and the slovenly practice of many who allow their seed tubers to send out long sprouts before planting, which are either broken off intentionally before or unintentionally during planting. This practice cannot be too strongly condemned.

The Trays may be made small to hold 40 pounds of potatoes, with a handle running lengthwise across the top, or to contain 80 to 100 pounds, and handled by two men, when the handles run across. The lum

ber for the trays, ready sawn in lengths, should be purchased at from five cents to ten cents per tray, according to size.

Whole Sets vs. Cut Sets.-Considerable attention has been given to the advisability of cutting seed tubers. The question is wholly a financial one, as in an average year with an ordinary late variety the weight of the seed planted is of more importance than whether it is whole or cut. Early varieties do not do so well when cut, and varieties with white flowers seem to be softer in texture and more liable to failure, if cut, than those with purple or colored blossoms. Some varieties cannot be cut with profit, owing to lack of bud-producing eyes.

The labor of cutting is often greater than the cost of the extra seed. When seed is expensive, as when a variety is new, it is wise to cut as far as possible to secure the largest possible yield in the least time, but this course must be followed by selection, or rapid deterioration of the variety will result. A potato cut into single-eye pieces, and each piece planted in a hill, will give a greater yield than it would had it been planted whole.

Time to Cut.-Formerly it was advised to cut the potatoes a few days before planting. Generally speaking, this is a mistake. Zavitz' reports as the result of hundreds of trials, during a period of eight years, that potatoes cut the day of planting gave 8 bushels per acre heavier yield than those cut four to six days before planting. Similar results were obtained at the Montana Experiment Station."

Size of Seed.—It is a matter of general observation, supported by experiments, that large seed usually insures a larger yield than small seed. This may be due to the greater amount of nourishment furnished to the young plants, which enables them to make stronger growth, and to the greater hereditary vigor possessed by such tubers. Good-sized seed is especially desiraable on light soils, and for early maturing varieties. Smaller seed from vigorous plants may be as satisfactory with late varieties, owing to their longer period of growth. The advisability of using large or small seed, cut or whole, depends largely upon the cost of the seed, the season, the culture given, and the price realized when harvested. Generally speaking, tubers weighing two to three ounces make the most profitable seed, as they are worth less for consumption. The amount of experimental work which has been undertaken to decide the influence of the size of the seed tuber upon the yield is enormous, and only a few references can be given here.

Fischer, of Germany, advises (1) that under ordi

1 Ont. Agr. College and Farm Report, 1898, p. 158; 1902, p. 127.

2 Mon. Bul. 9, p. 21.

3 E. S. R., IX., p. 331; X., pp. 361–367.

nary conditions large seed should be used, (2) on good soils with heavy fertilizing small tubers and closer planting is advisable; but that the small tubers shall be the progeny of large tubers grown on well-cultivated and fertilized soil, to prevent degeneration. Tubers which are small because the parent plant had not sufficient vigor to produce any larger are worthless for seed.

At Arkansas Station' whole tubers 2 inches to 3 inches in diameter yielded 18 per cent. more than small whole tubers 3/4 inches to 14 inches in diameter, and large cut tubers 15.8 per cent. more than small cut tubers. At the Ontario Agricultural College' the largest yields for four years in succession were from planting large seed. Sets weighing one-sixteenth of an ounce and having one eye yielded, on an average, for the four years, 44.2 bushels, while two-ounce sets having one eye averaged 177.4 bushels per acre, and intervening sizes of sets yielded in proportion to their size. As the result of eight years' careful experiments, this station advises that large tubers be cut into pieces weighing about two ounces each for sets.3

J. C. Arthur, of Indiana, conducted an elaborate set of experiments for three years to ascertain the relation of the number of eyes on the seed tuber to the product. He found that within certain limits the yield will increase with an increase in the weight of the set, and that the exact number of eyes per cutting is relatively unimportant. With tubers of the same weight and variety the number of shoots does not perceptibly

1 Ark. Bul. 50, p. 28.

Ont. Agr. Col. Report, 1902, p. 126.

2 Ont. Agr. Col. Report, 1898, p. 156. 4 Ind. Bul. 42.

increase with the increase of eyes on the tuber. Seed tubers weighing 11⁄2 ounces and carrying 8 to 10 eyes sent up, on an average, 5.5 stalks per tuber, while seed tubers weighing 3 ounces and having 14 to 18 eyes sent up, on an average, 11.3 stalks per tuber. Bisecting an eye tends to increase the number of stalks, because each eye is usually a collection of buds, and some would be left uninjured on each piece. The number of stalks sent up tended to increase with the size of the seed tuber, and the yield increased with the increase in number of stalks.

The Virginia Experiment Station' reports that large seed cannot be used at a profit, while small seed is not recommended, but that sound tubers of the size of a hen's egg and upward are proper seed.

2

Green, of Ohio, found that crops from whole seed mature a few days earlier than from the same sized seed cut in two, and that small cuttings require the soil to be in better condition than large cuttings, or whole potatoes, in order to secure a good stand and a profitable crop.

Amount of Seed Per Acre-Cost and Influence on Yield.-Plumb,' of Tennessee Experiment Station, found the largest seed tubers to be most productive and the least profitable, while those varying in weight from one to three ounces were most profitable.

At Kentucky Experiment Station' the amounts planted varied from six bushels per acre when mediumsized seed were cut to two eyes to 48 bushels per acre where large whole potatoes were planted. At the

1 Va. Bul. 8, p. 3.

2 Ohio Second Series Bul., Vol. III., I., p. 14.

Tenn. Bul., Vol. III,, I,. p. 6. Ky. Bul. 22, p. 136.

Michigan Experiment Station' three varieties were tested, with results as shown in the following table:

The writer has found from seventeen to twenty bushels to be necessary to furnish a good seeding, and others have advocated the same amount,' although a less quantity is frequently mentioned as satisfactory.

A compilation' of experiments made at thirteen stations to determine the proper amounts of seed shows:

1. Within ordinary limits, an increase in seed produces a marked increase in total yield and marketable potatoes.

1 Mich. Bul. 57, p. 18.

2 Mich. Bul. 93, pp. 5, 6.