Nature 59: 178-181 (1898)
Curiosities of Orchid Breeding
C. C. Hurst

A few years ago the raising of hybrid orchids was entirely in the hands of a few English experts, but quite recently it has been taken up with great success by many growers of these rare and beautiful plants, not only in England, but also on the continent and in America.

Consequently orchid hybrids have become very numerous, while new ones are constantly being brought to light. Up to the year 1860, we find but four flowered hybrids recorded, raised from distinct crosses: during the thirty years following the numbers gradually increased, until in 1890 there were about 200 enumerated. But it is in the present decade that the most rapid strides have been made, the numbers having increased by leaps and bounds until at the present time there are no less than 800 orchid hybrids, raised by hand from distinct crosses. These have all flowered in gardens, and have been duly described or recorded in various journals. And when we remember that many of these 800 crosses have been repeated in the same garden, and in other gardens, and that many individual plants may have been obtained from each capsule, we can form an idea of the large numbers of hybrid orchids that have been raised by hand.

Primary Hybrids

Of the 800 hybrids from distinct crosses, now on record, some 500 are primary hybrids, i.e. hybrids of the first generation between two different species. A careful examination of these primary hybrids shows that, as a rule, they are fairly intermediate between their parents, partaking of the characters of both, and at the same time perfectly distinct from either. The latter fact is most remarkable, and at once serves to distinguish primary hybrids clearly from all other hybrids. For example, Cypripedium x Leeanum is a typical primary hybrid, and has perhaps been raised more frequently and in larger numbers than any other orchid hybrid. It was obtained originally by crossing C. insigne (Wallich), with C. Spicerianum (Rchb. f.), both well marked and distinct species. The hybrid is fairly intermediate both in its outward characters and in its inner structure (see Prof. Macfarlane on "Minute Structure of Plant Hybrids," Trans. Roy. Soc. Edin., 1892, xxxvii, p. 245), and although it has innumerable minor and individual varieties, yet it is always perfectly distinct from both of its parents.

Nor has any intrinsic difference been observed in the reciprocal cross, the same forms appearing equally in the obverse and reverse crosses. Often many varieties are obtained from the same capsule, differing slightly in colour, form and size, but all are specifically C. x Leeanum, and cannot be mistaken for anything else. As it is with C. x Leeanum so it is with other primary orchid hybrids, so far as experiments have been made. In short, we find that primary hybrids are, as a rule, so intermediate between their parent species, and so comparatively uniform in character, that they are specifically distinct from both parents.

Generic Hybrids

Of the 500 primary hybrids on record, about 100 are generic hybrids, i.e. the parents belonging to different genera. In this respect alone these hybrids are interesting, though no doubt the systematic botanist views them with mixed feelings. It is not so long since generic hybrids were looked upon as anomalies, some of the older naturalists even regarding them as impossible, and forthwith proceeded to beg the whole question by classing those genera which are fertile with one another as species of one. Orchid growers, at all events, have almost ceased to regard generic hybrids with curiosity, familiarity with them having bred a certain amount of indifference.

A list of the generic hybrids on record up to the end of 1897, together with a diagram showing how twenty-six different genera have been linked together by artificial hybridisation, has been prepared by the writer and was published in the Journal of the Royal Horticultural Society (vol. xxi., April 1898), and to which those interested in the details of generic orchid hybrids may be referred.

Generally speaking, primary generic hybrids follow the rule of specific hybrids in bearing the intermediate characters of their parents, with a narrow range of variation.

But there are a few remarkable exceptions to this rule.

  1. Epiphronitis x Veitchii, a hybrid out of Sophronitis grandiflora (Lindl.) by Epidendrum radicans (Pav.)
  2. Epi-Cattleya x matutina, a hybrid out of Cattleya Bowringiana (Veitch) by Epidendrum radicans (Pav.)
  3. Epi-Laelia x radico-purpurata, a hybrid out of Laelia purpurata (Lindl.) by Epidendrum radicans (Pav.)
  4. Epi-Laelia x Charlesworthii, a hybrid out of Laelia cinnabarina (Lindl.) by Epidendrum radicans (Pav.)

Epi-Cattleya x radiato-Bowrigiana, out of Cattleya Bowringiana by Epidendrum radiatum (not listed by Hurst)

These four generic hybrids are very curious indeed, inasmuch as all agree in reproducing the generic characters only of the pollen parent Epidendrum, without the slightest trace of the peculiar structure of the seed-parents, Sophronitis, Cattleya and Laelia. Yet in minor characters, in colour, form and size, the four hybrids distinctly differ from one another and from their Epidendrum parent.

A close examination reveals the fact that these minor differences correspond with the peculiar differences in the parentage, thus showing that the crosses have really been effected: yet as the same time it must be candidly admitted that did we not know the parentage, we could never have determined it, so overwhelming is the influence of the predominant partner Epidendrum radicans. It will, no doubt, be observed that E. radicans is the pollen parent in each of the above cases. Curiously enough, when this reed-like E. radicans is crossed with the pseudo-bulbous E. vitellinum (Lindl.), (female), a similar result is obtained, the offspring E. x radico-vitellinum being scarcely distinguishable from the reed-like E. radicans. Again, when the reed-like E. x O'Brienianum — itself a hybrid out of E. evictum (Hook. f.) by E. radicans— is crossed with the pseudo-bulbous E. vitellinum, (female), a similar result is obtained, the offspring being reed-like in habit as in the pollen parent. Yet when the reed-like E. radicans is crossed with other reed-like species of Epidendrum, and again when the pseudobulbous species of Epidendrum are crossed with species of Laelia, in every case normal hybrids are produced intermediate between their parents.

Though we cannot pretend to unravel this tangled skein, yet, so far as experiments have been made, it seems quite clear that (1) the species of Cattleya, Laelia, Sophronitis, and the pseudo-bulbous species of Epidendrum, when intercrossed, produce normal hybrids intermediate in character.

(2) The same result is obtained with the reed-like species of Epidendrum are united with one another.

(3) But when the reed-like species of Epidendrum are united with the pseudo-bulbous species of Epidendrum, or with species of Cattleya, Laelia and Sophronitis, abnormal hybrids are produced, having the essential characters of the reed-like Epidendra. From these facts it might easily be argued that a reed-like Epidendrum was the ancestor not only of the pseudo-bulbous Epidendra, but also of the more highly specialised genera Cattleya, Laelia and Sophronitis. In that case the pseudo-bulbous Epidendra would form an interesting connecting link between the lowly reed-like Epidendra and the gorgeous aristocratic Cattleya and Laelia.

Prepotent Generic Crosses

Perhaps the strangest curiosity in the history of orchid hybridisation is the remarkable prepotency of the genus Zygopetalum over the three genera Odontoglossum, Oncidium and Lycaste, so far as experiments have been made. Zygopetalum Mackayi (Hooker) has been crossed with four distinct species of Odontoglossum, viz. O. Pescatorei (Linden), O. crispum (Lindl.), O. grande (Lindl.) and O. bictonense (Lindl.), also with one species of Oncidium, O. unguiculatum, and one species of Lycaste, L. Skinneri (Lindl.), by more than one hybridist, and the result has always been the same, namely Zygopetalum Mackayi pure and simple, without a trace of the peculiar structure of the pollen parent in any case. This result is very perplexing and exceedingly difficult to account for. I have made careful inquiries into the details of all these cases, and have satisfied myself that accidental self-fertilisation is out of the question, the pollen of the mother parent having been carefully removed in every case previous to pollination. Parthenogenesis, too, is evidently a broken reed to lean upon, for the seedlings from the same seed-pod differed among themselves in colour and other minor characters, which would hardly have been the case had they arisen from parthenogenetic seed-buds. Neither is Z. Mackayi naturally prepotent over other species when crossed, there being at least three cases to the contrary. Nor is the genus Zygopetalum naturally prepotent over other genera, as two distinct and intermediate hybrids between Zygopetalum and Colax testify. As in the case of the Epidendrum crosses, mentioned above, it may be suggested that Zygopetalum is the ancestral genus of Odontoglossum, Oncidium and Lycaste, and that the characters of the ancestral genus remain latent in the more recent genera, with the result that when the latter are crossed with the former, the mixing of the germ plasms causes these original characters to dominate, the outcome being a reversion to the ancestral genus Zygopetalum.

A rather interesting fact has come to light which certainly lends colour to the above speculation:— The seedling Odontoglossums, raised in the gardens of Baron Rothschild, of Paris, during the first eighteen months of their growth, are said to have resembled Zygopetalum more than they did Odontoglossum. This coincides with the established fact that living beings tend to resemble their ancestors in the early stages of their development. However, for the present we must suspend our judgment, and wait patiently for further facts: it is to be hoped that future experiments will throw more light upon these curious generic crosses.

Another curious fact in connection with generic crosses may perhaps be of interest, and that is the remarkable crosses between the East Indian species of Cypripedium and the South American species; these two sections of the old genus Cypripedium have recently been raised to generic rank, under the names of Paphiopedium and Phragmipedium respectively, by Mr. R. A. Rolfe, of Kew, and seem to form two distinct and natural groups. Hybrids between these two new genera are peculiarly interesting, inasmuch as the former has a one-celled ovary, while the latter has a three-celled one, showing that this condition is no barrier to fertilisation. More than twelve distinct crosses between different species of these two genera are on record, and many plants have been raised, but so far all resolutely refuse to flower, notwithstanding the many inducements that have been put in their way, and many of them are now large vigorous plants, long past the usual flowering age. One plant of these crosses is recorded to have flowered in the United States, but as it flowered exactly the same as the mother plant in genus, species and variety, one cannot be quite sure that the cross was really effected.

Many of the remaining plants (the writer has several in his collection) are distinctly intermediate in their foliage and habit of growth, and clearly bear the stamp of their recorded origin.

Secondary and Tertiary Hybrids

Of the 800 distinct crosses mentioned in the foregoing, some 270 are secondary hybrids, i.e. hybrids of the second generation, one or both parents being a primary hybrid; while thirty are tertiary hybrids, i.e. hybrids of the third generation, one of the parents, at least, being a secondary hybrid. So far no hybrid orchids are recorded to have flowered beyond the third generation, but perhaps it may not be premature to mention that unflowered hybrids of the fourth generation are known to be in existence. The writer has in his collection six hybrids of the fourth generation, five years old, all raised from the same capsule, and which contain in their pedigree five distinct species and three distinct hybrids. So that in the near future there will be ample material in this direction at the disposal of the student of heredity; with this distinct advantage, that orchids being individually valuable, their pedigree is carefully and systematically recorded, which unfortunately is more than one can say of the great majority of garden hybrids.

A careful examination of secondary hybrids, shows them to be very different from primary hybrids in their range of variation. As we have already seen, primary hybrids are comparatively uniform in their characters; so much so that, as a rule, they are quite distinct from their parents. On the other hand, secondary hybrids have a much wider range of variation, often approaching either parent, sometimes even reverting wholly to one or the other. For instance, to take the simplest form of a secondary hybrid, i.e. a hybrid crossed with one of its parent species. We find that the offspring, as a rule, are very variable, a few reverting to the parent species and a few to the parent hybrid; but the great majority are intermediate forms approaching either parent, the whole forming a series of links between one parent and the other. In short, we find that secondary hybrids have a far wider range of variation than have primary hybrids.

Natural Hybrids

The existence of natural hybrids was formerly thought by some naturalists to be highly improbable, if not actually impossible. But now, when absolute facsimiles of supposed natural hybrids have been raised by hand in gardens, from the same two species among which they grow, they can no longer be regarded as pious speculations, but are indeed accomplished facts. The number of proved hybrids in orchids alone is now very considerable, with the result that many intermediate and doubtful forms, hitherto classed as distinct species, are now placed in their proper position as natural hybrids. Mr. R. A. Rolfe, of Kew, has done yeoman service in reducing the chaos of natural hybrid orchids to something like order. And so it has come to pass that artificial hybridisation, which it was supposed would lead systematic botany into the direst confusion, by the irony of fate, seems destined to be the only trustworthy means of saving systematic botany from its own confusion; and the systematist, however orthodox he may be, can no longer afford to ignore artificial hybrids.

Fertility of Hybrids

The question of the fertility of hybrids is a highly interesting one, and especially important to the student of evolution; and I venture to think that recent experiments in orchid hybridisation have added considerably to our knowledge of the subject.

One of the principal objections to Darwin's theory of the origin of species was the supposed general sterility of hybrids.

Darwin fully appreciated this difficulty, and, after a careful and most elaborate survey of the whole question of hybridism came to the following conclusions:— "First crosses between forms sufficiently distinct to be ranked as species, and their hybrids are very generally, but not universally sterile . . . . The sterility is of all degrees" (Origin of Species, 6th ed., p. 262). Again:— "The sterility of distinct species, when first united, and that of their distinct hybrid offspring, graduates by an almost infinite number of steps from zero (when the ovule is never impregnated, and a seed capsule is never formed) up to complete fertility . . . . This high degree of fertility is, however, rare (Animals and Plants, 2nd ed., vol. ii. p. 163). Some fifteen years later, Dr. Alfred Russel Wallace took up a similar but more definite position. He writes:— "One of the greatest, perhaps we may say the greatest, of all the difficulties in the way of accepting the theory of natural selection as a complete explanation of the origin of species, has been the remarkable difference between varieties and species in respect of fertility when crossed. Generally speaking, it may be said that the varieties of any one species, however different they may be in external appearance, are perfectly fertile when crossed, and their mongrel offspring are equally fertile when bred among themselves; while distinct species, on the other hand, however closely they may resemble each other externally, are usually infertile when crossed, and their hybrid offspring absolutely sterile" (Darwinism, 1890, p. 152). Since that time, hundreds of hybrid orchids have been raised in gardens; as we have already seen, there are now on record some 500 distinct primary hybrids raised from distinct species, also some 300 secondary and tertiary hybrids from distinct crosses, raised from parents themselves hybrids.

In the face of these facts, therefore, we can hardly maintain that "distinct species . . . are usually infertile when crossed," and still less can we assert that "their hybrid offspring are absolutely sterile." As it is with orchids, so it is with other garden plants that have been hybridised, e.g. roses, rhododendrons, dahlias, chrysanthemums, carnations, begonias, pansies, &c.: the wonderful forms seen in gardens at the present day are all hybrids of many generations, being the known product of more than one species; and these are all fertile in the production of seeds. Darwin seemed to attach much importance to the different degrees of fertility in hybrids ranging from complete fertility down to absolute sterility. At present we have no means of ascertaining the relative fertility of hybrid orchids with that of ordinary species, owing partly to the fact that very few crosses have been made in gardens between varieties of the same species, and partly to the fact that many thousands of seeds are contained in one capsule, being practically innumerable. But through the kindness of Mr. Reginald Young, of Liverpool— an enthusiastic connoisseur in Cypripediums— the writer has been able to work out certain statistics bearing on the question of the comparative fertility of hybrids (see Journ. Roy. Hort. Soc., vol. xxi., April 1898).

The voluminous records in Mr. Young's stud-book cover a period of about five years, and were specially selected by the writer on account of the reputation of the breeder as a careful observer and a precise recorder. Out of the 577 crosses made by Mr. Young among thirty distinct species and fifty-three distinct hybrids of the new genus Paphiopedium, no less than 78.3 per cent. proved fertile, i.e. produced good seeds.

Of these, the crosses between distinct species only, 95.2 per cent. were fertile, while of the crosses in which a hybrid was concerned in the parentage 71.8 per cent. proved fertile.

This seems to show that crosses between distinct species are almost if not quite as fertile as crosses between varieties of the same species (taking the latter at complete fertility, i.e. 100 per cent.); while crosses with hybrids, though fertile to a high degree, are yet rather less fertile than crosses between species. A further analysis of the figures shows that while hybrids crossed with pollen of pure species give 89.5 per cent. fertile, yet pure species crossed with pollen of hybrids give but 56.7 per cent. fertile. This points to the conclusion that the slight decline in the fertility of hybrids is due in a large measure to the loss of power in the pollen of hybrids.

Why the male element in hybrids should be so much less potent than the female element I cannot pretend to say, but I venture to think that the matter is worthy of consideration as a possible factor in the evolution of species.

Evolution of Species by Natural Hybridisation

The experimental demonstration of natural hybrids shows clearly that intercrossing between different species is carried on in a state of nature, to a far greater extent than was formerly supposed; and the comparative fertility of these natural hybrids would be of vital importance to them in their struggle for life.

If, as seems highly probable from the above experiments with orchids, it is the pollen only of hybrids that is impaired, and the capacity of the hybrid to bear seed crossed with the pollen of pure species remains practically unimpaired, it is quite clear that the natural hybrid has a part to play in the evolution of new species.

As we have already seen, hybridisation tends to increase variation especially beyond the first generation, and, naturally, the more variable the offspring the better fitted would they be to adapt themselves to changed conditions of life.

If the circumstances changed rapidly and considerably, the variable offspring of the hybrids would stand a better chance in the struggle than the more uniform offspring of the parent species, which were themselves specially adapted to the old conditions. In this way, as conditions changed, new species would be evolved more fitted to the new conditions of life than the old species, which they would gradually replace, and I venture to suggest that natural hybridisation is the most rapid of nature's means towards that end.