tion. In a series of experiments during four consecutive years at the Wisconsin Station a corn cultivated 3 inches deep gave better yields than that cultivated 1 inches deep. The deeper cultivation also left the soil more moist below the stirred surface soil, but the higher soil temperatures were in favor of the land receiving shallow cultivation. Certain plats at the New Hampshire Station were given no cultivation, others five, and others fourteen cultivations. Some of the plats were cultivated deep and others shallow. On the plats not cultivated the weeds grew luxuriantly and the yield was reduced to 17.1 bushels of shelled corn per acre. The plats cultivated shallow fourteen times yielded at the rate of 80.6 bushels of shelled corn per acre; cultivated shallow five times the yield was 79.1 bushels, and cultivated deep five times, 69.7 bushels per acre. The amount of stover yielded in each instance stood in about the same ratio as the grain production.

Plats receiving shallow cultivation at the Georgia Station averaged 2.4 bushels of shelled corn per acre more and 38 pounds of cured fodder less than those receiving deep culture. In another tillage test, cultivating thrice in succession every three weeks, twice every two weeks, and once every week resulted in yields of 32.65, 31.25, and 32.55 bushels of shelled corn, respectively. The results of later work along this same line indicated that cultivation once a week for corn on unfertilized land is better than less frequent but more thorough working of the soil, while on well-fertilized land the more thorough tillage at longer intervals is the most advantageous. The results of cultivation experiments at the Alabama Station indicate that the first cultivation should be deep and that frequent cultivation should be continued late into the season. At the South Carolina Station corn planted on the level without cultivation, except the cutting out of grass and weeds with the hoe, gave practically as good yields as corn planted in furrows and cultivated shallow the first time and deep afterwards. The Mississippi Station " examined the records of one hundred and sixteen tests of shallow and deep culture made at different stations and found that sixty-one tests of deep cultivation gave an average yield of 64.9 bushels per acre, while fifty-five tests of shallow cultivation gave an average of 74.7 bushels, a difference of 9.8 bushels per acre, or more than 15 per cent in favor of shallow cultivation.

a Wisconsin Sta. Rpt. 1894, p. 266. New Hampshire Sta. Bul. 71.

e Georgia Sta. Bul. 10.

Georgia Sta. Bul. 55.

e Georgia Sta. Bul. 62.

f Alabama College Sta. Bul. 111.

9 South Carolina Sta. Bul. 44.
h Mississippi Sta. Bul. 33.

DETASSELING.

The effects of detasseling were studied by the stations several years ago, but recently no further attention has been given to the subject. In over 50 per cent of the experiments reported the results indicated that detasseling had been either injurious or without effect. The most marked results in favor of the practice were obtained in experiments by the New York Cornell Station," the increase in total yield being as follows: In 1890, 50.6 per cent; in 1891, a very small gain; in 1892, 21 per cent, and in 1893, 19.3 per cent. On the other hand, the Nebraska Station obtained 528 pounds of corn from ten detasseled rows, 1,220 pounds from ten normal rows alternated with the detasseled rows, and 2,369 pounds from twenty normal rows elsewhere. in the field. At the Kansas Station the results were also decidedly against detasseling. On plats from which the tassels were removed from alternate rows the detasseled rows yielded 114.55 pounds of corn, and the rows with tassels entire, 185.75 pounds. Where only the first tassels were removed and the later appearing ones allowed to develop the yield per plat was 329.57 pounds, and where the tassels were removed from alternate stalks the detasseled stalks yielded 71.77 pounds and the stalks not detasseled 151.61 pounds. Unfavorable weather at a critical period in the growth of the crop caused a scarcity of pollen for the fertilization of the ears, and detasseling made the matter worse. The Illinois Station believes that if detasseling is beneficial at all it is most likely to be so on poor soil or in dry seasons, and that the practice sometimes reduces the yield.

IRRIGATION.

In connection with a series of investigations in soil physics, begun by King at the Wisconsin Station in 1889 and continued until 1901, the water requirements of corn and the influence of irrigation on the yield were studied. It was found that corn is able to draw upon the permanent water in the ground when it lies at least 73 feet below the surface in the case of a subsoil of rather coarse sand, and that the crop may reduce the water in a subsoil of sand to 7 per cent of the dry soil at a depth of 40 inches below the surface when the water table is only 42 inches lower. The observation was also made that on May 13 recently planted corn ground contained 23.33 pounds of water to the 100 pounds of dry soil in the surface 6 inches, while clover on the same kind of soil near by had reduced the water content to 8.59 pounds to the 100 pounds of dry soil.

a New York Cornell Sta. Bul. 61.

Nebraska Sta. Bul. 25.

c Kansas Sta. Bul. 45.

d Illinois Sta. Bul. 37.

e Wisconsin Sta. Rpt. 1889, p. 195.

In 1890 the water content of soil upon which corn was grown was determined at the time of planting and when the corn was cut to ascertain the weight of water required for a pound of dry matter, as indicated by the diminished soil moisture and the precipitation between planting and harvesting. The results apparently showed that about 414 pounds were required to produce a pound of corn. A similar experiment the following year showed that in one case 309 pounds of water for 1 pound of dry matter had been required on manured ground and 333 pounds on unmanured ground. The difference in results is believed to be probably due to the difference of percolation in the soil during the two seasons. The average quantity of water required to produce a pound of dry matter in 1891 and 1892 was 309.2 pounds. In two tests dent corn required an average of 309.84 pounds of water and flint corn in a single experiment 233.9 pounds for the production of 1 pound of dry matter, or, in other terms, the former required 1.8 times and the latter 2.2 times the season's rainfall. In 1894 and 1895 surface irrigation and subirrigation of corn were tested, and in all cases the yield from the irrigated land was much larger than from the unirrigated land, and surface irrigation yielded both years much better than subirrigation. In 1896 a rainfall of 15.02 inches, from May 1 to August 31, was well distributed through the season, and still the yield on irrigated ground exceeded that on ground not irrigated by about 1 ton per acre. The yield of shelled corn was also in favor of the irrigated ground, the difference being greater for thickly than for thinly planted corn.

In 1897 these experiments had been conducted on the same plats without the use of fertilizers for four years. The rainfall during the growing season had ranged from 4.48 inches in 1895 to 15.02 inches in 1896, and the water supply by irrigation from 5.7 inches in 1897 to 26.6 inches in 1895. In every season the yields were increased by irrigation. The largest quantity of feed was obtained from three stalks per hill, at distances of 15 inches, in rows 44 inches apart, when grown with only the natural rainfall, and also when a little more than 7 inches of water in addition was supplied by irrigation. The smallest yields of dry matter were obtained where only one stalk per hill was grown. The yields of shelled corn were largest from thin planting receiving only the natural rainfall, and these yields were nearly as large as where the plats were irrigated, indicating that

a Wisconsin Sta. Rpt. 1891, p. 124.

Wisconsin Sta. Rpt. 1892, p. 94; 1893, p. 152.

e Wisconsin Sta. Rpt. 1895, p. 237.

& Wisconsin Sta. Rpt. 1896, p. 195.

e Wisconsin Sta. Rpt. 1897, p. 222.

with a thin stand the rainfall was nearly sufficient for a maximum yield. In determining the cost of irrigation in 1897 it was found that with an ordinary farm engine, using coal at $5 per ton, water was raised 26 feet through a 6-inch pipe at a cost of $17.32 per acre-inch, and with a 23-horsepower gas engine, using gas costing $1.25 per thousand feet, the cost of raising water 12.85 feet was $15.75 per acre-inch. After the experiments had been carried on for five years the results indicated that the fertility of the irrigated ground was being perceptibly reduced. The sixth year 37 per cent more dry matter was produced on the irrigated than on the unirrigated plats." In 1900 b the average yield of water-free substance per acre for the entire period since 1894 was 1,993 pounds greater for irrigated ground than on ground not irrigated. The first four years of the experiment the average gain due to irrigation was 3,543 pounds of water-free substance per acre, but for the last three years it was only 62.2 pounds. In 1901 the increase in yield of corn by supplementing the rainfall by irrigation amounted to 4.2 tons of silage, 1.9 pounds of dry matter, and 35.16 bushels of ear corn per acre.

Irrigation in addition to the natural rainfall increased the yield of corn at the Louisiana Stations by 100 per cent. Soil-moisture investigations for the seasons of 1901 and 1902 at the New Mexico Station showed that irrigations made at the time of planting, before tasseling, and when the grain was forming, which is the ordinary Mexican method, will produce a fair crop of corn, but that in the ordinary seasons of that region from one to three additional irrigations will prove beneficial. At the Utah Station the percentage of protein in corn kernels was increased from 12.05 to 15.08, as the amount of irrigation was decreased from 37.26 inches to 7.50 inches. The proportion of ears to stover in these experiments increased regularly with the increased application of water. Late irrigations did not affect the growth and yield of the crop unfavorably. It is estimated that the best amount of water for corn lies between 20 and 25 inches.

RATE OF GROWTH.

The relation of meteorological conditions to the development of corn was studied by the Pennsylvania Station." Corn planted May 8 attained its average maximum height of 81 inches on August 8, an interval of ninety-two days, during which time there was a daily

a Wisconsin Sta. Rpt. 1899, p. 206.
Wisconsin Sta. Rpt. 1900, p. 185.

c Wisconsin Sta. Rpt. 1901, p. 197.

d Louisiana Stas. Bul. 62, 2d ser.

e New Mexico Sta. Bul. 46.

f Utah Sta. Bul. 80.

9 Pennsylvania Sta. Rpt. 1888, p. 167.

mean temperature of 67° F.; a precipitation of 11.5 inches falling on twenty-eight days, of which thirteen days occurred before June 1; and a mean daily cloudiness of 4.9, on a scale of 10.

A similar study was conducted at the Illinois Station." The experiments extended over four years, and the average growth in height for the first week after the plants were 1 foot high was 11.6 per cent; for the second, 8.3; for the third, 11.3; for the fourth, 14; for the fifth, 20.3; for the sixth, 12; for the seventh, 12.7, and for the eighth, 8.2 per cent. The corn reached its maximum height eight weeks after it was 1 foot high, but in one season, when it was planted as late as June 3, it continued to increase in height for ten weeks after it was 1 foot high. The most rapid growth in height was made when the plants were between 3 and 6 feet tall. They grew 2 feet per week for two weeks in succession the last of June in the season of 1890. Excessive rains in April, May, and June, 1892, interfered with the growth during these months, but about 3 inches in height per day was made in the last week of July. Later work of this kind at the same station led to the conclusion that the rate of growth is to a considerable degree independent of the temperature and dependent upon the stage of development which the corn plant has reached.

HARVESTING.

The experiments with reference to harvesting were largely conducted for the purpose of determining the stage of ripeness at which corn is most profitably harvested, either for the grain alone or for both grain and fodder. At the Kansas Statione dent and flint corn were cut on August 8, 15, and 20, and September 4. An additional cutting of flint corn was made August 24. At the time of the first cutting the corn was in the milk stage, and at the last cutting it was hard and the stalks were drying. The average yield of dry corn increased with both varieties from one date of cutting to the next. The yield of grain from the earliest cutting was 15.36 and 29.69 bushels, and from the latest cutting 39.42 and 61.44 bushels per acre for the dent and flint corn, respectively.

In similar work of this kind the Pennsylvania Station found, by analyses made of silage, dent, and flint corn varieties at different stages of growth and degrees of ripeness, that while the percentage of water decreased as the plant grew the absolute amount of water increased, and that there was a rapid increase of dry matter per acre as the plants approached maturity. In many instances the amount of dry matter in mature stover alone was equal to or more

a Illinois Sta. Bul. 31. Illinois Sta. Bul. 42.

c Kansas Sta. Rpt. 1888, p. 42; Bul. 30. d Pennsylvania Sta. Rpt. 1888, p. 26.