Our summer is a short one, compared with the long and beautiful summer of California, and it was too late to cut off the faded and the open flowers, and await new ones, which might be purely fertilized after the destruction of all minor plants. So I had to gather the seed from flowers, which might have been partially fertilized by the wrong pollen. This however, is not so great a drawback in selection experiments as might be supposed at first sight. The selection of the following year is sure to eliminate the offspring of such impure parentage.

[497] A far more important principle is that of the hereditary percentage, already discussed in our lecture on the selection of monstrosities. In our present case it had to be applied only to the six selected plants of 1895. To this end the seeds of each of them were sown separately, the ray-florets of the terminal heads of each of the new generation were counted, and curves and averages were made up for the six groups. Five of them gave proof of still being mixtures and were wholly rejected. The children of the sixth parent, however, formed a group of uniform constitution, all fluctuating around the desired average of 21. All in all the terminal heads of over 1,500 plants have been subjected to the somewhat tedious work of counting their ray-florets. And this not in the laboratory, but in the garden, without cutting them off. Otherwise it would obviously have been impossible to recognize the best plants for preservation. I chose only two plants which in addition recommended themselves by the average number of rays on their secondary heads, sowed their seeds next year separately and compared the numerical constitution of their offspring. Both groups averaged 21 and were distributed very symmetrically around this mean. This result [498] showed that no further selection could be of any avail, and that I had succeeded in purifying the 21-rayed grandiflorum variety.

It is from this grandiflorum that I have finally produced my double variety. In the year 1896 I selected from among the above quoted 1,500 plants, 500 with terminal heads bearing 21 or more rays. On these I counted the rays of all the secondary heads about the middle of August (1896) and found that they had, as a rule, retrograded to lower figures. On many thousands of heads only two were found having 22 rays. All others had the average number of 21 or even less. I isolated the individual which bore these two heads, allowed them to be fertilized by insects with the pollen of some of the best plants of the same group, but destroyed the remainder.

This single exceptional plant has been the starting point of my double variety. It was not remarkable for its terminal head, which exhibited the average number of rays of the 21-rayed race. Nor was it distinguished by the average figure for all its heads. It was only selected because it was the one plant which had some secondary heads with one ray more than all the others. This indication was very slight, and could not have been detected save by the counting of the rays of thousands of heads.

[499] But the rarity of the anomaly was exactly the indication wanted, and the same deviation would have had no signification whatever, had it occurred in a group fluctuating symmetrically around the average figure. On the other hand, the observed anomaly was only an indication, and no guarantee of future developments.

Here it should be remarked that the indication alluded to was not the appearance of the expected character of doubling in ever so slight a measure. It was only a guide to be followed in further work. The real character of double flower-heads among composites lies in the production of rays on the disk. No increase of the number of the outer rays can have the same significance. A hasty inspection of double flower heads may convey the idea that all rays are arranged around a little central cluster of disk-florets, the remainder of the original disk-florets but a closer investigation will always reveal the fallacy of this conclusion. Hidden between the inner rays, and covered by them, lie the little tubular and fertile florets everywhere on the disk. They may not be easily seen, but if the supernumerary rays are pulled out, the disk may be seen to bear numerous small florets at intervals. But these intervals are not at all numerous, showing thereby that only a relatively small number of tubes has been [500] converted into rays. This conversion is obviously the true mark of the doubling, and before traces of it are found, no assertion whatever can be given as to the issue of the pedigree experiment.

Three more years were required before this first, but decisive trace was discovered. During these years I subjected my strain to the same sharp selection as has already been described. The chosen ancestor of the race had flowered in 1896, and the next year I sowed its seeds only. From this generation I chose the one plant with the largest number of rays in its terminal head, and repeated this in the following year.

The consequence was that the average number of rays increased rapidly, and with it the absolute maximum of the whole strain. The average came up from 21 to 34. Brighter and brighter crowns of the yellow rays improved my race, until it became difficult and very time consuming to count all the large rays of the borders. The largest numbers determined in the succeeding generations increased by leaps from 21 to 34 in the first year, and thence to 48 and 66 in the two succeeding summers. Every year I was able to save enough seed from the very best plant and to use it only for the continuance of the race. Before the selected plants were allowed to open the flowers from which the seed [501] was to be gathered, nearly the whole remaining culture was exterminated, excepting only some of the best examples, in order to have the required material for cross-pollination by insects. Each new generation was thereby as sharply selected as possible with regard to both parents.

All flower-heads were of course closely inspected. Not the slightest indication of real doubling was discovered, even in the summer of 1899 in the fourth generation of my selected race. But among the best the new character suddenly made its appearance. It was at the commencement of September (1899), too late to admit of the seeds ripening before winter. An inspection of the younger heads was made, which revealed three heads with some few rays in the midst of the disk on one plant, the result of the efforts of four years. Had the germ of the mutation lain hidden through all this time? Had it been present, though dormant in the original sample of seed? Or had an entirely new creation taken place during my continuous endeavors? Perhaps as their more or less immediate result? It is obviously impossible to answer these questions, before further and similar experiments shall have been performed, bringing to light other details that will enable us to reach a more definite conclusion.

[502] The fact that the origination of such forms is accessible to direct investigation is proven quite independently of all further considerations. The new variety came into existence at once. The leap may have been made by the ancestor of the year 1895, or by the plant of 1899, which showed the first central rays, or the sport may have been gradually built up during those four years. In each case there was a leap, contrasting with the view which claims a very long succession of years for the development of every new character.

Having discovered this first trace of doubling, it was to be expected that the new variety would be at once as pure and as rich as other double composites usually are. Some effect of the crossing with the other seed-bearing individuals might still disturb this uniformity in the following year, but another year's work would eliminate even this source of impurity.

These two years have given the expected result. The average number of the rays, which had already arisen from 13 to 34 now at once came up to 47 and 55, the last figure being the sum of 21 and 34 and therefore the probable uttermost limit to be reached before absolute doubling. The maximum numbers came as high as 100 in 1900, and reached even 200 in 1901. Such heads are as completely double as are the [503] brightest heads of the most beautiful double commercial varieties of composites. Even the best white camomiles (Chrysanthemum inodorum) and the gold-flowers or garden-marigolds (Calendula officinalis) do not come nearer to purity since they always have scores of little tubular florets between the rays on their disks.

Real atavists or real reversionists were seen no more after the first purification of the race. I have continued my culture and secured last summer (1903) as many and as completely doubled heads as previously. The race has at once become permanent and constant. It has of course a wide range of fluctuating variability, but the lower limit has been worked up to about 34 rays, a figure never reached by the grandiflorum parent, from which my new variety is thus sharply separated.

Unfortunately the best flowers and even the best individuals of my race are wholly barren. Selection has reached its practical limit. Seeds must be saved from less dense heads, and no way has been found of avoiding it. The ray-florets are sterile, even in the wild species, and when growing in somewhat large numbers on the disk, they conceal the fertile flowers from the visiting insects, and cause them also to be sterile. The same is the case with the best cultivated forms. Their showiest individuals are [504] barren, and incapable of the reproduction of the race.

This last is therefore, of necessity, always continued by means of individuals whose deviation from the mean average is the least. But in many cases the varieties are so highly differentiated that selection has become quite superfluous for practical purposes. I have already discussed the question as to the actual moment, in which the change of the grandiflorum variety into the new plenum form must be assumed to have taken place. In this respect some stress is to be laid on the fact that the improvement through selection has been gradual and continuous, though very rapid from the first moment. But with the appearance of the first stray rays within the disk, this continuity suddenly changed. All the children of this original mutated plant showed the new character, the rays within the disk, without exception. Not on all the heads, nor even on the majority of the heads on some individuals, but on some heads all gave clear proof of the possession of the new attribute. This was present in all the representatives of the new race, and had never been seen in any of their parents and grandparents. Here there was evidently a sudden leap, at least in the external form of the plants. And it seems to me to be the most simple conception, [505] that this visible leap directly corresponded to that inner change, which brought about the complete inheritability of the new peculiarity. It is very interesting to observe how completely my experience agrees with the results of the observations of breeders at large. No doubt a comparison is difficult, and the circumstances are not adequate to a close study.

Isolation and selection have been applied commonly only so far as was consistent with the requirements of practical horticulture, and of course a determination of the hereditary percentage was never made. The disregard of this feature made necessary a greater length of time and a larger number of generations to bring about the desired changes. Notwithstanding this, however, it has been seen that double varieties are produced suddenly. This may have occurred unexpectedly or after a few years' effort toward the end desired. Whether this sudden appearance is the consequence of a single internal differentiating step, or of the rapid succession of lesser changes, cannot yet be made out. The extreme variability of double flowers and the chance of their appearance with only slight indications of the previous petaloid alterations of a few stamens may often result in their origin being overlooked, while subsequent generations may come in for full notice. [506] In the greater number of cases recorded it remains doubtful whether the work said to be done to obtain a new double variety was done before the appearance of these preliminary indications or afterward.

In the first case, it would correspond with our selection of large numbers of florets in the outer rays, in the second however, with the ordinary purification of new races from hybrid mixtures.

In scientific selection-experiments such crosses are of course avoided, and the process of purification is unnecessary, even as in the Chrysanthemum culture. The first generation succeeding the original plant with disk-rays was in this respect wholly uniform and true to the new type.

In practice the work does not start from such slight indications, and is done with no other purpose in view than to produce double flowers in species in which they did not already exist. Therefore it is of the highest importance to know the methods used and the chances of success. Unfortunately the evidence is very scanty on both points.

Lindley and other writers, on horticultural theory and practice assert that a large amount of nourishment tends to produce double flowers, while a culture under normal conditions, [507] even if the plants are very strong and healthy, has no such effect. But even here it remains doubtful whether it applies to the period before or after the internal mutation. On the other hand success is not at all to be relied upon, nor is the work to be regarded as easy. The instances of double flowers said to be obtainable at will, are too rare in comparison with the number of cases, where the first indication of them was found accidentally.

Leaving all these doubtful points, which will have to be cleared up by further scientific investigation, the high degree of variability requires further discussion. It may be considered from three different points of view according to the limit of the deviation from the average, to the dependency on external conditions and to periodicity. It seems best to take up the last two points first.

On a visit to a nursery at Erfurt I once inspected an experiment with a new double variety of the common blue-bottle or blue corn-flower. The plants were dependent on the weather to a high degree. Bad weather increased the number of poorly filled flower-heads, while warm and sunny days were productive of beautiful double flowers. The heads that are borne by strong branches have a greater tendency to become double than those of the weaker ones, [508] and towards the autumn, when all those of the first group are faded away, and only a weak though large section of the heads is still flowering, the whole aspect of the variety gradually retrogrades. The same law of dependency and periodicity is prevalent everywhere. In my own cultures of the improved field-marigold I have observed it frequently. The number of the ray-florets may be considered as a direct response to nourishment, both when this is determined by external circumstances, and when it depends on the particular strength of the branch, which bears the head in question. It is a case exactly similar to that of the supernumerary carpels of the pistilloid poppy, and the deductions arrived at with that variety may be applied directly to double flowers.

This dependency upon nourishment is of high practical importance in combination with the usual effect of the doubling which makes the flowers sterile. It is a general rule that the most perfect flowers do not produce seed. At the height of the flowering period the external circumstances are the most favorable, and the flowering branches still constitute the stronger axes of the plants. Hence we may infer that sterility will prevail precisely in this period. Many varieties are known to yield only seeds from the very last flowers, as for instance some [509] double begonias. Others bear only seed on their weaker lateral branches, as the double camomile, or become fertile only towards the fall, as is often the case with the above quoted Erfurt variety of the blue-bottle. As far as I have been able to ascertain, such seeds are quite adequate for the reproduction and perpetuation of the double varieties, but the question whether there are differences between the seeds of the more or less double flowers of the same plants still remains open. It is very probable, from a theoretical point of view, that such differences exist, but perhaps they are so slight, as to have practically no bearing on the question.

On the ground of their wide range of variability, the double varieties must be regarded as pertaining to the group of ever-sporting forms. On one side they fluctuate in the direction towards such petalomanous flowers as are borne by the stocks and others, which we have previously discussed. Here no trace of the fertile organs is left. But this extreme is never reached by petaloid double flowers. A gap remains which, often overlooked, always exists, and which sharply separates the two types. On the other hand the alteration of the stamens gradually relapses to perfectly single flowers. Here the analogy with the pistillody of the poppies and with the "five-leaved" clover is obvious.

[510] This conception of the inner nature of double flowers explains the fact that the varietal mark is seldom seen to be complete throughout larger groups of individuals, providing these have not been already selected by this character. Tagetes africana is liable to produce some poorly filled specimens, and some double varieties of carnations are offered for sale with the note that the seed yields only 80% of doubles. With Chrysanthemum coronarium and blue-bottles this figure is often announced to be only about 50%. No doubt it is partly due to impurities, caused by vicinism, but it is obviously improbable that the effect of these impurities should be so large.

Some cases of partial reversion may be interpreted in the same way. Among the garden anemones, Anemone coronaria, there is a variety called the "Bride," on account of its pure white dowers. It is for sale with single and with double flowers, and these two forms are known to sport into one another, although they are multiplied in the vegetative way. Such cases are known to be of quite ordinary occurrence. Of course such sports must be considered as partial, and the same stem may bear both types of flowers. It even happens that some particular flower is partly double and partly single. Mr. Krelage, of Haarlem, had the kindness to [511] send me such a curious flower. One half of it was completely double, while the other half was entirely single, bearing normal and fertile stamens in the ordinary number.

The same halfway doubling is recorded to occur among composites sometimes, and from the same source I possess in my collection a head of Pyrethrum roseum, bearing on half of its disk elongated corolla tubes, and on the other half the small disk-florets of the typical species.

It is a current belief, that varieties are improved by continued culture. I have never been able to ascertain the grounds on which this conviction rests. It may be referred either to the purity of the race or to the complete development of the varietal character. In the first case it is a question of hybrid mixtures from which many young varieties must be freed before being placed on the market. But as we have already seen in a former lecture, this requires only three or four years, and afterwards the degree of purity is kept up to the point which proves to be the most suitable for practical purposes. The complete development of the varietal character is a question restricted to ever-sporting varieties, since in white flowers and other constant varieties this degree is variable in a very small and unimportant measure. [512] Hence the double flowers seem to afford a very good example for this discussion.

It can be decided by two facts. First by a consideration of the oldest double varieties, and secondly by that of the very youngest. Are the older ones now in a better condition than at the outset? Have they really been gradually improved during the centuries of their existence? Obviously this can only be answered by a comparison of the figures given by older writers, with the varieties as they are now in culture. Munting's drawings and descriptions are now nearly two centuries and a half old, but I do not find any real difference between his double varieties and their present representatives. So it is in other cases in which improvements by crossing or the introduction of new forms does not vitiate the evidence. Double varieties, as a rule, are exactly the same now, as they were at the time of their first introduction.

If this were otherwise one would expect that young double varieties should in the main display only slight grades of the anomaly, and that they would require centuries to reach their full development. Nothing of the kind is on record. On the contrary the newest double sorts are said to be not only equal to their predecessors, but to excel them. As a rule such claims may be exaggerated, but not to any great extent. [513] This is proven in the simplest way by the result of our own experiment.

In the double field-marigold we have the very first generation of a variety of pure and not hybrid origin. It shows the new attribute in its full development. It has flower-heads nearly as completely filled as the best double varieties of allied cultivated composites. In the second generation it reached heads with 200 rays each, and much larger numbers will seldom be seen in older species on heads of equal size. I have compared my novelty with the choicest double camomiles and others, but failed to discover any real difference. Improvement of the variety developed in the experiments carried on by myself seems to be excluded by the fact that it comes into conflict with the same difficulty that confronts the older cultivated species, viz.: the increasing sterility of the race.

It is perfectly evident that this double marigold is now quite constant. Continuously varying about a fixed average it may live through centuries, but the mean and the limits will always remain the same, as in the case of the ever-sporting varieties.

Throughout this lecture I have spoken of double flowers and double flower-heads of composites as of one single group. They are as nearly related from the hereditary point of [514] view, as they are divergent in other respects. It would be superfluous to dwell any longer upon the difference between heads and flowers. But it is as well to point out, that the term double flowers indicates a motley assemblage of different phenomena. The hen-and-chicken daisy, and the corresponding variety of the garden cineraria (Cineraria cruenta), are extremes on one side. The hen-and-chicken type occurs even in other families and is known to produce most curious anomalies, as with Scabiosa, the supernumerary heads of which may be produced on long stalks and become branched themselves in the same manner.

Petalody of the stamens is well known to be the ordinary type of doubling. But it is often accompanied by a multiplication of the organs, both of the altered stamens and of the petals themselves. This proliferation may consist in median or in lateral cleavages, and in both cases the process may be repeated one or more times. It would be quite superfluous to give more details, which may be gathered from any morphologic treatise on double flowers. But from the physiologic point of view all these cases are to be considered as one large group, complying with previously given definitions of the ever-sporting varieties. They are very variable and wholly permanent. Obviously this [515] permanency agrees perfectly with the conception of their sudden origin.

[516]

LECTURE XVIII

NEW SPECIES OF OENOTHERA

In our experiments on the origin of peloric varieties and double flowers we were guided in the choice of our material by a survey of the evidence already at hand. We chose the types known to be most commonly produced anew, either in the wild state or under the conditions of cultivation. In both instances our novelty was a variety in the ordinary sense of the word. Our pedigree-culture was mainly an experimental demonstration of the validity of conclusions, which had previously been deduced from such observations as can be made after the accidental birth of new forms.

From these facts, and even from these pedigree-experiments, it is scarcely allowable to draw conclusions as to the origin of real species. If we want to know how species originate, it is obviously necessary to have recourse to direct observation. The question is of the highest importance, both for the theory of descent, and for our conception of the real nature of [517] systematic affinities at large. Many authors have tried to solve it on the ground of comparative studies and of speculations upon the biologic relations of plants and animals. But in vain. Contradiction and doubt still reign supreme. All our hopes now rest on the result of experiments.

Unfortunately such experiments seemed simply impossible a few years ago. What is to guide us in the choice of the material? The answer may only be expected from a consideration of elementary species. For it is obvious that they only can be observed to originate, and that the systematic species, because they are only artificial groups of lower unities, can never become the subject of successful experimental inquiry.

In previous lectures we tried to clear up the differences existing between nearly related elementary species. We have seen that they affect all of the attributes of the plants, each of them changing in some measure all of the organs. Nevertheless they were due to distinct unities and of the lowest possible degree. Such unit-steps may therefore be expected to become visible some time or other by artificial means. On the other hand, mutations as a rule make their appearance in groups, and there are many systematic species which on close inspection [518] have been shown to be in reality composite assemblages. Roses and brambles, hawkweeds and willows are the best known examples. Violets and Draba verna, dandelions and helianthemums and many other instances were dealt with in previous lectures. Even wheat and barley and corn afford instances of large groups of elementary species. Formerly mixed in the fields, they became separated during the last century, and now constitute constant races, which, for brevity's sake, are dealt with under the name of varieties.

In such groups of nearly allied forms the single members must evidently be of common origin. It is not necessary for them to have originated all in the same place or at the same time. In some cases, as with Draba verna, the present geographic distribution points to a common birthplace, from whence the various forms may about the same period have radiated in all directions. The violets on the other hand seem to include widely diffused original forms, from which branches have started at different times and in different localities.

The origin of such groups of allied forms must therefore be the object of our research. Perhaps we might find a whole group, perhaps only part of it. In my opinion we have the right to assume that if Draba and violets and [519] others have formerly mutated in this way, other species must at present be in the same changeable condition. And if mutations in groups, or such periodic mutations should be the rule, it is to be premised that these periods recur from time to time, and that many species must even now be in mutating condition, while others are not.

It is readily granted that the constant condition of species is the normal one, and that mutating periods must be the exception. This fact does not tend to increase our prospect of discovering a species in a state of mutability. Many species will have to be tested before finding an instance. On the other hand, a direct trial seems to be the only way to reach the goal. No such special guides as those that led us to the choice of pelories and double flowers are available. The only indication of value is the presumption that a condition of mutability might be combined with a general state of variability at large, and that groups of plants of very uniform features might be supposed to be constant in this respect too. On the contrary, anomalies and deviations if existent in the members of one strain, or found together in one native locality of a species, might be considered as an indication in the desired direction.

Few plants vary in the wild state in such a [520] measure as to give distinct indications. All have to be given a trial in the garden under conditions as similar as possible to their natural environments. Cultivated plants are of course to be excluded. Practically they have already undergone the experience in question and can not be expected to change their habits soon enough. Moreover they are often of hybrid origin. The best way is to experiment with the native plants of one's own country.

I have made such experiments with some hundred species that grow wild in Holland. Some were very variable, as for instance, the jointed charlock (Raphanus Raphanistrum) and the narrow-leaved plantain (Plantago lanceolata). Others seemed more uniform, but many species, collected without showing any malformation, subsequently produced them in my garden, either on the introduced plants themselves or among their offspring. From this initial material I have procured a long series of hereditary races, each with some peculiar anomaly for its special character. But this result was only a secondary gain, a meager consolation for the negative fact that no real mutability could be discovered.

My plants were mostly annuals or biennials, or such perennials as under adequate treatment might produce flowers and seeds during their [521] first summer. It would be of no special use to enumerate them. The negative result does not apply to the species as such, but only to the individual strain, which I collected and cultivated. Many species, which are quite constant with us, may be expected to be mutable in other parts of their range.

Only one of all my tests met my expectations. This species proved to be in a state of mutation, producing new elementary forms continually, and it soon became the chief member of my experimental garden. It was one of the evening primroses.

Several evening-primroses have at different times been introduced into European gardens from America. From thence they have spread into the vicinity, becoming common and exhibiting the behavior of indigenous types. Oenothera biennis was introduced about 1614 from Virginia, or nearly three centuries ago. O. muricata, with small corollas and narrow leaves, was introduced in the year 1789 by John Hunneman, and O. suaveolens, or sweet-scented primrose, a form very similar to the biennis, about the same time, in 1778, by John Fothergill. This form is met with in different parts of France, while the biennis and muricata are very common in the sandy regions of Holland, where I have observed them for [522] more than 40 years. They are very constant and have proven so in my experiments. Besides these three species, the large-flowered evening-primrose, or Oenothera lamarckiana, is found in some localities in Holland and elsewhere. We know little concerning its origin. It is supposed to have come from America in the same way as its congeners, but as yet I have not been able to ascertain on what grounds this supposition rests. As far as I know, it has not been seen growing wild in this country, though it may have been overlooked. The fact that the species of this group are subject to many systematic controversies and are combined by different writers into systematic species in different ways, being often considered as varieties of one or two types, easily accounts for it having been overlooked. However, it would be of great interest to ascertain whether O. lamarckiana yet grows in America, and whether it is in the same state of mutability here as it is in Holland.

The large-flowered evening-primrose was also cultivated about the beginning of the last century in the gardens of the Museum d'Histoire Naturelle, at Paris, where it was noticed by Lamarck, who at once distinguished it as an undescribed species. He wrote a complete description [523] of it and his type specimens are still preserved in the herbarium of the Museum, where I have compared them with the plants of my own culture. Shortly afterwards it was renamed by Seringe, in honor of its eminent discoverer, whose name it now bears. So Lamarck unconsciously discovered and described himself the plant, which after a century, was to become the means of an empirical demonstration of his far-reaching views on the common origin of all living beings.

Oenothera lamarckiana is considered in Europe as a garden-plant, much prized for parks and ornamental planting. It is cultivated by seed-merchants and offered for sale. It has escaped from gardens, and having abundant means for rapid multiplication, has become wild in many places. As far as I know its known localities are small, and it is to be presumed that in each of them the plant has escaped separately from culture. It was in this state that I first met with this beautiful species.

Lamarck's evening-primrose is a stately plant, with a stout stem, attaining often a height of 1.6 meters and more. When not crowded the main stem is surrounded by a large circle of smaller branches, growing upwards from its base so as often to form a dense bush. These branches in their turn have numerous lateral [524] branches. Most of them are crowned with flowers in summer, which regularly succeed each other, leaving behind them long spikes of young fruits. The flowers are large and of a bright yellow color, attracting immediate attention, even from a distance. They open towards evening, as the name indicates, and are pollinated by humble-bees and moths. On bright days their duration is confined to one evening, but during cloudy weather they may still be found open on the following morning. Contrary to their congeners they are dependent on visiting insects for pollination. O. biennis and O. muricata have their stigmas in immediate contact with the anthers within the flower-buds, and as the anthers open in the morning preceding the evening of the display of the petals, fecundation is usually accomplished before the insects are let in. But in O. lamarckiana no such self-fertilization takes place. The stigmas are above the anthers in the bud, and as the style increases in length at the time of the opening of the corolla, they are elevated above the anthers and do not receive the pollen. Ordinarily the flowers remained sterile if not visited by insects or pollinated by myself, although rare instances of self-fertilization were seen.

In falling off, the flowers leave behind them a stout ovary with four cells and a large number [525] of young seeds. The capsule when ripe, opens at its summit with four valves, and contains often from two to three hundred seeds. A hundred capsules on the main stem is an average estimate, and the lateral branches may ripen even still more fruits, by which a very rapid dissemination is ensured.

This striking species was found in a locality near Hilvers, in the vicinity of Amsterdam, where it grew in some thousands of individuals. Ordinarily biennial, it produces rosettes in the first, and stems in the second year. Both the stems and the rosettes were at once seen to be highly variable, and soon distinct varieties could be distinguished among them.

The first discovery of this locality was made in 1886. Afterwards I visited it many times, often weekly or even daily during the first few years, and always at least once a year up to the present time. This stately plant showed the long-sought peculiarity of producing a number of new species every year. Some of them were observed directly on the field, either as stems or as rosettes. The latter could be transplanted into my garden for further observation, and the stems yielded seeds to be sown under like control. Others were too weak to live a sufficiently long time in the field. They were discovered by sowing seed from indifferent plants [526] of the wild locality in the garden. A third and last method of getting still more new species from the original strain, was the repetition of the sowing process, by saving and sowing the seed which ripened on the introduced plants. These various methods have led to the discovery of over a dozen new types, never previously observed or described.

Leaving the physiologic side of the relations of these new forms for the next lecture, it would be profitable to give a short description of the several novelties. To this end they may be combined under five different heads, according to their systematic value. The first head includes those which are evidently to be considered as varieties, in the narrower sense of the word, as previously given. The second and third heads indicate the real progressive elementary species, first those which are as strong as the parent-species, and secondly a group of weaker types, apparently not destined to be successful. Under the fourth head I shall include some inconstant forms, and under the last head those that are organically incomplete.

Of varieties with a negative attribute, or real retrograde varieties, I have found three, all of them in a flowering condition in the field. I have given them the names of laevifolia, brevistylis and nanella.

[527] The laevifolia, or smooth-leaved variety, was one of the very first deviating types found in the original field. This was in the summer of 1887, seventeen years ago. It formed a little group of plants growing at some distance from the main body, in the same field. I found some rosettes and some flowering stems and sowed some seed in the fall. The variety has been quite constant in the field, neither increasing in number of individual plants nor changing its place, though now closely surrounded by other Lamarckianas. In my garden it has proved to be constant from seed, never reverting to the original lamarckiana, provided intercrossing was excluded.

It is chiefly distinguished from Lamarck's evening-primrose by its smooth leaves, as the name indicates. The leaves of the original form show numerous sinuosities in their blades, not at the edge, but anywhere between the veins. The blade shows numbers of convexities on either surface, the whole surface being undulated in this manner; it lacks also the brightness of the ordinary evening-primrose or Oenothera biennis.

These undulations are lacking or at least very rare on the leaves of the new laevifolia. Ordinarily they are wholly wanting, but at times single leaves with slight manifestations of this [528] character may make their appearance. They warn us that the capacity for such sinuosities is not wholly lost, but only lies dormant in the new variety. It is reduced to a latent state, exactly as are the apparently lost characters of so many ordinary horticultural varieties.

Lacking the undulations, the laevifolia leaves are smooth and bright. They are a little narrower and more slender than those of the lamarckiana. The convexities and concavities of leaves are said to be useful in dry seasons, but during wet summers, such as those of the last few years, they must be considered as very harmful, as they retain some of the water which falls on the plants, prolonging the action of the water on the leaves. This is considered by some writers to be of some utility after slight showers, but was observed to be a source of weakness during wet weather in my garden, preventing the leaves from drying. Whether the laevifolia would do better under such circumstances, remains to be tested.

The flowers of the laevifolia are also in a slight degree different from those of lamarckiana. The yellow color is paler and the petals are smoother. Later, in the fall, on the weaker side branches these differences increase. The laevifolia petals become smaller and are often not emarginated at the apex, becoming ovate [529] instead of obcordate. This shape is often the most easily recognized and most striking mark of the variety. In respect to the reproductive organs, the fertility and abundance of good seed, the laevifolia is by no means inferior or superior to the original species.

O. brevistylis, or the short-styled evening primrose, is the most curious of all my new forms. It has very short styles, which bring the stigmas only up to the throat of the calyx tube, instead of upwards of the anthers. The stigmas themselves are of a different shape, more flattened and not cylindrical. The pollen falls from the anthers abundantly on them, and germinates in the ordinary manner.

The ovary which in lamarckiana and in all other new forms is wholly underneath the calyx-tube, is here only partially so. This tube is inserted at some distance under its summit. The insertion divides the ovary into two parts: an upper and a lower one. The upper part is much reduced in breadth and somewhat attenuated, simulating a prolongation of the base of the style. The lower part is also reduced, but in another manner. At the time of flowering it is like the ovary of lamarckiana, neither smaller nor larger. But it is reached by only a very few pollen-tubes, and is therefore always incompletely fertilized. It does [530] not fall off after the fading away of the flower, as unfertilized ovaries usually do; neither does it grow out, nor assume the upright position of normal capsules. It is checked in its development, and at the time of ripening it is nearly of the same length as in the beginning. Many of them contain no good seeds at all; from others I have succeeded in saving only a hundred seeds from thousands of capsules.

These seeds, if purely pollinated, and with the exclusion of the visits of insects, reproduce the variety, entirely and without any reversion to the lamarckiana type.

Correlated with the detailed structures is the form of the flower-buds. They lack the high stigma placed above the anthers, which in the lamarckiana, by the vigorous growth of the style, extends the calyx and renders the flower bud thinner and more slender. Those of the brevistylis are therefore broader and more swollen. It is quite easy to distinguish the individuals by this striking character alone, although it differs from the parent in other particulars.

The leaves of the O. brevistylis are more rounded at the tip, but the difference is only pronounced at times, slightly in the adult rosettes, but more clearly on the growing summits of the stems and branches. By this character, the plants [531] may be discerned among the others, some weeks before the flowers begin to show themselves. But the character by which the plants may be most easily recognized from a distance in the field is the failure of the fruits. They were found there nearly every year in varying, but always small numbers.

Leaving the short-styled primrose, we come now to the last of our group of retrograde varieties. This is the O. nanella, or the dwarf, and is a most attractive little plant. It is very short of stature, reaching often a height of only 20-30 cm., or less than one-fourth of that of the parent. It commences flowering at a height of 10-15 cm., while the parent-form often measures nearly a meter at this stage of its development. Being so very dwarfed the large flowers are all the more striking. They are hardly inferior to those of the lamarckiana, and agree with them in structure. When they fade away the spike is rapidly lengthened, and often becomes much longer than the lower or vegetative part of the stem.

The dwarfs are one of the most common mutations in my garden, and were observed in the native locality and also grown from seeds saved there. Once produced they are absolutely constant. I have tried many thousands of seeds from various dwarf mutants, and never observed [532] any trace of reversion to the lamarckiana type. I have also cultivated them in successive generations with the same result. In a former lecture we have seen that contrary to the general run of horticultural belief, varieties are as constant as the best species, if kept free from hybrid admixtures. This is a general rule, and the exceptions, or cases of atavism are extremely rare. In this respect it is of great interest to observe that this constancy is not an acquired quality, but is to be considered as innate, because it is already fully developed at the very moment when the original mutation takes place.

From its first leaves to the rosette period, and through this to the lengthening of the stem, the dwarfs are easily distinguished from any other of their congeners. The most remarkable feature is the shape of the leaves. They are broader and shorter, and especially at the base they are broadened in such a way as to become apparently sessile. The stalk is very brittle, and any rough treatment may cause the leaves to break off. The young seedlings are recognizable by the shape of the first two or three leaves, and when more of them are produced, the rosettes become dense and strikingly different from others. Later leaves are more nearly like the parent-type, but the petioles remain short. The bases of the blades are frequently [533] almost cordate, the laminae themselves varying from oblong-ovate to ovate in outline. The stems are often quite unbranched, or branched only at the base of the spike. Strong secondary stems are a striking attribute of the lamarckiana parent, but they are lacking, or almost so in the dwarfs. The stem is straight and short, and this, combined with the large crown of bright flowers, makes the dwarfs eminently suitable for bed or border plants. Unfortunately they are very sensitive, especially to wet weather.

Oenothera gigas and O. rubrinervis, or the giant, and the red-veined evening-primroses, are the names given to two robust and stout species, which seem to be equal in vigor to the parent-plant, while diverging from it in striking characters. Both are true elementary species, differentiated from lamarckiana in nearly all their organs and qualities, but not showing any preponderating character of a retrograde nature. Their differences may be compared with those of the elementary species of other genera, as for instance, of Draba, or of violets, as will be seen by their description.

The giant evening-primrose, though not taller in stature than O. lamarckiana, deserves its name because it is so much stouter in all respects. [534] The stems are robust, often with twice the diameter of lamarckiana throughout. The internodes are shorter, and the leaves more numerous, covering the stems with a denser foliage. This shortness of the internodes extends itself to the spike, and for this reason the flowers and fruits grow closer together than on the parent-plant. Hence the crown of bright flowers, opening each evening, is more dense and more strikingly brilliant, so much the more so as the individual flowers are markedly larger than those of the parents. In connection with these characters, the flower-buds are seen to be much stouter than those of lamarckiana. The fruits attain only half the normal size, but are broader and contain fewer, but larger seeds.

The rubrinervis is in many respects a counterpart to the gigasv, but its stature is more slender. The spikes and flowers are those of the lamarckianav, but the bracts are narrower. Red veins and red streaks on the fruits afford a striking differentiating mark, though they are not absolutely lacking in the parent-species. A red hue may be seen on the calyx, and even the yellow color of the petals is somewhat deepened in the same way. Young plants are often marked by the pale red tinge of the mid-veins, but in adult rosettes, or from lack of sunshine, this hue is often very faint.

[535] The leaves are narrow, and a curious feature of this species is the great brittleness of the leaves and stems, especially in annual individuals, especially in those that make their stem and flowers in the first year. High turgidity and weak development of the mechanical and supporting tissues are the anatomical cause of this deficiency, the bast-fibers showing thinner walls than those of the parent-type under the microscope. Young stems of rubrinervis may be broken off by a sharp stroke, and show a smooth rupture across all the tissues, while those of lamarckiana are very tough and strong.

Both the giant and the red-veined species are easily recognized in the rosette-stage. Even the very young seedlings of the latter are clearly differentiated from the lamarckiana, but often a dozen leaves are required, before the difference may be seen. Under such circumstances the young plants must reach an age of about two months before it is possible to discern their characters, or at least before these characters have become reliable enough to enable us to judge of each individual without doubt. But the divergencies rapidly become greater. The leaves of O. gigas are broader, of a deeper green, the blade more sharply set off against the stalk, all the rosettes [536] becoming stout and crowded with leaves. Those of O. rubrinervis on the contrary are thin, of a paler green and with a silvery white surface; the blades are elliptic, often being only 2 cm. or less in width. They are acute at the apex and gradually narrowed into the petiole.

It is quite evident that such pale narrow leaves must produce smaller quantities of organic food than the darker green and broad organs of the gigas. Perhaps this fact is accountable partly, at least, for the more robust growth of the giant in the second year. Perhaps also some relation exists between this difference in chemical activity and the tendency to become annual or biennial. The gigas, as a rule, produces far more, and the rubrinervis far less biennial plants than the lamarckiana. Annual culture for the one is as unreliable as biennial culture for the other. Rubrinervis may be annual in apparently all specimens, in sunny seasons, but gigas will ordinarily remain in the state of rosettes during the entire first summer. It would be very interesting to obtain a fuller insight into the relation of the length of life to other qualities, but as yet the facts can only be detailed as they stand.

Both of these stout species have been found [537] quite constant from the very first moment of their appearance. I have cultivated them from seed in large numbers, and they have never reverted to the lamarckiana. From this they have inherited the mutability or the capacity of producing at their turn new mutants. But they seem to have done so incompletely, changing in the direction of more absolute constancy. This was especially observed in the case of rubrinervis, which is not of such rare occurrence as O. gigas, and which it has been possible to study in large numbers of individuals. So for instance, the "red-veins" have never produced any dwarfs, notwithstanding they are produced very often by the parent-type. And in crossing experiments also the red-veins gave proof of the absence of a mutative capacity for their production.

Leaving the robust novelties, we may now take up a couple of forms, which are equally constants and differentiated from the parent species in exactly the same manner, though by other characters, but which are so obviously weak as to have no manifest chance of self maintenance in the wild state. These are the whitish and the oblong-leaved evening-primroses or the Oenothera albida and oblonga.

Oenothera albida is a very weak species, with whitish, narrow leaves, which are evidently incapable [538] of producing sufficient quantities of organic food. The young seedling-plants are soon seen to lag behind, and if no care is taken of them they are overgrown by their neighbors. It is necessary to take them out, to transplant them into pots with richly manured soil, and to give them all the care that should be given to weak and sickly plants. If this is done fully grown rosettes may be produced, which are strong enough to keep through the winter. In this case the individual leaves become stronger and broader, with oblong blades and long stalks, but retain their characteristic whitish color.

In the second year the stems become relatively stout. Not that they become equal to those of lamarckiana, but they become taller than might have been expected from the weakness of the plants in the previous stages. The flowers and racemes are nearly as large as those of the parent-form, the fruits only a little thinner and containing a smaller quantity of seed. From these seeds I have grown a second and a third generation, and observed that the plants remain true to their type.

O. oblonga may be grown either as an annual, or as a biennial. In the first case it is very slender and weak, bearing only small fruits and few seeds. In the alternative case however, it [539] becomes densely branched, bearing flowers on quite a number of racemes and yielding a full harvest of seeds. But it always remains a small plant, reaching about half the height of that of lamarckiana.

When very young it has broader leaves, but in the adult rosettes the leaves become very narrow, but fleshy and of a bright green color. They are so crowded as to leave no space between them unoccupied. The flowering spikes of the second year bear long leaf-like bracts under the first few flowers, but those arising later are much shorter. Numerous little capsules cover the axis of the spike after the fading away of the petals, constituting a very striking differentiating mark. This species also was found to be quite constant, if grown from pure seed.

We have now given the descriptions of seven new forms, which diverge in different ways from the parent-type. All were absolutely constant from seed. Hundreds or thousands of seedlings may have arisen, but they always come true and never revert to the original O. lamarckiana type. From this they have inherited the condition of mutability, either completely or partly, and according to this they may be able to produce new forms themselves. But this occurs only rarely, and combinations of more than one [540] type in one single plant seem to be limited to the admixture of the dwarf stature with the characters of the other new species.

These seven novelties do not comprise the whole range of the new productions of my O. lamarckiana. But they are the most interesting ones. Others, as the O. semilata and the O. leptocarpa are quite as constant and quite as distinct, but have no special claims for a closer description. Others again were sterile, or too weak to reach the adult stage and to yield seeds, and no reliable description or appreciation can be given on the ground of the appearance of a single individual.

Contrasted with these groups of constant forms are three inconstant types which we now take up. They belong to two different groups, according to the cause of their inconstancy. In one species which I call O. lata, the question of stability or instability must remain wholly unsolved, as only pistillate flowers are produced, and no seed can be fertilized save by the use of the pollen of another form, and therefore by hybridization. The other head comprises two fertile forms, O. scintillans and O. elliptica, which may easily be fertilized with their own pollen, but which gave a progeny only partly similar to the parents.

The Oenothera lata is a very distinct form [541] which was found more than once in the field, and recently (1902) in a luxuriant flowering specimen. It has likewise been raised from seeds collected in different years at the original station. It is also wholly pistillate. Apparently the anthers are robust, but they are dry, wrinkled and nearly devoid of contents. The inner wall of cells around the groups of pollen grow out instead of being resorbed, partly filling the cavity which is left free by the miscarriage of the pollen-grains. This miscarriage does not affect all the grains in the same degree, and under the microscope a few of them with an apparently normal structure may be seen. But the contents are not normally developed, and I have tried in vain to obtain fertilization with a large number of flowers. Only by cross-fertilization does O. lata produce seeds, and then as freely as the other species when self-fertilized. Of course its chance of ever founding a wild type is precluded by this defect.

O. lata is a low plant, with a limp stem, bent tips and branches, all very brittle, but with dense foliage and luxuriant growth. It has bright yellow flowers and thick flower-buds. But for an unknown reason the petals are apt to unfold only partially and to remain wrinkled throughout the flowering time. The stigmas are slightly divergent from the normal type, [542] also being partly united with one another, and laterally with the summit of the style, but without detriment to their function.

Young seedlings of lata may be recognized by the very first leaves. They have a nearly orbicular shape and are very sharply set off against their stalk. The surface is very uneven, with convexities and concavities on both sides. This difference is lessened in the later leaves, but remains visible throughout the whole life of the plant, even during the flowering season. Broad, sinuate leaves with rounded tips are a sure mark of O. lata. On the summits of the stems and branches they are crowded so as to form rosettes.

Concerning inheritance of these characteristics nothing can be directly asserted because of the lack of pollen. The new type can only be perpetuated by crosses, either with the parent form or some other mutant. I have fertilized it, as a rule, with lamarckiana pollen, but have often also used that from nanella and others. In doing so, the lata repeats its character in part of its offspring. This part seems to be independent of the nature of the pollen used, but is very variable according to external circumstances. On the average one-fourth of the offspring become lata, the others assuming the type of the pollen-parent, if this was a lamarckiana or [543] partly this type and partly that of any other of the new species derived from lamarckiana, that might have been used as the pollen-parent. This average seems to be a general rule, recurring in all experiments, and remaining unchanged through a long series of successive generations. The fluctuations around this mean go up to nearly 50% and down nearly to 1%, but, as in other cases, such extreme deviations from the average are met with only exceptionally.

The second category includes the inconstant but perfectly fertile species. I have already given the names of the only two forms, which deserve to be mentioned here.

One of them is called scintillans or the shiny evening-primrose, because its leaves are of a deep green color with smooth surfaces, glistening in the sunshine. On the young rosettes these leaves are somewhat broader, and afterwards somewhat narrower than those of O. lamarckiana at the corresponding ages. The plants themselves always remain small, never reaching the stature of the ancestral type. They are likewise much less branched. They can easily be cultivated in annual generations, but then do not become as fully developed and as fertile, as when flowering in the second year. The flowers have the same structure as those of the lamarckiana, but are of a smaller size.

[544] Fertilizing the flowers artificially with their own pollen, excluding the visiting insects by means of paper bags, and saving and sowing the seed of each individual separately, furnishes all the requisites for the estimation of the degree of stability of this species. In the first few weeks the seed-pans do not show any unequality, and often the young plants must be replanted at wider intervals, before anything can be made out with certainty. But as soon as the rosettes begin to fill it becomes manifest that some of them are more backward than others in size. Soon the smaller ones show their deeper green and broader leaves, and thereby display the attributes of the scintillans. The other grow faster and stronger and exhibit all the characteristics of ordinary lamarckianas.

The numerical proportion of these two groups has been found different on different occasions. Some plants give about one-third scintillans and two-thirds lamarckiana, while the progeny of individuals of another strain show exactly the reverse proportion.

Two points deserve to be noticed. First the progeny of the scintillans appears to be mutable in a large degree, exceeding even the lamarckiana. The same forms that are produced most often by the parent-family are also most ordinarily [545] met with among the offspring of the shiny evening-primrose. They are oblonga, lata and nanella. Oblonga was observed at times to constitute as much as 1% or more of the sowings of scintillans, while lata and nanella were commonly seen only in a few scattering individuals, although seldom lacking in experiments of a sufficient size.

Secondly the instability seems to be a constant quality, although the words themselves are at first sight, contradictory. I mean to convey the conception that the degree of instability remains unchanged during successive generations. This is a very curious fact, and strongly reminds us of the hereditary conditions of striped-flower varieties. But, on the contrary, the atavists, which are here the individuals with the stature and the characteristics of the lamarckiana, have become lamarckianas in their hereditary qualities, too. If their seed is saved and sown, their progeny does not contain any scintillans, or at least no more than might arise by ordinary mutations.

One other inconstant new species is to be noted, but as it was very rare both in the field and in my cultures, and as it was difficult of cultivation, little can as yet be said about it. It is the Oenothera elliptica, with narrow elliptical leaves and also with elliptical petals. It repeats [546] its type only in a very small proportion of its seed.

All in all we thus have a group of a dozen new types, springing from an original form in one restricted locality, and seen to grow there, or arising in the garden from seeds collected from the original locality. Without any doubt the germs of the new types are fully developed within the seed, ready to be evolved at the time of germination. More favorable conditions in the field would no doubt allow all of the described new species to unfold their attributes there, and to come into competition with each other and with the common parents. But obviously this is only of secondary importance, and has no influence on the fact that a number of new types, analogous to the older swarms of Draba, Viola and of many other polymorphous species, have been seen to arise directly in the wild state.

[547]

LECTURE XIX

EXPERIMENTAL PEDIGREE-CULTURES

The observation of the production of mutants in the field at Hilversum, and the subsequent cultivation of the new types in the garden at Amsterdam, gives ample proof of the mutability of plants. Furthermore it furnishes an analogy with the hypothetical origin of the swarms of species of Draba and Viola. Last but not least important it affords material for a complete systematic and morphologic study of the newly arisen group of forms.

The physiologic laws, however, which govern this process are only very imperfectly revealed by such a study. The instances are too few. Moreover the seeds from which the mutants spring, escape observation. It is simply impossible to tell from which individual plants they have been derived. The laevifolia and the brevistylis have been found almost every year, the first always recurring on the same spot, the second on various parts of the original field. It is therefore allowable to assume a common [548] origin for all the observed individuals of either strain. But whether, besides this, similar strains are produced anew by the old lamarckiana group, it is impossible to decide on the sole ground of these field-observations.

The same holds good with the other novelties. Even if one of them should germinate repeatedly, without ever opening its flowers, the possibility could not be excluded that the seeds might have come originally from the same capsule but lain dormant in the earth during periods of unequal length.

Other objections might be cited that can only be met by direct and fully controlled experiments. Next to the native locality comes the experimental garden. Here the rule prevails that every plant must be fertilized with pollen of its own, or with pollen of other individuals of known and recorded origin. The visits of insects must be guarded against, and no seeds should be saved from flowers which have been allowed to open without this precaution. Then the seeds of each individual must be saved and sown separately, so as to admit of an appreciation, and if necessary, a numerical determination of the nature of its progeny. And last but not least the experiments should be conducted in a similar manner during a series of successive years.

[549] I have made four such experiments, each comprising the handling of many thousands of individual plants, and lasting through five to nine generations. At the beginning the plants were biennial, as in the native locality, but later I learned to cultivate them in annual generations. They have been started from different plants and seeds, introduced from the original field into my garden at Amsterdam.

It seems sufficient to describe here one of these pedigree-cultures, as the results of all four were similar. In the fall of 1886 I took nine large rosettes from the field, planted them together on an isolated spot in the garden, and harvested their seeds the next year. These nine original plants are therefore to be considered as constituting the first generation of my race. The second generation was sown in 1888 and flowered in 1889. It at once yielded the expected result. 15,000 seedlings were tested and examined, and among them 10 showed diverging characters. They were properly protected, and proved to belong to two new types. 5 of them were lata and 5 nanella. They flowered next year and displayed all the characters as described in our preceding lecture. Intermediates between them and the general type were not found, and no indication of their appearance was noted in their parents. [550] They came into existence at once, fully equipped, without preparation or intermediate steps. No series of generations, no selection, no struggle for existence was needed. It was a sudden leap into another type, a sport in the best acceptation of the word. It fulfilled my hopes, and at once gave proof of the possibility of the direct observation of the origin of species, and of the experimental control thereof.

The third generation was in the main a repetition of the second. I tried some 10,000 seedlings and found three lata and three nanella, or nearly the same proportion as in the first instance. But besides these a rubrinervis made its appearance and flowered the following year. This fact at once revealed the possibility that the instability of lamarckiana might not be restricted to the three new types now under observation. Hence the question arose how it would be possible to obtain other types or to find them if they were present. It was necessary to have better methods of cultivation and examination of the young plants. Accordingly I devoted the three succeeding years to working on this problem.

I found that it was not at all necessary to sow any larger quantities of seed, but that the young plants must have room enough to develop into full and free rosettes. Moreover I observed [551] that the attributes of lata and nanella, which I now studied in the offspring of my first mutants, were clearly discernible in extreme youth, while those of rubrinervis remained concealed some weeks longer. Hence I concluded that the young plants should be examined from time to time until they proved clearly to be only normal lamarckiana. Individuals exhibiting any deviation from the type, or even giving only a slight indication of it, were forthwith taken out of the beds and planted separately, under circumstances as favorable as possible. They were established in pots with well-manured soil and kept under glass, but fully exposed to sunshine. As a rule they grew very fast, and could be planted out early in June. Some of them, of course, proved to have been erroneously taken for mutants, but many exhibited new characters.

All in all I had 334 young plants which did not agree with the parental type. As I examined some 14,000 seedlings altogether, the result was estimated at about 2.5%. This proportion is much larger than in the yields of the two first generations and illustrates the value of improved methods. No doubt many good mutations had been overlooked in the earlier observations.

As was to be expected, lata and nanella [552] were repeated in this third generation (1895). I was sure to get nearly all of them, without any important exceptions, as I now knew how to detect them at almost any age. In fact, I found many of them; as many as 60 nanella and 73 lata, or nearly 5% of each. Rubrinervis also recurred, and was seen in 8 specimens. It was much more rare than the two first-named types.

But the most curious fact in that year was the appearance of oblonga. No doubt I had often seen it in former years, but had not attached any value to the very slight differences from the type, as they then seemed to me. I knew now that any divergence was to be esteemed as important, and should be isolated for further observation. This showed that among the selected specimens not less than 176, or more than 1% belonged to the oblonga type. This type was at that time quite new to me, and it had to be kept through the winter, to obtain stems and flowers. It proved to be as uniform as its three predecessors, and especially as sharply contrasted with lamarckiana. The opportunity for the discovery of any intermediates was as favorable as could be, because the distinguishing marks were hardly beyond doubt at the time of the selection and removal of the young plants. But no connecting links were found.

[553] The same holds good for albida, which appeared in 15 specimens, or in 0.1%, of the whole culture. By careful cultivation these plants proved not to be sickly, but to belong to a new, though weak type. It was evident that I had already seen them in former years, but having failed to recognize them had allowed them to be destroyed at an early age, not knowing how to protect them against adverse circumstances. Even this time I did not succeed in getting them strong enough to keep through the winter.

Besides these, two new types were observed, completing the range of all that have since been recorded to regularly occur in this family. They were scintillans and gigas. The first was obtained in the way just described. The other hardly escaped being destroyed, not having showed itself early enough, and being left in the bed after the end of the selection. But as it was necessary to keep some rosettes through the winter in order to have biennial flowering plants to furnish seeds, I selected in August about 30 of the most vigorous plants, planted them on another bed and gave them sufficient room for their stems and branches in the following summer. Most of them sent up robust shoots, but no difference was noted till the first flowers opened. One plant had a much larger crown of bright blossoms than any of the others. [554] As soon as these flowers faded away, and the young fruits grew out, it became clear that a new type was showing itself. On that indication I removed all the already fertilized flowers and young fruits, and protected the buds from the visits of insects. Thus the isolated flowers were fertilized with their own pollen only, and I could rely upon the purity of the seed saved. This lot of seeds was sown in the spring of 1897 and yielded a uniform crop of nearly 300 young gigas plants.

Having found how much depends upon the treatment, I could gradually decrease the size of my cultures. Evidently the chance of discovering new types would be lessened thereby, but the question as to the repeated production of the same new forms could more easily and more clearly be answered in this way. In the following year (1896) I sowed half as many seeds as formerly, and the result proved quite the same. With the exception of gigas all the described forms sprang anew from the purely fertilized ancestry of normal lamarckianas. It was now the fifth generation of my pedigree, and thus I was absolutely sure that the descendants of the mutants of this year had been pure and without deviation for at least four successive generations.

Owing partly to improved methods of selection, [555] partly no doubt to chance, even more mutants were found this year than in the former. Out of some 8,000 seedlings I counted 377 deviating ones, or nearly 5%, which is a high proportion. Most of them were oblonga and lata, the same types that had constituted the majority in the former year.

Albida, nanella and rubrinervis appeared in large numbers, and even scintillans, of which I had but a single plant in the previous generation, was repeated sixfold.

New forms did not arise, and the capacity of my strain seemed exhausted. This conclusion was strengthened by the results of the next three generations, which were made on a much smaller scale and yielded the same, or at least the mutants most commonly seen in previous years.

Instead of giving the figures for these last two years separately, I will now summarize my whole experiment in the form of a pedigree. In this the normal lamarckiana was the main line, and seeds were only sown from plants after sufficient isolation either of the plants themselves, or in the latter years by means of paper bags enclosing the inflorescences. I have given the number of seedlings of lamarckiana which were examined each year in the table below. Of course by far the largest number of them were [556] thrown away as soon as they showed their differentiating characters in order to make room for the remaining ones. At last only a few plants were left to blossom in order to perpetuate the race. I have indicated for each generation the number of mutants of each of the observed forms, placing them in vertical columns underneath their respective heads. The three first generations were biennial, but the five last annual.

PEDIGREE OF A MUTATING FAMILY OF OENOTHERA LAMARCKIANA IN THE EXPERIMENTAL GARDEN AT AMSTERDAM

Gener: O.gig. albida obl. rubrin. Lam. nanella lata. scint. VIII. 5 1 0 1700 21 1 VII. 9 0 3000 11 VI. 11 29 3 1800 9 5 1 V. 25 135 20 8000 49 142 6 IV. 1 15 176 8 14000 60 73 1 III. 1 10000 3 3 II. 15000 5 5 I. 9

It is most striking that the various mutations of the evening-primrose display a great degree of regularity. There is no chaos of forms, no indefinite varying in all degrees and in all directions. Quite on the contrary, it is at once evident that very simple rules govern the whole phenomenon.

I shall now attempt to deduce these laws from [557] my experiment. Obviously they apply not only to our evening-primroses, but may be expected to be of general validity. This is at once manifest, if we compare the group of new mutants with the swarms of elementary forms which compose some of the youngest systematic species, and which, as we have seen before, are to be considered as the results of previous mutations. The difference lies in the fact that the evening-primroses have been seen to spring from their ancestors and that the drabas have not. Hence the conclusion that in comparing the two we must leave out the pedigree of the evening-primroses and consider only the group of forms as they finally show themselves. If in doing so we find sufficient similarity, we are justified in the conclusion that the drabas and others have probably originated in the same way as the evening-primroses. Minor points of course will differ, but the main lines cannot have complied with wholly different laws. All so-called swarms of elementary species obviously pertain to a single type, and this type includes our evening-primroses as the only controlled case.

Formulating the laws of mutability for the evening-primroses we therefore assume that they hold good for numerous other corresponding cases.

[558] I. The first law is, that new elementary species appear suddenly, without intermediate steps.

This is a striking point, and the one that is in the most immediate contradiction to current scientific belief. The ordinary conception assumes very slow changes, in fact so slow that centuries are supposed to be required to make the differences appreciable. If this were true, all chance of ever seeing a new species arise would be hopelessly small. Fortunately the evening-primroses exhibit contrary tendencies. One of the great points of pedigree-culture is the fact that the ancestors of every mutant have been controlled and recorded. Those of the last year have seven generations of known lamarckiana parents preceding them. If there had been any visible preparation towards the coming mutation, it could not have escaped observation. Moreover, if visible preparation were the rule, it could hardly go on at the same time and in the same individuals in five or six diverging directions, producing from one parent, gigas and nanella, lata and rubrinervis, oblonga and albida and even scintillans.

On the other hand the mutants, that constitute the first representatives of their race, exhibit all the attributes of the new type in full display at once. No series of generations, no selection, [559] no struggle for existence are needed to reach this end. In previous lectures I have mentioned that I have saved the seeds of the mutants whenever possible, and have always obtained repetitions of the prototype only. Reversions are as absolutely lacking as is also a further development of the new type. Even in the case of the inconstant forms, where part of the progeny yearly return to the stature of lamarckiana, intermediates are not found. So it is also with lata, which is pistillate and can only be propagated by cross-fertilization. But though the current belief would expect intermediates at least in this case, they do not occur. I made a pedigree-culture of lata during eight successive generations, pollinating them in different ways, and always obtained cultures which were partly constituted of lata and partly of lamarckiana specimens. But the latas remained lata in all the various and most noticeable characters, never showing any tendency to gradually revert into the original form.

Intermediate forms, if not occurring in the direct line from one species to another, might be expected to appear perhaps on lateral branches. In this case the mutants of one type, appearing in the same year, would not be a pure type, but would exhibit different degrees of deviation from the parent. The best would then have to [560] be chosen in order to get the new type in its pure condition. Nothing of the kind, however, was observed. All the oblonga-mutants were pure oblongas. The pedigree shows hundreds of them in the succeeding years, but no difference was seen and no material for selection was afforded. All were as nearly equal as the individuals of old elementary species.

II. New forms spring laterally from the main stem.

The current conception concerning the origin of species assumes that species are slowly converted into others. The conversion is assumed to affect all the individuals in the same direction and in the same degree. The whole group changes its character, acquiring new attributes. By inter-crossing they maintain a common line of progress, one individual never being able to proceed much ahead of the others.

The birth of the new species necessarily seemed to involve the death of the old one. This last conclusion, however, is hard to understand. It may be justifiable to assume that all the individuals of one locality are ordinarily intercrossed, and are moreover subjected to the same external conditions. They might be supposed to vary in the same direction if these conditions were changed slowly. But this could of course have no possible influence on the plants of the [561] same species growing in distant localities, and it would be improbable they should be affected in the same way. Hence we should conclude that when a species is converted into a new type in one locality this is only to be considered as one of numerous possible ones, and its alteration would not in the least change the aspect of the remainder of the species.

But even with this restriction the general belief is not supported by the evidence of the evening-primroses. There is neither a slow nor a sudden change of all the individuals. On the contrary, the vast majority remain unchanged; thousands are seen exactly repeating the original prototype yearly, both in the native field and in my garden. There is no danger that lamarckiana might die out from the act of mutating, nor that the mutating strain itself would be exposed to ultimate destruction from this cause.

In older swarms, such as Draba or Helianthemum, no such center, around which the various forms are grouped, is known. Are we to conclude therefore that the main strain has died out? Or is it perhaps concealed among the throng, being distinguished by no peculiar character? If our gigas and rubrinervis were growing in equal numbers with the lamarckiana in the native field, would it be possible to decide [562] which of them was the progenitor of the others? Of course this could be done by long and tedious crossing experiments, showing atavism in the progeny, and thereby indicating the common ancestor. But even this capacity seems to be doubtful and connected only with the state of mutability and to be lost afterwards. Therefore if this period of mutation were ended, probably there would be no way to decide concerning the mutual relationship of the single species.

Hence the lack of a recognizable main stem in swarms of elementary species makes it impossible to answer the question concerning their common origin.

Another phase of the opposition between the prevailing view and my own results seems far more important. According to the current belief the conversion of a group of plants growing in any locality and flowering simultaneously would be restricted to one type. In my own experiments several new species arose from the parental form at once, giving a wide range of new forms at the same time and under the same conditions.

III. New elementary species attain their full constancy at once.

Constancy is not the result of selection or of improvement. It is a quality of its own. It can neither be constrained by selection if it is absent [563] from the beginning, nor does it need any natural or artificial aid if it is present. Most of my new species have proved constant from the first. Whenever possible, the original mutants have been isolated during the flowering period and artificially self-fertilized. Such plants have always given a uniform progeny, all children exhibiting the type of the parent. No atavism was observed and therefore no selection was needed or even practicable.

Briefly considering the different forms, we may state that the full experimental proof has been given for the origin of gigas and rubrinervis, for albida and oblonga, and even for nanella, which is to be considered as of a varietal nature; with lata the decisive experiment is excluded by its unisexuality. laevifolia and brevistylis were found originally in the field, and never appeared in my cultures. No observations were made as to their origin, and seeds have only been sown from later generations. But these have yielded uniform crops, thereby showing that there is no ground for the assumption that these two older varieties might behave otherwise than the more recent derivatives.

Scintillans and elliptica constitute exceptions to the rule given. They repeat their character, from pure seed, only in part of the offspring. I have tried to deliver the scintillans from this [564] incompleteness of heredity, but in vain. The succeeding generations, if produced from true representatives of the new type, and with pure fertilization, have repeated the splitting in the same numerical proportions. The instability seems to be here as permanent a quality as the stability in other instances. Even here no selection has been adequate to change the original form.

IV. Some of the new strains are evidently elementary species, while others are to be considered as retrograde varieties.

It is often difficult to decide whether a given form belongs to one or another of these two groups. I have tried to show that the best and strictest conception of varieties limits them to those forms that have probably originated by retrograde or degressive steps. Elementary species are assumed to have been produced in a progressive way, adding one new element to the store. Varieties differ from their species clearly in one point, and this is either a distinct loss, or the assumption of a character, which may be met with in other species and genera. laevifolia is distinguished by the loss of the crinkling of the leaves, brevistylis by the partial loss of the epigynous qualities of the flowers, and nanella is a dwarf. These three new forms are therefore [565] considered to constitute only retrograde steps, and no advance. This conclusion has been fully justified by some crossing experiments with brevistylis, which wholly complies with Mendel's law, and in one instance with nanella, which behaves in the same manner, if crossed with rubrinervis.