If we were to take many books about trees and shrubs or general gardening as a guide, one might be led to think that only one way of increasing a tree or shrub existed, and that by grafting; but, as we have pointed out elsewhere, it is a mischievous practice when indiscriminately applied. It is _not_ contended by this that grafting and budding are utterly needless, as in many instances these methods may be rightly adopted, but the four natural ways of increase are by layers, seeds, suckers, and cuttings. Many trees and shrubs are much better when grafted upon other stocks.
Of these, practise seed-raising whenever possible; but if seeds cannot be procured, then adopt other ways, and the man is wise who tries to keep a plant on its own roots. Neither budding nor grafting should be resorted to, unless other means fail absolutely. When standard trees cannot be got true from seed, budding or grafting must be practised, and the evils of these methods of propagation are not so pronounced in such cases as with dwarf plants. With the former, suckers, or growths from the stock, are easily seen as soon as they appear, but with dwarf plants a perfect forest of suckers may seriously weaken the plant before they are noticed.
SEEDS.--These can be sown at almost any time, but the spring is the best, as those which germinate quickly have time to form strong young plants before the following winter. Some take two years to come up, and should be left in the ground. This refers more to seeds sown outdoors, and few hardy trees and shrubs require heat to assist germination. When sown in the open the beds should be made on a fairly rich, moist piece of ground, protected from cold winds, but fully exposed to the sun. After the seeds are sown, cover them with light tiffany shading, fir branches, or heather, but the first is best, as it is easily removed to attend to the bed. Conifers especially should be sown in beds, whether indoors or outdoors, as pot-culture results in the roots taking the shape of the pot, and never afterwards recovering from their cramped condition. It must be remembered, however, that varieties cannot be depended upon to come true from seed, though by careful selection for a few years many varieties will almost reproduce the characteristics of the parents. Hybrids, such as _Berberis stenophylla_, _Hypericum moserianum_, and many others, also do not come true from seed, so that cuttings, layers, or division of the old plants must be the practice chosen.