> "K % ™Q Division of Agricultural Sciences UNIVERSITY OF CALIFORNIA $ v WM l mmii&MMjii^ MP JAMES R. TAVERNETTI H. F. MILLER, JR. CALIFORNIA AGRICULTURAL EXPERIMENT STATION BULLETIN 747 is California's most valuable field crop. In 1952, it accounted for 42 per cent of the cash farm value of all field crops, and 12 per cent of the state's total cash farm income. Mechanization of the various operations involved in growing and harvesting cotton is a fairly recent development. As late as 1948, over 90 per cent of the crop was still harvested by hand labor — an expensive and time-consuming operation. Use of the mechanical picker has been the major factor in reducing labor requirements for harvesting. In addition, labor needed for thinning 1948 1949 MAN-WEEKS OF LABOR PER BALE PRODUCED or chopping and for weed control has also been lowered because of increased use of machines. In 1948, a cooperative project was organized to study the different phases of mechanization. The work included design, building, and testing of new equipment; testing and altering of equipment already in use; and experiments with cultural practices that would reduce labor or make the use of machinery more fea- sible. While seeking ways to reduce hand labor, the project did not overlook the equal importance of maintaining yield and quality of fiber. This bulletin describes the mechanization experiments and sum- marizes their results. PER CENT OF ACREAGE MECHANICALLY PICKED MECHANIZATION STUDIES SHOWED THAT: HIGH- and LOW-BED PLANTING gave similar results in yield and in picker efficiency. . . . Low-bed planting was better for early weed control 7 A SEED-PRESS WHEEL on the cotton planter resulted in earlier and greater plant emergence in sandy loam soil 9 PLANT POPULATIONS of 20,000 or more per acre gave the greatest yields. . . . Efficiency of the mechan- ical picker increased with greater numbers of plants 10 WEED CONTROL by late cultivation (after lay-by time) was possible with special equipment. . . . Flaming also gave good control. . . . Selective oil sprays were effec- tive in small cotton (2 to 6 inches) 15 MACHINE TOPPING helped prevent lodging of rank growing cotton . . . did not reduce the yield 21 DEFOLIATION improved the grade of mechanically harvested cotton slightly . . . had little effect on picker efficiency 23 MECHANICAL COTTON PICKERS vary in efficiency, depending on type of plants, condition of picker, and skill of operator. . . . Cost for mechanically picked cot- ton was from $15 to $20 less, per bale, than for hand picked 24 APPENDIX 32 THE AUTHORS: James R. Tavernetti is Agricultural Engineer in the Experiment Station, Davis. H. F. Miller, Jr., is Agricultural Engineer, Agricultural Research Service, United States Depart- ment of Agriculture. NOVEMBER, 1954 STUDIES ON MECHANIZATION OF COTTON FARMING IN CALIFORNIA 1,2 JAMES R. TAVERNETTI H. F. MILLER, JR. In 1948, less than 10 per cent of the California cotton crop was harvested mechanically. In 1952, machines picked 67 per cent of the crop. This increase in mechanization has brought a correspond- ing decrease in the labor required to pro- duce a bale of cotton — 1*4 man-weeks in 1948, as compared with % man-weeks in 1952. Over the same period of time, cotton acreage has risen from 804,000 to 1,400,000 acres (table 1 ) , while the maxi- mum number of harvest laborers has de- creased from a peak of 117,800 to 80,000. Some work on cotton mechanization was done from 1928 to 1932 by H. B. Walker and E. J. Stirniman of the De- partment of Agricultural Engineering, University of California. Experiments were conducted with stripper type har- vesters, but it was concluded that these machines could not compete with hand labor at that time because of field losses, lower grades, and cost of the equipment. Following reactivation of the work on mechanization in 1946, a project was be- 1 Submitted for publication April 15, 1954. 2 These studies were conducted in cooperation with the Agricultural Engineering Research Branch, Agricultural Research Service, United States Department of Agriculture. gun, in 1948, for more detailed study of the problem. This project was a joint effort, with the University of California Agricultural Experiment Station and the (then) Bureau of Plant Industry, Soils and Agricultural Engineering of the Uni- ted States Department of Agriculture cooperating. (This bureau is now the Fig. 1. Types of cotton seed. Top: fuzzy, or gin-run. Lower left: mechanically delinted; lower right: acid delinted. [5] c o o u c o o o o w w »■■ c s 0) 2 o (A c s (A *- s o E o u "5 o o 8 s c .2 w 3 o O O O O O CO 9. 9 *H O 00 O iH ^ 00 o © o •<* O O OS t- © © • • o o" o »o O CO © O »> O O CO O lO o io 00 CO o o -^ O O 00 O O CO o o o O O : O O : OS o w (N CD CO t- o o O CO o -tf o CO i-l o o o oo o o m • io t> oo" eo~ 00 I-. a CO o IO (J> I-l oooooot-oooco OOCOOOCOOOiOCO 9 9 th' 9. 9 rA rl w . "* H 00 O U) © iH 00 t- O t- 00 ro © o a o CO CM o> o> o o t> o o O O CO o o 9 9 i-i 9 9 io oo o"n OJ CO O CD 3 is o -^ p. CO OJ CD CD I— I "o s CD i-l c$ o c3 (-1 CD c3 ■i o O etl B CD Sh CD ft ,5 ^ CD A S_ ^H ._ 01 o c€ <3 M to hn CD M O X> CD CO Cd (-1 »-H ~* o «d +» « ja h «-. 2; CD CD ^ ft ft (S CO u A M M a » Js ° o cd § o M CD *s s CD C JD CD £ CD o i G a o o ■ 4) 3 . •C G be v «s oo ea d) G gw rt C ft*" 05 G Q CO a a) -3 SO CD to"" O J) 03 ™ 1" +a » M ® O hi hi 9 CO 23 CO 3*. C co bo boo * -i * <■ Agricultural Engineering Research Branch, Agricultural Research Service, U. S. D. A.) Until 1949, the experiments were con- ducted on a private farm in Fresno County, but since then, they have been carried on almost entirely at the United States Cotton Field Station near Shafter in Kern County. Cotton for the studies at Shafter was grown on beds with 40-inch row spacing, in Hesperia sandy loam soil. Acid-de- linted seed (fig. 1), averaging about 3,500 seeds per pound, was planted dur- ing the first part of April. Unless other- wise indicated, all the cotton was planted to a stand at a seeding rate of 15 to 20 pounds per acre. All irrigation was by the furrow method (fig. 2). One preplanting irrigation was made, followed by irriga- tions after planting at one- to two-week intervals from about June 1 to about Sep- tember 1. All harvesting was done with a barbed spindle type, single-row picker. First picking was usually made in Octo- ber, and the second in November. All plots were fertilized with 80 pounds of available nitrogen (400 pounds of am- monium sulfate) . The fertilizer, as a side dressing at the time of planting, was ap- plied on one side of the row, 2 inches be- low and 6 inches to the side of the seed. In calculating yields, 1,350 pounds of clean seed cotton were figured as one bale of lint. Picking efficiency was calculated by dividing the weight of seed cotton harvested by the total pickable cotton on the plants. The remaining cotton after harvest was hand-gleaned from several small portions of each plot to obtain the total yield. Classing for grades was done by the regular classers in the United States Department of Agriculture Class- ing Office at Bakersfield. The cotton was ginned in a relatively new, modern gin equipped with a drier, seed cotton clean- ers, and a lint cleaner. [6] Fig. 2. Test plots showing type of beds and method of irrigation used. Water is turned into furrows through small pipes in ditchbank. Each pipe serves two furrows, but only one is irrigated at a time. HIGH- and LOW-BED PLANTING gave similar results in yield and in picker efficiency. . . . Low-bed planting was better for early weed control. A three-year study was made to de- termine the effect of different seed place- ments on yield and picking efficiency. In high-bed planting, only enough dirt was scraped from the top of prepared beds to allow the seed to be planted in firm, moist soil. This left the irrigation fur- rows practically undisturbed so that, dur- ing subsequent cultivations, only a small amount of dirt was thrown into the drill row. In low-bed planting, most of the dirt was removed, seed was planted near Table 2 — Comparison of High- and Low-bed-planted Cotton Type of planting Factor High Low 1948 1949 1950 1948 1949 1950 Plants per acre (thousands) .... Yield (bales per acre) 23.8 2.39 93.6 7.0 30.0 3.26 97.0 5.7 2.17 91.9 5.1 24.3 2.45 93.2 7.0 47.4 3.31 97.1 5.1 2.30 Picker efficiency (per cent) Trash, first picking (per cent) . . . 91.6 6.1 [7] LOW-BED PLANTING HIGH-BED PLANTING 20" IO" O" IO" 20" After planting O" eo H 10" o" 10" ao" After first cultivation Final row profile Fig. 3. Row profiles at different stages with low- and high-bed planting. the bottom of the bed, and the bed and irrigation furrows were re-formed dur- ing cultivation (fig. 3). Wings attached to the planter runners were used to scrape soil from the beds (fig. 4). The spread of these wings should be narrower for low than for high planting. Spreads of 12 to 16 inches for the low, and 18 to 24 inches for the high are most common. The results showed a slight tendency toward greater yield with low planting, but there was no difference in efficiency of mechanical picking nor in the trash content of the harvested cotton (table 2). Low-planted cotton was better for early weed control — the dirt thrown onto the drill row covered small weeds. In soils where the moisture content is too great, some difficulty may be en- countered with low-bed planting. This may be overcome by scraping off some of the dirt a few hours or a day ahead of planting, to allow the beds to dry. Rotary hoes have also been used to loosen the soil and hasten drying. Table 3 — Comparison of Emergence of Cotton Planted With and Without a Seed-press Wheel on the Planter Treatment Seed-press wheel . No seed-press wheel Seed-press wheel. . . No seed-press wheel Seed-press wheel. . . No seed-press wheel Seed-press wheel. . . No seed-press wheel Planting depth inches 1 1 i l A 2 2 2V 2 2 l A Plant emergence counts 7 days 1 45 22 31 17 18 8 10 days 3 83 59 110 94 91 71 14 days 5 1 90 69 124 116 104 97 * Number of plants for 39 feet of row, or .003 acres. [8] A SEED-PRESS WHEEL on the cotton planter resulted in earlier and greater plant emergence in sandy loam soil. A comparison was made between emer- gence of cotton planted at different depths, in sandy loam, with two kinds of planters — a runner type, having only an open-center, steel, surface-press wheel, and a similar type, but with a seed-press wheel in addition. This extra wheel was rubber-covered, about 1 inch wide, and 8 inches in diameter. It was located im- mediately behind the runners so that it rolled directly over the seeds and pressed them into firm, moist soil on the bottom of the furrow (fig. 4). Depth of plant- ing was adjusted by the depth of the run- ners below the wings attached to them. At the time of planting (May 4), there were about 2 inches of dry soil on the beds. This was scraped off by the wings, and the seed was placed in firm, moist soil. Acid-delinted seed was planted at a rate of 15 pounds per acre. This was slightly over 50,000 seeds per acre, or an average of about 4 per foot. Results showed that the seed-press wheel was a definite help in obtaining faster and greater plant emergence (table 3). A planter with seed-press wheels has been used for several years, for most of the planting at the Shafter Experiment Station, with consistently good results. Fig. 4. Planter used in seed-press wheel studies. A. open-center, steel-surface press wheel; B. seed- press wheel; C. runner openers; D. wings to scrape dry soil from tops of beds. [9] PLANT POPULATIONS of 20,000 or more per acre gave the greatest yields. . . . Efficiency of the mechanical picker increased with greater numbers of plants. Thinning or chopping of cotton to single plants spaced from 8 to 12 inches apart was the common practice for many years. This operation was done by hand hoeing, and usually required from 5 to 7 man-hours per acre. Studies were made of plant population (plants per acre) to determine whether thinning could be done mechanically or could be elimin- ated, by planting the cotton to a stand, without affecting yield and grade. Opti- mum population for best yield, grade, and adaptability for mechanical harvest- ing was also determined. Comparisons were made between planting to a stand by continuous drilling and by hill drop- ping, and between various types and makes of choppers for mechanical thin- ning. Summaries of the results are given in tables 4, 5, and 6. Yield. Results show that, for condi- tions under which these tests were con- ducted, thinning can be eliminated or done mechanically without detrimental effect on yield, provided the plant popu- lation is above a certain minimum. With hand thinning, this minimum is about 20,000 plants per acre (8-inch spacing), while with mechanical thinning or plant- ing to a stand, it is nearer 30,000 plants per acre. The reason for this difference is that, with the latter methods, the plants are not so uniformly spaced, and there are more clumps with two or more plants together. Yield tended to decrease with populations above 60,000 per acre, with all three methods. Plant characteristics. Fruiting node height is the distance from the soil sur- face, at the base of the plant, up the main stalk to the node (base) of the first branch having a fruit or boll. This height varied directly with differences in the plant populations, in all the tests. It ranged from as little as 2 inches, with populations of less than 10,000, to as much as 10 inches with populations of more than 60,000 plants per acre. Type of plant growth also varied with different populations (fig. 5). With low popula- Fig. 5. Effects of spacing on plant characteristics. Left: 16-inch spacing; right: 4-inch spacing. [10] iH m as O 00 co a> "«tf t- co ^ in co o oo U3 OS ■^ ^ «tf CO CO IO ^ ^ (N ^ ^ ^ tC P3 M rH as o as a> as as as as as OS OS 05 OS OS S M o S O (3 O rH 00 tO tJ( lO« H O in rH 00 o t> in m ■*# Tf CN CO CM . co co m m H OS OS as as as as as as . os as as as .2 a i H OS OS rH . a> ^ H rj t> i> in as as as CM CO CO ; co in id . as as as c a tH O 00 CO to rH t* ^ m l> O OS 00 00 CO in O t» CO ^ N u r-i rH bo c *> c a "en © rH W to as CO ^ CN 00 S OS O 00 co m t- co m co +■ 5.9 *H rH «d^ o .a bo D & Q. OS ■ 00 to t- o • co as co H OS rH ; to CO CN CN ; CO co" CO *■ C fcu> 8 « Q. rH co t> CO 00 ■ '3 in co co o oo O N CO t CD to OS rH to t> co to • s lO CO CO ^ H TJ co co co t- t- •^ CN CN cm* c4 • l-t ,£5 CN tN CN CN M as CN CM CN CN CN IA "as" 3 •»* f W "O n O > • cm oo m co "aj 0> S M 00 to t- w -3 CO t- ^ CO • as oo t- co It UJ C HB 3 OS rH 3 cn cn cn c4 3 -•■» xn cn cn cm* cm* o •a : CM CN CN CM O -*> CO o OS • to w cn ic 3 • in o co HI (A OS rH : cm oo m in cn cm cm • t> cq in ; cn cn cn s 4) H" ^^ Ph % Ml !■ s >- bo o. 3 O. CO k W 0) 1 t- CM «■ < rH OC rH* - C £ «s ft W w 2 ■» O ca «5 cm 00 00 O M ^ • rt< m rH m tr CS CO rH 0C t- in CN rH ■ w Tjt co cn as a. as co* a: 1 as in rH co oc ) o co t> as as w co r-i rH in CO CN rH t- ^ CO CN rH Table 5 — Comparison of Planting Cotton by and by Hill Dropping Continuous Drilling Factor 1949 1950 Hilldrop Continuous Hilldrop Continuous 11.7 7.1 35.4 21.6 4.4 8.1 2.74 2.44 6.4 5.2 92.2 93.1 5.6 6.1 Planting rate (pounds seed per acre) 12.6 48.6 14* 2.60 6.2 96.3 8.3 13.5 48.9 3.2 2.75 6.3 96.2 8.2 7.6 28.1 5.6 2.60 4.8 95.6 8.1 11.7 36.0 12* 2.49 5.9 93.8 5.1 Plants per acre (thousands) .... Average plant spacing (inches). . Yield (bales per acre) Height of first fruiting node (inches) Picker efficiency (per cent) Trash (per cent) . * Center to center of hills. tions, the plants were bushy and had rela- tively large main stalks and lateral branches. With larger populations, the plants were more spindling, and had fewer and shorter lateral branches. Efficiency of mechanical pickers. Picker efficiency was calculated by divid- ing the amount harvested by the total yield, and was highest with the greatest plant populations. While the differences were not large, they were consistent in all tests. The increased efficiency is attrib- uted to the greater height of the lower bolls and to the smaller lateral branches on the plants. Trash and grade. Trash content of the seed cotton in the various tests ranged from 4 to 8 per cent, for first picking, but did not vary consistently with the plant populations. In general, the most Table 6 — Comparison of Hand and Mechanical Thinning of Cotton Chopper 1949: Hand thinned . . Dixie Mayco 1950: Hand thinned . . Mayco B.&P.Weeder Winter Weiss . . Dixie B. & P. Chopper Eversman Plants per acre 19,300 31,600 25,800 23,600 45,700 45,100 37,100 30,800 29,100 19,200 Average plants per hill 1.3 2.2 2.1 1.2 2.6 2.3 2.3 2.1 1.9 1.8 Spaces under 12 inches per cent 100 95 99 92 92 87 77 Yield bales per acre 3.09 3.05 3.07 2.84 2.92 2.92 2.86 2.72 2.78 2.63 [12] trash was obtained with the higher popu- lations. This was partially due to the fact that less defoliation was accomplished be- cause of the denser growth. Continuous drilling versus hill dropping. Tests were made in 1949 and 1950 to compare continuous drilling with hill dropping to a stand. A seeding rate of 12 to 13 pounds per acre and hill spacings of 12 and 14 inches were used for the hill-drop planting. This gave 3 to 5 seeds per hill. Length of hill varied from 1 to 4 inches. A rotary valve attach- ment in the seed tube of the regular planter was used for hill dropping. The continuous drilling was done at two seed- ing rates — one the same as that used for hill dropping, and the other about 60 per cent of that rate. The results showed that, with the same seeding rate .and plant population, continuous drilling gave a higher yield than did hill dropping. Con- tinuous drilling with the lower seeding rate and plant population gave the same yield (table 5). Results of these tests in- dicated that there is no advantage in hill- drop planting from the standpoint of yield and mechanical harvesting. Mechanical choppers. Tests com- paring mechanical thinning (by differ- ent makes and types of choppers) with hand thinning were conducted during 1949 and 1950. The choppers were both ground- and power-driven types, some with rotating and others with oscillating blades (figs. 6 and 7). No attempt was made to set the choppers to give the same lengths of spaces and hills. Results showed that the plant population left after thinning, rather than any effect of the chopper, determined the yield (table 6). While some of the choppers did a cleaner job of thinning than did others, there was no indication that one chopper gave better results than the others when the plants per acre after thinning were • "W W i /ft + 6. Cotton chopper that is operated by contact with the ground. Chopper is set at an angle with the plant row, as shown. [13] Fig. 7. Two-row, power-operated mechanical thinners. Top: tractor-mounted thinner driven from power take-off. Bottom: trailer type thinner driven by power from wheels. the same. With hand thinning there were no spaces over 12 inches. With mechan- ical thinning, when the remaining plant population (20,000 per acre) was ap- proximately the same as the hand- thinned, more than 20 per cent of the spaces were over 12 inches. When over 30,000 plants per acre were left by mechanical thinning, less than 10 per cent of the spaces were over 12 inches. A thick, uniform stand is essential for mechanical thinning to insure a desir- able final stand with a minimum of long spaces between plants. (See Appendix, p. 33, for a method of determining plant population.) [14] WEED CONTROL by late cultivation (after lay-by time) was possible with special equipment. . . . Flaming also gave good control. . . . Selective oil sprays were effective in small cotton (2 to 6 inches). Weed control remains the principal problem to be solved in the achievement of complete mechanization of cotton growing. The studies on this problem were concentrated on the development and testing of equipment and methods for late cultivation, flaming, and spray- ing with oil and chemicals. Late cultivation. Cultivation is still the cheapest method of eliminating weeds wherever and whenever it can be done. Its use is limited to the space between the plant rows, and to the period up to "lay- by" time when the plants become so large that they would be damaged during the operation. Normal lay-by time, with reg- ular equipment, is usually the latter part of July. Since irrigation may continue up to the first part of September, late-grow- ing weeds sometime become a problem. Weeds and grasses are particularly un- desirable with mechanical harvesting as they may cause "grassy" cotton. It may also be desirable, in some cases, to loosen the soil before irrigating to get better water penetration. Special shielded equipment, consisting of a high-clearance tractor (36" under axle) and cultivator was designed and tested for late cultivation (after normal lay-by time). This equipment had 22- inch sweeps with special shanks mounted on the rear tool bar, for cultivating, and burners on the front cultivator gangs for flaming (see flame weed control) at the same time (fig. 8). The sweeps were set relatively flat, and at a depth of 2 inches or less. They were equipped with rod ex- tensions for smoothing the bed (figs. 9 and 10). All test plots were cultivated with regu- lar equipment until normal lay-by time, but all cultivating after that was with the high-clearance, shielded equipment, at 3.1 miles per hour. Results showed that the yield was not significantly reduced nor was the picker efficiency affected by the late cultivation (table 7). Fig. 8. High-clearance, shielded cultivator and flamer. Front view (left) shows where burners are mounted under small shields that extend from bottom of main shields. Rear view (right) shows fuel tank for flamers and tool bar equipped with 22-inch sweeps. 3* 5 \N '^^S^T^: Standard 22" sweeps Fig. 9. Left: twenty-two-inch cultivat- ing sweep with extension rods to smooth shoulders of beds (see Fig. 10, right). Fig. 10. Below: bed profiles made with 22-inch cultivating sweeps. O"- 5"-. 3£ I ^ ^ 3£3 22" sweeps with 10" rods \V 4£- Table 7 — Effects of Late Cultivation on Yield of Cotton Date of last cultivation 1950: August 1 * . . . August 16 . . 1951: July 14* August 7 . . . . 1952: July 31* August 7 . . . . , August 15 . . . September 2 1952: July 31* August 7 . . . . August 15 . . . September 2 1953: July 24* August 5 . . . . August 14 . . . August 31. . * Check plots. Plants per acre 20,000 20,000 35,000 35,000 57,000 57,000 57,000 57,000 17,600 17,600 17,600 17,600 27,000 27,000 27,000 27,000 Yield bales per acre 2.80 2.72 2.66 2.59 2.62 2.57 2.67 2.65 2.54 2.60 2.58 2.49 2.96 2.86 2.88 2.83 Picker efficiency per cent 96.3 95.3 94.4 94.5 97.0 97.2 96.8 96.3 95.0 95.5 95.1 94.9 95.3 95.3 95.5 94.4 [16] Flame weed control. The use of flame to control weeds in the plant row was tried a number of years ago, but never became popular because of damage to the cotton. About 1948, a new type burner was developed which revived in- terest in this method of control (fig. 11) . The new burner gives a relatively short, flat flame. When mounted at a 45-degree angle, the burner practically eliminates the tendency for the flame to hit the ground, or clods, and deflect up into the cotton plant. Liquefied petroleum gas (propane and butane) is used for fuel at rates of about 2 to 4 gallons per acre. The nozzle in the burner is a standard, flat-spray type with a rating of 0.4 gal- lons of liquid per minute at 40 pounds per square inch pressure. One burner is mounted on each side of the row (stag- gered) so that the flames strike the ground about 1 to 2 inches from the plants. The tips of the burners are ap- proximately 8 inches out from the plant row and 8 inches above the base of the plants. Other equipment needed for a Fig. 11. Burners for flaming, mounted on cultivator gangs. Curved rod attachments lift cotton branches and allow better access to weeds. complete flaming setup includes a special fuel tank with the necessary gauges and valves, a vaporizer to change the fuel from liquid to gas, a pressure regulator, control valves, and special fuel lines and fittings. Tests were conducted in 1949, 1950, and 1951 on the flaming of cotton, using constant pressure (30 p.s.i.) and tractor Table 8- —Effects of Flame for Weed Control on Cotton Yield* Date of first flaming Number of flamings Yield 1949: June 7 9 8 7 3 3 5 4 bales per acre 2.21 June 24 2.22 Julyl 2.00 Check 2.11 1950: July 8 2.56 July 24 2.57 Check 2.63 1951: June 5 2.85 June 18 2.74 Check 2.70 All flaming in these tests was done with 30 p.s.i. fuel pressure and a speed of 3 m.p.h. [17] speed (3 m.p.h.), to determine the effect on yield. In these tests, the time of the first flaming and the number of flamings were varied. The earliest applications were made when the plants were 8 to 10 inches tall and about %g inch in diam- eter. The results showed no reduction in yield because of the flaming (table 8) . Tests to determine the effect of dif- ferent intensities of flame were conducted in 1952 and 1953. A constant tractor speed of 3.1 miles per hour was used, and the flame intensity was varied by chang- ing the fuel pressure (20 to 50 p.s.i.) at the burner nozzle. Six flamings were ap- plied in 1952, beginning June 13 and ending August 7. Five flamings were applied in 1953, beginning June 30 and ending August 4. Results showed no sig- nificant difference in yield among the various treatments (table 9). With the high flame intensities, there was some damage to the lower leaves in small cot- ton, but this did not seem to affect the growth or yield of the plants. Flaming was a definite help in con- trolling weeds and grasses, especially when they were in the seedling stage. Older weeds were more difficult to kill, and required high flame intensity or re- peated applications. The theory of flam- ing is not to burn the weeds, but to apply enough heat so that the liquid in the weed cells expands and ruptures the cell walls, causing the plant gradually to wilt and die. For best results, flaming must be done with the right equipment prop- erly set up and adjusted, careful oper- ation, and proper timing of the applica- tions. Weed control with oil in young cotton. Selective herbicidal oils for weed control in small cotton (2 to 6 inches high) have been used in some southern states (National Cotton Council, 1951 ). 3 These oils are specially made to have a certain boiling temperature and naph- thene and aromatic content. These char- acteristics make it possible to kill weeds but not harm cotton plants when the oils are applied only to the stems while they are still in the waxy stage. Experiments were conducted in 1953 to determine the possibility of using this type of oil under California irrigated conditions. The seedbeds in these tests were specially prepared by leveling with the wings on the planter runners and by smoothing with rollers (fig. 12). The rollers were run over the beds a day after planting so that the surface soil had time to dry and not form a crust. The seeds were planted high on the beds so that the 3 See "Literature Cited" for citations, re- ferred to in the text by author and date. Table 9 — Effects of Intensity of Flame for Weed Control on Cotton Yield* Fuel pressure Fuel per acre Tear and yield Treatment 1952 1953 6 flamings 5 flamings p.s.i. gallons bales per acre bales per acre Check 2.13 3.08 Flamed 20 2.15 2.21 3.07 Flamed 30 2.86 2.19 3.10 Flamed 40 3.69 2.03 3.12 Flamed 50 4.21 2.11 3.13 All flaming done at 3.1 m.p.h.; first flaming when plants were 8 to 11 inches high. [18] irrigation furrow could be developed and cultivation could be done without throw- ing dirt onto the portion of the bed where the oil was applied. Special shoes with shields also helped to keep dirt off the oiled portion (fig. 13). The oil was ap- plied to a band 8 to 10 inches wide, centered on the plant row, by two spray nozzles mounted on parallel acting shoes 8 inches apart (fig. 14) . The nozzles were a type which made an 80-degree, flat-fan spray, and each was capable of deliver- ing .05 gallons of liquid per minute at 40 pounds per square inch pressure. They were operated, however, at 20 pounds pressure. Tests were first conducted to determine the best nozzle arrangement for a com- bination of greatest weed control and least plant damage. In these tests, four different nozzle arrangements were used. With one arrangement, the nozzles were set at a height of 1 inch and perpendic- ular to the plant row so that the entire spray pattern was directed across the row. With the other three arrangements, the nozzles were mounted at heights of 1, 2, and 3 inches, respectively, and were directed to the rear and in at an angle of 30° toward the plant row. With this arrangement, only a portion of the spray pattern was directed across the drill row; consequently, less oil was applied to the plant stems. With all the different ar- rangements, the sprays were directed slightly downward so that the oil struck the ground about 6 inches from the nozzles and was applied to the plant stems at a height of less than 1 inch. Two applications were made, the first when the plants were an average of 2 Fig. 12. Top: rollers smooth planted seedbed to prepare for weed control with oil. Fig. 13. Center: shields used during cultiva- tion to keep dirt from being thrown onto oiled part of plant row. Fig. 14. Bottom: spray nozzles for applying oil to small cotton are mounted on parallel- acting shoes. VLa*«? ■d&'>* & mR. 4 v# *t ^ Table 10 — Effects of Nozzle Arrangements for Applying Oil in Small Cotton — 1953 Nozzle arrangement Weeds killed Plants damaged Apparent after 5 days* Definite after 13 days Perpendicular to row, 1 " high per cent 92.5 90.0 71.0 65.0 per cent 65.0 38.0 16.0 6.0 per cent 25.0 4.0 3.5 6.0 30° to row, 1 " high 30° to row, 2 "high 30° to row, 3 " high * Showing discoloration or possible damage 5 days after second oil application. inches tall, and the second five days later when the plants were 2% inches high. Nutgrass was the main type present in the plots, but there were also some other weeds and grasses. A count was made five days after the second oil application to determine the percentage of grass killed. Cotton plants were inspected five days and 13 days after the second appli- cation to determine the percentage of plants damaged. The results showed that directing the spray perpendicular to the row gave the best grass kill, but also caused the most plant damage (table 10) . The combination of best weed kill and least plant damage was with the nozzles to the rear, in at a 30-degree angle to the row, and 1 inch above the ground. To determine the effect of the early oil applications on final yield, two tests were made using the nozzle arrangement de- scribed above. In these tests, oil was ap- plied at rates of 5 and 7 gallons per acre. Three applications were made — the first when the plants were about 2 inches tall, and the last when the plants were 5 to 6 inches high. Results showed no reduction in yield or picker efficiency with any of the oil treatments (table 11) . Early weed control by this method requires not only a special oil, but also extreme care in application. Only oils recommended for this purpose by the manufacturer should be used, and they should be applied ac- cording to his recommendations. Chemical weed control. During 1949 to 1951, several chemicals were tested for both preemergence and mid- Table 11 — Effects of Oil Applications for Weed Control on Yield in Small Cotton Factor Test A Test B Check 5 gals. oil per acre 7 gals. oil per acre Check 5 gals. oil per acre 7 gals. oil per acre Plant population (per acre) Yield (bales per acre) Picker efficiency (per cent) 35,000 3.09 95.2 33,500 3.10 94.4 36,000 3.15 95.4 32,000 2.99 95.0 32,000 3.00 94.7 31,000 2.99 95.2 [20] season weed control. The preemergence applications gave little control and did not prove practical under the conditions of the trials. One of the main difficulties was that, because of dry conditions, the chemicals remained inactive. The mid- season applications gave poor to good control, depending on the chemical, but they also damaged the cotton plants in some cases. More information is needed concerning different chemicals, methods and time of applications, and the eco- nomic feasibility of this type of weed control. Work on these problems is con- tinuing under a project conducted by weed control specialists. MACHINE TOPPING helped prevent lodging of rank- growing cotton . . . did not reduce the yield. Tall, rank-growing cotton has a tend- ency to lodge (fall over) so that it is diffi- cult to harvest by either machine or hand labor. Because tall cotton is also difficult to defoliate, yield is sometimes reduced by boll rot in the heavy growth near the ground. To prevent lodging, some grow- ers practice topping. This consists of Table 12 — Comparison of Yield and Picker Efficiency in Topped and Untopped Cotton Treatment 1951: Check (untopped) Hand-topped Machine-topped — 48 " 1952: Check (untopped) Hand-topped Machine -topped — 46 " 1952: Check (untopped) Hand-topped Machine-topped — 42 " 1953: Check (untopped) Hand-topped Machine-topped — 48 " 1953: Check (untopped) Machine-topped — 42 " Machine-topped — 48 " Machine-topped — 54 " Yield Picker efficiency bales per acre per cent 2.33 92.7 2.33 93.3 2.29 93.0 2.94 92.7 2.97 95.5 2.87 95.5 3.19 95.0 3.20 95.0 3.00 94.7 2.04 91.0 2.36 93.6 2.22 93.7 2.22 93.7 2.32 94.1 2.26 94.6 2.23 92.6 [21] Fig. 15. Four-row cotton topper mounted on specially built tractor. Topping is done by four horizontal revolving blades under frame on front. cutting off the terminal bud of the main stalk to prevent tall growth. Several dif- ferent types of machines have been built to do topping mechanically, two of which are shown in figures 15 and 16. In hand topping, only the terminal bud on the main stalk is removed, whereas with ma- chine topping, all the lateral branches are cut off above the height for which the machine is set. Tests were conducted in 1951, 1952, and 1953 to determine the effect of both hand and machine topping on lodging, yield, and mechanical har- vesting. The results are given in table 12. 1951 tests. One set of plots was used to compare the effects of no topping, hand topping, and machine topping at a height of 48 inches. Topping was done on August 8 when the plants were 4 to 5 feet tall. The results showed no signifi- cant difference in yield or in picker effi- ciency among the three tests. There was no lodging in the topped cotton, and only about 10 per cent in the untopped, since the latter did not reach a height sufficient to cause lodging. 1952 tests. Two sets of plots were used in these tests to compare no topping, hand, and machine topping. In one set, Fig. 16. Two-row cotton topper on the picker, made by removing the picker head and mount- ing a stalk cutter with a horizontal revolving blade. machine topping was done at a height of 46 inches when the plants were 4 to 5 feet tall. In the other set, the plants were about 4 feet high and were machine topped to 42 inches. The average plant population in all the plots was about 50,000 per acre. Topping was done on August 8. Where the cotton was machine topped to 46 inches, there was little dif- ference in yield among the three treat- ments, but the topped cotton was better from the standpoint of picking efficiency and lodging. Approximately 75 per cent of the untopped cotton lodged, while no lodging occurred in any of the topped plots. Where machine topping was done at a 42-inch height there was some re- duction in yield. No lodging occurred with any of the treatments in this test, and there was practically no difference in picking efficiency. 1 953 tests. Two tests were conducted. The first compared cotton untopped, hand topped, and machine topped to 48 inches. In the second, untopped cotton was compared with machine topped at heights of 42, 48, and 54 inches, respec- tively. The cotton for these tests averaged about 35,000 plants per acre. Topping [22] was done on August 3 when the plants were a maximum of 5 feet in height. In the first test, growth was quite rank, and the plants in the untopped plots ranged up to 6 feet in height at the time of harvest. Approximately 50 per cent of the untopped, about 25 per cent of the hand-topped, and less than 10 per cent of the machine-topped cotton lodged. The lodging reduced yield and picker effi- ciency in the untopped treatment. De- foliation (defoliants applied by airplane) was very poor in the lodged cotton (about 25 per cent defoliated). In the second test, growth was vari- able, and only a small portion was rank enough to cause lodging. None of the cotton topped to 42 and 48 inches lodged, but lodging occurred in about 20 per cent of the untopped and of that topped to 54 inches. Part of the lodging in the 54-inch was caused by the machine, which knocked down some of the plants during the topping operation. Those plants remained down. There was little difference in yield and in picker effi- ciency among the various treatments. The studies indicated that topping is a definite help in reducing lodging, and that it can be done mechanically without reducing yield. Further studies are being conducted to determine the best time and height for topping. Indications are that topping at about 4 feet should be done when the majority of the plants are be- tween 4 and 5 feet high and when bloom- ing begins near the 4-foot level — usually the last week in July or first week in August, in the Shafter area. DEFOLIATION improved the grade of mechanically harvested cotton slightly . . . had little effect on picker efficiency. Defoliation of cotton before harvest has become a common practice. This con- sists of causing the leaves to drop off the plants by use of various chemical de- foliants applied either as dusts or sprays. Results with different materials have varied widely, ranging from good defoli- ation to very poor. Some of the materials (dessicant, or drying, types) cause leaves to dry but not drop off. Tests were conducted in 1952 and 1953 to determine the effect of defoliation on grade and picker efficiency of mechani- cally harvested cotton. Three treatments were tested: (1) cotton that was not de- foliated (check) ; (2) cotton defoliated as completely as possible; and (3) cotton partially defoliated — the leaves were dry, but most of them remained on the plant. The defoliants (sodium chlorate and penta chlorophenol) were applied by a ground sprayer (fig. 17). Cotton in the 1952 tests was planted to a stand, and averaged about 50,000 plants per acre. The plants at time of harvest (first picking October 10) were 3 to 4 feet high, and standing erect. Samples for grade were taken at the time of first picking, but were not ginned until a month later. This gave the green trash time to dry. In 1953 the cotton was also planted to a stand, and averaged about 35,000 plants per acre. The plants were 3 to 4 feet high at the time of harvest (first picking November 9). Grade samples were taken the day after harvest when the cotton was ginned. Results showed that, with good defolia- tion, the grade of the cotton was some- what higher but there was little effect on the efficiency of mechanical harvesting under conditions of the tests (table 13). These results were similar to those ob- tained in tests at the Delta Branch Ex- periment Station, Stoneville, Mississippi (Colwick, 1953). [23] Table 13 — Effects of Defoliation on Picker Efficiency Mechanically Harvested Cotton and Grade of Treatment First picking results Final results Picker efficiency Tield Trash Grade* Picker efficiency Yield 1952: No defoliant, green leaves per cent 93.8 94.3 93.4 91.2 92.5 92.3 bales per acre 2.13 2.09 2.03 2.32 2.24 2.32 per cent 7.3 7.0 4.6 5-M, 3-SLM + 5-M, 3-SLM + 8-M 4-M 1-M, 3-SLM 1-M +, 3-M per cent 95.4 93.9 95.4 95.0 92.5 94.0 bales per acre 2.27 2.22 2.09 2.45 2.46 2.48 Partial defoliation, dry leaves Defoliated 80 to 90 per cent 1953: No defoliant, green leaves Partial defoliation, dry leaves Defoliated 70 to 80 per cent * 8 samples in 1952, 4 .< samples in 1953. MECHANICAL COTTON PICKERS vary in efficiency, de- pending on type of plants, condition of picker, and skill of operator. . . . Cost for mechanically picked cotton was from $15 to $20 less, per bale, than for hand picked. The number of mechanical cotton pickers in California has increased from a few experimental machines, in 1945, to an estimated 5,500 in 1952. The pick- ers have been designed and developed almost entirely by four manufacturers. All four makes employ the same picking principle (revolving spindles), but differ in type and arrangement of spindles. Two use gear-driven, cone-shaped, barbed spindles (fig. 18) that are mounted on bars arranged to form a vertical, revolv- ing cylinder. The other two use round rod spindles that are either smooth or fluted by lengthwise grooves. These spindles, mounted on vertical slats fas- tened to chains at top and bottom, are turned by friction between a roller on the spindle and a stationary track. Both types of pickers are available in one- or two-row models, some tractor- mounted, others self-propelled. Both of the barbed types have spindles that pick from both sides of the row. One of the rod types can be arranged as a one-row machine, with spindles picking from both sides of the row, or as a two-row machine with spindles picking from one side only. The other rod type is a two- row machine with spindles on only one side of each row. Studies on the pickers consisted of field tests to determine picking efficiency, grade of cotton from different types of plants, and operating characteristics. Some cost data were also obtained. [24] Fig. 17. Above and below: two types of high-clearance spray machines used for applying defoliants and for late insect control. [25] Fig. 18. Cotton picker spindles. Left: fluted rod; center: smooth rod; right: barbed cone. Picker tests, 1951. These tests were made with a single-row, barbed spindle picker and a single-row, smooth rod spindle type. Both machines were oper- ated in the same field at the same time. The plants were 3 to 4 feet in height, standing erect, and the plant population was between 35,000 and 40,000 per acre. The field was free of weeds and grass, and defoliation was good although there was some second growth. Tests were for first picking only, which yielded about 2^ bales per acre. Results showed the barbed spindle machine to have a higher picking efficiency than the smooth spindle (table 14) . However, if a second picking had been made, the difference probably would have been reduced since over half the cotton left was still on the plants. The trash was less and the grade higher with the smooth spindle machine. Picker tests, 1952. Two tests were conducted with all four makes of pickers (fig. 19). Test No. 1 was in uniform cotton that was 3 to 4 feet in height with all plants standing erect. The plant popu- lation was 40,000 to 45,000 per acre, and defoliation was about 80 per cent corn- Table 14 — Results of Cotton Picker Tests — 1951 (First picking only) Factor Picker efficiency (per cent) Cotton left on plant (per cent) Cotton knocked on ground (per cent) Trash (per cent) Grades (7 samples) : Classer No. 1 : Middling + Middling Classer No. 2: Middling Strict low middling + Strict low middling Type of picker Barbed cone spindle 94.9 3.0 2.1 7.4 3 4 2 1 4 Smooth rod spindle 90.4 5.1 4.5 4.4 6 1 [26] plete. The average yield was approxi- mately 2 bales per acre. In this test, the machines were checked on both first and second pickings — October 9 and 31, re- spectively. About 90 per cent of the yield was harvested in the first picking, from which all 11 samples for grade were taken. Samples were taken from the trailers when harvesting, from the press box during ginning, and from the bales after ginning. Test No. 2 was in cotton which varied from short (2% feet), on one end of the field, to tall (6 feet) , rank cotton on the other end. The short cotton was standing erect, and was about 90 per cent defoliated, whereas the tall cotton was about 40 per cent lodged and 75 per cent defoliated. The plant population averaged 45,000 per acre, and average yield was 2% bales per acre. The test was made only on first picking, on Oc- tober 31. Gleanings for field losses were made in both the short and rank cotton, and the picker efficiency was calculated from the average of the two gleanings. Samples for trash were taken from the trailer at the time of picking, and samples for grade, from the press box during ginning; thus they were a composite of both types of cotton. Results showed that, with good picking conditions, in short to medium high cot- w^p Fig. 19. Four pickers used in 1952 tests. Top: one- and two-row, barbed spindle types. Bottom: one- and two-row, rod spindle types. [27] Table 15 — Results of Cotton Picker Test No. 1 — 1952 (First and second pickings) Picker efficiency, two pickings First picking Type of picker Picker efficiency Loss on ground Trash per cent 3.8 4.7 6.1 9.3 Grades, 11 samples M SLM Round rod spindle, fluted* Round rod spindle, smooth Barbed cone spindle, 1-row Barbed cone spindle, 2-row per cent 95.4 95.8 95.3 96.5 per cent 90.5 94.3 93.7 93.7 per cent 3.6 3.8 3.2 2.7 6 11 11 3 5 8 * Spindles picking from one side of row only. All other pickers had spindles on both sides of row. ton, all the pickers had a high efficiency (tables 15 and 16). With tall, rank cot- ton, however, the barbed spindle pickers did a better job than did the rod spindle type. It was also noted, in running the tests, that there was less mechanical trouble with the barbed spindle type in the heavy, rank cotton. Cotton picked by the rod spindles had less trash than that picked by the barbed spindles, but the grade was determined by the make of picker rather than the type of spindles. Another study of pickers, made in 1952, was in cooperation with a grower who had four machines representing three different makes. Three of the pick- ers were the barbed spindle type, and one a fluted rod spindle type. They were used from September 26 to December 16 to harvest 2,108 bales of first and second picking cotton from approximately 1,150 acres. During first picking the machines were operated in the same fields at the same time; part of the cotton picked by each machine was ginned separately; and a record was kept of the bale classings. In addition, six checks were made on each machine during first picking, to de- termine picking efficiency. Because of mechanical trouble in heavy cotton, the bottom row of spindles was removed from the rod spindle picker. The pickers were operated by the regular employees of the ranch. Results of this study (which Table 16 — Results of Cotton Picker Test No. 2 — 1952 (First picking only) Average picker efficiency Field loss Trash per cent Grades, J J samples Type of picker Short cotton Rank cotton M SLM per cent lbs. seed cotton lbs. seed cotton per acre per acre Round rod spindle, fluted* 83.6 287 696 7.3 3 Round rod spindle, smooth 87.2 224 507 6.4 1 2 Barbed cone spindle, 1-row 90.3 168 387 9.5 1 2 Barbed cone spindle, 2-row 90.6 232 340 10.2 3 Spindles on one side of row only. All other machines had spindles on both sides of row. [28] was for first picking only) showed a rela- tively wide range of picking efficiencies, not only between different pickers, but also for individual machines (table 17). These variations in efficiency were more a result of the kind of cotton picked, the skill of the operator, and the condition of the picker, than of the type or make of machine. The rod spindle machine had the lowest efficiency, but this was partly due to operating without the bot- tom spindles and picking only from one side of the plant. The cotton picked with this machine averaged slightly higher in grade than did that picked by the other machines. Picker tests, 1953. Tests were con- ducted to compare two different sizes of the same make of barbed cone type spindle picker in cotton with various kinds of plant growth. One of the pickers had 20 spindles per bar and a throat height of 32 inches; the other had 14 spindles per bar and a throat height of 21 inches. The larger machine was mounted on a 40-horsepower tractor and had a picking speed of 2.2 miles per hour; the smaller one was mounted on a 22-horsepower tractor and operated at 2.6 miles per hour. Cotton for one test had been topped to a uniform height of 46 inches, and was all standing erect. It was well defoliated (90 per cent or bet- ter), had no green foliage, but did have some green bolls. The plant population varied from 30,000 to 40,000 per acre. Cotton for the other test varied in height from 3 feet, on one end of the field, to 5 feet or better on the other end. The plant population was 30,000 to 35,000 per acre, with rank growth, and about 60 per cent lodged. Defoliation was poor (about 50 per cent) , and there were con- siderable green leaves and bolls. Results showed no significant differ- ence in picking efficiency between the two pickers in either test (table 18). There was some difference in the trash content of the cotton picked by the two machines, but the grades were practically the same. When picking against lodged cotton, the small machine seemed to break up the Table 17 — Results of Cooperative Picker Study With Private Grower — 1952 (First picking only) Factor Plants per acre, range (thousands) . . Plants per acre, average (thousands) Yield, range (bales per acre) Yield, average (bales per acre) Picker efficiency, range (per cent) . . . Picker efficiency, average (per cent) . Bales (classed) Grades (per cent of bales ) : Middling Strict low middling Low middling Strict good ordinary Barbed spindle, 1-row 25-37 31 2.1-2.7 2.3 90-95 92.5 158 10.8 60.8 25.3 3.1 Type of picker Barbed spindle, 1-row 20-39 29 2.1-3.3 2.5 84-96 91.3 264 7.6 64.4 23.9 4.1 Barbed spindle, 2-row 23-39 30 1.9-3.1 2.6 77-93 86.8 533 1.2 64.3 25.5 9.0 Fluted rod spindle, 2-row* 22-28 26 2.1-3.0 2.5 80-86 82.7 348 9.8 75.3 14.9 * Spindles picking from one side of row only. [29] Table 18 — Results of Cotton Picker Tests — 1953 (Comparison of same make of barbed cone spindle pickers of different size. First picking only) Factor Picker size (spindles per bar) Yield (bales per acre) Picker efficiency (per cent) Trash (per cent) Grades : Strict middling Middling Strict low middling + Strict low middling Green bolls knocked from plant (per cent) Test 1 Test 2 14 20 14 2.57 2.58 2.23 93.9 94.2 91.3 5.6 4.4 10.4 1 3 4 3 1 10 16 14 20 2.25 91.4 7.4 11 plants more than did the large one. No mechanical trouble or clogging was ex- perienced with either machine. Harvesting costs. Cost studies of mechanical harvesting were not possible at the Shafter Experiment Station be- cause of the small acreage harvested and the intermittent use of pickers. A sep- arate, detailed study was made, however, during 1948, 1949, and 1950 in the San Table 19 — Costs per Bale of Cotton, for Mechanical and Hand Harvesting, as Reported by Three Growers Factor Grower A B c 1948 1949 1949 1950 1949 Number of mechanical pickers used 2 332 4 283 10 295 13 164 13 232 Average bales harvested per picker Labor $ 4.30 1.10 8.70 14.10 4.10 9.00 8.90 $ 4.40 1.30 6.00 11.70 4.80 10.60 5.10 $ 4.20 1.20 3.50 8.90 4.60 8.10 11.20 $ 4.30 1.10 8.60 14.00 8.30 10.60 5.10 $ 6.80 1.80 7.10 15.70 6.10 4.50 2.60 Fuel and lubricants Repairs and overhaul Total operating cost Depreciation and interest Field loss Grade loss Total machine picking cost. . $36.10 $32.20 $32.80 $38.00 $28.90 Hand picking cost $49.00 $47.00 $57.00 $45.00 [30] Joaquin Valley (Burlingame and Bailey, 1950; Bailey and Hedges, 1954). Some cost information was obtained from three large-scale cotton growers who kept records on their own pickers (see table 19). All of the mechanical pickers used by these growers were the same make, single-row, barbed spindle type. All figures given in the table, except those for depreciation and interest, were obtained from the growers. Depreciation and interest were calculated by the au- thors on an assumed picker life of eight years and interest, at 5 per cent per year, on half of the initial cost which was be- tween $9,000 and $9,500 per unit. Another comparison between the cost of mechanical and hand harvesting of cotton in the San Joaquin Valley was obtained by the price charged for custom picking. In 1951 and 1952, this price ranged between 2 and 2% cents per pound of seed cotton, for mechanical harvesting, and between 3 and 4 cents for hand picking. In 1953, mechanical har- vesting charges were between iy± and 1% cents, while hand picking ranged from 2% to 3% cents per pound. LITERATURE CITED Bailey, Warren R., and Trimble R. Hedges 1954. Economics of mechanical cotton harvesting. California Agr. Exp. Sta. Bui. 743. Burlingame, Burt B., and Warren R. Bailey 1950. Cost of harvesting cotton with mechanical pickers — 1948. California Agr. Ext. Serv. (Mimeo.) Colwick, Rex (compiler) 1953. Mechanization of cotton production. Mississippi Agr. Exp. Sta. Southern Coop. Ser. Bui. 33. Meek, Wm. E., and B. B. Ewing 1948. Line diagram method of setting farm implements. Mississippi Agr. Exp. Sta. Cir. 138. National Cotton Council 1952. Weed control in cotton. Prog. Rept. SELECTED BIBLIOGRAPHY Creasy, L. E., and H. L. Barr 1950. Flame cultivation guide. Louisiana State College Agr. Exp. Sta. Cir. 38. Crowe, Grady, and John T. Holstun, Jr. 1953. The economics of weed control in cotton. Mississippi Agr. Exp. Sta. Cir. 179. Hinkle, D. A., W. F. Buchele, and H. S. Stanton 1951. Mechanized production of cotton. Univ. of Arkansas Rept. Ser. No. 22. Holecamp, E. R., W. I. Thomas, and K. R. Frost 1951. Cotton cultivation with tractors. Arizona Agr. Exp. Sta. Bui. 235. Humphries, R. T., J. M. Green, and E. S. Oswalt 1952. Mechanizing cotton for low cost production. Oklahoma Agr. Exp. Sta. Bui. B-382. Smith, Gordon L., H. G. Miller, Jr., and T. F. Leigh 1953. Low volume, low pressure sprayer for cotton insect and spider mite control. California Agriculture. Univ. of California, Berkeley. May. Smith, H. P., et al. 1953. Mechanical harvesting of cotton. Texas Agr. Exp. Sta. Prog. Rept. 1527. Smith, H. P., D. L. Jones, and H. F. Miller, Jr. . 1950. The cleaning of mechanically harvested cotton. Texas Agr. Exp. Sta. Bui. 720. Smith, H. P., and H. F. Miller, Jr. 1952. Performance of stalk cutter-shredders. Texas Agr. Exp. Sta. Prog. Rept. 1444. (Appendix appears on next page) [31] APPENDIX Line diagram method of setting farm implements Row crop farming requires accurate setting of implements for such operations as listing, planting, cultivating, and the like, to obtain greatest efficiency and speed. Setting in the field is often diffi- cult and time consuming because of the uneven conditions. The line diagram method was developed in 1948 at the Delta Branch Experiment Station, Stone- ville, Mississippi (Meek and Ewing, 1948) . It requires a smooth, level surface marked with parallel lines representing the plant rows and the middles between the rows (fig. 20) . For example, with 40- inch row spacing there would be lines 20 inches apart. The equipment is run onto this surface with the wheels exactly over the lines representing the middles. The ground working tools to be used are then set as desired in relation to the plant rows. When the implements are properly set, little if any final adjustment is neces- sary in the field. A wooden floor or concrete slab is best for laying out the diagram with painted lines. A smooth, hard dirt area can also be used by driving spikes into the ground and stretching string or wire between them to represent the lines. The diagram should be large enough so that all the equipment to be set will fit over the lines, which should be accurately spaced. It is helpful to have the lines representing rows a different color from those repre- senting middles. When equipment is used for two different row spacings (for ex- ample, 40-inch and 32-inch), two dia- grams may be laid out on the same area. One can be represented by solid lines and the other by broken or dotted lines — each in a different color. Fig. 20. Line diagram for setting equipment before going into field. [32] A method for determining plant population It is often desirable to know the ap- proximate plant population in a stand of cotton. A simple and quick method of determining the number of plants per acre is to count the plants in a length of row equal to 1/1,000 of an acre, and multiply that number by 1,000. For ex- ample, if the number of plants counted is 32, then the plants per acre would be 32 x 1,000, or 32,000. The following are lengths of row equal to 1/1,000 of an acre, for the common cotton row spac- ings: ROW SPACING LENGTH OF ROW FOR 1/1000 ACRE inches feet inches 36 14 6 38 13 9 40 13 1 42 12 5 A stick or light chain of the proper length for the row spacing to be meas- ured can be used to mark off the length of row for the count. Counts should be made on a number of rows and in several locations in the field to get an average of the plant population. The average spacing of plants in the row, where the distribution is reasonably uniform, can be determined from the following : AVERAGE PLANT SPACING inches 2 4 6 8 10 12 ROW SPACING AND PLANTS PER ACRE 36-inch 38-inch 40-inch 42-inch 87,000 82,500 78,400 74,700 43,500 41,200 39,200 37,300 29,000 27,500 26,200 24,900 21,700 20,600 19,600 18,700 17,400 16,500 15,700 15,000 14,500 13,700 13,100 12,500 ACKNOWLEDGMENTS The authors wish to acknowledge and give credit for the work and cooperation of the following persons: J. P. Fairbank, formerly with the De- partment of Agricultural Engineering and now Regional Director of the Agri- cultural Extension Service, Berkeley, who was Project Leader from 1946 through 1949; K. 0. Smith, formerly Agricultural Engineer with the United States Depart- ment of Agriculture, who worked on the project from July, 1948, to August, 1950 ; B. B. Ewing, formerly Agricultural Engineer with the United States Depart- ment of Agriculture, who was located at the Shafter Station from October, 1950, to December, 1951; Donald Little, Lab- oratory Mechanic for the Mechaniza- tion Project since 1950; Marvin Hoover, Extension Cotton Specialist, Shafter; George Harrison, John Turner, E. Gor- don Smith, Gordon L. Smith, John Pyle, Homer Craig, Allen Robbins, Roy Rob- bins, and Oscar Jones, at the Shafter Ex- periment Station; and Waldo Weeth, farmer, Coalinga. The authors also wish to acknowledge the help and advice of the following organizations: California Planting Cot- ton Seed Distributors; California Cotton Mechanization Advisory Committee; In- ternational Harvester Co., John Deere Co., Allis-Chalmers Co., and Ben Pearson Co., who loaned cotton pickers; and the other companies and dealers who loaned or donated equipment or materials. [33] ->4 In order that the information in our publications may be more intelligible it is sometimes necessary to use trade names of products or equipment rather than complicated descriptive or chemical iden- tifications. In so doing it is unavoidable in some cases that similar products which are on the market under other trade names may not be cited. No endorsement of named products is intended nor is criticism implied of similar products which are not mentioned. Co-operative Extension work in Agriculture and Home Economics, College of Agriculture, University of California, and United States Department of Agriculture co-operating. Distributed in furtherance of the Acts of Congress of May 8, and June 30, 1914. J Earl Coke, Director, California Agricultural Extension Service. 15m-ll,'54(6265)LL * !*£* 1 "stat". . . This is probably the world's largest plow — it was built about 1910. It plowed an acre in four and one-quarter minutes. A swath 60 feet wide was turned under by 55 bottoms, pulled by three oil-burning tractors. The monster plow was built in sections and assembled for several test runs in the midwest. Impractical? ... do! This "stunt" yielded new knowledge about hitches . . . knowledge that agri- cultural engineers have used in designing many of today's farm implements. For more than 40 years agricultural engineering has offered opportunity to young men of mechanical bent with an interest in agriculture. And as mechanization increases on farms, opportunities in agricultural engineering expand . . . with the GOOD JOBS going to those who are WELL TRAINED. Many leaders in the field were trained at the University of California at Davis. The staff at Davis is recognized nationally and internationally for its accomplishments in teaching and in research. The Department of Agri- cultural Engineering is accredited ... a graduate is eligible for examina- tion for a Professional Engineer's license, or he may continue study toward a master's or doctor's degree. The growing College of Letters and Science on the same campus broadens the student's educational and social back- grounds. For further information . . . about courses and careers in agricultural engineering, write tAr. Roy Bainer, Chairman, Department of Agricultural Engineering, University of California, Davis. Or . . . See the College Entrance Advisor in the office of your local Farm Advisor.