Propagate fruits (grape, fig, strawberry) of various kinds; also ornamental plants. How long does it take them to root? Geraniums rooted in the spring will bloom in the fall. Do you know any one who selects seed potatoes properly? Make a careful selection of seed at the next harvest-time. SECTION XVIII. PLANT SEEDING In propagating by seed, as in reproducing by buds, we select a portion of the parent plant--for a seed is surely a part of the parent plant--and place it in the ground. There is, however, one great difference between a seed and a bud. The bud is really a piece of the parent plant, but a piece of _one_ plant only, while a seed comes from the parts of two plants. You will understand this fully if you read carefully Sections XIV-XVI. Since the seed is made of two plants, the plant that springs from a seed is much more likely to differ from its mother plant, that is, from the plant that produces the seed, than is a plant produced merely by buds. In some cases plants "come true to seed" very accurately. In others they vary greatly. For example, when we plant the seed of wheat, turnips, rye, onions, tomatoes, tobacco, or cotton, we get plants that are in most respects like the parent plant. On the other hand the seed of a Crawford peach or a Baldwin apple or a Bartlett pear will not produce plants like its parent, but will rather resemble its wild forefathers. These seedlings, thus taking after their ancestors, are always far inferior to our present cultivated forms. In such cases seeding is not practicable, and we must resort to bud propagation of one sort or another. While in a few plants like those just mentioned the seed does not "come true," most plants, for example, cotton, tobacco, and others, do "come true." When we plant King cotton we may expect to raise King cotton. There will be, however, as every one knows, some or even considerable variation in the field. Some plants, even in exactly the same soil, will be better than the average, and some will be poorer. Now we see this variation in the plants of our field, and we believe that the plant will be in the main like its parent. What should we learn from this? Surely that if we wish to produce sturdy, healthy, productive plants we must go into our fields and _pick out just such plants to secure seed from as we wish to produce another year_. If we wait until the seed is separated from the plant that produced it before we select our cotton seed, we shall be planting seed from poor as well as from good plants, and must be content with a crop of just such stock as we have planted. By selecting seed from the most productive plants _in the field_ and by repeating the selection each year, you can continually improve the breed of the plant you are raising. In selecting seed for cotton you may follow the plan suggested below for wheat. [Illustration: FIGS. 49 AND 50. CHRYSANTHEMUMS AND ASPARAGUS]