stalks are broken off by drawing a heavy drag over the surfacean old railroad rail being frequently used for this purpose. In many cases-and this practice is growing-the maize is cut and shocked before the proper time to sow the wheat. Then the wheat is sown as in the standing maize, or the more common practice on the heavier soils is to cut out the maize stubs with a disk harrow and harrow down with some suitable levelling instrument, preferably a spring tooth harrow. These methods make it possible to follow maize with winter wheat and the expense of putting in the wheat is small. It is thought also that the stalks are some protection to the wheat at times in preventing the snow from drifting off the wheat. The effect of this practice upon yield is hardly subject to determination experimentally except where the maize is cut before seeding. The experiments which have been made under the latter conditions indicate that the relative yield will depend upon the character of the soil. Where the soil is mellow and light, it should not be plowed; where it is heavy clay, plowing will be found desirable. In the latter case rotation is generally such that wheat does not follow maize.

In the spring wheat region, land that has previously been in oats or wheat is sometimes prepared without plowing, by using a disk harrow or similar instrument. Minnesota1 found disking as good as plowing on burned stubble field; while North Dakota found that plowing gave the best results. Among the objects to be attained in preparing the seed bed are the prevention of the growth of weeds and the conservation of the soil moisture, and whichever method most nearly accomplishes these results will probably be best. Plowing is not necessary for root penetration in the friable soils of the spring wheat region.

129. Time of Sowing. The proper time to sow wheat depends upon climatic conditions, the fertility of the soil, the 1 Minn. Bul. 46.

2 N. Dak. Bul. 10.

preparation of the seed bed, the liability to injury from the Hessian fly, and perhaps slightly upon variety.

It is possible to sow later as we go south, and necessary to sow earlier as we go north. When sown too late, the wheat has not sufficient vitality to stand the cold weather. When sown too early, its growth is so rank and succulent as to be injured by freezing. Experiments indicate the best average time of seeding in Ohio, Indiana and Illinois on the fortieth parallel to be about two weeks earlier than in Tennessee upon the thirty-sixth parallel; while the results at Columbus, Ohio, on the fortieth parallel and Wooster on the forty-first parallel indicate a difference of about one week. Doubtless differences in the fertility of the soil as well as temperature and rainfall have affected the results.

In some localities, early sown wheat is subject to attack from the Hessian fly. When such attacks are imminent, they may be avoided, by concerted action among the farmers of a neighborhood, by later sowing, especially if delayed until there is a killing frost, and also by sowing early some strips of wheat where the Hessian flies will congregate, and may be destroyed by plowing under the wheat. Generally speaking, delay until killing frosts occur is too late for the best growth of wheat in the fall, except on fertile soils. Where it is necessary, therefore, to delay the seeding of wheat to escape the ravages of the Hessian fly, the seed bed should be put in the best possible condition both as to fertility and physical properties.

The results of the various stations show clearly that there is no best time for any given locality. Some seasons quite early sowing gave the largest yield, while other seasons late seeding gave the best results. Very much depends upon the season prior to and after seeding. It may be said as a general rule, although late sowing is often as good as early sowing, it is seldom better, while early sowing is often better than late sowing. The more fertile the soil, the later the

seeding may be done with safety, as the rich soil produces the growth needed in a shorter time. Wheat often suffers in the fall from lack of rainfall. It is seldom injured from an excess of rainfall. As the time and manner of preparing the seed bed materially affects the moisture of the soil, the preparation of the seed bed may have a decided influence upon the time of sowing. The earlier and better the seed bed is prepared the later the seeding is permissible. On the fortieth parallel at an altitude of 500 to 1,000 feet, winter wheat should be sown generally about September 20th, with variations of a week either way, depending upon various factors indicated above.

While obviously not as many factors enter into the time of seeding of spring wheat as winter wheat, climatic and seasonal variations necessitate as wide variations perhaps in the former as in the latter. It may be laid down as a general rule that spring wheat should be sown as early as the ground can be got in fit condition for seeding. In both North Dakota and Minnesota the earlier sown spring wheats gave best results, while in Utah a medium date gave the best yields. Delay of two or more weeks in sowing caused marked losses where conditions were those of Ontario and Quebec. In other provinces the loss from delay in sowing was less marked. Seeding should be finished by May 1st in Ontario and Quebec, and in other provinces from: May 15th to 25th.1

130. Depth of Sowing. This will vary with the kind of soil, the moisture, and the levelness and the firmness of the seed bed. Wheat may be sown deeper in a sandy soil than in a clay soil. It is necessary to sow deeper in a dry than in a wet soil. Variations in rainfall often materially modify the depth of seeding. It is reasonably well established that, under ordinary conditions, the nearer the seed is covered with one inch of moist soil, the better. An uneven and cloddy soil would require that some be planted deeper than is desirable 1 Cent. Expt. Farm, Canada, Bul. 21.

in order that all may be covered. A summary of the work of eight stations, mostly in the Mississippi and Ohio Valleys, aggregating twenty years' results, shows that in some instances four inches was at least as good as shallower depth, but in most instances one to three inches gave the best results, and indicates that usually it is not safe to go beyond these extremes.

131. Drilling Compared with Broadcasting. Stations of thirteen States have made experiments to compare drilling wheat with broadcasting it. The number of years' trial at a station varied from one to nine years and aggregate thirtythree years. Only two stations (Iowa and South Carolina) report, as the result of one year's trial, in favor of broadcasting. While in individual years broadcasting has produced the best results, at other stations the average of two or more years was in favor of drilling whether for fall or spring seeding. For fall seeding, the Ohio Station found as the result of nine years' trial two bushels in favor of drilling; Indiana in four years' trial reports eight bushels gain, and Kentucky in three years' trial reports four bushels gain. For spring seeding, Minnesota in three years' trial reports two bushels gain; North Dakota in two years' trial reports five bushels gain, and South Dakota in two years' trial reports two bushels gain. While these differences are not great, they generally amply pay for any extra cost of drilling, which is almost the universal practice for fall seeding.

A number of reasons may be given for this practice, not all of which, however, will apply in any given locality. The wheat is more uniformly distributed and covered and is sown at a more even depth. Quick germination is insured by having the seed in moist soil. It is believed also to be less easily winter killed either by freezing or heaving. The drill makes little furrows in which the snow lodges and is prevented from being blown away. It has been abundantly proved that the amount of snow held in the furrows is sufficient to modify the temperature

of the soil considerably. The wheat is less likely to be heaved out from freezing and thawing. The soil at the bottom of the furrow offers greater resistance to the heaving than does that at the top of the ridge. The movement of the soil will take place at the point of least resistance, which will be at the top of the ridge, thus increasing the chances of the plant at the bottom of the furrow to remain undisturbed. At the same time the loosened soil, aided by the rains, tends to fall into the furrows and thus further protect the plant. Just how much effect this has one year with another is not known, but in some trials during one year by the writer, where the furrows were obliterated by rolling, the yield was not materially affected.

In the spring wheat districts, the winds tend to lay bare the seeds when broadcasted, while drilling rather tends to deepen the covering by partially filling up the furrows. Practice seems to show also that weeds are less troublesome in spring wheat when drilling is practiced, doubtless because it insures quicker germination of the wheat.

132. Quantity of Seed per Acre.-The quantity of seed to be sown per acre will vary with the character of the soil, climate, time of seeding, seed bed, size, quality and variety of seed, and method of seeding. If sown early, less would be required than when sown late, because each plant would become larger, tiller more, and thus cover more ground. If the seed bed is well prepared, and the vitality of the seed good, a larger percentage of the seed will grow than if the seed bed and seed are poor. Fertile soil requires a less number of plants per acre than a poor soil because each plant tillers more and grows larger and thus occupies more room. A bushel of one variety may contain three times as many grains as another. A variety which tillers profusely could be sown thinner than one that does not. If drilled, a less quantity could be sown than if sown broadcast.

The yield will not be at all in proportion to seed sown. The wheat plant adjusts itself to its surroundings. If sown