Buying seeds is largely a matter of experience. So glowing are the descriptions in the numerous catalogues sent out that one may easily be led into ordering many worthless novelties, and many desirable ones for which there is neither room nor sufficient knowledge of their wants to grow them successfully. Cheap collections, where one is requested to send ten cents for a catalogue and twelve packages of seeds, are worst of all.

In buying flower-seeds, as in everything else, one never gets “something for nothing”—not even experience, and cheap seeds usually prove a very poor investment; the only safeguard is to buy of trustworthy dealers whose seeds are offered at fair prices. Take, for instance, Pansy seed: one ought not to expect to buy a fine mixture in a full-size packet—two hundred and fifty seeds—for less than twenty-five cents, while such varieties as Giant Trimardeau and Giant Cassier should be fifteen cents or more, according to the place of purchase. A few of the old favourites—like the Snow Queen—may be bought in packets costing from three to five cents and prove good value. But generally speaking, packets selling for less than five cents are to be looked upon with suspicion. Seeds bought at the grocer’s or other local dealer’s are likely to be old or inferior stock—the better class of seedsmen selling direct to the consumer only. At the end of every season seedsmen have a large stock left over. The best concerns destroy these, but the less scrupulous put them up as prizes for advertising purposes, or sell them to the local dealers.

Some varieties of seeds retain vitality for years, like the Morning-glory, while others lose vitality in a few months or weeks, like the perennial Phlox, which must be sown as soon as ripe. From this it can be readily seen how worthless some of the seed carried over from year to year must be. As an experiment, I planted under glass in the house, with more than usual care, a part of one of these trial packages of twelve for a quarter. From five of the packages I got just one plant—a Lantana, while from the sixth I grew a number of inferior Petunias. In striking contrast to this I planted, under the same conditions, a packet of Bush Hill Pioneer Cyclamen containing twelve seeds, for which I paid fifty cents, and secured eleven plants, every one of which lived, making four and a half cents apiece for a very choice variety of Cyclamen, while the one little common Lantana cost me ten cents.

The prices of one firm will frequently be much higher than those of another equally trustworthy, the difference being caused, in most cases, by the number of seeds in the packets. A few firms now state the number of seeds in their price-lists. This should become the general practice. There is manifest injustice in selling packets containing a quantity which is entirely problematical, and often much below what the purchaser has a right to expect. To advertise seeds much under the usual price and reduce the number accordingly is a common practice.

It is better to buy a few good seeds, and by careful cultivation bring the plants as near perfection as possible, than to buy a quantity of cheap seeds. One can join with a friend or neighbour and thus secure a larger variety at less cost than would otherwise be possible. Generally speaking I think it a great mistake to save one’s own seed. Few choice varieties are fixed and so tend to run out in two or three seasons. From fine strains of Asters you may save seed that will give good results the first year, but flowers grown from seed saved from these will show more or less centre, so that fresh seed should be purchased every other year at least. As the bees do not visit the Asters so assiduously as other flowers, they are less given to hybridisation and come quite true from seed; the trouble with them being deterioration. The first flower on the plant is usually the best, and should be reserved for seed by tying a string around it and labelling it, if it is desired to keep the varieties separate. Remove all other blooms on the plant, as they fade, allowing only the one selected to ripen.

Pansy seed should never be saved, as the ripening of it weakens the plant and checks its blooming. Pansies growing in a mixed bed never come true from seed and deteriorate rapidly. Growing Pansy seed is work for the Pansy specialist, but if it must be saved a few plants should be grown in a bed by themselves and covered with a netting to keep off the bees, which work incessantly over the Pansy bed. Not more than two or three of the finest flowers on a plant should be allowed to seed, all other buds being removed as they form, and the plants kept in a high state of cultivation and supplied with liquid manure at least once a week during the season of seeding.

Seed may be saved freely from Poppies, as they do not deteriorate appreciably, and the hybridisation constantly going on results in a wonderful variety of form and colouring. The ripening of one or two capsules of seed does not materially affect the period of bloom; but it is well to mark one blossom on each plant, giving a preference to the terminal blossom on the main stalk, as being usually the finest. Remove all other blossoms as they fade. Perennial Poppies will do better, however, if no seeds are allowed to form. Morning-glories, Cosmos and Verbenas sow themselves better than the gardener can, the plants being much more vigorous than from the hand-sown seeds. Verbenas mix badly, and are apt to hark back to the original purple strain, or come striped. The seeds of the pink variety come true, I find, and some of the finest pinks I have seen were from self-sown seed. They range in colour through all the shades from pink to deep rose. A fair proportion of white may also be expected, but scarlet and red with a white eye are shy and rarely appear.

Nasturtiums rarely come true from seed, and if one desires a particular colour or shade, she must either buy fresh seed or grow that particular variety alone. If, however, one simply desires an abundance of bright flowers with a preponderance of yellow, orange and scarlet, it is worth while to save seed, as its vitality is remarkable and every seed may be expected to germinate.

No variety of Petunia can be depended upon when grown in a mixed bed. A particularly fine variety may be lifted and planted on the opposite side of the house, either in the ground or in a window-box, or in a pot in the house. From one exceedingly fine specimen of the large ruffled variety, lifted and grown in a window-box on the opposite side of the house from the Petunia bed, I saved a single pod of seed from which I grew, the following year, sufficient plants for a large bed. No two of these were alike in colour, but they retained the immense size, ruffled edges, velvety texture and beautifully marked throats of the parent strain; some were a velvety crimson, nearly black, some white with crimson blotches, and some showed rare shades of grey.

Sweet Alyssum sows itself, and the trouble is usually not to save but to get rid of it. Candytuft may be saved, but the fewer flowers allowed to seed the better for the continuation of bloom. Where one does not actually need the seed all flowers should be removed as fast as they fade. No one thing will add so much to the appearance of the garden as this, as nothing is more untidy than flowers gone to seed.

Perennials are greatly benefited by having the withered flowers removed; indeed so noticeable is the advantage that some, like the Monkshood, will give a second crop of flowers, and those that will not respond in this way will be more robust and bloom more freely the next year.

Vincas, too, and Hibiscus bloom more freely if not allowed to seed, while Coleus should not even be allowed to bloom. Sweet-peas must on no account be permitted to seed unless grown for the purpose. Seed may be gathered from the Thunbergia, Maurandya, Adlumia, common Morning-glory and Hollyhock without risk of deterioration or harm to the vine. Antirrhinums suffer severely from seed-bearing, while the Ricinus, Lantanas and Salvias may be allowed to seed freely. Ageratums look rusty when ripening seed, and it pays to go over them frequently with the shears; this will keep them fresh and full of bloom until frost. One flower-head will furnish all the seed one is likely to want.

All seeds should be gathered when ripe and the foliage dry. They should be exposed to the air in any convenient receptacle, except in the case of the Thunbergias, Balsams, and others having pods which snap open and expel the seeds to a distance. These should be saved in paper bags until perfectly dry, when they may be freed from their husks by sifting or winnowing, and stored in paper bags, envelopes, or boxes. It pays to save the florists’ envelopes for this purpose, as they are already labelled. Never put seeds away unlabelled, and it is a good idea to add to the label the date, period of germination, with the colour of the bloom, height, and other data of interest. It is still better to keep a book of such information to which one may refer from year to year. There are frequent doubts concerning some operation—the season of starting the hotbed, the time it took a certain class of bulbs to bloom after potting, and the like. In all such cases the note-book offers unimpeachable evidence.