deep, rich prairie soil. The corn was planted in hills 3 ft. 8 in. apart each way. A meterological summary for the years 1889-195, inclusive, and tabulated data for each experiment are given.

Test of varieties (pp. 163-173, 179). This trial test of 81 varieties occupied 100 fortieth-acre plats, 80 plats having been in corn the previous year and 20 in wheat. The plats were fall-plowed and well harrowed before planting.

One variety, planted on 13 different plats scattered throughout the trial area, ranged from early to late, and in yield from 45.8 to 100.8 bu. per acre. The authors say "It will not do to credit all [the] variations in yield to varietal differences so unaccountable are these extreme variations in yield that it seems necessary to resort to systematic duplication . . . to give anything like a just comparison for a single season. It is therefore only after a long series of years and from plantings in a variety of plats that anything like a true comparison can be established."

Usually medium maturing varieties have given slightly the highest average yield; but in 1895 the highest average was from the late sorts.

Time of planting (pp. 173-175, 179).-Nine plantings of the same variety of corn were made one week apart between April 22 and June 17. Medium plantings gave highest yields.

Thickness of planting (pp. 175, 176, 179).-Plantings at successive dates were thinned to 4 rates of seeding, 2, 3, 4, and 5 kernels to the hill. Thicker plantings gave the higher yields with smaller ears and more nubbins; for all varieties about 11 per cent of the stalks were usually barren; but this varied greatly with both variety and season. With maximum yields there are generally about 10,000 ears per acre.

Rotation experiment (pp. 176-179).-Corn grown continuously, with and without manure, was compared with corn in rotation with oats, and with oats and clover.

The yield of corn on unmanured land decreased in rotation with corn and to a less degree in rotation with oats. Land under a rotation containing clover gives decidedly superior yields. No effect was observed from commercial fertilizers.

Rate of growth (p. 178).-This is to a considerable degree independ ent of the temperature and dependent upon the stage of development which the corn plant has reached.

Lucern in Argentina and England, H. GIBSON and W. FREAM (Indian Agr., 21 (1896), No. 3, pp. 82, 83).-In Argentina lucern, or alfalfa, is grown both for hay and pasture. In the province of Buenos Ayres nearly every "Estancia" (stock-breeding estate) has its field of 10 to 40 acres of alfalfa. The land is plowed, cross-harrowed, and 3284-No. 11———4

sown with 26 to 36 lbs. of seed per acre. Sowing in autumn, spring (which begins September 24), and midsummer is practiced. The crop is cut 4 or 5 times each year, and the total hay crop from it averages from 6 to 10 tons per acre. The first cutting is generally gathered by the end of November, and the last by the end of March or beginning of April. These alfalfa fields last a long time. Fields 30 years old are known, still well covered and yielding a good crop.

In the provinces of Buenos Ayres, Santa Fé, and part of Cordoba the soil is a rich, black loam of great depth. Here the alfalfa follows wheat and is pastured.

"With regard to the possible extent of laud capable of carrying lucern for grazing purposes, it is difficult to establish the limit. Alfalfa has done equally well in the rich, deep soil of Santa Fé and the sandy, loose earth of the Pampas formation which begins in Cordoba.

"It is to be found thriving in the far West in Mendoza and San Luis, and experiments in the south of the province of Buenos Ayres have been attended with suecess. . . . I have found it to grow well in latitude 36 to 37° S. . . . The zone where, by reason of the conditions of soil and climate, alfalfa can be permanently laid down, embraces an area of not much less than 150,000 sq. miles."

Exporting alfalfa hay has become a business of considerable importance. In the province of Santa Fé the roots of alfalfa have been found at a depth of 40 ft. The greatest enemy of alfalfa is the migra tory locust.

In England, in the course of 10 years, 1885-'95, the area of cultivation of alfalfa has extended from 13,754 acres to 23,853 acres. This is mainly in the eastern and south central parts of England.

Oat culture, R. J. REDDING (Georgia Sta. Bul. 30, pp. 383-385).-Ten acres, manured with 200 lbs. acid phosphate, 50 lbs. muriate of potash, and 400 lbs. cotton-seed meal per acre, were sown to oats in October and November. This sowing was killed by frost. During the last week in February the piece was resown to the "Burt,” an early rapid growing variety of oats. Early in April the crop was fertilized with 50 lbs. of nitrate of soda per acre, except one-tenth acre in one of the best acres of the field. This tenth acre yielded at the rate of 42 bu. per acre, and the adjacent tenths of an acre yielded 60 bu., a gain of 18 bu. in favor of the nitrated portions.

The whole piece averaged 40 bu. per acre.

Some cultural directions for oats are given in which fall or winter sowing is recommended in preference to spring sowing, on land fertilized with a complete manure.

Experiments with oats, 1888-1895, E. DAVENPORT and W. J. FRASER (Illinois Sta. Bul. 11, pp. 156-160).-This is a continuation of work published in Bulletin No. 34 of the station (E. S. R., 6, p. 408). For the years 1888 to 1895 notes and tabulated data are given on the quantity of seed per acre between 1 and 4 bu., the time of sowing between March 14 and May 10, on the depth of sowing between 1 and 6 in., on broadcast vs. drill sowing, and on a test of 28 varieties.

The authors give the following summary of results:

"Within reasonable limits thickness of seeding has little influence upon yield, with a preference decided though slight in favor of 24 bu. per acre.

"From different thickness of seeding the yield of grain and its weight per bushel are quite independent of the weight of straw.

"The most favorable time for sowing is the very last of March, with a tendency favoring an earlier date.

"Yields of grain and straw from seedings at different dates fluctuate more nearly together than do those from different rates of seeding.

"Comparatively shallow seeding is necessary.

"Maximum yields may be had with medium weights of straw.

"The low yields of the season of 1895 are due to the remarkably dry season."

Variety tests of oats, H. J. WATERS and E. II. HESS (Pennsylvania Sta. Rpt. 1894, pp. 290, 291).—In 1894 16 varieties of oats were grown on duplicate twentieth-acre plats. The yields of grain and straw and the weight per struck bushel for the years 1888 to 1894 inclusive are tabulated. The highest average yields were given by Japan, Baltic White, Improved American, and German, all having produced an average of over 40 bu. per acre.

Variety tests of potatoes, H. J. WATERS and E. H. HESS (Pennsylvania Sta. Rpt. 1894, pp. 292, 293).—This is a continuation of work published in the Annual Report of the Station for 1893 (E. S. R., 6, p. 722). In 1894 16 varieties of potatoes were grown, in rows 42 in. apart, pieces 12 in. in the row. Clean, shallow, and level culture was practiced. The yields tabulated are for the years 1889 to 1894 inclusive. The highest average yields were given by Stone-Road No. 2, Early Puritan, Ben Harrison, Burpee Superior, and Green Mountain.

Saltbushes, J. G. SMITH (U. S. Dept. Agr., Div. of Agrostology Circ. 3, pp. 4, figs. 3).—The Australian saltbushes (Atriplex semibaccata and A. leptocarpa) and "Winter fat" or sweet sage (Eurotia lanata) are figured and a general statement given of the practical value of saltbushes. There are over 30 different kinds of American saltbushes. In dry seasons and during severe winters they supplement the native grasses. They are natives of the arid and semiarid uplands and grow where better forage plants will not, and they should be preserved from extinction. Foreign varieties should be introduced with caution.

Fertilizer experiments on tobacco, W. FREAR (Pennsylvania Sta. Rpt. 1894, pp. 169-189).

Size of leaves of cured crop (pp. 174–176).—Tabulated data are given for the length and breadth of the cured leaves grown with different applications of fertilizers. The leaves grown at Rocky Spring were 10 per cent larger than those grown at Donegal.

Sweated tobacco (pp. 176-185).-Tabulated data are given for loss in sweating, weight of leaves, and proportion of rib, and for thickness of leaves.

The average loss in sweating at Donegal was 10.47 per cent, and at Rocky Spring 12.82 per cent. Tobacco grown with stable manure lost

more than any other; that grown with the carbonates of potash and magnesia lost more than that with sulphates.

The thinnest leaves grew on the unfertilized plats.

Duration of glow (pp. 185–188).-"The test was made upon the selected wrappers already examined as to thickness and density. The dried leaf was touched with a glowing cigar tip, and the time elapsing till the running spark died out was noted. The results reported are the average of 5 such tests upon the tobacco from each plat." Tabulated data are given. There was no distinctive difference between the results from tobacco grown with different commercial fertilizers.

Experiments for 1891 (pp. 188, 189).-This is a report of progress, since the results from the cured crop had not been obtained. The plan of operation was the same as in 1893, except that sulphate of ammonia was used instead of nitrate of soda as a substitute for half the cotton seed meal.

Experiments on the curing of tobacco, W. FREAR and E. J. HALEY (Pennsylvania Sta. Rpt. 1891, pp. 189-201, 206-257).-A general discussion of the manner of curing tobacco is given, in which the different processes used in this and foreign countries are reviewed.

Air-curing process with whole stalks and with separate leares (pp. 208-218). This trial was conducted in an old-fashioned tobacco shed 30 by 60 ft., 15 ft. to the plate. It was ventilated by vertical shutters on 2 sides, by large doors at one end, and by the openings between the shrunken siding. The floor of this shed was carefully cleared of all scraps of tobacco and limed. August 28, when the middle leaves were mature, one-fourth acre of tobacco, harvested on the stalk and slightly wilted, was hung in the shed on 4-foot laths 18 to 24 in. apart, 7 stalks to a lath. The laths were placed in 3 horizontal tiers, which were a little over a foot apart.

At the same time the ripened lower and middle leaves from another fourth-acre were broken from the stalks, brought to the shed, and strung upon the wires of the Snow lath. During the whole period daily observations of the temperature and humidity were made on the east and west sides and on the floor of the shed, and a continuous record of the temperature among the curing leaves was kept by a selfrecording thermometer. The meteorological and other data are tabulated. In the shed the temperature ranged from 44 to 88° F. and the humidity from 56 to 100°, the humidity as a rule being below that outside.

"The leaf began to yellow in 60 hours after the curing was begun, beginning at the tips and borders and gradually working up toward the midrib and base of the leaf. By the time the body of the leaf had attained the yellow color, the tips and borders had begun to turn brown; this browning was accompanied by a curling and drying of the tip and edges. During this stage of the process, while the leaf was sweating, the temperature ranged from 70 to 750, and the relative humidity increased from 70 to 88 per cent. . . . The browning of the web of the leaf progressed slowly and was not completed until the twenty-third day of the curing."

The whole process of curing required 40 days. Similar observations. were also made at Lititz and Rocky Spring. Among the leaves harvested separately "the first yellowing appeared after 42 hours. The borders and tips of the leaves were yellow in 66 hours. The first trace of brown appeared after 90 hours. The web of the leaf was cured in 16 days and the rib of the leaves first introduced into the curing shed 20 days later, making the total period required for the cure 36 days." The loss of weight was 77.55 per cent with the whole stalk and 84.51 with the leaves cured separately.

The Snow process (pp. 218-230).-The unpatented lath invented by Mr. Snow, and which gave the name to the process, consists of a stick 4 to 4 ft. long, with a rectangular cross section and with wires projecting on opposite sides, each long enough to hold about 5 leaves, 70 leaves to the lath. The shed should be tight, with ventilators that may be opened at will, and with furnaces for heating. The harvesting of the one-fourth acre set apart for this process began August 27; the ripest or bottom leaves from each stalk, not exceeding 5 or 6, were removed, placed in baskets, and carried to the barn. The laths when filled were placed 10 in. apart. Heat was supplied artificially and regulated according to the outside temperature and the condition of the curing leaves. Tabulated data are given: "The average of 5 days' observations at 4 p. m., taken in the front, middle, and rear of the barn, lower tier, and in the second tier back, give the following temperature averages: Front, 104.2°; middle, 101.4°; back, 106.6°; second tier, 103°. The tobacco over the ovens and over the drum received, therefore, the greatest amount of heat."

The time required for first yellowing was 22 hours; for first browning, 24 hours; for development of tobacco odor, 60 hours; for complete curing of leaf tissue, 18 hours; for complete drying of rib, 34 hours; in all, 158 hours, or 6 days and 14 hours.

"The curing having been completed, the fires were allowed to go out; 20 hours afterwards the temperature was like that outside, and water was sprinkled over the basement floor. The next day the leaves were ready to handle, were taken down, sorted, and made up into hands.

"The total curing operation, including the filling of the barn and the removal and sorting of the leaf, occupied only 9 days."

Modified Snow process (pp. 230-234).-In this process the temperature was not allowed to go above 100° F., the beginning of the rise was gradual and the draft through the barn greatly increased. By accident smoke escaped into the building, and the fires were drawn for 18 hours. By this time the leaf was completely cured. This process required 42 hours to the appearance of yellow spots; to appearance of brown color, 28 hours; to development of tobacco odor, 24 hours; to end of sweating, 72 hours, and to completion of curing, 216 hours; in all, 382 hours, or 15 days and 22 hours.

Cuban process (pp. 234-237).-Two trenches, 2 ft. deep, were dug in dry ground, filled with hickory wood, and this set on fire at noon