From the foregoing discussion the reader will conclude that propagation by grafts or other mode of division will always reproduce the variety. This is so nearly true that the whole art of nursery propagation may be said to be based upon it. However, it is also true that under certain unknown conditions any one bud may change its character so as to give rise to a new variety sufficiently distinct to be worthy of a separate name. This is termed "bud variation," and is a prolific source of new varieties in certain plants, such as chrysanthemums and roses. Sometimes the change is only slight, and this may be termed a sub-variety. In Northwestern nurseries, for instance, two Wolf plums are now recognized, the freestone and the cling; it is not known, however, whether this is a bud variation or a seedling mixture. The Wealthy apple has apparently varied somewhat under propagation, and recently gave rise to remarks at the Minnesota State Horticultural Society meeting by the son of the originator, who distributed sprouts from the original tree to help settle the matter. Experienced fruit-growers have long noticed that in a large number of apple-trees of one variety in the same orchard some trees are more productive than others. The most advanced thinkers in this line now recognize this fact as due to bud variation, and take advantage of it by cutting scions for grafting from the best and most productive trees only. Florists long ago learned the necessity of care in taking cuttings from the best and most productive individuals, or even parts of the plant.