Bradley, about a century later, in 1718, is believed to have been the first author who speaks of the accomplishment of cross-breeding, which he describes as having been effected by bringing together the branches of different trees when in blossom. But the gardeners of Holland and the Netherlands were the first to put it into practice.[8] The following extract is given to explain the manner in which Mr. Knight conducted his celebrated experiments on fruits, which rewarded him with some varieties that were highly esteemed:--"Many varieties of the apple were collected which had been proved to afford, in mixtures with each other, the finest cider. A tree of each was then obtained by grafting upon a Paradise stock, and these trees were trained to a south wall, or if grafted on Siberian crab, to a west wall, till they afforded blossoms, and the soil in which they were planted was made of the most rich and favorable kind. Each blossom of this species of fruit contains about twenty chives or males (stamens,) and generally five pointals or females (pistils,) which spring from the center of the cup or cavity of the blossom. The males stand in a circle just within the bases of the petals, and are formed of slender threads, each of which terminates in an anther. It is necessary in these experiments that both the fruit and seed should attain as large a size and as much perfection as possible, and therefore a few blossoms only were suffered to remain on each tree. As soon as the blossoms were nearly full-grown, every male in each was carefully extracted, proper care being taken not to injure the pointals; and the blossoms, thus prepared, were closed again, and suffered to remain till they opened spontaneously. The blossoms of the tree which it was proposed to make the male parent of the future variety, were accelerated by being brought into contact with the wall, or retarded by being detached from it, so that they were made to unfold at the required period; and a portion of their pollen, when ready to fall from the mature anthers, was during three or four successive mornings deposited upon the pointals of the blossoms, which consequently afforded seeds. It is necessary in this experiment that one variety of apple only should bear unmutilated blossoms; for, where other varieties are in flower at the same time, the pollen of these will often be conveyed by bees to the prepared blossoms, and the result of the experiment will in consequence be uncertain and unsatisfactory." * * * In his Pomona Herefordiensis, he says:--"It is necessary to contrive that the two trees from which you intend to raise the new kind, shall blossom at the same time; therefore, if one is an earlier sort than the other, it must be retarded by shading or brought into a cooler situation, and the latest forwarded by a warm wall or a sunny position, so as to procure the desired result."