As in budding, the possible limits are not yet known. As a rule, close botanic affinity must exist between stock and scion, such as apple upon apple, pear upon pear, and plum upon plum, etc. But this rule is not wholly adhered to in practice. As an instance, in order to secure a dwarfed growth and earlier bearing, certain varieties of the pear are worked on quince roots. Also to secure a dwarfed habit and earlier bearing, certain varieties of the plum are worked on sand cherry {Prunus pumila) stocks. As to the need of both budding and grafting, it happens that budding will succeed often on stocks not wholly congenial where grafting fails to secure a good union. As an instance, the uncolored juiced varieties of the cherry seem to unite well with the wood of Mahaleb-cherry stocks. On the other hand, grafting seems to give a poor union and usually the scions fail to grow. Again, the cultivated apples seem to make a good union with stocks of the cherry crab (Pyrus baccata) by budding, while the root grafts usually fail to grow, and when they start growth the union is far from perfect.