Journal of the Royal Horticultural Society 1: 57-82 (1866)

XIV. Abstract of Max Wichura's Observations on Hybridization*.
By the Rev. M. J. BERKELEY, M.A., F.L.S.

*Die Baztardbefruchtung im Pflanzenreich, erläutert an den Bastarden der Weiden; von Max Wichura. Mit zwei. Tafeln. 4to. Breslau, 1865.

IN the former series of this Journal I gave an abstract of Gaertner's important work on vegetable hybrids; and a translation of Naudin's more recent observations appears in the present volume of the New Series, together with some interesting remarks of M. Decaisne's on the effects of intercrossing amongst different varieties of pears. Very recently a most instructive memoir appeared at Breslau, by Max Wichura, which contains so much weighty matter that a copious abstract cannot fail to be acceptable to our readers. I am indebted to Mr. Darwin for the loan of the book, and have profited by his marginal notes pointing out the principal points of interest.

Herr Wichura informs us in his preface that he was led to the subject by the fact of his friend Dr. Wimmer having arrived at the conclusion, from numerous observations, that a great part of the doubtful forms in the difficult genus Salix were hybrids. From a desire to confirm the correctness of this conclusion, of which he was, however, already convinced, he determined to institute a series of experiments with a view to find out what would be the actual results of artificial impregnation in this perplexing genus. These experiments were commenced in 1852, and continued uninterruptedly till 1858; but unfortunately, partly from illness and partly from a long absence in Japan, they were discontinued, and as there seems no probability of their being resumed, he has thought it best to publish at once the results of his experiments, comparing them, where requisite, with those of Koelreuter and Gaertner, and pointing out their connexion with Darwin's views on the origin of species.

In all experiments on hybridization, the one great desideratum is to exclude the possibility of access of every kind of pollen except that which is the subject of experiment. As willows have a singular propensity to form hybrids, and are, at the same time, dioecious, a circumstance which obviates all difficulties about doubtful extirpation of anthers in the female parent, they are peculiarly fitted for the purpose, and as their pollen is so heavy that it falls at once to the ground when ripe, instead of floating in the air, and is, moreover, covered with a raised delicate network, it is peculiarly adapted to adhere to the hairs of insects, without whose help it scarcely ever reaches the female plant, which is often at a great distance from the male. Multitudes of insects, however, greedily search out the blossoms of either sex, and in consequence impregnation seldom fails in wild plants.

In this appetency for the flowers, it is of the greater importance in the course of experiments to find some effectual way of preventing the access of insects. For this purpose cylinders were made of thin tarlatan, 2-3 inches wide, and 6-12 inches long, furnished with a string at either end to tie them closely to the branches, and strengthened in the centre with two or three bands to prevent them collapsing. To show that the insulation was perfect some flowers were left to themselves; and not a capsule set. In the case of those flowers which were artificially impregnated, so soon as the stigma dried and the ovary began to swell, the cylinders were removed as no longer necessary, end replaced when the seed was nearly ripe, to prevent its being carried away by the wind. The spikes having the male blossoms, moreover, were separated as soon as the anthers began to burst, and placed in water to prevent the access of insects. The pollen was applied with a camel's-hair pencil, and a separate pencil used for each kind of pollen, which was always taken from the same individual and, as nearly as possible, normal plant.

For the impregnation of early flowering species with those which blow later in the season, the fact was of great importance that the pollen of willows retains its potency for some time. In some eases pollen ten days old was efficient, while the vitality was still further prolonged by steeping it in a solution of honey, made of as much as will lie on the point of a knife, mixed with two ounces of water. Fresh pollen, placed in this mixture, frequently began, in the course of ten or twelve minutes, to put forth its tubes. Pollen of Salix silesiaca, eight days old, seemed almost as potent as ever; in twenty-eight days the traces of vitality were very slight, while that of Salix cinerea had become weak in sixteen days, though still capable, after immersion for three hours, of slowly developing its tubes. On the whole it should seem that pollen kept in a dry, cool, shady place, may, under favourable circumstances, be still efficient after fourteen days, while that eight days old may be trusted implicitly.

The seeds of willows germinate very rapidly, and lose their vitality with proportional rapidity. The pollen-grains, indeed, sometimes preserve their vitality longer than the seed. This quickness of germination arises from the fact that the embryo is already richly supplied with chlorophyll, and that the testa is very thin and transparent, while there is not a trace of albumen. They soon dry up if they are not in contact with moist earth; they should therefore be sown as soon as they are ripe, and in from twelve to twenty-four hours the cotyledons make their appearance. Care must be taken not to water the young plants too heavily, or the delicate seedlings may be washed away. Our author's plan was to raise the plants in pots, which could be supplied with water from below, and easily protected from the access of any strange seed. They were allowed to remain in this situation till they were some inches high, and then transplanted into the open air; and thus plants sown in May obtained a height of two or three feet by the end of summer. The smaller kinds generally flowered in three years, the larger, but not arboreous species, in about four. Scattered catkins, however, sometimes appeared on two-year-old plants, and in one instance a plant sown in June produced flowers when ten months old.

Our author adds a remark on the use which he made of natural hybrid willows during the course of his experiments. They seemed to be of the greatest importance as regards hybrids composed of more than two factors. By their help he was enabled to raise two hybrids (of which admirable nature-printed figures are given at the end of the volume) compounded of six different species, which he believes, after seven years' experience, to have been impossible with artificial hybrids.

An example is added of the composition of hybrids of more than two species, the parents being a spontaneous hybrid of S. Lapponum and S. silesiaca, a spontaneous hybrid of S. purpurea and S. viminalis, with an artificial hybrid of S. caprea and S. daphnoides. The circles in the following scheme indicate the male, the squares the female, and the triangles the father and mother plants of the free hybrids or species indifferently.

His notation to exhibit the whole history of the hybrid is expressed thus, ♀ indicating female, ♂ male, spont. spontaneous, art. artificial

S. (♀(♀(Lapponum + silesiaca) spont.+ ♂(purpurea + viminalis) spont. + (♂ ♀caprea + ♂daphnoides) art.),

which, expressed at length, denotes that the senary hybrid was raised between a female parent derived from the intercrossing of two spontaneous hybrids, of which the female parent was derived from S. Lapponum and silesiaca, and the male from S. purpurea and viminalis; while the male parent was derived from an artificial cross of S. caprea impregnated with the pollen of S. daphnoldes. In the names of artificial hybrids, the female parent always comes first, while in spontaneous hybrids the parents are placed in alphabetical order.

A complete list of the author's experiments follows, giving in one column an account of those which proved successful, and in another those which did not succeed, with the date of the experiments. The former, comprising binary, ternary, quaternary, quinary, and senary hybride, amount to thirty-five; the unsuccessful, amounting to eighty-one, include the three first heads; no unsuccessful experiments were made in the fifth and sixth; but an attempt to obtain an octonary with two quaternary hybrids did not succeed. Both lists are well worth the study of every one engaged in the scientific determination of willows. The proportion of successful experiments is far more unfavourable in binary than in more complicated combinations. Nothing, however, can be concluded from this; for in binary hybrids an endeavour was made to effect heterogeneous combination, or certain unions already existing in free nature, and that with old pollen, while in complicated combinations the object was to combine as many species as possible; the species, therefore, most nearly related to each other were chosen, and fresh pollen was used. It is not to be doubted, moreover, that the more favourable results amongst complicated hybrids depended on the more favourable circumstance selected for their production; besides, it should seem that willows have a great tendency towards the production of complicated hybrids, which must be regarded as a very peculiar circumstance. Exceptional cases doubtless often depend upon unfavourable accidents; for combinations which failed one year succeeded the next, and it is only after a large continuous series of observations that we can arrive at any perfectly satisfactory result.

Wichura does not consider it necessary to repeat in willows the experiments of Koelreuter, Gaertner, and others, which show that plants have a superior appetency for their own pollen to that of any other species. This depends probably on the circumstance that the strange pollen does not push forth the pollen-tube so rapidly as that which is native, so that if both kinds of pollen are applied together, the tube of the strange pollen, supposing it to be properly protruded, finds the micropyle occupied on its arrival. Gaertner informs us that in different plants a different number of generations must pass by to secure the conversion of one species into another. In willows, it appears from Wichura's experiment that it requires only four years.

A small-leaved form of S. (purpurea+viminalis) was used for this experiment, which was continued for two generations. A female hybrid between S. purpurea and viminalis was impregnated by S. viminalis, the result of which was a hybrid scarcely distinguishable from S. viminalis. Again, a female hybrid, the result of a hybrid between S. purpurea and viminalis, impregnated by S. viminalis, impregnated in its turn with S. viminalis, produced a hybrid still more near S. viminalis, and a further crossing would in all probability have perfected the change. It was a matter of interest to observe how far the peculiar smallness of the leaves would be recognizable in the result of the first-mentioned experiment. This was distinctly the cam, and it seems clear that the propagation of individual peculiarity may be effected in a hybrid.

A hybrid fertilized with its own pollen produces plants the same as or resembling both the parents. The peculiarities of the parents are moreover repeated in the hybrid. Quaternary hybrids fertilized with their own pollen, produced seed which germinated, but the young plants soon failed.

Gaertner believed that in such combinations, amongst many individuals resembling the hybrid, some plants would occur which reverted very nearly to the maternal, or less frequently to the paternal type. No such reversions, however, occurred in the course of Wichura's experiments. Particular individuals showed indeed many variations, but no reversion. He believed, therefore, that the pollen of either parent must have had access without Gaertner's knowledge; and Mr. Darwin thinks it probable that the instances alleged by Naudin were due to the circumstance of his plants being cultivated. Wichura had no opportunity of examining the truth of the fact asserted by both Koelreuter and Gaertner, that hybrids prefer the pollen of one of the parents to their own, or that the pollen of some strange species may make the pollen of the hybrid powerless, though he thinks both statements quite probable. He remarks, moreover, that his quinary and senary hybrids are quite new in the history of hybridization. He believes that the possibility of still higher combination is only prevented by the constantly increasing sterility of the produce. The senary hybrid whose formula is given above was still so fruitful that it would scarcely refuse union with an heterogeneously combined senary, ternary or quaternary hybrid, while the other senary hybrid, between a quaternary of Salix Lapponum and silesiaca ☐ and purpurea, viminalis ◯, and fertilized with an artificial hybrid of caprea ☐ and daphnoides ◯,and then finally fertilized with pure daphnoides ◯, did not give an opportunity of carrying the experiment further, inasmuch as all the plants perished before they produced flowers.

Impregnation between two quaternaries failed as stated above, in consequence of the far advanced sterility of the female plant. He doubts, moreover, whether it is possible to raise a hybrid of the sixteenth degree between two octonaries. He has nothing to say as regard willows in confirmation of Gaertner's remark, that while species will not combine with some other species, they will do so with a hybrid of that species, as, for example, Nicotiana glutinosa, Tabacum, and Langsdorfii, will not combine with N. rustica, though they will do so with a hybrid of N. paniculata and rustica.

We now come to a very important chapter on the imperfect nature of hybrids. Our author divided his experiments into two series, successful and unsuccessful. There are, he says, certain degrees of success, which, however, cannot be used as the stepping-stones for a new division, since they pass into each other, but deserve mention as characteristic phenomena in hybridization.

The following degrees of failure may be noted:—

  1. The catkins submitted to hybridization wither as soon as the flowering is complete, like those which have not been impregnated, and fall off.
  2. The ovaries swell and ripen, but do not contain a trace of seed.
  3. The ovaries are quite filled with the silky hairs which clothe the umbilical end of the seed, but contain no embryo.
  4. Seeds are present, but small, languid, and incapable of germination.
  5. Seeds apparently perfectly, developed, but do not germinate.
  6. Seeds germinate, but the young plants are weak and wither in a short time without further development.

The gradation in the number of seeds next demands our notice. The results of different experiments were very various. Sometimes the seeds were few, but fertile and active; sometimes, on the contrary, numerous, but with only a few fertile mixed with a number of abortive seeds; sometimes tolerably numerous, without any such admixture; but in general hybrids yielded on the whole a smaller number of seeds than plants impregnated with their own pollen.

This imperfection depends on the nature of hybrids, which, according to our author, are never so active as pure plants—a position which is perhaps doubtful. This weakness he believes to arise from a comparative want of potency in the pollen, though hybrids often exhibit a peculiarly luxuriant but rank growth.

It is not necessary to describe the structure of willow-pollen. It is sufficient to observe that in hybrid willows many modifications occur.

The imperfection of the pollen in hybrids is often adduced as a decidedly distinctive mark between them and pure species, but this is incorrect. There are hybrids whose pollen is scarcely less regular than that of the parents, as for example in Petunias, and there are pure species which have more irregular pollen than many hybrids. It is, however, true that in the greater number of cases the pollen of hybrids is less regular than that of pure plants.

In willows this is a law without exception. Amongst pure species, only S. fragilis (though not the nearly allied S. alba) and single individuals of S. triandra had a large number of irregular grains in their pollen. The pollen of all other European willows is remarkable for its great regularity. One grain is almost perfectly like another in size, colour, constitution, and form; and pure species are peculiarly distinguished by this character from hybrids, in whose pollen, amongst a greater or less number of large, regular grains, there are always some which are abnormal. Moreover the various hybrids differ greatly from each other in this respect.

Hybrids whose pollen contains no perfectly developed grains, and hybrids which amongst normal pollen-grains contain only a small percentage of irregular grains, stand at the two extremities of a series which numbers almost as many intermediate links as there are hybrids. It may be asked of what nature is the irregularity of hybrid pollen? How do the differently formed grains comport themselves with respect to potency? How are they developed? In what relation do the different degrees of irregularity which we recognize in hybrids stand to their composition?

These irregularities are of six kinds.

  1. Linear elongated bodies, about the size of an anther-cell, which contain within a membrane a large number of dark, dirty-yellow, round pollen-grains of somewhat more than the usual size.
  2. Irregular bodies, consisting of from two to four full-grown grains, sometimes light, sometimes of a dark dirty yellow, which contain a great mass of oil.
  3. Grains nearly twice the usual size, of a dark dirty yellow, scarcely transparent, otherwise of the regular orbicular form, and containing much oil.
  4. Globular bodies, three, four, or ten times smaller than the normal grains, colourless and almost translucent.
  5. Grains which differ only in a small increase or diminution of size from normal grains, with which, moreover, they agree in the pale colour, semitransparent aspect, and in the regular development into a bell when placed in watery fluids.
  6. More or less regularly plicate, dark, impellucid bodies, of the normal size or smaller, which, placed in water, are unaltered and do not assume a spherical form.

The last form is the most common, and exists in pure S. fragilis and triandra; 4 and 5 often occur together. No. 3 is uncommon, and 1 and 2 very rare. No. 1 has occurred only in a spontaneous hybrid between S. cinerea and incana, and No. 2 only in a very few instances.

Imperfections in the sexual organs of hybrids have, however, been shown by Gaertner and others to go further in other plants than amongst hybrid willows. But what most concerns us at present is the remark of Gaertner, that in the anthers of all fertile hybrids, as for example in Nïcotiana rustica panniculata, Malva mauritiana sylvestris, Aquilegia atropurpurea canadensis, Lychnis diurna vespertina, larger and smaller grains are mixed in different proportions, besides nail grains of different degrees of irregularity, as elongated, shrivelled, or mere arid sacs; and this is the case especially with those which are the least fertile, a fact which was observed also by Koelreuter. A tolerably correct opinion, therefore, may be formed of the comparative fertility of hybrids from an examination of the pollen, as appears also very distinctly from Wichura's observations.

Pollen-grains of willows were treated with the above-mentioned solution of honey; and it appeared, on microscopical examination, that the normally formed grains of hybrids, of a light colour and semitransparent aspect, constantly developed pollen-tubes; those like No. 5 usually did so; the irregular grains like No. 2, if of a light colour, occasionally; while the confluent grains, of a dark, dirty yellow hue, with those comprised under Nos. 1, 3, 4, and 6, never made pollen-tubes, and must be regarded as absolutely sterile.

It appeared also, on comparing the activity of the pollen-grains of hybrids with those of pure species, that there was a difference. The defect in hybrid pollen in some hybrids, and probably in a great many, though not in all, was twofold. The number of tube-producing grains is far smaller than in pure species, while the potency of these grains is comparatively weak.

The study of the pollen-grains in willows is comparatively easy, as the contents of the anthers in every stage are easily squeezed out on the stage of the microscope by the covering glass, so that they are ready for examination without any further preparation. Willow-branches, moreover, are readily developed in water, and the different stages of development of the pollen observed without much trouble.

In pure willows the mother cells of the anthers are of equal size, and divide, with almost mathematical precision, into four tetrahedric pockets, in which the contents form yellowish shining pollen-grains of exactly the same size; while in hybrids the course of development is far less regular, and subject to the above-mentioned derangements.

The first form arises from the fact that the component cells of the substance of the anthers (Collenchymzellen), contrary to the established rule, do not dissolve, but still hang together, and thus prevent the separation of the mother cells. In mature anthers these confluent pollen-grains form yellow linear bodies; when pressed out of the immature anther they have an appearance like that of frog-spawn.

The four next forms arise from an irregular division of the mother cells. In hybrids in which these forms occur, sometimes we find the mother cells only uniseptate, and in consequence two large pollen-grains are formed instead of four half the size. Sometimes, though the cells are quadriseptate, the division is irregular, and one larger grain with three smaller ones is often the consequence. A further cause of irregularity arises from the fact that the dissepiments of the mother cells are so imperfect that the pollen-grains of the different component sacs are confluent with each other. In this way the compound grains of No. 2 arise, consisting of two or more confluent individuals. More rarely the cells divide into a larger number, from sixteen to twenty, and the grains are consequently minute. In No. 6 the division of the mother-cell is at first apparently normal, as are also the pollen-grains, and the malformation first appears when the pollen-grains begin to contract. The triple folding does not take place. One half of the globe remains convex, the other is pushed in, so that a purse-like body is formed, or, if the triple folding takes place, it is very irregular. In either case, the dark colour and want of transparency show that the fluid contents of the grains are more or less deficient.

It is worthy of remark that the irregular division of the mother-cells does not necessarily induce sterility. On the contrary, the pollen-grains of a greater size than usual, which arise from the irregular division of the mother cell, are frequently fertile. Sterility attaches rather to those recklings which are below the normal size and which are generally colourless and transparent, to those which are not expanded when in contact with fluids, or, finally, to those which are extravagantly large and confluent, if of a dark brownish yellow. This change of colour depends probably upon some anomalous composition of the fluid contents.

It is scarcely possible to classify hybrids according to this imperfection of the pollen, because the gradations are so gentle; certain laws, however, seem to prevail.

  1. The anomaly of the pollen increases with the succession of generations arising from the fertilizing of hybrids with their own pollen. This is the case with hybrid willows; and Gaertner has shown that, with respect to hybrids of Aquilegia, Dianthus, Lavatera, and Geum, there was a perpetual decrease of fertility and a general decrepitude, though he does expressly state that this depended on imperfection in the pollen.
  2. Different individuals of the same hybrid generation nearly resemble each other in the degree of imperfection in the pollen. This was found to be the case in all experiments with artificial hybrids which for the most part yield a large number of male plants. In wild hybrids, especially of the formula

♀ (♀(a + b)+ ♂(a + b)+ ♂(a + b), &c.,

it is very probable that individuals may be found in which the different degrees of imperfection of pollen may be found in one and the same hybrid. It is possible also that different individuals of the same two species, when hybridizing, may yield different kinds of irregular pollen.

  1. Distantly related species of willow, when combined in hybridization, give a more irregular pollen than nearly allied species.
  2. The anomaly of pollen increases with the number of intermingled species.

S. (♀(Lapponum + silesiaca) spont. + ♂ (purpurea+viminalis) spont.) has far more sterile grains than S. (purpurea+viminalis) and S. (Lapponum+silesiaca) spont., which has proportionally regular pollen, while even some ternary hybrids are quite sterile though the parents are fertile. Again, the quinary ♀ S. (♀(purpurea+repens) + ♂ aurita) art.+(caprea + viminalis) art.) has by far a greater number of sterile pollen-grains than all the binary hybrids which are known of its five parents.

We now arrive at the female flowers of the hybrids; but unfortunately this part of the subject has engaged Wichura's attention much less than the former. It is indeed overwhelmed with difficulties, and would require immense patience to follow it out properly; for the contents of the embryo-sac could not be observed so easily as those of the anthers. Professor Henslow went as far as could be expected in this direction when he wrote his model treatise on a hybrid Digitalis; but more is required to satisfy the existing state of science, and a thorough examination of any differences that may exist in the more intimate characters of the ovules in hybrid plants is one of our great desiderata.

Wichura, however, remarks that the female sexual apparatus suffers in many cases a certain degree of imperfection. The ovaries of certain willows, which appear to be binary hybrids of S. viminalis, alba, and cinerea with S. triandra, or ternary hybrids of S. viminalis, caprea, and cinerea, are perfectly sterile. They do not show the slightest trace of any development of the seed, though abundant opportunity was afforded, either in the way of artificial application, or through the natural agency of insects in the open air. The female catkins of the above-mentioned hybrids of S. triandra wither quickly after expansion, and fall off prematurely, those of S. longifolia, on the contrary, are apparently luxuriant, ripen and burst, but they contain no seed. As the male plants of S. (triandra+viminalis) and S. longifolia are distinguished by a far advanced irregularity of pollen, a certain connexion between the imperfect formation of the male and female organs of this hybrid is not to be mistaken. Other examples might be adduced. In the greater number, indeed, of female hybrids of willows there is, in comparison with the pure species, a less degree of fertility, which appears only in the diminished number of seeds; while many hybrid willows, examined superficially, appear to be quite as fruitful, if attention is paid only to the woolly contents of the capsules. Even in these, Wichura believes that an attentive examination would show a diminished number, in confirmation of Gaertner's assertion "that even the most fruitful hybrids yield constantly a smaller number of seeds than the parent species fertilized with their own pollen."

But not only in the organs of reproduction, but also in their vegetation, hybrids exhibit many phenomena by which they are more or less decidedly distinguished from pure species. Koelreuter and Gaertner both agree that the greater number of the hybrids which they raised artificially were distinguished by rankness of growth. The plants were taller than their parents, spread more on every side, had a longer duration (being biennial or perennial when the parents were annual or biennial), were more capable of withstanding cold, and blossomed more freely, luxuriantly, and precociously than the parents; something of which appears also from Naudin's experiments, so far as extreme luxuriance is concerned, as witness his hybrid Mirabilis. On the contrary, Gaertner speaks of other hybrids which succeeded only in very favourable weather, and were intolerant of cold. These were hybrids whose parents were only slightly related, and whose seedlings were delicate from the first.

Similar examples occur amongst hybrid willows; but luxuriant growth is by no means the normal character. It is doubtful whether the hybrids of S. caprea and viminalis, S. cinerea and incana, with their tall and spreading habit, are to be reckoned as examples, since they are to be explained by the union of tendency to an arboreous habit in S. caprea and incana on the one aide, and to a fruticose habit in S. viminalis and cinerea on the other. The rank growth of the senary hybrid from S. viminalis, purpurea, Lapponum, silesiaca, caprea, daphnoides was indeed very striking; and in this case, at the age of ten mouths a single catkin was produced—a circumstance which agrees with Gaertner's instances of the early blossoming of hybrids.

Traces of a defective feeble development are far more frequent amongst willow hybrids than rankness of growth. S. (♀ arbuscula + purpurea) art., a union it is to be observed of distantly related species, put forth every year a strong shoot, which, however, in the course of the summer began to wither. Not one of the numerous plants which were raised from the intercrossing came into blossom, and all gradually perished. A senary hybrid impregnated with the pollen of one of its parents afforded a quantity of plants. Three of these, 6 1/2 years old, had attained a height of six feet, forming miserable little trees. Buds were formed every autumn which promised well for the following spring, but the wood did not get ripe enough to give strong shoots, so that they always remained weak; and though they had the best place in the garden, they never produced any blossoms; and the result even of a hot summer was the same. S. (♀ viminalis + daphnoides) art., in itself a tolerably strong hybrid, had a rather unfavourable place near the garden fence. Of fifteen plants, eight were alive on Wichura's return from his travels, and all of these sickly. A plant of pure S. viminalis, which had sprung up with the hybrids, and had the same difficulties to contend against, was healthy, a proof of its steady enduring nature. S. (♀ caprea + ♂ daphnoides) art, grew vigorously to a height of from ten to twelve feet, but, though planted in a favourable spot, gradually failed, so that out of six original examples only one was left.

S. (♀(♀(Lapponum +silesiaca) + ♂(purpurea + viminalis)) + pruinosa) art. was from the first weak and miserable, and all the specimens were overgrown by other stronger hybrids. Fries remarks of S. rubra, that is S. (purpurea +viminalis), Wimm., that in Sweden all the branches freeze, while this is not the case with S. purpurea or S. viminalis. He concludes, therefore, in contradistinction to Wimmer, that it cannot be a hybrid. Wichura proved that it is so by actual experiment. The circumstance, therefore, mentioned by Fries is exactly applicable to what has been said of the comparatively feeble nature of hybrids, and is so much the more interesting since S. (purpurea+viminalis), Wimm., independent of the imperfection of the pollen, exhibits no symptoms of weakness in its outward appearance. The possibility must therefore be allowed that the vegetative growth of other seemingly strong hybrids (as, for example, by far the greater number of hybrid willows) is essentially weak, so as to prevent them in the Darwinian struggle for life from competing with their parent species.

If all these circumstances are combined with the imperfection of the pollen and the partial sterility of the ovaries, the comparative defect of vital energy in hybrids may be considered as proved. Nor does the luxuriant growth of some contradict this, as we know that it often accompanies suppressed fertility. Koelreuter's views seem then to be confirmed, that the luxuriance of hybrids arises from sterility. The most fruitful hybrids are always less so than their parents. The weakness of the generative organs in very luxuriant hybrids induces an increase of vegetative growth, while this is not the case in others which are too weak to exercise such a reaction.

The relation of the sexes is modified in hybrids. The proportion between the male and female plants is different from what it is in pure species. This indeed requires further proof, the confined limits of his garden not allowing the author to make as many experiments as he otherwise would have done with pure species. His observations were therefore confined to the proportions which exist respecting hybrid willows in free nature, where the difference in the proportion of the sexes in hybrid and pure species is very striking. Pure species, however, are not quite equal in the number of their male and female plants. S. fragilis, alba, pentandra, and triandra, which are so strikingly distinguished from other European willows by the double nectary of their male blossoms, bear, at least in the neighbourhood of Breslau, a greater number by far of male individuals than the species with a single nectary, as S. cinerea, viminalis, purpurea, repens, &c. In the former the males are more numerous than the females, in the latter this is not the case. Males and females are in nearly equal proportions in these, though with a slight preponderance of females, while in the other the females are far more numerous, In artificial hybrids with one nectary, there are about ten females to one male; but when more than two species are combined, it should seem that there are great differences; but observations are at present not sufficiently numerous to establish any decided law. Wichura considers his observations on this matter to be quite isolated, and not apparently to agree with those of Girou de Buzareingues on the proportion of male and female individuals in Cannabis sativa, Rosa cinnamomea, Rumex acetosella, Spinacia oleracea, and Lychnis dioica.

Since willows have no petals, they will teach us nothing respecting the different variations of colour which are so striking in many hybrids. We must therefore look rather to the different forms which they assume. When we consider the form of a hybrid with reference to that of its ancestor, we must distinguish three categories of characters.

1. Constant characters in which the parent species agree. These enter unaltered into the hybrid.

2. Constant characters in which the parent species are distinguished from each other. These enter only by halves into a hybrid, so that it is intermediate between them.

3. Variable characters. In these the hybrid is equally variable. If the parents agree with each other in their variable characters, these are not necessarily inherited by the hybrid; and if they differ in their variable characters, the hybrid is not always intermediate.

As regards the first and second, the following may serve as an example. S. purpurea has two stamens whose threads and anthem are so intimately united that they seem to have only one thread and a four-celled anther. The stamens in the other European willows are free, S. incana only being exceptional, in which the threads are united halfway up. If a hybrid is formed between S. purpurea and another willow, the threads are confluent below in accordance with the first rule, while they are free above and divergent. S. purpurea, moreover, has sessile stigmas. The style of other species, as for instance S. viminalis, is rather long. The hybrid has the style about half as long as that of the second species. Similar examples may be adduced as to the position of the stigma, the leaves, the pubescence, the rough under surface of the leaves, the bark, the stature, the time of blooming, and finally the quantity of salicin in the bark. Wichura has seen only a single exception to this law in S. (♀ arbuscula + purpurea) art. As regards foliage, it was exactly intermediate; but it had nothing in common with the upright habit of S. purpurea, but rather lay completely prostrate on the ground, being in this respect an exaggeration of S. arbuscula. It is possible, however, that this was only a sign of the weakness inherent in hybrids, and so the exception is only apparent. More complicated hybrids exhibit the same law, but it is of course more difficult to indicate the points of resemblance and difference.

Gaertner indeed supposes that in genera which are rich in species, there are some which have a prepotent influence when hybridizing, so that in some hybrids the type either of the male or female parent prevails. Amongst the various hybrid willows, though the genus is so rich in species and so prone to hybridizing, Wichura has never seen a prepotent type, and doubts Gaertner's statement, especially as he makes it in very qualified terms. With respect to the critical examination of types—whether, for example, a hybrid is more like the mother or father—the perfect distinction is subject in many cases to great difficulties, since very much depends on the subjective view of the observation; for in consequence of the frequent intermelting of both characters, the one observer finds in a hybrid the maternal type, while another thinks the paternal type prevalent.

The question is of more importance whether the paternal or the maternal parent has most influence on the form of the hybrid. Gaertner has made many experiments in this direction, and says, "The most important and interesting phenomenon in the intercrossing of plants is the perfect equality of either production; seeds arising from the impregnation of either parent produce plants of the most complete resemblance; so that the different origin, on the most careful examination of either kind of hybrid in respect of formation and type, does not induce the slightest difference." Slight differences were said to occur in the genus Digitalis only, but it is doubtful whether these depend on intercrossing or on some other conditions. S. (♀caprea + ♂viminalis) and S. (♀viminalis + ♂caprea) did not present the slightest difference in the number of normal pollen-grains or in their outer form, but their resemblance was so perfect that they might be taken for products of one and the same crossing.

* It is very doubtful, however, whether this be a hybrid at all. It is believed to be an accidental variety produced by grafting C. purpureus on C. Laburnum; and it is conjectured that a portion of the cells of the one plant must have become so intimately connected with the cells of the other, that when cell-division took place part of the plant assumed the type of C. Laburnum, another of C. purpureus, while others were exactly intermediate between the two.

Since the variable characters of species maintain their peculiarity as such in their progeny, a wide margin is left for the formation of varieties in hybrids which comprise the variable characters of two species, notwithstanding the law of intermediation. The most remarkable examples are afforded by those hybrids which exhibit the prevailing characters in different parts of the same individual. The occurrence of such types is placed beyond doubt by such productions as Cytisus Adami, a hybrid of C. Laburnum and purpureus*. Mixed types are, however, very rare, and nothing of the kind has ever occurred amongst willows.

Wichura confirms Gaertner in the assertion that where hybrid pollen is used for the impregnation of simple or complicated hybrids, as also in pure species, there is a great predominance of individual forms, while hybrid ovules impregnated by the pollen of pure species, even in the most complicated combinations, give very uniform products.

On the whole, our author's experiments have far more generally exhibited variety of form in the produce where the pollen of hybrids, and, on the contrary, uniformity where the pollen of pure species was employed. We may therefore attribute to the pollen of hybrids a tendency to produce varieties, while, since the ovules of the same, if fertilized with pure pollen, yield as uniform produce as those of true species, there does not appear in general to be any inherent propensity in them to produce varieties. There is but one observation of Gaertner's which gives any support to the assertion that the varieties from the ovules of hybrids with pure pollen may be more numerous than from those of pure species. Tolerably constant forms arose from a hybrid fertilized with the pollen of the male parent, if the female hybrid was fruitful, but variable forms if the productive powers were weak. This could not, however, depend upon the pollen, which in either case was the same, but on the incipient sterility of the female, which induced a malformation of the ovules; so that the rule that the product of hybrid pollen is more polymorphous than that of pure pollen remains unshaken.

The circumstances which favour the spontaneous production of hybrids in willows, seem to be nearly the same as those which facilitate artificial intercrossing. Dioecious plants, which are subject to fertilization by insects, must necessarily produce hybrids, if they comprise, like willows, a great number of nearly related species, which grow in company with each other, and which have the same or nearly the same time of flowering.

Wimmer believes that there are thirty-four undoubtedly pure European species of willow, or perhaps thirty-five if S. helvetica is not a hybrid. If, however, S. pedicellata, Desf. and S. arctica, Br., whose hybrids in the extreme southern and northern latitude of Europe have not yet been studied, be excluded, we have thirty-two species which may concur in the formation of hybrids. From these sixty-six indubitable binary hybrids are known which occur here and there mixed with their parent species. Of these a complete list is given, which deserves attentive study. Ternary hybrids also are said to occur in nature, of which nine are enumerated; but these must be considered doubtful unless reproduced artificially.

Hybrids are distinguished from pure species, as regards their occurrence, by certain peculiarities. They can only arise where the parent species are associated. This is especially observable with respect to widely distributed species hybridized by species of rare occurrence. The intermediate form of S. Lapponum and myrtilloides is peculiar to northern swamps, where both species grow together. The hybrid S. (aurita+myrtilloides) is confined to Silesia and the few places where S. myrtilloides grows.

A second more important difference consists in the comparative number of individuals.

* Wichura's opening observations as to the peculiar facility for hybridizing in consequence of the agency of insects being requisite to carry the pollen to the females, often situated at a considerable distance, must be regarded as modifying this observation.—ED.

1. As plants prefer their own pollen to that of strange species, and willows often grow in groups, there is a much greater chance for the success of their own than of strange pollen, especially when the times of shedding of the pollen do not coincide. Hybrids must therefore be comparatively rare*.

2. Hybrid willows are generally in a position for self-impregnation; and as this induces a gradual decay of vitality, hybrids have not the same chance of propagation by seed as pure species.

3. Hybrids, however, take the pollen of the parent species more readily than their own. The male plants are rarer in hybrid willows than the female; the pollen, moreover, of hybrids, when fertile, contains a certain number of impotent grains. The hybrid therefore has a greater chance of being fertilized with parent pollen, and, as this process is repeated, it is more probable that the hybrid should revert to one of its parents than that it should be preserved unaltered for any considerable time.

4. The difficulty is scarcely less with the female hybrid. Some hybrids have perfectly sterile ovaries and cannot be propagated by seed; others produce only, a few seeds; and even the most fruitful are not so fruitful as pure species. Here, then, is another ground for hindrance of their diffusion.

5. Others, finally, are weak as regards vegetative power. If, then, they grow in company with their parents, they are soon overgrown, and so gradually die out.

From all these circumstances it is clear that spontaneous hybrids must be rare. If we except those which, like hawkweed, sedges, or brambles, are propagated in certain localities by suckers, they are of rare occurrence. In Silesia perhaps there is one hybrid amongst 500 individuals. There is perhaps of S. (alba+ fragilis) one per cent. of the parent species, 1/3 per cent. of S. (purpurea+viminalis), 1/5 of S. (aurita+repens), 1/10 of S. (purpurea+repens), 1/50 of S. (cinerea +purpurea), 1/500 of S. (triandra+viminalis). These are, it is to be observed, only rough approximations. Complex hybrids are so rare that no estimate is made respecting them.

In some confined localities, min swampy wastes, sandbanks, river-beds, &c., certain hybrids sometimes grow together in considerable numbers. This is the case with S. (pupurea + viminalis) on the Breslau river-banks, and with S. (aurita + repens) and S. (purpurea + repens) in swamps. According to Andersson, S. (Lapponum +myrtilloides) (S. versifolia, Wahl.) surpasses its parents in number in some parts of Lapland. This depends probably on a little-noticed biological peculiarity in willows. They cannot bear being overshadowed, but become sickly, while the young plants perish, even amongst herbs of small stature. Moist places in districts bare of vegetation, either from artificial or natural circumstances, are the favourite places of the development of the light seeds. All therefore in a particular spot have vegetated in the same year; and so it may happen that, if these have come from some hybrid, the hybrids may surpass the pure parents in number. In other spots of the same locality not a single hybrid can be found.

Finally, hybrids have been widely cultivated in some districts by means of cuttings, as S. (alba+fragilis) and S. (fragilis+pentandra), which are used for enclosing the roads, and S. (purpurea+viminalis) for basket-work.

We pass over the systematic chapter, which, however excellent, is not especially suitable to the objects of this Journal. It is, however, remarkable that the great similarity between S. caprea, cinerea, and aurita is rather apparent than real. Though they do not absolutely refuse intercrossing, like the Cucurbita Pepo, maxima, melanosperma, and moschata, they comport themselves differently towards different species. Hybrids of these three species are by no means more frequent than other hybrids; and while many hybrids of S. caprea and aurita, with other uninectariferous willows, have a tolerably regular pollen, there is in hybrids of S. cinerea a marked irregularity which reaches its maximum in S. (cinerea+incana). While S. caprea rather combines with S. viminalis, Lapponum, and silesiaca, S. aurita does so with S. livida and repens, and S. cinerea, as S. incana and purpurea, seems to occupy a somewhat isolated position.

* This view is distinctly stated in Darwin, Or., p. 288, ed. 3.

We come now to the general conclusions. After first very briefly stating Mr. Darwin's theory of the origin of species by variation and elimination, our author proceeds to some general consideration of hybrids. The habitual weakness, however, which he supposes in males is not universal, though perhaps in every male there is some weak point, however strong it may be in some particulars. The hybrid then comes with no new peculiarities into the world; it has those of its parents, and generally in a decidedly less complete degree. It is a perfect whole, so far as the peculiarities of the parents agree with each other; so far, however, as they differ it is intermediate, so that no speciality is completely reproduced; and since the differing parts are combined in one intermediate whole, since these require different conditions, the hybrid can never be perfectly accommodated to outward circumstances*. A hybrid between a fish and a bird, were such a creature possible, would never either swim or fly well with its imperfect fins and feathers; and something of the kind must be the case with all males. We cannot, however, go so far with the author as to say that no hybrid can be exactly fitted for its place in nature, because, though it may not suit the circumstances of either of its parents, its intermediate nature may be suited to some intermediate conditions, as species of different climates form hybrids.

As no two parent species agree entirely with each other, though the hybrid may be equally strong with the parent as regards one or more characters, there will always be characters in which it is weaker, and therefore as a whole it will be less perfect. This is quite the case with willows.

The most imperfect simple hybrids are those which S. viminalis and S. cinerea form with S. triandra; and if we compare their peculiarities, we find that they differ from each other more than all other willows which form hybrids in free nature. S. arbuscula, an alpine willow of low stature, with the leafy stem of the catkin, is very different from S. purpurea, a low-ground species. The artificial hybrid between them is so imperfect, that it dies every year down to the ground, and never bears blossoms, S purpurea with confluent stamens, and S. viminalis with free stamens, form a hybrid with very irregular pollen. On the contrary, the pollen of S. (caprea+viminalis), both of the parents of which have free stamens, has much more regular pollen. The more different the parent species, the more imperfect the hybrids. Examples might, moreover, be adduced of more complicated hybrids in confirmation of this position.

It follows that those species only can combine to form hybrids which agree in a proportionally large number of peculiarities and the relative biological conditions, which accords with the fact that only cognate species or nearly allied genera can combine.

Were it necessary to prove by experiment that every species, in order to maintain itself in certain vital conditions, requires all the peculiarities with which it is endowed, one could think of nothing more appropriate than hybridizing, which calls into action in a weak and impaired intermediate condition all the constant differences of the parents. Were the hybrid as vivid and vigorous, it would be a contradiction to this hypothesis.

The constantly increasing sterility in hybrids, and their dying out when fertilized with their own pollen, belongs probably to another class of phenomena. It is notorious that families which have the seeds of disease in them, and yet intermarry, die out after some generations; and the raisers of varieties are well aware that all abnormal peculiarities in plants and animals increase, if attention is paid to them in successive generations, so that propagation is confined to these abnormal individuals. If a hybrid is fertilized for successive generations with its own pollen, individuals come together which have the same weak point, viz. that of reproduction. The increase of weakness and sterility, and the rapid dying out of hybrids by continual impregnation with their own pollen, agrees perfectly with the above-mentioned circumstances. This is exactly Darwin's view of interbreeding causing sterility in successive generations.

AEgilops speltaeformis, Jordan (♀(♀AE. ovata, L. + ♂Triticum vulgare, L.) + ♂T. vulgare) is an apparent exception to this rule. Esprit Fabre raised in 1858, from a spontaneous hybrid (AEgilops ovata + Triticum vulgare), AEgilops triticoides, Requien, and at a later period Godron bred an artificial hybrid from the same parents, and fertilized this again with the pollen of Triticum vulgare, which is now commonly multiplied in gardens, under the name of AEgilops speltaeformis, while the primary hybrid is very unfruitful. In this second hybrid it is to be observed that there are three parts of Triticum vulgare, and in consequence it is nearly accommodated to the vital conditions of T. vulgare. We must remember, however, that it is propagated only artificially. The hybrid has not yet been found wild, and is therefore rare, if it exists at all, and of no great power of endurance.

So long as hybrids, like stars, were regarded as freaks of nature, they added nothing to our knowledge; but when it was ascertained that the same laws existed in the formation of monsters, though differently directed, they became a fertile source of information respecting morphology; and so hybrids, if looked upon as products of a normal fertilization under extraordinary circumstances, may teach us important lessons respecting the generation of plants When both parents belong to the same species, we cannot tell what part the male and female parent take respectively in the formation of their progeny. But dissimilar factors are united in hybrids, and an intermediate form is the consequence. The products which arise from reciprocal crossing in plants, unlike those which are formed amongst animals, are perfectly alike. It is of no consequence which is the male and which the female parent. It is therefore a mathematical necessity that the pollen-cells must have just the same part in the act of generation as the ovules. The following observations in the form of aphorisms are to be considered conjectural, and require to be submitted to proof:—

1. Setting out from the fact that every branch, with a few exceptions, repeats perfectly the type of the plant from which it springs, and that the origin of every branch is referable to the development and division of a single cell, it must be allowed that the cells of plants partake of their specific peculiarities, so that, under favourable circumstances, they can reproduce new individuals.

2. Embryo-cells and pollen-tubes are cells. They must contain therefore the type of a certain form, which will appear pure and unmixed on the growth of the cells into new branches.

3. The secret of reproduction depends on the combination of two different cells into a common whole. Accepting this as an axiom, it is natural and necessary in the union of two cells if they belong to differently constituted individuals, that a more or less perfect intermediate form should arise, whose form is not altered, whether a or b supply the pollen; for each of the two cells, whether embryo-cell or pollen-tube, bears in itself the type of the individual from which it was taken, and each of the two species gives to the new structure a numerically equal part, namely one cell. Both united in opposite crossing must give the same intermediate form, in which both species have entered in equal proportions.

4. Embryo-cells and pollen-tubes give exclusively the subsequent peculiarities to the product in consequence of their containing within them the type. The relation of the mother plant to the embryo after the completion of impregnation is that of the stock to the graft. Both nourish a strange individual, and are intimately connected with it, without exercising any influence whatever on its typical peculiarities.

5. If the remaining cells, on their growth into branches, reproduce as a rule the same individual of which they, are the essential part, experience teaches, on the contrary, that in the process of production many individuals of abnormal structure—that is, varieties—make their appearance.

6. Since the sexual union of differently constituted individuals (that is, hybridizing) always produces a being intermediate between the type contained in the ovule and that in the pollen-tube, we may regard it as a law which has equal weight in the origination of varieties.

7. The existence of a variety is therefore a proof that the ovule or pollen-tube from whence it sprung, or both, must have had a type departing from that of the normal species.

8. The embryo-cells and pollen-tubes have therefore not merely the function of reproducing the individual, but also of new abnormal forms.

9. In hybrids the power of forming varieties exists especially in the pollen, in a less degree in the ovules; and this is probably also the case in pure species.

10. If we reflect that the new individual arising from impregnation is intermediate between the type of the pollen of the male and the embryo-cell of the female, we must, in order to explain the form of the variety, assign to the variety-producing sexual cells a tolerably wide departure from the ancestral type.

11. If a variety-forming ovule combines with a variety from pollen-grain, so abnormal a form may arise from the union that we may perhaps explain in this way the origin of Gaertner's exceptional types, assuming the correctness of his observation.

12. It is doubtful whether the variety-forming power of the pollen can be seen from outward inspection; this certainly is not possible in all cases. There are very variable plants with regular pollen, as Salix nigricans, S. Lapponum, and the bistigmatic Carices. On the contrary, we have plants with irregular pollen, like hybrids, which are distinguished by the greater or less stability of their productions, as Salix fragilis, Trifolium montanum, Barbarea vulgaris, but not B. stricta, Potentilla incana, Hierochloè, and many others. But in the greater number of cases, as, for example, amongst hybrids, in many cultivated plants, in many indigenous plants conspicuous for white or light-coloured blossoms, and finally in a great part of the very variable wild plants, as hawkweeds, roses, the shrubby brambles (but not Rubus caesius and idaeus, which yield no varieties and have regular pollen), multiformity in the pollen and great variability are combined. It is also possible that in the multiformity of the pollen of these plants, we see the variability of their offspring pointed out as it were in embryo, or in other words, the increasing variability of the progeny of hybrid pollen must be referred to an irregular division of the mother pollen-cell as its proximate cause. Mr. Darwin believes that there is some connexion between sterility and variability; and there is suggestive matter in these observations in other directions.

* It should be observed that in a state of nature plants are vary frequently found under conditions which are not the most suited to their nature, but in places where they are able to maintain the struggle of life against others by which they would be overwhelmed. This is a fact which is too little attended to by cultivators. Some excellent observations on this subject by the Dean of Manchester will be found in the first volume of the former series of this Journal, p. 44.—ED.

Finally, a parallel may be drawn between hybrids and many cultivated plants. A common point in both consists in the fact that they are not fully accommodated to the conditions under which they live. Hybrids are not so because, in consequence of their abnormal generation, they have inherited only a portion of the peculiarities which belong to their accommodation, and cultivated plants because, from artificial treatment, they are kept in climatic and local conditions for which they were not destined*. The history of all our cultivated plants, so far as it is known, shows this. Transported from free nature into the garden, from a warm into a cold climate, the plant at first preserved its peculiarities for a time; then slight changes crept in; more followed, till at last, by repeated generation, scarcely one of the individuals from seed is like the other. In this state the pollen of many plants is quite like that of hybrids, and all the forms are found in it which have before been adduced as phenomena in the pollen of hybrid willows. Most cultivated varieties of Primula auricula, Hyacinthus orientalis, Tulipa Gesneriana, Solanum tuberosum, Brassica oleracea, Mathiola incana, Antirrhinum majus, Cineraria cruenta, and Verbenas have very strikingly irregular pollen. In a white variety of Cineraria cruenta a tetrahedric confluence of the pollen-grains was found like that in S. (cinerea+incana); Koelreuter therefore says rightly, "The nature of plants and beasts is in a certain degree like that of hybrids, as soon as in any way they are removed from that destination for which they are especially fitted." Where culture and hybridizing concur, the consequences of disaccommodation are naturally quicker and more extensive than where only one of these is at work. Thus we find in the Fancy Pelargoniums, the Giant Pansies, the Calceolarias, and the Fuchsias, variability and multiformity of pollen in the highest degree. All these plants have irregular pollen, and in many individuals to a surprising extent. It is probable, though it has not yet been proved, that, as in hybrids, irregularity of pollen in cultivated plants favours variability. If gardeners, in the raising new varieties, would have recourse to the microscope, and let those individuals remain for seed which have the most irregular pollen, or if they would use the most irregular pollen in artificial impregnation, they would in all probability materially expedite the accomplishment of their wishes.

At all events this remarkable position arises from our discussion, that imperfect accommodation gives to an organism an increased tendency to form varieties. Does, then, the same law prevail in nature? Plants are subject to the most different local and climatic conditions. Organisms which at any former time were adapted to climate and locality, must, when change of condition takes place, gradually cease to be accommodated. Had they in this state of transition possessed only the degree of variability which the greater number of wild plants now exhibit, their persistence would have been placed in question. If variability, however, increased with increasing disaccommodation, there might well be one among the many, varieties which, suited to the new condition, would have full scope, while the other less adapted forms would be displaced, whether change in combination with "natural selection" be the agent, or whether matter endowed with life, in consequence of an inherent necessity, accommodates itself to a law of conformity suited to outward circumstances. Teleologically considered, there is, in the variability of unaccommodated species arising from generation, a struggle to produce a new form suited to the altered conditions; and hence it follows that the dying out of species and the origination of new forms are links of the same conditions, just as when an individual dies to perfect and mature the seeds from which it arises with new life in a numerous offspring.

CybeRose note: I emphasize the comment, "It is therefore a mathematical necessity that the pollen-cells must have just the same part in the act of generation as the ovules."