Herbertia 1:52-60 (1934)
Reprinted with the kind permission of the Author and publishers from
Gardeners' Chronicle (London) 29: 37-38;53;71-72;89-90;111-112. 1901

HYBRIDIZATION IN AMARYLLEAE

A. WORSLEY, Middlesex, England

I purpose in this article to deal with hybrids that have been raised in gardens between individuals of this sub-order, from warm, temperate, tropical, and equatorial regions, differing from each other specifically (or in a higher degree), and especially with the few that have come under my own observation in a living state.

I have purposely excluded the Narcissi, and two other genera hardy in Great Britain (Galanthus and Sternbergia), partly because I have not made a special study of them, and partly because the subject of hybridization within such limits has already been dealt with by expert writers.

One generalization, however, is not out of place here. The habitat of the Narcissi is over a comparatively restricted area, and the fact that so many alleged hybrids have originated in a natural state without human interposition, coupled with the ease with which hybridization is effected in gardens, indicates that the whole genus has within comparatively recent times sprung from one prototype. Even between the extreme types of this genus exist regular gradations of individuals, each differing in some respect from each other—there is hardly one link missing in the chain.

This, again, points to a period of time, by no means lengthy, antecedent to which the genus was represented by only one, or at most very few, types or species; for this period has not been long enough for any number of the intermediate types to die out. Hence it follows that if the bulk of species of Narcissus are of recent establishment, the relationship existing between them is closer, and their individual characters not so irrevocably fixed as in the case in other genera in which no break has occurred for an enormous period of time.

The hybridization of Narcissi is, therefore, an occupation in which the chances of success are so great as to amount to a practical certainty. But when attempts are made to hybridize, say, the Crinums of Asia or Africa with those of America, the chances are equally great against a successful issue. Yet the prize is great, too, for success establishes a fresh epoch in that branch of horticulture, and opens fresh possibilities and fresh avenues of enterprise for all to profit by.

Among the genera specially dealt with in this article, forty-seven in number, only one, Hippeastrum, has become generally recognised among florists as worthy of special treatment.

It is a matter of general belief that the first hybrid raised was H. johnsoni, about 1799, and from this and other subsequent crosses the present race of mongrels has been evolved.

Yet this fact has for many years stood out in my mind as of paramount importance, that if a botanist were to find growing wild all the Hippeastrums cultivated to-day at our most notable professional establishments, he could not, at the outside, make more than two species out of them. Out of these, H. vittatum, remains just as it was nearly one hundred years ago in all its principal specific characters; it is certainly bigger, wider in the segments, more expanded in the flower-it has, in short, been "improved," but it has not been altered. 1 can see no evidence of hybridization, no evidence of anything further than selection and good cultivation would naturally produce.

I have only once seen a form of H. vittatum which in my estimation showed a probable hybrid origin. This came to me from the West Indies under the name of "Defiance," but no information as to origin or parentage were forth-coming. It is a small-flowered, brilliantly-coloured form, resembling that figured in Bury, Selection of Hexandrian Plants (1831), tab. 31, as "superba," but remarkable for having the typical vittate marks obscured towards the base of the upper segments by red suffusion. The style is also ascending to an unusual degree, and carries the stigma as high as the tip of the upper segment at maturation. I have raised hundreds of seedlings from dozen of alleged hybrids bearing the typical vittate markings, and these have all come as true from seed as any good species would do.

In these experiments I have obtained the alleged hybrids from a variety of sources, both British and continental. I have also attempted, times without number, to implant the vittate markings upon self-coloured forms, such as H. equestre, and vice versa: but without any results so far, although I have now some promising supposed hybrids with H. aulicum.

The other species which the botanist could form has no exact counterpart in nature. It has a short, wide-segmental, well expanded, large, substantial flower with the hairy throat of H. equestre major, the colour most commonly of H. equestre and H. rutilum (rarely of H. aulicum) and the keel markings of H. reginae. The colours are now very varied and the rarer colours, especially the coppery-reds and those nearest white, have of late been diligently sought and selected, and hence have increased in collections at the expense of the eliminated colour varieties.

Recently some pure self reds have appeared in collections, and these at first caused some speculation, although it is admitted that in all parti-coloured or "marked" flowers there is a tendency for selfs to appear at times spontaneously among the seedlings. (I had an example last summer, where out of some thousand "Cloth of Gold" Marigolds, three plants reverted to pure yellows.)

I have always held to the theory that if we could only look back far enough into the past we should see that the first parents of all our parti-coloured or marked flowers were selfs, and inconspicuous-coloured at that. Hence the appearance of selfs among our collections of seedling Hippeastrums did not surprise me in the least, although I was unaware of any coloured selfs among good species.

In the summer of 1895, however, I received from Brazil a box of bulbs which proved to be those of H. stylosum, and among these were a few bulbs of a self-red species hitherto undescribed, for which Mr. Baker, on examination of the flowers, suggested the very appropriate name of H. tricholepis, described and figured in M.S.S., Feb., 1890. This was au interesting incident, not as tending to disprove the assertion of alleged hybridizers —that they had introduced something new to the genus in their self-coloured reds, but as putting forward a tenable proposition that the result of their efforts had caused a colour reversion to some ancestral type, such as might have occurred in any seedling.

On the same hypothesis, it is easy to account for the occasional appearance of a few whitish, or greenish-white, seedlings.

Having described these two species, a botanist could do no more. Out of the twenty species which, according to Mr. Baker, belong to the subgenera (Macropodastrum, Omphalissa, Aschamia, and Lais), and which may be said to constitute the true Hippeastrums, seventeen have been at some time common in cultivation. Of these seventeen common species, the alleged hybridizers have succeeded in perpetuating one (H. vittatum), and four others in one composite form. No traces of the remaining twelve exist. Where can we find the habit of colour of H. procerum, the habit of H. solandriflorum, or H. cybister (so remarkably distinct), the spreading stamens of H. calyptratum, the spots of H. pardinum, the double flowers of H. alberti, or the marvellous markings of H. leopoldi or H. reticulatum?

Some say that the pre-potency of certain species has extinguished the weaker blood of the rest (?). If so, that is an effect which intelligent persons would be expected to fight against. It is a certain fact that the species of which all traces have disappeared are not long-lived as individuals, nor good seed-bearers under cultivation.

If on top of this the efforts of our hybridizers to impregnate other species with their pollen were ineffectual, their ultimate disappearance is easily accounted for, without having to call in any speculative ideas. My belief is that there has not been so much genuine hybridization among the Hippeastrums as we have been led by some to think; yet the fact remains that something. has been done, and that the blood of some four species runs in our present mongrel race.

Messrs. Veitch have done great work among the garden forms, and have produced a race of large-flowered mongrels. They were the introducers of H. pardinum and of H. leopoldi, and many people lay great stress on the improvement caused by these introductions. But to my mind it is very doubtful to what extent, if at all, their blood runs into the existing garden mongrels. No doubt many attempts were made, and seedlings raised, but Mr. Harry Veitch himself records in his contribution to the issue of the Journal of the R.H.S., July, 1890, that "many of them," the supposed hybrids, "came so near the species as to be practically the same thing, or the same but slightly varied, yet we are able to select several distinct new forms showing a marked improvement on their progenitors in breadth and substance of segment, size and symmetry of flower, &c." This certainly shows careful selection, but disproves hybridization.

When we try to discover when and by whom hybridization has been effected, we are met by grave difficulties. Mrs. Bury's work, published in 1831, just when the early hybridizers were in full swing, figures four supposed hybrids, and these figures may reasonably be held to give us either the whole of the ascertained hybrids that had up to then been flowered, or at least the most remarkable of them. For it cannot be supposed that a work got up at such expense and with such care would have figured the least noteworthy forms. Yet among the four figured there is not one solitary case in which specific hybridization had undoubtedly taken place. At best we have to deal with guesses and suppositions, which subsequent writers have treated as though they were dealing with ascertained facts.

In dealing with these earliest hybrids figured by Mrs. Bury, attention should be drawn to the fact that the Liverpool Botanic Gardens, where many of the plants were drawn, gave, at that period, unrivalled facilities for studying the subject with which I am dealing.

A Mr. Harrison, a native of Liverpool, but living in Brazil, was then sending home many species of Hippeastrum, mostly of great beauty, and all of them new to cultivation, to his friends and relatives in Liverpool.

As, in those days, few private individuals possessed efficiently-heated structures to grow such plants in, the botanic gardens became the recipients of most, if not all, of these importations. Hence, it was here, under the care of Messrs. Shepherd, that a great opportunity occurred for the hybridization of these plants. To begin with Johnsoni, we are at once plunged into a sea of doubt and speculation. Mr. Baker says that Johnson was a watchmaker of Prescot, in Lancashire, when he effected the first hybridization in 1799; but Mrs. Loudon, in 1841, writes that he was "a person named Johnson who had a small garden at Mitcham in 1810;" no other information was apparently forthcoming about him at that date. Mrs. Bury's figure was certainly drawn prior to 1830, yet she evidently felt that she was treading on dubious ground, for she says—

"The prototype is said to have been first raised about the year 1799, from the seed of vittata impregnated with formosissima, by Mr. Johnson . . . and the present specimen is from one of the original bulbs, presented by Mr. Johnson to the late E. Falkner, Esq., of Fairfield, near Liverpool."

At least she felt certain that this plant that she was drawing was indubitably the Johnsoni of Johnson himself. It bears a seven or eight-flowered scape, and is of a dark self crimson, banded to the apices of the segments, with a narrow, sharply defined, deep red keel, turning suddenly to whitish-green in the lower third.

The plant called Johnsoni nowadays is an entirely different thing; it has three or four flowers, is of a light brick-red colour, with a lighter-coloured (or white) keel in the lower two thirds, as described in Nicholson's Dictionary of Gardening. He only claims that it was one of the earliest hybrids. Only once have I seen the Johnsoni of Johnson alive, and that was about a dozen years ago, in the collection of the late Mr. James Backhouse, of York. I believe that the plant is now extinct. As to its parentage, formosissima, as we know the name, and as it is figured in Bury's work, belongs to the genus Sprekelia, which will not fertilise the stigma of any true Hippeastrum (so far as very extensive experiments teach me; Dr. Bonavia has reached the same conclusion as the result of his experiments). Secondly, there is no resemblance whatever to Sprekelia in the alleged offspring. The alleged female parent, H. vittatum, is one that I have never succeeded in hybridizing, nor does Johnsoni bear any resemblance whatever to H. vittatum.

Hence, I think we may dismiss the supposed parentage of Johnsoni as pure guesswork—and bad at that. It might pass as a hybrid between H. equestre, and some form of the rutilum-reginae group, or it might be simply a seedling of some variety of the latter group.

Mrs. Bury notes a "fringed nectary in the throat," and this is especially remarkable in H. equestre major, and in a less degree of some rutilum-reginae forms, such as H. subbarbatum (Bot. Mag., 2475). The facts, as they are presented by Mrs. Bury, would not lead me to class Johnsoni as a true hybrid, much less as one of ascertained parentage.

Mrs. Bury notes that "many learned disputes" had arisen on the subject, and that "many seedlings from reginae and others have obtained the appellation of Johnsonian." We are referred to the writings of Herbert, Gowen, and Lindley, for further information. Mr. Baker states that Johnsoni was raised out of H. reginae by H. vittatum.

To proceed with Mrs. Bury's hybrids. The beautiful and distinct white form called "picta," I have never seen alive. Had a hybrid origin been claimed for it, no one could have controverted the statement. But no claim to hybridization is put forward by Mrs. Bury. In figure 7, however, a very lovely unnamed seedling is figured (white, edged pink), raised out of H. solandriflorum by Johnsoni. It is a pity that such a charming and distinct form should so soon die out of cultivation, for there does not appear to be any subsequent record of this seedling or hybrid. Not improbably it was a mule, in which case it would die out in about five or six years, the average length of family-life of H. solandriflorum under cultivation in England.

The reverse cross is also given in fig. 46, showing a fine crimson flower with white star, and long tapering apices to the segments. I have seen nothing like this alive. If Johnsoni was not a hybrid, then these two beautiful seedlings were specific hybrids, and not improbably mules. Mr. Baker, in his invaluable Handbook of the Amaryllideae, pp. 47, 52, 53, gives a mass of information about the earlier and subsequent hybrids and mongrels.

H. griffini, of the Botanical Magazine, 3528, is cited as a hybrid between H. psittacinum and Johnsoni. Certainly the figure is not typical of H. psittacinum, and the plant is not improbably a hybrid. Yet it cannot be said that its parentage is undoubted, because it appears (in the letter-press) that it was raised by "W. Griffin in his hothouse at S. Lambeth previous to 1820," and did not flower till after sixteen years or more had elapsed. Seedlings usually flower in from eighteen months to four years from date of sowing, and in sixteen years or more there was certainly time for many things to happen.

Any attempt to follow out the 100 alleged hybrids named in Sweet's British Flower Garden in 1830, or those dealt with by the other authors referred to by Mr. Baker, cannot be made within the limits of this article. But with regard to those for which hybrid origin is claimed, mentioned in Mr. Baker's work, I would remark that I have received direct from different parts of South America plants indistinguishable from those figured in Redoute, 469 (Brasiliensis), and by Loddiges in Botanical Cabinet, 159 (spectabile) for which Loddiges raised no claim to hybrid origin. Mr. Baker has also included among alleged hybrids, plants such as ambiguum of Botanical Magazine, 3542, and Harrisoni of Bury, 27, of which latter Mrs. Bury distinctly states that it was imported from Peru and flowered for the first time in 1824. (Vide also Nicholson in Dictionary of Gardening.)

It may perhaps be claimed fifty years hence that the alleged new species. H. arechaveletae, is a hybrid, because it resembles some such alleged hybrid of the early years of last century. Some of the hybrids, or mongrel-bred plants, mentioned by him I have, or have had, alive in my collection-such as Carnarvonia.

Mr. James O'Brien figured in the Garden of July 12, 1879, the fine hybrid O'Brieni, raised out of H. pardinum by H. reticulatum, but this also seems to have proved a mule, and died out in due course.

M. Van Houtte also figured in his Fl. des Serres, 1277, an alleged hybrid between H. reticulatum and H. vittatum, but I have no information as to whether the parentage of this plant was ascertained beyond doubt. There seems nothing improbable in the parentage from the figure.

Sir Charles Strickland, Bart., informs me of a hybrid between H. aulicum and H. calyptratum which he has in cultivation. This extremely interesting plant should certainly be figured and put on record. I have visited the districts where both species grow in close proximity, although I never found them intermingled. Both species are prolific in varieties, but I found no trace of any intermediate forms, such as might be reasonably supposed to be natural hybrids, between them.

The following give some notable plants belonging to this genus, for which a hybrid origin has been claimed :—Ackermani pulcherrima (H. aulicum x Johnsoni), Garaway, Bristol, 1850; Splendidum (H. vittatum x H. reginae, or H. equestre), Herbert (Bot. Reg. App., 1924); Johnsoni II ( H. vittatum x H. reginae), Gowen, Hichclere; Acramanii (H. aulicum x H. psittacinum), Garaway Bristol, 1835.

I have put II after Johnsoni, to distinguish it from the Johnsoni of Johnson.

My own efforts have been directed for many years to raise Hippeastrum hybrids, hut I cannot claim to have met with success in any single instance. From this experience it would appear that raising inter-specific hybrids in this genus is a much more difficult matter than is generally supposed.

The labours of many hybridists during the past century have only produced, so far as I am aware, the three or four undoubted hybrids mentioned previously, viz: (1), H. solandriflorum x Johnsoni; (2), Johnsoni x H. solandriflorum; (3), H. pardinum x H. reticulatum, and possibly H. reticulatum x H. vittatum These must be reduced to two or three if Johnsoni is regarded as a hybrid. Probably, almost certainly, others have been raised, and their parentage left in doubt. I think this has been more especially the case with such species as H. equestre, H. aulicum, and the rutilum-reginae group. I have found all these species to be good seedbearers on their own pollen, or when crossed with mongrel forms. The fact that such mongrels have already in them the blood of these species, accounts to my mind for the ease with which they cross back with their pure bred relations. Dr. Bonavia also informs me that when in Lucknow he raised many crosses of H. equestre (presumably with garden mongrels?), which produced interesting and beautiful varieties. My own experience coincides with this. The epiphytal section, from the Organ Mountains, also set seed freely with me on their own pollen; and I have raised seedlings from H. solandriflorum, H. stylosum, and H. vittatum. In fact in every case where fair and prolonged trials were carried out with healthy bulbs, fertile seeds were produced. I experimented with twenty-five species and varieties of Hippeastrum, not including hybrids.

Among allied genera such as Habranthus, Zephyranthes, and Sprekelia, I can find no specific hybrids. Some Zephyranthes seed very freely with me, especially Z. brachyandra (Habranthus brachyandrus of some), Z. gracilfolia, Z. rosea, Z. elliptica (sp. nova), and Z. andersoni. Z. branchyandra is a most extraordinary plant as regards seed production. I have taken the greatest precautions to prevent self-fertilization, often cutting into the flower the day before expansion to remove the inert anthers; and have attempted to cross it with every form of Hippeastrum and Zephyranthes. On practically every occasion seed has formed, and with unvarying regularity the seedlings have proved true to type. This mystery is one I have been unable to solve.

To test the "prepotency" theory, I tried the pollen of Sprekelia, Placea, and Lycoris, with just the same result, or lack of result. Z. candida major never bears seed with me, though typical Z. candida does so freely in most places; Z. carinata also never carries seed, nor does Sprekelia. I experimented with thirteen species in these genera.

Among Hymenocallids, some three very interesting hybrids have been raised H. macrostephana was raised in Herbert's time, by crossing H. speciosa and H. calathina. Recently Mr. Hoog (of the firm of C. G. Van Tubergen, Junr., Haarlem), repeated the cross, taking fine forms of both species as parents. He has named his hybrid Daphne, and it is certainly both superior to, and distinct from, the H. macrostephana previously in commerce. It differs from Mr. Baker's description, in having foot-stalks 7 inches long to the leaves, and the flowers are larger and "toothed" differently. It appears to be a mule.

H. amancaes crossed with H. calathina by Herbert gave the sulphur-coloured hybrid of Botanical Register, 1665 equiposed between either parent. This hybrid (Sulphurea) is still living, and I received it a few years back from Sir Charles Strickland.

Colonel Trevor Clarke, by crossing H. calathina with Elisena longipetala, raised a hybrid which Mr. Baker could not differentiate from the Peruvian Ismene deflexa.

I cultivate twelve kinds of Hymenocallis and Elisena, and find them all to carry seeds at times, excepting the two hybrid mules, and a new sp. (schizostephana).

Among Eucharis and Urceolina, two hybrids are recorded. Far the most interesting is the generic hybrid (Urceocharis, Mast.) between Eucharis (species?) and Urceolina aurea raised by Messrs. Clibran. This, roughly, takes after Eucharis in colour, and Urceolina in form. Stevensii is a hybrid between E. candida and E. sanderi, of recent origin, 1 think.

Burfordiensis, raised in the garden of Sir Trevor Lawrence, but of unascertained parentage, has recently been claimed as a hybrid. Similar claims have been advanced to another form under the name of Elmetiana.

Mr. Krelage, in his short monograph of the genus Eucharis, listed both Mastersii and Bakeriana as hybrids, the former between E. grandiflora and E. sanderi; the latter between E. grandiflora and E. candida. I cultivate only five kinds of these genera, and the only one which bears seed with me is E. lehmanni, which does so regularly.

In the large cosmopolitan genus Crinum is wide field for hybridization. I know of three undoubted hybrids. Kunth, in his Enumeration, vol. v., p. 582, gives twenty-three alleged hybrids, and Sweet, B.F.G., p. 512, gives thirty-two.

Herbert claimed Govenianum as a hybrid, and had it figured, but the plant apparently died out shortly after.

Bury (fig. 30) gives an unnamed hybrid raised out of C. pedunculatum by C. zeylanicum. Curiously enough, the result, as appearing in the painting, shows a plant indistinguishable from C. amabile. However, there is no reason for hesitating to accept Bury's statement as to its hybrid origin.

C. Powelli was apparently raised some time not long prior to 1887, in Sir W. Bowman's garden, by crossing C. moorei and C. longifolium. The result is a beautiful and very useful plant equi-poised between the parents, but possessing originally a brilliant rosy-pink colour more intense than either. Since then Powelli has become widely spread in British gardens, and a very beautiful pure-white form has appeared, besides others bearing many shades of pink. The hybrid has all the hardiness of C. longifolium, and, I am informed, seeds freely, which is very rare in hybrids. I have, however, some doubt as to whether C. moorei is a good species. Seedlings raised by me, from plants flowering in the open ground, in a garden where no other open-air forms were grown, showed a wide divergence from the type, both in colour and shape of flower. The same splendid rosy-pink that Powelli has appeared in one seedling, combined with a flower widely divergent in shape from the parent. In fact, had I attempted to hybridize the parent, and had I succeeded in raising such a plant as this. I should have been quite satisfied that I had raised a true hybrid. This is, I believe, the way in which many alleged hybrids have been raised by really conscientious gardeners.

Only two of my seedling C. moorei have yet flowered, and neither of them was true to type.

Of course, if C. moorei comes "any way" from seed, we may entertain doubts as to whether the blood of longifolium enters into the composition of Powelli, which might be merely a seedling of C. moorei.

In any case, if C. moorei does not come true from seed, it is not a "species," and hence Powelli would not be a specific hybrid, nor the plant below (C. scabrum x C. moorei). Among the twenty-five kinds of Crinum cultivated in my garden I find C. moorei the most certain seed-bearer. C. giganteum, C. odorum (sp. nova), and C. purpurascens have never borne fertile seed.

Recently I raised a new hybrid Crinum out of C. scabrum by C. moorei schmidti. The female parent was the Jamaican variety, which is by far the finest form of C. scabrum. The seedlings flowered in three years and ten months, and the foliage partook of the characters of both parents. The flowers were most like C. moorei in shape, but of a brilliant crimson-pink colour, more intense than in any Crinum I have seen before. Another seedling from the same fruit was not so intense in colour. Both have refused to carry seed so far.

Another plant of tantalizing parentage is Amaryllis kewensis. When I first saw this plant at Kew, some dozen years or so back, it was labelled "Amaryllis hybrid, Arbuckle's var." Later on it got down to Amaryllis belladonna, Arbuckle.

At one time, as far as I could gather from various sources, it had come to the Royal Gardens as an unflowered supposed hybrid between Brunsvigia josephinae and Amaryllis belladonna. Its supposed parentage did not carry it through the critical examination to which it was subject on flowering. But still, there it was—a live thing that was not Amaryllis belladonna, and yet could claim no ascertained parentage.

And so it got called A. kewensis. And certainly, as a commemorative name, its splendid infloresence, unmatched among the Amaryllids for fragrance and beauty, entitled it to be associated with the Royal Gardens.

Yet I am of opinion that had this plant been claimed as an ascertained hybrid of Brunsvigia and Amaryllis by some authority, such claims could not have been overlooked: for it shares in many respects the characters of both suggested parents. Among the five kinds of Brunsvigia and Amaryllis in my garden, all bear seeds freely.

In the Gardeners Chronicle of Nov. 10, 1900, mention is made of an alleged hybrid between Vallota and Amaryllis, raised by Mr. Rix [Nix] of Truro, having cerise flowers. He states that he has never known Vallota to produce seed unless artificially impregnated. This is not my experience, although it is certainly a bad seed bearer in some districts. At Terrington, in N. Yorkshire, where I cultivated Vallota, it seeded freely every autumn without any artificial impregnation.

Mr. Rix himself says that, "The only noticeable difference is the thicker necks of the bulbs, and the variation in colour"—that is from the female parent. It would appear that in Mr. Rix's seedlings, 89 per cent. came true, 2 per cent. were anaemic, and 9 per cent. showed colour variety. This is not much beyond the allowed 5 per cent. of variation in true species. I would also remark that cerise coloured Vallotas were recorded many years ago, and constitute a well ascertained, but inconstant, colour variety. I have an old plate, marked "225, J. Andrews" (believed to be from the Floral Magazine) of this variety.

Among Vallota and Cyrtanthus I have only heard of one hybrid, raised by crossing Cyrtanthus (Gastronema) sanguineus and Vallota. This interesting plant is nearly equiposed between its parents but has not the showy points of either. It is named C. hybridus, and was raised apparently in 1885, and first described by Mr. N. E. Brown, but I have never seen any statement as to who raised it, or which was the female parent.

I have on several occasions repeated the cross both ways, but without raising seed except on one occasion; in fact, I can never get C. sanguineus to carry seed of any sort, and of recent years, from some cause (probably smoke), Vallota has refused to thrive in the London neighbourhood.

Among Nerines some twelve hybrids have been raised dating from Herbert's time. He himself raised seven: N. curvifolia x N. undulata gave Mitchamia and Veriscolor; N. curvifolia x N. flexuosa gave Haylocki; N. sarniensis x N. undulata gave Spofforthiae; N. flexuosa x N. undulata, and x N. humilis; N. humilis x N. undulata, and N. curvifolia x N. sarniensis, all gave unnamed hybrids.

Mr. Baker also tells us (Handbook of Amaryllideae, p. 103) that within recent years the following hybrids have been raised by Messrs. O'Brien, Leichtlin, Cam, and others:—N. pudica x N. humilis gave Amabilis; N. flexuosa x N. sarniensis gave Elegans; N. sarniensis x N. curvifolia gave Meadowbanki; N. flexuosa x N. curvifolia gave Manselli (said to be hardy); N. undulata x N. flexuosa gave Roseo-Crispa. I have the two latter.

Some of these show a great advance on the parents, but a really critical examination by an expert would, I feel sure, reduce this list. I think that the facility with which the species of Nerine hybridize is due to the same causes which have admitted of a similar state of things among the Narcissi.

Personally I do not claim much knowledge of this genus, and should be interested in learning whether all the alleged species of Nerine come true from seed?

Between Clivia and Imantophyllum one hybrid has been raised. I. cyrtanthiflorum (Flore des Serres, t. 1887), I. miniata x C. nobilis.

In Haemanthus I can only find one admitted hybrid, Clarkei, raised by Colonel Trevor Clarke by crossing H. coccineus and H. albiflos, but I have heard of several recently raised but not yet distributed, such as H. King Albert (H. katherinae x H. puniceus). Among the six species I cultivate H. katherinae, H. Kalbreyri, and H. puniceus bear seed freely.

Among the genera which have produced no hybrids whatever, I find the following carry seeds freely besides those I have incidentally mentioned above):—Pancratium, P. canariense, P. illyricum, and P. maritimum; Lycoris cyrtanthiflora (sp. nova); Acis autumnalis. The following plant I have frequently tested, but without ever raising seed: Lycoris squamigera. The following hybrids are said to have been recently raised, but have not yet flowered:—Crinum giganteum x C. longifolium (Elwes) : C. scabrum x C. erubescens (Elwes); Vallota x Amaryllis (Kew); Brunsvigia x Amaryllis (Kew).

Synopsis of Results
Number of admitted (specific or generic) hybrids of ascertained parentage:—

1 Hippeastrum   4  
2 Hymenocallis } 3  
3 Elisena  
4 Eucharis } 2  
5 Urceolina  
6 Crinum   3  
7 Vallota } 1  
8 C'yrtanthus  
9 Nerine   12  
10  Haemanthus   1  
11  Clivia   1  
Total in 11 genera   27 hybrids
36 genera produced no hybrids.
47 genera dealt with above.

These results have been reached by a process of elimination. They do not claim to give the total number of hybrids raised in these genera, but only the ascertained hybrids of undoubted parentage.

As for those which have appeared from time to time in gardens by some fortuitous process, by some forgotten labours, or as the purely natural result of grouping together in one house the various members of one genus, these may be many or few, but it is beyond the reach of human forces to tabulate them in any way.

ATTEMPTS AT RAISING INTER-SPECIFIC HYBRIDS AT ISLEWORTH.

—A bare record of the few successes that have attended efforts at hybridization give no idea of the totality of effort required to produce even such meagre results. I have therefore tabulated a few of the crosses which I have attempted, and which were registered at the time. These do not represent one-half of the attempts I have made, because no register was kept, except in cases in which the swelling fruit gave promise of seed, and in hundreds of cases this did not occur. I have also practically eliminated the huge record of generic crosses attempted, none of which have, up to the present succeeded beyond possibility of dispute. I have defined for these purposes "a species" to be a collection of individuals bearing evidence of a common parentage, in which all the important and easily recognizable attributes of the inflorescence and seeds are fixed, and which reproduce such characters in their seed progeny.

Variations in the leaves or in the colour or markings of the flowers constitute varieties.

A specific hybrid, therefore, in my view, must differ specifically from the female parent.

A varietal divergence is not enough to prove hybridization, as self-fertilised seedlings of many pure species will show varietal divergence from the type in perhaps one per cent. to five per cent, of the seedlings.

In many cases where seeds have been raised after attempted hybridization, sufficient time has not elapsed for them to flower; in other cases they have flowered and shown no specific divergence from the female type.

In the former case SMALL CAPITALS indicate the supposed male parent; in the latter case bold face is used in the subjoined tabular matter; and italics indicate that no fertile seed was formed.

The total of results gives 159 registered attempts, of which 146 are absolute failures, and thirteen possible successes. Of these thirteen, I have hopes of success in three cases, and have undoubtedly succeeded in one case.

On this analysis it would appear, on the one hand, that the chances of really effecting hybridization is at the most not more than about two per cent., perhaps not more than 0.6 per cent.; and, by including unregistered attempts, these figures would be halved.

On the other hand, it should not be overlooked that my object throughout was not to register the correct percentage of possible hybrids between all the species in any genus, but rather to raise hybrids between species so far removed from each other as to make any offspring possess horticultural merit. In short, my aim was to do the difficult thing rather than the obvious.

A FEW INTER-GENERIC CROSSES ATTEMPTED.

SPREKELIA x HIPPEASTRUM

Inter-Specific Fertilisations
HIPPEASTRUM, ZEPHYRANTHES, AND SPREKELIA

CRINUM

PANCRATIAE

ZEPHYRANTHES