ascertain this, we might then work by rule. It is the opinion of some Vegetable Physiologists that the offspring assumes the foliage and habit of the male, while the flowers are influenced more by the female parent.* These may be the rules, but there are exceptions to them; and it would appear that there is nothing yet made known that can be taken as a correct guide in the matter. But if, in hybridizing, the operator follow the dictates of his own reason, and closely watch the results of his labour, he will, in all probability, not absolutely fail, and be at length enabled to found a theory of his own. Until he has done this, he must be content to work by the light of others, or grope his way in darkness.

That certain cultivators have acquired by practice sufficient knowledge to enable them to attain almost to a given object, is my firm belief; and this is founded on the frequent appearance of the kind of plant, or an approach to it, that has been pronounced a desideratum, A case occurs to me which will serve to illustrate this remark. Until lately we had no very dark or very light Bourbon Roses; nearly all were of a rose or lilac hue; but there was a cry raised for dark ones. Proserpine first arose, and by her beauty captivated every beholder. Next came Paul Joseph, darker still, and still more beautiful. Then it was noised abroad, and proved true, that one raiser was in possession of several very dark varieties, and some pale-coloured ones also appeared about the same time. Now what inference can we draw from these facts, when we consider that similar varieties proceeded from different quarters, unless it be that the skill of the cultivator was directed, and that successfully, towards originating them.

The dark Bourbon Roses, of which we have just spoken, are those introduced to England in 1843, under the names of Souchet, Charles Souchet, Dumont du Courset, Gloire de Paris, Princesse Clémentine, Souvenir du Dumont d'Urville, and Comte de Rambuteau. These were all raised from seed by one individual. I remember visiting the establishment of M. Souchet at Paris, where, alone, these Roses were to be seen, in the summer of 1842; and although a violent thunderstorm had just passed over the city, producing the usual consequences to the denizens of the garden, I could see from the wreck they were a splendid lot of Roses. There were at that time twelve varieties, the one a light-coloured one (Madame Souchet), but nine only were introduced to England. What became of the other three I could never learn. Probably they proved of little merit, and were therefore not offered to the public.

Now, with such an example as this before us, need we sigh over the improbabilities of improving or extending the range of colour in any other class of Roses? Surely not. Will not the same skill which produced dark and light Bourbon Roses prove adequate to any future reasonable demand?

But the plants are in flower, and there is no further time for talking: we are

* See Theory of Horticulture, p. 330.

now called upon to act. Keeping in mind the points recently advanced, let us next inquire, What is there desirable among Roses that we do not already possess? The answer to this question will be, the things we should endeavour to obtain. But shall we be satisfied with merely crossing the varieties, and gathering and sowing the seeds indiscriminately; or do we wish to know the results of the turn we are seeking to give to the workings of nature? I think this knowledge is desirable, and it certainly heightens the interest of the work. To carry it out with little trouble, it is a good plan to obtain some thin sheet lead, and cut it into strips an inch long and a quarter of an inch wide. On these may be stamped figures, from 1 to an indefinite number, and, as each flower is crossed, one of these numbered leads is wound round the flower-stalk. The number is then set down in a book, and the name of each parent, with the object in view, are entered opposite the number. Here is an extract from my note-book of 1846.

In No. 17 fulness and colour are the points to engage our attention. The colour of the Copper Austrian Rose is distinct and beautiful, but the flower is single. I want a double one. The Harrisonii is double, and nearly allied to the other. I choose it for the female parent, because it is the best seed-bearer. This seems to me the most reasonable means to pursue in order to accomplish this end.

In No. 21 the female parent, Général Allard, is a model in form. I am satisfied with the colour, but it is a delicate grower, except when young, and not a free autumnal bloomer. I am seeking to remedy these defects, and cross with Madame Laffay, which is nearly of the same tint, and has the desired properties, deficient in the other, abundantly developed.

The subject of No. 42, La Reine, is an extraordinary Rose. I hybridize it with Du Luxembourg, with the view of obtaining a large, red, globular-shaped Moss Rose. As one parent here is an autumnal bloomer, there is also a chance of some of the offspring becoming such.

But we may proceed from individuals to classes. Two very desirable classes of Roses in prospect are Hybrid Moss and Perpetual Moss; and I look more to the hybridizing of the species for future improvements of the Rose, than to mere cross-breeding. The latter has already been pushed so far that fresh sources must

be opened before any thing great and new can be accomplished. Thus it is, that while we view as doubtful certain things talked of, we hail with delight the prototypes of the Hybrid Moss and Perpetual Moss, which already appear in the horizon. We have some; and what appears the most reasonable means to pursue to increase their number? The Hybrid Moss, it would seem, may be obtained by hybridizing the Hybrid Chinese and Hybrid Bourbon with the Moss kinds, or vice versâ; the more double and mossy the parent Moss is the better, that the offspring may produce full flowers, and not lose the mossy characteristics.

The Perpetual Moss would appear easiest obtainable by hybridizing the Hybrid or Damask Perpetual with the varieties of Moss, or vice versa, using the Perpetual Moss kinds already obtained, on either side, according to whether they produce seeds or pollen.

Of other Roses wanted may be instanced striped Hybrid Perpetuals, which may probably be obtained by bringing the most constant flowering varieties of that group in union with the Rosa Mundi, or any of the striped French Roses. Then there are no striped Hybrid Chinese or Hybrid Bourbon Roses, which we should expect to obtain from the union of the striped French with the varieties of Bourbon or Chinese. There is no striped Moss Rose worthy of the name. Might not such be obtained by working various of the Moss kinds with the Rosa Mundi? There are no striped autumnal Roses; and few autumnals of growth sufficiently rapid to form high pillars. Here is a field for experiment! Is there not a fair chance of working out the former by crossing the striped summer kinds with the freest-blooming autumnals? and of obtaining the latter from between the most vigorous growers of both seasons? We hear talked of Yellow Moss and Yellow Bourbon. As an attempt to obtain the former, cross the palest Moss Roses with Emerance (Provence); for the latter unite the yellowest tints of the Tea-scented with the buff and most colourless Bourbons. But we have not space to pursue this argument further. We need not confine ourselves to the instances above quoted: they are merely given as examples. The work may be varied ad infinitum any thing that reason may suggest the head and hands may work for.

It is desirable, before crossing or hybridizing, to see if the flowers about to be crossed have any stamens. If so, they should be cut away with a pair of roundpointed scissors, just as the flowers expand. It is not known for certain whether superfœtation can occur in plants, but it is well to provide against it, especially where it can be done with little trouble. The plan I adopt when crossing is, to bring a flower of the male parent to the seed-bearing tree. If the weather be calm, I cut away the petals of the former, holding a finger over a flower of the latter, upon which I strike the flower deprived of its petals. The sudden shock drives the pollen into the other flower, and the work is done. But if the wind be high, this plan will not do. It is better then to collect the pollen on the end of a camel-hair pencil, and convey it thus to the styles of the other flower.

When the flowers have passed away we shall soon have the satisfaction of seeing the seed-pods swell; and as they ripen it will be well to devise some means to protect them from birds. I have no direct proof to adduce that they eat them; but I have often seen the greenfinch feasting on the Sweet-Brier hips, and should these, by any chance, become scarce, or fail, he probably might not object to this slight change of diet. It is important to leave the seed-vessels on the trees as long as possible, and they never should be gathered until quite ripe. It is not enough that they are red: they should, if possible, hang till they grow black. So soon as gathered, let them be buried in the earth, and, if they have been numbered, pack them separately in small pots, interring the whole in a large pot, or in the ground.

CHAPTER VII.

ON CLEANING AND SOWING THE SEED, AND THE SUBSEQUENT TREATMENT OF THE SEEDLINGS.

In the last Chapter we left the seeds stored up for winter, that drear season at which the Flower Garden presents but few attractions. Its denizens, stripped of their gay attire, are sunk in repose; often bound fast in icy chains; all vegetation sharing in one general imprisonment, waiting for the balmy breath of spring to release them from their slumbers, to burst forth endued with fresh life and vigour. It is for this latter state of things we must watch; for it is advantageous to sow the seeds in the earliest of spring. The end of February or beginning of March is a good time; but should the frost be too severe then to admit of the work being done well, we must delay the operation until a fitting opportunity occurs.

When removing the seed-pods from the boxes or pots in which they have been stored, the seeds should be rubbed out between the hands previous to sowing. Some pods that are hard, or were not fully ripened when gathered, will require stronger measures to separate them. These may be rubbed through a coarse wire sieve, the hard coating of the seeds protecting them from injury in the process; and if any are found proof against this, they may be crushed beneath a rollingpin, or slight taps of the hammer.

So soon as the seeds are broken up they should be laid out in the sun and air to dry; and when sufficiently dried it is easy to rid them of their pulp and external covering by sifting and winnowing in pans. The seeds then are sufficiently clean for sowing; and in what manner shall this be performed?

The French growers sow principally in the open ground, either in beds or in drills by the sides of walks; but M. Hardy, of the Jardin du Luxembourg, sows largely in frames. I remember seeing there, about four years since, a vast number growing in a wooden frame, the most of them apparently varieties of ROSA INDICA, and their healthful appearance and vigorous growth sufficiently attested the aptitude of the treatment.

If a frame cannot be spared for the purpose, the tender kinds, at least, should be sown in pans, thoroughly drained, and filled with equal parts of leaf mould and loam well mixed together. After the seeds are sown they may be watered, and covered over with about half an inch of the same soil, sifted, and mixed with