Third International Conference on Genetics (1906) 396-400


GEORGE KERSLAKE, of Sydney, Australia.

PROBABLY there are few countries which offer better natural climatic advantages for cross-breeding plants than New South Wales. The range of subjects that can be grown without the aid of expensive glass structures is very large, though at some seasons difficulty is experienced in procuring a supply of pollen, on account of the dryness of the air and hot winds, which frequently render it impotent in a few hours. However, the difficulty is easily overcome by artificial development in a more suitable atmosphere.

In relating a few incidents in my experience here, I do not know how far my methods will agree with those practised in other countries, as in a great measure I have been playing a lone hand in this far-off country, having few of the advantages of intercourse with those of similar tastes and inclinations, which the more populous centres naturally afford. It is my intention to only casually refer to a few peculiarities noted in connection with some of the popular races of flowers, such as the rose, chrysanthemum, dahlia, &c., which have engaged my attention for many years, as in my opinion they offer very few difficulties to any intelligent operator. It is when we get off the beaten track of those subjects which have been crossed and recrossed for a considerable time, that problems so frequently occur which theory may explain.

But their practical solution is quite another matter, for the simple reason that without practical experiment we are ignorant as to whether we are working in a grove which nature will respond to, or one which will defy our most persistent efforts under all circumstances. A case in point may be of interest. A number of blooms of Opuntias were emasculated, and as the period is short during which the unopened flower can be safely operated upon (about ten hours) before the stigma arrives at the receptive stage, a fine jet of water under moderate pressure is directed into the flowers, to remove the stamens after being severed, which is rather difficult in the immature flower. This method is perfectly safe and effective in this dry climate, and is adopted with most Cacti, excepting Epiphyllums and a few others that flower in winter.

In the case of Opuntia Piccolomini x Cereus Spachianus and of O. Piccolomini x Phyllocaclus Schlimmii, the union was complete in every instance; while O. elatior and O. senilis, operated upon with the same male parents and under the same conditions, resulted in an absolute refusal. A certain amount of irritation or swelling of the fruit was noticeable for a short time, but neither reached fertility. This behaviour is difficult to explain, as the plants were of equal age and had been fruiting regularly for some years. These experiments are most interesting as showing bow some members of the same family will at once respond whilst others positively refuse to be influenced by sexual intercourse.

Lilium tigrinum x L. elegans Wallacei resulted in every flower operated upon producing huge pods of seed. But when Lilium speciosum album x L. tigrinum, and L. speciosum rubrum x L. tigrinum, and L. tigrinum x L. speciosum album were tried, the results were very different, as the female organ in a very short space of time showed unmistakable signs of decay, and in a couple of days had quite withered. This was so not in one instance only, but the whole forty blooms used in the experiment showed the same symptoms. As evidence that this was no fault of the prospective seed parent, some later flowers were tried, L. speciosum album x L. speciosum rubrum, when they at once returned to fertility. The same thing occurred, only under less favourable conditions, with L. speciosum album x L. auratum, and the same x L. sulphureum. However, under the circumstances I am reluctant to conclude, without farther extended trial, that a union of speciosum and auratum cannot be effected, as so much depends upon what may be termed the seed-bearing mood of the plant, which is often absent when the reproductive organs show the most perfect development.

The family of the amaryllids undoubtedly offers a wide field for investigation. In 1903 a rather extended trial was made to induce a union between Crinum Moorei and Crinum yemense x Vallota purpurea, and Brunsvigia x Vallota. At the very commencement the indications were not encouraging. However, to make the tests conclusive, they were continued through the whole flowering season, more especially to see whether any atmospheric conditions prevailing during the time would be more favourable for inducing fertility than others: bright and hot days; cool, dewy and cloudy; the cool atmosphere of early morning; the dry heat of midday; and the cool dry evening air, were all alike in failing to produce what appears to me to be a forbidden union.

Crinum yemense x Hymenocallis macrostephana crossed without any difficulty, as did also Brunsvigia Baptisii x Lycoris aurea. But the most interesting was that effected between an unnamed variety of Hippeastrum and Agapanthus umbellatus. The Hippeastrum flowers here in October and November, but in favourable seasons a stray scape is again thrown up in January and February. It was one of these that made this cross possible; otherwise the pollen of Agapanthus would have to be preserved for a period of ten or eleven months, which is far too long. There were four flowers on the scape when they were pollinated with Agapanthus umbellatus: of these four, one was injured by being too early emasculated, a week before they reached the receptive stage (probably through being produced in the off season). Two flowers perished, but the remaining one produced a full pod of seeds. At this stage, though the seeds were of normal appearance I doubted their fertility. However, they were sown, and nine plants resulted. Three died in the seed pan; but six still remain, which have not yet reached the flowering stage, but are strong and healthy and appear to be evergreen. This being a case of overlapping botanical divisions I never intended divulging it until the plants had flowered. But a recent report that a gentleman near Sydney had effected the same cross has led me to believe that there is nothing remarkable about it.

I will now briefly refer to the unreliability of sports, as male parents, to impart colour in the chrysanthemum. One or two instances must suffice to illustrate what has proved invariably the rule in practice. In a general way colour can be well controlled in the chrysanthemum. Two varieties raised by me and well known in England may serve to explain my meaning. 'Lady Trevor Lawrence' x 'H. Cannell' resulted in 'Oceana,' and 'Edwin Molyneaux' x 'Stanstead Surprise' produced 'Australie.' This is a fair example of control of colour. But when 'Mme. Carnot' x 'Charles Davis' were used, there was not the slightest indication that 'Charles Davis' was the male parent, as all the resultant plants were either shades of deep mauve, or amaranth, which pointed conclusively to 'Vivian Morell' from which 'Charles Davis' had sported. 'Mrs. Barclay' x 'G. I. Warren' behaved much in the same way as if 'Mme. Carnot' had been the pollen parent. Many other instances could be mentioned which show more conclusively than in any other way that these sportive characters, at least in the chrysanthemum, are only superficial changes which are incapable of being imparted to another. The same may apply to the numerous sports among roses; but as far as my experience goes it cannot be so easily proved, as with very few exceptions the rose appears to be the most uncontrollable, as regards colour, among any of the popular flowers that I have manipulated.

I now pass on to the Bouvardia. It is now more than ten years since I commenced to manipulate the Bouvardia, which, by the way, does remarkably well in this climate, and treated as "cut-backs" with good culture frequently attains the height of six to eight feet in one season, and flowers profusely for at least six months in the year. I thoroughly believe in an hybridist or plant-breeder having preconceived and definite aims, well thought out, before any action is taken. This involves the choice of parents, which is perhaps the most important factor in assuring success in any future operation, whether it concern fruits, flowers, or vegetables. To a very large extent, on a female parent that can be relied upon to produce good stock, hinges the success or otherwise of future crosses. And when one is found (and there are not too many), work it in all directions, until another is proven to be better. This does not always involve the highest type in these respective kinds, as we shall presently see in reference to the Bouvardia. In going through the best cultivated varieties of Bouvardia I was struck with the very large sweet-scented flowers of Humboldtii corymbiflora in comparison with such varieties as 'President Cleveland,' Hogarthii, Vreelandii, 'Priory Beauty,' and others of similar size and form. My aim was thus fixed to get the large flowers of H. corymbiflora, combined with the larger truss and more free-flowering qualities of the smaller flowered section. With this object in view I chose H. corymbiflora as the seed parent on account of its naturally seeding freely, which very few of the other varieties do at all. I do not say they are sterile, as the composition of the soil and environment are such important factors in bringing about fertility, that it is a difficult matter to decide when anything is definitely sterile. For five consecutive years I crossed H. corymbiflora with all the best commercial varieties known to me. Large numbers were raised and tested every year, with the result that I was surrounded with a host of mere weeds, none being equal to any of the parents used. At this stage I ban to seriously consider whether to abandon the enterprise or adopt some other course to gain my object. The plants previously raised were carefully scrutinised, and one white variety was found, which was raised three years previously, which seemed promising; and as it was already bearing seed this decided it to be the variety to operate upon, as I consider it is waste of time to work upon anything that shows no natural disposition to fruit. This variety, though of no commercial value, was named 'Progenitor' merely for the sake of identification. It proved a difficult variety to emasculate, as the stamens were set deeper in the tube than any variety I have met with; it necessitated laying open the tube nearly down to the junction; a delicate operation as the pistil will not bear exposure without injury, even when fully developed. 'Progenitor' was crossed the first season with 'Priory Beauty,' 'President Cleveland,' Hogarthii, and 'Laura.' About one hundred plants were raised from the seeds obtained from these crosses, and eighty reached the flowering stage the following autumn. The control of colour was more complete in this instance than in any other of the numerous crosses my notes record. As I have before mentioned, the seed parent was white, but not one white variety appeared in the whole batch of plants. Those pollinated with 'President Cleveland' and Hogarthii were various shades of red, and those with 'Priory Beauty' shades of pink, while the 'Laura' influence resulted in varied shades of salmon. The blooms of most were very large, and in some instances the trusses were immense, while a good many showed a leaning towards the sparsely flowered heads of H. corymbiflora. As most of the best still remain here, I can only point you to one which has recently been distributed in England, viz. 'King of Scarlets,' which was among the first batch of meritorious varieties raised. Nearly all produced the sweet jasmine perfume which characterises H. corymbiflora, though in 'King of Scarlets' there appears to be a departure, as in it there is a decided leaning towards vanilla.

By the above remarks on the Bouvardia I have no doubt the hybridist will perceive a deep object lesson, insomuch that a revolution may be effected when all outward appearances are most discouraging. In the above instance there was three years' loss of energy by persistently following one course, instead of following up any little break of character at first obtained, as this is clearly a case that needed the second generation to accomplish what was only partially done by the first. And the first step in this instance had more the appearance of a retrograde character than an advanced stage. More recently these crosses have been repeated, and have brought about a notable increase in the segments from the normal four to five and six, which gives the flower a rounder and much fuller appearance. I may mention one exceptionally fine form named 'Magnificent' which has outdistanced all others. This usually has six segments, and as the pips are from an inch and a quarter to an inch and a half across, and as about fifty of these go to make an ordinary truss, its appearance at a casual glance is most unlike any Bouvardia known in cultivation. However successful one may be in one direction, failure is sure to follow the plant-breeder in another, or at least such is my experience. For many years past I have been repeatedly trying to obtain an improved yellow Bouvardia, but I have failed completely to bring about the fertility of B. flava or to impart its colour to any other.

In concluding this paper I can only express an oft-occurring thought, how little we know by actual experiment in this vast field of research, as to where Nature is willing to open her rich storehouse to the hand of man, and where, on the other hand, she effectually defies any intrusions, simply because we are ignorant of her natural ways, and have in a certain sense to grope about until we find them.

Bouvardia Humboldtii corymbiflora