Breeding new lilies

New lilies can be obtained by cross-fertilization, mutation, and through polyploidy.

The term ‘cross-fertilization’ refers to the fusion of male and female gametes of two specially selected but different varieties or species. Seed resulting from such a cross produces a hybrid lily, which may have two parents utterly different in appearance, character, and growing habits.

Mutations are usually spontaneous, but can also be artificially induced; they are transmissible to the progeny. If, however, a whole complex of cell tissue changes genetically, it is termed a ‘sport’. Artificially induced mutations are created by subjecting plant tissue to gamma-rays or x-rays: results are mostly negative, I.e., plants or their offspring are deformed or sterile.

A polyploid plant has more than the usual number of chromosomes; again, the process is artificially induced, and usually leads to stronger and more vigorous plants as well as flowers of larger than normal size.

The cross-fertilization of lilies

The ever-1ncreasing popularity of lilies in our gardens during recent years is undoubtedly due to the development of hybrids, their enhanced beauty, and their universal adaptability – wild lilies do not unfortunately possess the latter virtue, and have individual and often difficult requirements which can only rarely be satisfied in the average garden.

Wild types continually and frequently crossed with each other produce more suitable lilies for garden culture, because they have lost some of their parents’ undesirable characteristics, but breeders can only obtain such results provided they rigorously select the progeny of these crosses and ruthlessly discard any which fall short of requirements. Thousands of lilies are discarded in this way, certainly many thousands more than those few which eventually reach the stage of a variety.

Selection takes account of every characteristic of the plant – not only the shape, size and colour of the flowers, but also their number and spacing, the length of pedicels, whether the flowers are pendulous, erect or horizontal, the number of leaves, their colour, the height of the main stem. Every detail is considered – many of them never noticed by a non-specialist. The essential aims are vigour of growth, health, and disease resistance. The ability of a lily to multiply quickly is also of the greatest commercial importance; a lily which multiplies by axil bulbils and/or bulb scales is of greater value than a plant without bulbils or one which produces relatively few bulbs from scales. Only if a breeder is satisfied that his new plants possess the maximum number of desirable attributes will he make arrangements for large-scale multiplication.

The raising of new lilies is an expensive process – the success of one cross has to compensate the professional breeder for the heavy expenses incurred for the thousands of initial crosses and subsequent raising of plants in expensive greenhouse space, most of them ruthlessly discarded for failure of one or another characteristic.

Relationship and incompatibility

Not every lily crosses with every other lily. Lilies only cross with each other if they are closely related, and even then success is not inevitable; the only means of determining compatibility is by trial – visual observation or academic knowledge is of no use in this matter. Whether a cross ‘takes’ or not is often influenced by uncontrollable circumstances. As a rule fair and dry weather and a warm summer contribute greatly to a cross’s success whereas many a cross has been spoilt by a cold and wet summer.

One failed cross between two lilies does not necessarily prove that they are incompatible, and many crosses of the two particular varieties need to be made before the case is proved one way or the other; to be absolutely certain, numerous parent plants of the same variety or species need to be used (also reciprocally) and, if possible, under differing circumstances and weather conditions.

The use of the word ‘reciprocal’ in this context means the interchange of plants as male and female parents. If in the first case plant A was fertilized with pollen from plant B, then in a reciprocal cross plant B must be fertilized with pollen from plant A. The use of reciprocal crosses often enables breeders to overcome anatomical and genctical insufficiencies, e.g., sterile pollen of one parent.

Successful crosses between widely different species are rare, and are mostly brought about by sheer good luck; E. Debras of Orleans succeeded in 1925 with L. x aurelianense from L. sargentiae x L. henryi.

To cross two lilies, the pollen (male) of one plant is transferred to the stigma (female) of another. The pollen absorbs the nutrient secreted by the stigma, and the protoplasm in the pollen grains produce tubes which grow through the style to the ovary, where they fuse with the female nucleus. To obtain the maximum quantity of seed in the autumn, enough pollen is necessary to fertilize the several hundred available egg cells in the ovary.

Two lilies of the same species cross in the same manner as closely related species. It is possible to cross, for example, L. davidii, not only with another plant of L. davidii, but also with its near-related L. davidii var. willmottiae. But a cross between L. davidii and L. auratum or L. speciosum will not succeed, because the relationship between them is too slight, and they belong to another section.

Self-pollination – that is, the fertilization of a plant with its own pollen – is only rarely possible. This barrier is described as self-1ncompatibility or self-sterility. If self-sterile plants are multiplied by bulb scales to the clone stage, all further plants obtained from the clones are self-sterile, I.e., they are unable to cross-pollinate, and therefore produce no seeds. The term ‘clone’ refers to all those plants which are vegctativcly reproduced from one original parent.

In addition to the genetical prerequisites, the pollen must always be viable, and capable of fertilizing the female egg cells; old or rain-damaged pollen is of little use.

For the female organs, too, certain conditions are essential to successful fertilization. The stigma must be ripe to take the pollen. This stage is reached after the flower has opened, and only when the stigma has secreted a gummy, sugar-containing substance, although previously deposited pollen also ‘takes’. Rain spoils the receptivity of the stigma.

Once the stigma has been dusted with pollen, it is useless to attempt to fertilize again; but a self-stcrile plant, although already dusted with its own pollen, is able to produce seeds if pollinated with another plant.

Another possibility is the fertilization of various lilies with a mixture of pollen. This method is of advantage in an attempt to break through quickly the incompatibility barrier between lilies only distantly related, and speeds determination of the fruitful pollen, which can be identified later by single crosses between individuals.

The use of pollen mixtures, combined with the fact that pollination fertilizes a great number of ovules, always results in nonuniform seed.

Cross-1ncompatibility and self-1ncompatibility can occasionally be overcome by the use of growth hormones. According to Dr S. L. Emsweller, the ovary is treated with growth hormones previously dissolved in lanolin. One per cent of naphthalene acetamide is dissolved in previously heated and liquefied lanolin until the mixture turns dark in colour. The cold fat is applied to the base of the ovary, which in trumpet lilies can only be reached by first removing one of the petals.

This technique, limited in use and impracticable for large-scale work, induced L. longifloruni Croft and some other lilies to set seeds through self-fertilization.

Pollination technique

Crossing lilies artificially is not difficult; the flowers and reproduction organs are large and accessible. The style, topped by the stigma, grows from the ovary and emerges through the centre of the open flower. Around the ovary and the stigma stand six filaments, each one with an anther pivoted at the top which contains the pollen. If the flower opens during the early morning, the anthers are still closed, and remain so until warm sunshine releases their pollen. At the same time the secretions of the stigma are brought into action, making it ready to receive the pollen, I.e., receptive for the pollen tubes to grow into it.

Insects, wind, and physical shaking of the plant are the natural means of pollination.

Artificial cross-pollination should be timed to take place earlier than it would do in the natural way – this ensures that only the chosen pollen is used for fertilization, and so prevents stray and undesired pollen and the plant’s own pollen from spoiling the cross.

Most lily-breeders use their own preferred and often personally evolved pollination method. The following describes my own technique of controlled pollination:

The flower winch is to receive the pollen is first emasculated, that is, the still-closed anthers are removed with tweezers before or as soon as the flower opens, usually during early morning. An anther is next removed (again with tweezers, but now they must be previously disinfected with alcohol) from the partner plant and used to dust the stigma of the first flower. To avoid further and unwanted cross-pollination, the dusted stigma is protected with a small aluminium foil hood (chocolate wrapping formed around the point of a pencil), and gently sealed with finger pressure around the style. The cross (mother and father) is entered in the breeding book and preceded with the number of the cross, e.g.

No. 147 18.6.1966 L. martagon cattaniae x L. hansonii

A label bearing the same number is fastened to the stem of the pollinated flower. Paper labels are soon spoiled by the weather, and do not provide as permanent a record as the foil of an empty toothpaste tube which has the number ‘engraved’ on it by a typewriter.

Two to three weeks after the flower has faded, the ovary begins to swell, becomes erect, and so confirms that fertilization has taken place.

Other breeders use fine paint-brushes or pipe-cleaners for the transfer of pollen, but risk contamination from unwanted pollen unless a new instrument is used for every cross. The Americans use artificially produced drinking straws, instead of foil hoods, for the protection of the pollinated stigma; straws of various diameters are necessary if the stigma, particularly big in trumpet lilies, is not to be damaged.

The pollen employed for a cross can either be freshly gathered from a flower or be a few days or even weeks old provided it has been air-dried. Pollen keeps for a month if stored in a desiccator at 28-6 deg F (—2 deg C). Lily pollen packed in glass tubes or aluminium foil travels well, and facilitates not only the crossing of home-grown lilies with those overseas but also the cross-fertilization of lilies of different flowering periods.

Other methods are open to breeders who wish to cross early and late-flowering lilies. Certain varieties are greenhouse-forced for early flowering like their partners; in the case of others, the bulbs are kept in cold storage for longer than the usual period in order to delay flowering.

If breeding is to be successful, several essentials must be observed: cleanliness during pollination, use of flowers only just opened, rich cover of pollen on the stigma, pollen uncontaminated with that from other flowers, cleanliness of pollen storage containers and their correct labelling, and complete and safe protection of the stigma to prevent ‘foreign’ pollen from reaching it. The penalty for insufficient attention to these important details is wasted time and seedlings unrepresentative of the original cross.

Purpose of artificial pollination

Artificial pollination achieves two aims. Firstly, it ensures the true reproduction of a lily species without being spoiled by ‘foreign’ pollen. Secondly, it enables hybrids to be produced by crossing two different lily species or varieties. The second possibility, called hybridization, opens the door to an influx of a host of lilies of various colours and shapes which have been unknown and which grow better in cultivation than their wild forms.

The characteristics inherited as a result of a cross are as per Mendclian segregation. The first cross between two different lilies produces fairly uniform plants (FJ). Back-crossing of the Fj generation with their parents or their Fj brothers and sisters produces a great number of varied and interesting form and colour variants (F2). Generations obtained from further crosses are termed F2 , F3, and so on (F stands for filial).

The back-crossing (one or more times) of the Fj generation with their parents increases and strengthens the desired characteristics in the following generations.

The art of the breeder lies in the selection of plants which through further breeding produce the characteristics which he aims at and which he hopes to strengthen and to fix (selection), or to breed a lily which combines the many and various attributes obtained from crossing differently constituted types (combination).


We know that lilies normally have 24 chromosomes, a pair each of 12 types, contained in each cell nucleus. We call such a plant a diploid. In 1935, four American research workers, A. F. Blakeslee, A. G. Avery, B. R. Nobel, and M. L. Ruttle, were successful in doubling the number of chromosomes in each cell nucleus by treating seed or seedlings with a colchicine solution. Colchicine is a highly poisonous alkaloid obtained from the autumn crocus (Colchicum), which acts only on dividing cells or tissues in growth. The action of this poison during the cell division, and especially during division of the chromosomes, ensures the successful doubling of their numbers. If during subsequent cell-splitting the double number of chromosomes can be retained, a tetraploid is formed.

To ensure the production of tetraploid plants from even a single tetraploid cell, it is essential that the colchicine contacts the growing point as cell division takes place there, as indeed does the whole growing process. If other cells are treated, it could happen that only one leaf or even only part of one leaf develops tetraploidy. Treatment with colchicine is therefore a shot in the dark, a shot which very seldom hits the target.

A tetraploid lily has twice as many chromosomes as a diploid plant, namely 2 x 24 = 48 chromosomes. The difference is noticeable in the bigger and thicker leaves, the taller and stronger stems and the blooms, which are appreciably larger with significantly improved flowers -exactly what the breeder aims for, stronger plants with bigger blooms.

Several tetraploids with noticeably larger and more vigorous flowers have already been bred from L. longiflorum (Easter Lily). S. L. Emsweller has treated Ins diploid hybrids Cavalier, Mega, and Brandynrine with colchicine, and produced significantly stronger tetraploids.

The doubling of chromosome numbers is also used to induce fertility in sterile hybrids. It usually makes the plant fertile and set seed, if dusted with the pollen of another tetraploid plant. The pollen from tetraploid plants is bigger than that of diploid lilies, and so are the stomata on the undersurface of leaves. Crossing diploid lilies with tetraploids can also result in seed being produced, which when grown produce triploids, usually sterile, of 36 (3 x 12) chromosomes.

To induce polyploidy artificially, previously separated bulb scales are treated with colchicine. To stimulate cell division, the bulb scales are first stored under warm conditions for several days. The broken edge of the bulb scales, where a callus always forms, is immersed in a colchicine solution (0-2 per cent) for approximately two hours. After drying and dusting with a fungicide, normal bulb-scale multiplication methods are followed under damp and warm conditions. Treatment with colchicine, as already mentioned, is a shot in the dark; comparatively few targets are hit – less than 1:100. To make sure of securing some success, it is advisable to treat as many bulb scales as possible.

In addition to bulb scales, sprouting seeds or seedlings can also be subjected to colchicine treatment. An exposure of between two and 10 hours in a 0-2 per cent colchicine solution has proved the most suitable.


A basic knowledge of inheritance is essential to the serious breeder. In fact, it is only during the present century that science has begun to grapple with the problem and discovered the chromosomes in the cell nucleus (and the genes in the chromosomes) which are the carriers of inherited characteristics. Each gene is responsible for the transmission of a particular characteristic to the future generation. The almost countless number of characteristics collectively make up the picture of the plant and influence every part of it. Examples are: shape of bulb, plant and flower; number and colour of bulb scales; pendant, erect or horizontally held flowers; small, broad, pointed or blunt petals; colour of flower -red, dark-red or mahogany-red; colour of the centre; speckles on the flower petals – their colour, size, number, pattern of distribution – and so on. The number of these and other characteristics reaches into many thousands, and also includes the less visible attributes of a plant, such as ability to withstand disease, susceptibility to frost, preference or dislikes for certain types of soil, etc.

At least one gene is responsible for each characteristic, and often more than one, as is the case in increased intensity of colour.

The breeder must learn to recognize and differentiate between even the smallest characteristics, whether advantageous or disadvantageous, and to try and select parents for further crosses so that they are always superior to previous generations. These small, almost imperceptible changes often lead to greater overall improvement than a spectacular colour variation or other big change: for instance, the length of pedicels (long pedicels give the plant an air of elegance), or if petals are flat and broad instead of reflexed (flowers with flat, broad petals seem larger). Details such as these enable new crosses to supersede older varieties and enhance their value – but only if they lend themselves to comparatively trouble-free multiplication. A most beautiful lily may be capable of taking many prizes at flower shows but be commercially useless and unlikely to ever reach the stage of distribution, if for any reason it is only capable of slow and therefore costly multiplication.

Breeding necessitates keeping meticulous records and numbering, in sequence, every seedling, otherwise it is quite impossible to establish and to review, with hindsight, from which parents desirable characteristics have been inherited and, even more important, to eliminate those which have only contributed negative characteristics.

Breeders devote particular care to the raising of seedlings, as undue losses nullify much of the laborious work of crossing. For example, E. Debras obtained only two seeds from his successful cross L. sargentiae x L. henryi. Although both seeds germinated, one seedling was later lost; if both had succumbed we would probably have had to wait for many more years before obtaining a hybrid likcL. x anrelianense.

It is possible that freshly harvested seeds from a cross are not capable of germination even though they may contain embryos. Dr S. L. Emsweller found this when crossing L. speciosum with L. auratum, but-solved the problem by removing (from the same cross) the embryo from the still-green seed. The embryo was treated like orchid seed and germinated in previously sterilized test-tubes filled with an agar-agar growth medium. The lack of viability in seeds of this cross was due to growth inhibitors forming tumours which, in turn, killed the germ; subsequent analysis established that the growth inhibitors consisted of three organic acids which can be partially removed by washing the seed for 15 hours in continuously changing water.

Hybridization is undoubtedly the most interesting part of lily-breeding. The lifelong work of the man who is the best-known lily-breeder not only in the United States, but in the world, Jan de Graaff, demonstrates what remarkable success can be achieved. Not only has de Graaff produced lilies less demanding in their individual requirements which grow well in gardens under a wide range of conditions; he also dominates the world lily market through planned and scientifically controlled large-scale breeding.

The enthusiast can also create new lilies. However many beautiful lilies there may be in our gardens, it is difficult to contain one’s patience for the first home-crossed lily to open and to sec it with amazement one sunny morning for the first time – face to face! The feeling of having created something new gives real satisfaction.

The breeding on a wide front as practised by Jan de Graaff, the largest commercial producer, has advanced tremendously. But we must also be grateful to nonprofessional lily enthusiasts for both the small advances and indeed for some of the big forward steps in lily-breeding. E. Debras, of Orleans, is to be thanked for L. x aurelianense – our gardens would be very much poorer without it and the many thousands of its varieties. Australia’s Roy M. Wallace bred the unique Auratum-Speciosum hybrid Jillicw Wallace, which combines the beauty and majestic colour of two Japanese lilies, and excited world-wide admiration among breeders. Professor Patterson, of Saskatoon, Canada, was the first to cross successfully L. cernuuni with the Davidii-Umbellatum-Elegans complex and thus brought new colours and pastel shades of mother-of-pearl into the colour range of lilies.

Hundreds of breeders set themselves the task of producing a completely yellow Regal Lily, but none succeeded until L. N. Freimann, of Bellingham, United States, raised one in his small garden. His professional knowledge of breeding and selection, a sure intuition, the will to succeed, and the patience of the born breeder, combined with many years of continuous crosses, helped him to produce Golden Regal after 14 years of work!

The potential of new crosses is vast, and certainly not yet exhausted. The hereditary characteristics dormant in lilies make countless new combinations possible. Amateur breeders have thousands of possibilities open to them and the chance, through controlled and purposeful breeding, of producing a lily which is as good as any now available; nor will the more serious and scientific find any difficulty in satisfying their creativity in this field.

Sorry, comments are closed for this post.