Boughton was dibbled in-shows the highest ratio of straw to grain.

In comparing the two varieties, it will be seen that with four pecks sown broadcast the Boughton proved superior, and with one pack, dibbled in, equal to the Champion Amber, but that in all the other tests (6) it proved decidedly inferior.

In tabulating the results for the last three years, it is necessary to exclude all the tests with quantities of seed over four pecks, as this was the only amount of seed tested in 1880.

TABLE D.

EXPERIMENTS OF SECTION IV. FOR 1889, 1881 AND 1882.

Straw and chaff per acre

in pounds.

Grain per Acre in Bus.

2,720 1,408 1,264 1,797 14.90 14.40 16.26 15.18

From this table it appears that in the Boughton four pecks sown broadcast gave not only a better general average crop of grain than four pecks drilled in, but also a better crop each year. Drilling gave, however, rather the best average of straw. In the Champion Amber experiments, drilling gave much the best average crops of grain and straw. In 1881 broadcasting gave a little more grain, but in 1880 and 1882 greatly less. On comparing the two varieties we see that when sown broadcast the Boughton gave the better yield; when drilled, the Champion Amber.

The following diagram shows the conflicting nature of the resnlts:

It is plain from the lines representing the yields of straw and grain, crossing midway between the verticals, that the difference in favor of the Boughton sown broadcast was about equal to that of the Champion Amber drilled in.

In summing up the results of the experiments of the past three years, all we are justified in saying is, that it would appear, when four pecks of seed per acre are used, that broadcasting suits the Boughton best, and drilling the Champion Amber.

If we compare the results of the experiments of this year and last, with six pecks per acre, we find that, as far as both varieties. are concerned, drilling gave uniformly better results than broadcasting.

SECTION V.

Experiments Testing Modes of Preparation.

I explained in my preceding reports why I considered the experiments of this Section necessary. The preparation of the proper seed-bed for wheat is certainly a matter of very great importance, and unfortunately, there yet exists considerable difference of opinion in regard to this subject. Some hold that the land can not be too thoroughly prepared, nor too deeply. Some would make the surface very fine and smooth, by repeated harrowings and rollings. Others, again, advocate a shallow seed-bed, and as little preliminary cultivation as possible. For example, as far back as 1833 an English writer observes:

"Almost all seeds require a well consolidated bed after being well ploughed or digged, wheat and pulse particularly; and this for two reasons: first, because the seeds are equally affected by the close contact of the medium in which they are laid, germinate more regularly, and take a firmer hold of the soil, and secondly, because seed weeds are less likely to rise in a compact surface than in one which is loose, nor can slugs or grubs work their way so well in a close surface. Hence the use of rollers, pressers, and treading loose ground with sheep among farmers, and the practice of treading in seed by the gardener."* A recent American writer, on the other hand, declares: "I condider too much cultivation a detriment to the crop, making the surface run together and forming a crust on the land, which prevents a proper circulation of the atmosphere through the soil, weakens the wheat plant in its earlier stages of growth, from which it never fully recovers."+

In speaking of the influence which the seed-bed exerts on the harvest, Tanner, in one of his valuable little text-books, deplores the empiricism of the prevalent methods of preparing it. "The proper condition of the seed-bed may therefore be regarded as of the greatest importance. Although this is so self-evdient that no one would think of disputing the fact, it must still be admitted that we are too much in the habit of neglecting many of the conditions which are calculated to promote the results we desire. We almost unconsciously drift into certain courses of procedure, without

*Main's Vegetable Physiology. London, 1833. Page 181. †Michigan Board of Agriculture, 1880. Page 217.

taking into consideration the objects in view. Custom and established usage prescribe a certain depth as being necessary; this quantity of earth is consequently turned over, and the natural conclusion is, that we have thereby provided a satisfactory seed-bed.* The experiments of this year were made to correspond as closely as possible those of last season, the only differences between them being with the following: 1st. Only one variety of wheat was experimented with this year, the Boughton. Two plats sown with this variety received a thorough preparation and two an inferior. The experiments were therefore made in duplicate. 2nd. The preparation of the plats marked "inferior" was even rougher than that of the corresponding plats last year. The plats were merely ploughed once, and received neither harrowings nor rollings. The seed was drilled in, and the surface left rough and cloddy. The preparation of the plats marked "thorough" was the same as that of all the other plats of the series. The land was well prepared by repeated harrowings and rollings, after being ploughed and cross-ploughed. A reference to the plan will show that the four plats named in the Table were situated among the plats of Series II, but the experiments are properly discussed under Series I. The past treatment (ir preceding years) of each of the four plats was of course the same. results of the experiments are given in Table V:

TABLE V.

EXPERIMENTS TESTING MODES OF PREPARATION.

The

Plats 1-16 Acre Each. Seeded October 29th. 4 Pecks per Acre. Unfertilized.

The results present several points of interest. Between the best "thorough" plat and the worst there was a difference of 1.33 bush

*Elementary Lessons in Agricultural Practice. London, 1881. Page 72.

els per acre.

Between the best "inferior" plat and the worst the difference was practically the same, 1.34 bushels. Between the best "inferior" plat (12.00 bushels) and the best "thorough" plat (11.73 bushels), there was a difference of only 0.27 bushel in favor of the "inferior." Between the worst "inferior" (10.66 bushels) and the wórst "thorough" plat (10.40 bushels), there was again practically the same difference (0.26 bushel) in favor of the "inferior." And if we take 11.33 bushels, the average of the two "inferior" plats, and compare it with the 11.06 bushels, the average of the two "thorough" plats, we find again the same difference of 0.27 bushel in favor of inferior preparation. This is a remarkably close agreement in the duplicate experiments, and the results show that a very rough preparation gave equally as good a crop of wheat as a very thorough one. Of course economy was on the side of the former. It should be noticed that there was not the same correspondence in the crops of

straw.

In comparing the results of the experiments of this section for the last three years, we, of course, have to throw out the experiments with the Champion Amber last year and the year before, as they were not continued this.

TABLE E.

EXPERIMENTS OF SECTION V. FOR 1880, 1881 AND 1882.

It will be seen that the inferior preparation gave slightly the best average crop of grain, but the thorough preparation greatly the best crop of straw. In 1880 and 1882 the inferior gave the best yield of grain, but the thorough in 1881. In 1880 and 1881 thorough preparation gave much the best crop of straw; in 1882 inferior preparation gave slightly the best. The diagram brings out the great difference between the average crops of grain and straw for each mode of preparation :