Rawson says ("Success in Market Gardening," p. 57) : "Perhaps we might truthfully say that the most important of all points in gardening is the right selection of seed; for without good seed the care and expense devoted to selecting and fitting the land, or procuring and using implements, fertilizers, etc., are all bestowed in vain".

It is easily possible, however, to select seeds for years without making any advancement. This actually happened in the experience of Livingstone. For 15 years he labored in vain, eager to improve all varieties, but no progress was made, because wrong methods were employed; the largest and finest specimens of tomatoes were selected, year after year, with little or no regard for the plant. Then the plant instead of the individual tomato was made the unit, and Livingstone soon became a prolific producer of important varieties; no other man has accomplished so much for the improvement of the tomato.

The securing of good seed is not so much a question of selection as it is of careful and intelligent breeding. Starting with the plant as the unit, the grower must decide what he wants and what his market demands; for he himself might be very well satisfied, and his market very much dissatisfied. Suppose he is growing tomatoes and the plants are yielding well, but the fruits are generally rough and ill-formed, and yet, in looking over the field a few plants may be found which are highly prolific, and also produce better-shaped fruit than the hundreds or thousands of other plants growing under the same conditions. Seed should be saved from each of these plants, kept separately in numbered packets, and the next year the plants from each lot of seed set in different rows or plats. One of the selected plants may possess greater power to perpetuate its good qualities than any other, but this important discovery cannot be made if the seed from these plants be mixed.

This method of procedure is just as important with every other vegetable. The grower who wants better seed must have well-defined ideas before attempting any work of breeding or selection. If the cantaloupes are too coarsely netted, select for fineness of markings; if the cabbage lacks uniformity in time of ripening, select with this in view; if the onions are too flat, deepen the bulbs by selecting with this idea prevailing. Many other illustrations might be given. By intelligent selection it is possible to make improvements in size, color, form, flavor, texture, number of seeds, habit of growth, resistance to drouth, cold, heat and disease. Many valuable new varieties have been developed by the method indicated. The purpose of this chapter, however, is not to encourage the creation of new varieties, important though this may be, but to assist growers in the betterment of old varieties.