With reference to garden roses, the new multiflora hybrids open out a wide field for experiments, especially as there are indications of a fresh break-perpetualflowering pillar roses, Thalia Remontant and Trier being the first. Cross the Multifloras and Wichuraianas with Hybrid Perpetuals and Hybrid Teas, and should the first results fail to produce perpetuals, do not be discouraged but cross these seedlings, and in their offspring our object may be attained.

Treatment of Seed.-A seed is a reproductive offspring of the parent plant, possessing within itself a germ of life which, after arriving at maturity, may remain dormant for years, or may be called into activity at the earliest possible moment. Dryness delays this activity, moisture hastens it. The cause of inactivity is the superabundance of carbon which, while serving as a preservative of life, prevents that life from development. Before the seed can grow the greater part of the carbon must be thrown off in the form of carbonic acid gas, and to effect this a supply of oxygen is necessary. But the seed cannot obtain sufficient oxygen to do this because it is enveloped in pulp and encased within the thick skin of the hep, so the first thing to be done as soon as the hep is ripe is to liberate the seeds from the hep, and either sow at once the better plan or keep the seeds moist in damp fibre or sand for a time. Mice are passionately fond of rose seed; it is therefore not advisable to sow it in the open ground, but in pots under glass, where it will be better protected from mice and other foes. Each seed should be sown separately, not sprinkled or rough

cast over the surface, but laid at equal distances from the others, and then covered with half an inch of soil, and the pot labelled with the registered number.

The soil should be light and well drained, for aeration is essential; it should not be allowed to become hard or caked either from insufficient moisture or from being watered too heavily. What has been previously stated in the chapter on soils with reference to the action of water as the messenger boy holds good here; water conveys oxygen and hydrogen to the seed by which the excessive carbon is speedily liberated. When a sufficient amount of this has been thrown off the seed germinates, heat sets the vital parts in action, and the seed proceeds first to throw a radical or rootlet downwards, and then to send a stem upwards, which, pushing through and above the soil to about 2 inches, terminates in two rudimentary leaves not unlike mustard and cress, but bearing no resemblance whatever to rose foliage.

At this stage great care is necessary to prevent damping off, and to guard against attacks from such pests as slugs, woodlice, green-fly, and mildew. Should the seedlings come up too close together it is advisable, after they have developed the third leaf, to prick them out separately into small pots.

Treatment of Seedlings.-Seed sown in the autumn will, for the most part, germinate the following spring; some, however, will lie dormant for another year. As soon as the seedling has made its first growth, and is sufficiently ripened, no matter how small the wood may be, it should be cut up, and every eye budded;

so much the better if the seedling bears a flower bud, since by the shape of the calyx as well as the foliage we may hope to ascertain its value.

As to the stock: since the buds from seedlings will naturally be very small, it is not well to have the stocks too large; ordinary Manetti cuttings will probably be so; medium-sized brier cuttings or seedling briers are preferable. These stocks may be grown in pots so as to be ready for those seedlings which mature early, but by far the best plan is to grow them in the open, when they may be budded in the summer.

It, however, may be asked, why take the trouble to bud a seedling before you have seen it flower, or at any rate know whether it is worth propagating? The reason for doing so is this: the quality of a seedling cannot be accurately gauged in its early state. There are certain secretions necessary to its full perfection which are developed gradually as the seedling grows in age, and until this period is reached we can never be quite sure as to its possibilities. Naturally we must expect many failures, and some eminent raisers are satisfied if only two per cent. of the seedlings propagated turn out well, and once more let us remember that in the first crop of seedlings we are laying the foundation for further developments; it is from the seedlings of seedlings that we may look for fresh breaks, and therefore the least promising seedling may become a parent of most beautiful offspring.

CHAPTER XIV

GROWING FOR EXHIBITION

Staking and Tying Dwarf Maidens-Staking and Tying Standard Maidens-Thinning-Disbudding-Tying the Blooms-Shading.

As has already been stated, roses are classified under two distinct heads: (1) Roses whose value lies in the specimen bloom, and (2) roses wherein masses of flowers are the chief feature. In order to distinguish the one from the other, the roses of the former class are technically termed "exhibition" roses, and those of the latter "garden" or 'decorative" roses. But one must admit that this nomenclature is faulty, and to the uninitiated somewhat misleading, because both classes are exhibited at the rose shows, and a so-called "exhibition" variety is often equally suitable for the garden and house decoration. No true rosarian, whether he be an exhibitor or nonexhibitor, will be content with inferior flowers; his appreciation of the beautiful will not permit him to sacrifice quality for quantity. He will not therefore grow roses of the exhibition section and expect from them a mass of flowers; if he does he will certainly be disappointed. With reference to those varieties whose beauty lies in the individual blooms, the difference between the exhibitor and non-exhibitor is this: the chief object of the exhibitor is to have specimen flowers during the exhibition season-if the variety gives him

a second or third crop of roses so much the better, but it is not the first consideration; whereas the non-exhibitor desires first and last to have specimen blooms throughout the summer and as late in the autumn as possible. The point of difference then lies in the kinds to grow; the treatment in both cases will be the same. So let not the non-exhibitor imagine that the following observations do not concern him; both have the same object in view, the production of specimen blooms.

I. STAKING AND TYING DWARF MAIDENS

Dwarf stocks budded the previous year should be topped before the sap rises, say, early in February, to prevent bleeding. The stem of the stock should be cut through an inch above the dormant bud, and then as spring advances the buds will begin to swell and push; this is termed "breaking." As soon as possible after topping the stocks a stake must be affixed to each in readiness to support the shoot when it begins to grow. For the first twelve months the shoot of a newly budded rose, the maiden growth, is weak at the junction of bud with stock, and unless artificially supported will come off the stock. The best stakes are bamboo canes 4 feet long and inch in diameter; they last longer and are neater in appearance than wooden stakes. If the stocks have been planted in trenches, as already advised, the roots for the most part will be spreading the same way, and therefore the bamboo should be set close to the back of the stock away from the roots to avoid injuring them. As soon as the buds after breaking have grown about