Van Mons' theory was based upon certain assumptions and observations, some of which are well founded, others are not so firmly established. He claimed correctly that all our best fruits were artificial products, because the essential elements for the preservation of the species in their natural condition, are vigor of the plant and perfect seeds for the perpetuation of the race. It has been the object of culture to diminish the extreme vigor of the tree so as to produce early fruitage, and at the same time to enlarge and to refine the pulpy portion of the fruit. He claimed, as a principle, that our plants of culture had always a tendency to run back toward the original or wild type, when they were grown from seeds. This tendency is admitted to exist in many cases, but it is also claimed, that when a break is once made from the normal type, the tendency to improve may be established. Van Mons asserted that the seeds from old trees would be still more apt to run back toward the original type, and that "the older the tree, the nearer will the seedlings raised from it approach the wild state," though he says they will not quite reach it. But the seeds from a young tree, having itself the tendency to melioration, are more likely to produce improved sorts. He thinks there is a limit to perfection, and that, when this is reached, the next generation will more probably produce bad fruit than those grown from an inferior sort, which is on the upward road of progression. He claims that the seeds of the oldest varieties of good fruit yield inferior kinds, whereas those taken from new varieties of bad fruit, and reproduced for several generations, will certainly give satisfactory results in good fruit. He began with seeds from a young seedling tree, not grafted upon another stock; he cared nothing for the quality of the fruit, but preferred that the variety was showing a tendency to improvement or _variation_. These were sowed, and from the plants produced, he selected such as appeared to him to have evidence of improvement, (it is supposed by their less wild appearance), and transplanted them to stations where they could develop themselves. When they fruited, even if indifferent, if they continued to give evidence of variation, the first seeds were saved and planted and treated in the same way. These came earlier into fruit than the first, and showed a greater promise. Successive generations were thus produced to the fourth and fifth, each came into bearing earlier than its predecessor, and produced a greater number of good varieties, and he says that in the fifth generation they were nearly all of great excellence. He found pears required the longest time, five generations; while the apple was perfected in four, and stone fruits in three.