 lb. of straw to .56 lbs. of grain.
produced 1 lb. of straw to .45 lbs. of grain.
produced 1 lb. of straw to .47 lbs. of grain.
produced 1 lb. of straw to .55 lbs. of grain.
produced 1 lb. of straw to .49 lbs. of grain.
produced 1 lb. of straw to .51 lbs. of grain.
produced 1 lb. of straw to .52 lbs. of grain.
Plats unfertilized
Plats fertilized
averaged 1 lb. of straw to .55 lbs. of grain.
averaged 1 lb. of straw to .48 lbs. of grain.

It has been recognized by zoologists43 that "checks to nutrition, especially in the form of sudden scarcity, will favor sexual reproduction." I think I may safely enlarge this statement, and say that any cause which retards uniform progress in the development of an animal or plant favors reproduction. By this is meant that after such a check occurs the organism will develop the reproductive parts of its structure faster and more fully than the other parts, and in the case of crops the yield of seed will be greater proportionately, than of the leaves and stems.44

  1. Geddes and Thompson, Evolution of sex, p. 218.
  2. I have developed this proposition more fully, and shown its application in another direction, in an article entitled: "A new factor in the improvement of crops." Agric. Sci., vii (1893), pp. 340-345.

Enough has doubtless been said to show that the deviations in development, which arise when unripe seeds are used, drop into a general category of changes dependent upon the available energy of the plant and the uniformity of its development. In general, the change is a tendency toward reproduction at the expense of the vegetative parts of the plant.

Purdue University, Lafayette, Ind.


The following are the chief works treating of the subject of the growth of unripe seed. Additional citations have already been made to brief or incidental references to interesting information in this connection.