In raising seedling trees we cannot expect much more than a reproduction of the species, and not that of the parent tree. Plants that have been subjected to unnatural conditions and surroundings, as usual under cultivation, are far more likely to show a wider range of variation in the seedlings than those growing wild in their native habitats; but even the latter cannot be depended upon to reproduce exact types from seed. In other words, there is nothing certain about seedling nut trees; the large nuts may produce trees bearing very small ones, the early-ripening give late ones, the tall dwarf trees and the precocious fruiting some of the most tardy varieties; and yet, with all this uncertainty, we still think it best to select for planting the best nuts obtainable, _i. e._, best and most promising for the conditions under which the seedlings are to be grown. For the multiplication and perpetuation of choice varieties we must resort to artificial modes of propagation, mainly by budding and grafting. These modes, however, while the best at present known, are so difficult and uncertain in cool climates,--even in the hands of the most skilful propagators,--that grafted walnut trees have never been very plentiful in the nurseries of this or other countries with which we have commercial relations. In the south of France nurserymen appear to have been more successful in the propagation of walnuts by budding and grafting, than elsewhere; but in the northern provinces, as well as in Great Britain, we hear little of this mode of propagation. So difficult has this mode of propagating the walnut been considered in England, that Thomas Andrew Knight, president of the London Horticultural Society, early in the present century discouraged all attempts to propagate this tree by such means; but later, in a paper read before the Society April 7, 1818, he admits to having changed his mind, especially in regard to budding the walnut, and says: