Division of Agricultural Sciences UNIVERSITY OF Economics of MECHANICAL COTTON HARVESTING CALIFORNIA AGRICULTURAL EXPERIMENT STATION BULLETIN 743 Machines picked more than 16.2 per cent of California's 1,268,000-bale cotton crop in 1949. This bulletin presents the findings obtained from interviews with 63 San Joaquin Valley cotton growers, each of whom owned and operated one mechan- ical cotton picker during that year. The study was undertaken to determine the range of situations in which mechanized cotton harvesting is economically sound. Mechanical cotton picking was found to be practicable and profitable when Acreage is 100 acres or more, Yield is 1 bale per acre or more, and Full-season use is made of the machine. Typical operation was estimated as: Full season = 60 working days, or 500 hours Acreage = 200 acres of 1.5-bale cotton Yield = 300 bales Total economic costs of machine picking were computed, analyzed, and compared with the cost of hand picking in a wide variety of situations. Total economic cost includes : Labor Overhead and operating expenses Field waste Reduced sale value, due to reduction in quality In all cases where the basic requirements of acreage, yield, and use of the machine were met, machine harvesting was advantageous. When total economic costs of the 1949 crop were added up and compared, it was found that Per bale, hand picking averaged $45.00 machine picking averaged 25.76 Saving $19.24 Mechanical cotton pickers can be operated most efficiently and grade loss kept at a minimum if cotton is planted and fields are cared for with mechanical harvest- ing in mind. The grower should Space rows 40 inches apart Make the stalk-row ground higher than that between rows Plant stalks 4 to 8 inches apart Keep fields clean of weeds and grass Keep the ground surface smooth and free of clods The Long-Range View 1949 was a year of prosperous conditions and high employment. The advantage of machine over hand harvesting, in the average situation, was considerable. To deter- mine the advantage, if any, when wage costs are lower compared with machine costs, 1938, a year of less prosperous conditions and intermediate employment, was selected as a basis for comparison. When 1949 machine-harvesting costs were con- verted to 1938 prices, it was found that savings, while less, would still be real and substantial. CONTENTS Summary 4 Introduction 9 Effectiveness of Man Labor in Cotton Production 12 The Situation prior to 1940 12 Developments 1941-1950 13 The Situation in 1949 17 Use and Performance of Mechanical Harvesters 18 Summary of machine use 18 Summary of performance rates 19 Discussion 19 Effect of Machine Picking on Gin Turnout 20 Machine-Picking Costs 22 Average picking cost per machine 22 Cost per bale, per hour, per acre, and per 100 pounds of seed cotton . . 27 Discussion 28 Effect of Machine Picking on Grade of Lint Cotton 29 How machine picking can affect grades of cotton 29 Comparative grades in the 1949-1950 season 30 Trends in machine-picked cotton grades 1948-1950 33 Picking Efficiency of Mechanical Harvesters 36 Discussion 37 Economic Analysis of Mechanical Cotton Harvesting 38 Summary of economic costs 38 Effect of changing prices on machine- versus hand-harvesting costs . . 40 Effect of acreage harvested on picking costs 42 Effect of yield per acre on picking cost per bale 45 Interrelationship of acreage harvested, yield, and picking cost .... 46 Comparison of cost in first and second picking 46 Later Developments in Mechanical Cotton Harvesters 48 Cost of picking with mechanical picker Number 2 49 Appendix 52 Acknowledgments 53 Literature Cited 53 Appendix Tables 54 THE AUTHORS: Trimble R. Hedges is Associate Professor of Agricultural Economics, Associate Agricultural Economist in the Experiment Station and on the Giannini Foundation, Davis. Warren R. Bailey is Agricultural Economist, Production Economics, Research Branch, Agricul- tural Research Service, United States Department of Agriculture, Washington, D.C. Assistance in collecting the data and in certain phases of the analysis was provided by Chester O. McCorkle, Jr., Assistant Professor of Agricultural Economics, Assistant Agricultural Economist in the Experiment Station and on the Giannini Foundation, Davis. APRIL, 1954 Economics of MECHANICAL COTTON HARVESTING TRIMBLE R. HEDGES WARREN R. BAILEY A Report of Studies Made in the San Joaquin Valley* SUMMARY California growers are rapidly mech- anizing their cotton picking. Between 1946 and 1951 the number of machines increased from 25 to 3,700, and the por- tion of the crop picked by machine in- creased to 54 per cent. With this rapid development of a new harvesting method many technical and economic problems have arisen. The economic costs of har- vesting cotton mechanically as compared with those of hand harvest include: 1) the cost of machine picking; 2) the re- duction (if any) in grade and value of the cotton; and 3) the additional field waste (if any) attributable to the machine. This study was undertaken to get information on these factors and to * Submitted for publication May 25, 1953. determine the range of situations in which mechanical cotton harvesting is economically feasible. Information on use of machines, cost of picking, and gin turnout was ob- tained by interviews with 63 growers who used machines in 1949. Data on grade and value of machine- and hand- picked cotton were obtained from repre- sentative gins for three seasons, 1948 to 1950. Estimates of field waste from each method of picking were based on experi- mental work at the Shafter Cotton Field Station. In 1949 machine picking did not begin in the San Joaquin Valley until mid- October, when about a fifth of the cotton crop had already been picked by hand. The average use and output of the 63 Table 1. Average Use and Output of 63 Mechanical Harvesters During 1949 Season First pick Second pick Total season pick Days operated Hours operated Acres picked Bales picked Hundredweight seed cotton 26 243 145 182 2,482 21 164 139 47 701 47 407 284 229 3,183 [4] Table 2. Average Performance Rates per Hour and per Day, 1949 Acres picked per hour Pounds of seed cotton per hour Bales picked per day First pick 0.60 1,021.00 7.00 Second pick 0.82 429.00 2.30 Total season pick 0.70 782.00 4.90 mechanical harvesters for that season are shown in table 1. Use and therefore output per harvester varied considerably from farm to farm because some growers who were using machines for the first time did not at- tempt first picking. Others used ma- chines merely to supplement hand pick- ing or bought machines after the season had begun. Thirty-one machines were used at full season's capacity. They picked an average of 292 bales on 356 acres of picking in 520 hours of opera- tion. From this, a grower can expect a machine to harvest completely 200 acres (150 acres picked a second time) or about 300 bales of cotton in a typical season. Growers reported they had fewer breakdowns and kept their machines op- erating more of the time in 1949 than in previous years because they under- stood better the mechanical require- ments. Some moderate improvement can be expected in the performance rates shown in table 2. Gin turnout of machine-picked cotton was 36.5 per cent. This was less than one percentage point below the turnout of hand-picked cotton. It required 1,370 pounds of machine-picked seed cotton, compared with 1,348 pounds of hand- picked, to turn out a 500-pound bale of lint. Turnout of the machine-picked cot- ton was relatively low early in the sea- son; it equaled the hand-picked in mid- season; and it was actually higher than the turnout of hand-picked cotton in late- season picking. Cost of machine picking (excluding grade loss and field waste) for the 63 growers in 1949 averaged $14.65 per bale harvested, $11.72 per acre of pick- ing, and $8.25 per acre of machine op- eration. Highest cost per bale, $20.72, was in the Madera-Merced area where an average of only 158 bales was picked per machine. The lowest cost, $11.71, was in Westside where the average was 292 bales per machine. Overhead items dominate the cost pic- ture. Cost per bale included $7.57 for overhead on harvester and tractor, $4.43 for operating expenses, and $2.64 for labor. The purchase price of approxi- mately $9,500 (in 1949) for harvester and tractor largely explains the high overhead costs. Annual depreciation charges were $1,483, and annual interest on investment was $217 for the equip- ment. The average cost of $14.65 a bale was based on a season pick of 229 bales (somewhat less than full machine ca- pacity in the Valley) and a service life of five years for the harvester (somewhat less than farmers can reasonably expect today). Therefore, the cost of picking was recalculated for a season pick of 300 bales and a service life of six years. On that basis, the cost of picking would be $12.29 a bale. Since overhead is a major item of cost, full use of the machine is important. The effect of annual use on cost of picking per bale (in 1.5-bale cotton) was calcu- lated as shown in the table at the top of page 6. [5] When annual use is: Picking cost acres bales per bale is : 50 75 $21.55 100 150 $14.94 150 225 $13.04 200 300 $12.29 300 450 $11.91 When season use falls below 100 acres, the cost rises sharply, but pushing use beyond 200 acres results in only small additional savings. One hundred acres seems to be the minimum for which it pays a grower to own a harvester. With less than 100 acres, he can better afford to hire custom machine picking. The effect of yield on cost of machine picking per bale, when a harvester op- erates 500 hours a season, would be: When yield per Picking cost acre (in bales) is: per bale is : 0.75 $20.15 1.25 $13.56 1.75 $11.32 2.25 $10.28 The interrelationship of yield and sea- son use as it affects cost of machine pick- ing per bale is shown in table 3. If the yield is high, the picking cost per bale is reasonable even though the season pick is only 100 acres. If yield is low, the cost is comparatively high even though the season pick is 300 acres. Second picking, especially near the end of the season, often is very light. Whether this "scrapping" should be picked by hand, by machine, or not at all is an economic question that faces many farmers. Hand picking, in this sit- uation, often costs $4.00 or more per 100 pounds of seed cotton. The cost of ma- chine picking varies with the pick per acre as shown in table 4. If the grower includes overhead in his machine-picking cost, he would find machine picking less economical than hand picking when the pick is 150 pounds or less. If he con- siders only the direct operating costs, costs of both methods would be about equal when the pick is around 100 pounds. Reduced grades of lint cotton repre- sent one of the economic costs of ma- Table 3. Interrelationship of Yield and Season Use as it Affects the Cost of Machine Picking Season pick Cost per bale 1-bale cotton 2-bale cotton 100 acres $20.96 15.71 $ 12.22 10.39 300 acres Table 4. Cost of Machine Picking in Relation to Pick of Seed Cotton per Acre Pick per acre, in pounds Machine-picking cost per 100 pounds Total costs Operating costs 50 $15.48 7.81 5.21 3.94 $9.64 4.86 3.24 2.46 100 150 200 [6] chine picking. Machine-picked cotton in the 1949 season averaged slightly less than one full grade below hand-picked. Thus, machine bales were concentrated in the grade of Strict Low Middling and hand bales in Middling. Consequently, the machine-picked cotton averaged lower in value. The average government loan value was $132.52 a bale for the machine-picked and $142.84 for the hand-picked, a difference of $10.32 a bale. The difference was $19.75 a bale in the Madera-Merced area, but only $8.06 a bale in Westside. This variation between areas was doubtless due mainly to differences in cotton, weed conditions, and experience of machine operators. The difference in grades and values also varied widely among gins. Even in Westside, where machine-picked cotton more nearly approached the hand-picked in grade, the difference ranged from $13.34 a bale at one gin to $1.47 at an- other. The variation among gins was probably due to differences in equipment and experience of the gin operator with machine-picked cotton. Grades of machine-picked cotton also varied widely among growers in each subarea. But in each, at least one grower had seasonal average grades of Strict Low Middling or better. This suggests that good grades can be obtained with machine picking in all parts of the Val- ley, but that more careful operation is required in some areas. Seasonal trends in grades of machine- picked and hand-picked cotton at eight gins located in various parts of the Val- ley show that machine-picked cotton averaged lower in grade throughout the season. But the spread between grades Figure 1. Transferring the seed cotton from the picker basket to a cotton trailer. Capacity of th< basket is about 750 pounds of seed cotton, or the equivalent of about one-half bale. [7] tended to narrow in late-season picking, and the week-to-week fluctuations in grade were smaller in the hand-picked than in the machine-picked cotton. Although experimental results have shown that mechanical picking was 1.1 percentage point less efficient than hand picking, it is assumed that under general farm conditions the disparity would be 3.0 percentage points. Thus, in 1.5-bale cotton, a mechanical picker would leave 69 pounds more seed cotton per acre. The field value of that much seed cotton in 1949 was about $4.73; it amounted to about $3.15 per harvested bale. This additional field waste caused by the ma- chine is an economic cost of mechanical harvesting when compared with hand harvesting. Most growers said machines "now clean the fields" more thoroughly than in previous years; few were still concerned about field waste. They re- ported that field waste was relatively smaller in rank-growing, high-yielding cotton. The economic advantage or disadvan- tage of machine picking is found by add- ing together the costs of picking, value of grade loss, and value of field waste and comparing the sum with the cost of hand harvesting. For this purpose, the recalculated cost of picking, $12.29 per bale, representing full use and a six-year service life of the harvester, is used. It is assumed that cost of hand harvesting is $45.00 per bale (as in 1949) and that the average yield is 1.5 bales an acre. The total economic cost of machine pick- ing includes $12.29 for picking, $10.32 for grade loss, and $3.15 for field waste, a total of $25.76 per bale, which repre- sents a saving of $19.24 over hand har- vesting. This saving in economic costs varied among the subareas according to differences in grade loss (table 5). Cost of picking with a less widely used, newer make of machine harvester was also obtained by interviews with 18 growers for the 1950 season. The infor- mation obtained was deemed inconclu- sive because this make of machine had been used commercially in the Valley for such a short time. Data suggest, how- ever, that growers using this machine also found their picking cost well below the cost of hand picking. Table 5. Average Savings of Machine Harvesting in the Five Subareas of the San Joaquin Valley Area Economic cost of machine harvesting per bale Saving over cost of hand harvesting per bale Kern $ 26.29 30.16 27.12 23.50 35.19 $ 18.71 14.84 17.88 21.50 9.81 Eastside San Joaquin-Tranquillity Westside Madera-Merced [8] INTRODUCTION Mechanical cotton pickers began to find acceptance in California in 1945, although a few machines had been used in field trials as early as 1943 and 1944. In 1945 Venstrom (1946) noted 20 machines operating in the south and Westside of the San Joaquin Valley. Progress was slow at first, and two years later only 75 machines were being op- erated. By 1948, however, some 475 har- vesters were in use, and the possibilities of mechanizing the harvest looked more promising. Mechanical picking came of age in 1949 when more than 16.2 per cent of a 1,268,000-bale crop was ma- chine picked. Information for 1951 indi- cates that an estimated 3,700 machines picked more than half of the 1,800,000 bales of California cotton (table 6). Al- though mechanization of the California cotton harvest is approaching full reali- zation, many technical and economic problems remain to be solved. The 1949 season proved to many grow- ers that mechanical cotton picking is practicable and profitable. Improvements in machines, in operating techniques, and in growing the cotton crop have made successful machine harvesting possible. Individual growers had used mechanical pickers in previous years with varying degrees of success. Some said that high field waste and low grades made machine harvest economically impracticable ; others claimed unqualified success. Ob- viously, the true situation for most grow- ers would be somewhere between these two extremes. The study upon which this report is based was undertaken to determine the range of situations in which machine cotton harvesting is economically feasi- ble. Information for the study was drawn from many sources. Following the 1949 harvest season, enumerators interviewed 63 growers selected at random from among those who owned and operated one mechanical harvester. The machines were all one-row spindle type, mounted on a standard rubber-tired, general-pur- Table 6. Mechanical Cotton Picking in California, 1945- 1951 Total harvest* Mechanical harvesters in use! Machine picking! Year Acres Bales Average bales per machine Total bales picked Percentage of crop thousands 20 § 150 thousands 1945 317 353 3 0.8 1946 358 458 25 175 4 1.0 1947 534 772 75 200 15 2.2 1948 804 968 475 195 93 9.6 1949 957 1,268 900 229 205 16.2 1950 581 978 1,450 221 310 31.7 1951 1,331 1,800 3,700|| 264 975 54.2 * Reported by the California Crop and Livestock Reporting Service, Sacramento. f Estimates based on information furnished by ginning companies, data collected in earlier studies (Ven- strom, 1946; Burlingame and Bailey, 1948), and other information developed by the authors. t Estimates based on number of machines distributed by dealers and the estimated number purchased outside and shipped into the state. § It is reported that 5 machines were observed operating in 1943. U Includes about 400 machines shipped into the state for custom harvesting. [9] Figure 2. Subareas of the San Joaquin Valley included in studying mechanical cotton harvesting, 1949. (The 63 growers interviewed were located in these five subareas.) pose tractor. These machines studied in 1949 are identified as Mechanical Har- vester Number 1. Growers interviewed were located in five subareas considered representative of typical production sit- uations in the San Joaquin Valley. These [10 subareas, as defined on the map, are identified by geographic names, which henceforth are used in this report (fig. 2). Information furnished by the grow- ers included the number of hours ma- chines were operated, pertinent costs, and man-hours. Data were also taken from growers' gin statements regarding weight of seed cotton, weight of lint, and bale numbers. This information helped to identify the exact grades of each lot of cotton whether machine- or hand- picked. It also permitted a comparison of gin turnout and lint grades of cotton harvested by the two methods. Data were collected in 1950 from 18 operators of another kind of mechanical cotton harvester, here identified as Me- chanical Harvester Number 2. Twelve machines were one-row, tractor-mounted, while six were two-row, self-propelled. These data were analyzed similarly to the 1949 data for Mechanical Harvester Number 1. Two other makes of mechani- cal cotton pickers were also being used on a very limited scale in the Valley. Neither make was included in this study because fewer than a dozen of either were in use in 1950. The effect of machine picking on lint cotton grades was further studied by analyzing the grades of machine- and hand-picked cotton obtained from repre- sentative gins for each of three seasons, 1948-1950. The 1949 data were consid- ered in greater detail to ascertain varia- Table 7. Average Acres of All Land, Irrigated Land, and Principal Crops on Selected Farms Using Mechanical Cotton Pickers, San Joaquin Valley, 1949 Item San Joaquin Valley Subareas Kern Eastside San Joaquin- Tran- quillity Westside Madera- Merced Cotton Grain* Number of farms 63 15 15 9 9 15 Average acres per farm 349 210 62 17 22 5 2 11 5 61 25 290 49 44 18 1 7 304 38 70 26 18 17 4 25 242 70 8 26 30 13 645 1,196 38 104 32 1 388 40 340 86 103 13 9 8 3 56 Alfalfa hay Potatoesf Ricef Flax Sugar beets Orchard. . Pasture Fallow Miscellaneous § All irrigated land 769 19 409 3 502 54 389 2,444 19 615 16 Other land in farms || Total land in farms 788 412 556 389 2,463 631 * Mostly barley. t Mostly early potatoes. X Some acreage planted as first crop in land reclamation. § Mostly onions and milo in Kern ; milo in Eastside ; melons in Westside ; vegetables in Madera-Merced. || Includes some range pasture. [11] tions in grade loss among subareas in the valley. They were also analyzed at a few gins to ascertain week-by-week trends in grades during the season. Growers interviewed in all subareas had sizable cotton operations. In 1949 the average acreage of cotton per farm varied from 242 in San Joaquin-Tran- quillity to 645 in Westside. The average for all growers interviewed in the San Joaquin Valley, where practically all California cotton was grown in 1949, was 349 acres (table 7). Total land op- erated varied from 389 acres in San Joaquin-Tranquillity to 2,463 in West- side and averaged 788 per farm for the Valley. Grain and rice were important in the farming systems of cotton growers in Westside, grain occupying almost twice as much land as cotton. Potatoes and alfalfa hay were also found on many farms in Kern and Eastside and alfalfa in Madera-Merced (table 7). The proportion of irrigated land in cotton on these farms varied from one- fourth in Westside to three-fourths in Kern and averaged slightly under one- half for the San Joaquin Valley. EFFECTIVENESS OF MAN LABOR IN COTTON PRODUCTION Economic literature concerning cotton production in the United States abounds with references to its relatively high labor requirements. This is not surpris- ing, inasmuch as cotton annually ranks first or second in gross value among cash crops produced in the United States, is the major field crop requiring most labor in total and per acre, and consti- tutes the principal, in many cases the only, source of cash income for a large proportion of farmers. Economists and other research workers have recognized the entire process of cotton growing as a heavy user of labor. Harvesting in par- ticular, however, has received close at- tention because of both its weight in total labor requirements and its seasonal concentration. The Situation prior to 1 940 A cross-section evaluation of cotton labor requirements for harvesting and for the total production process is avail- able in Holley and Arnold's report (1938) .That report is particularly useful in that it preceded the commercial de- velopment of the mechanical picker and the spread of mechanization in cultural operations throughout the cotton-produc- * This section constitutes a review of litera- ture. ing areas. It also combines the 1938 data for various areas with a digest of earlier information. Data for four selected seasons — 1909, 1919, 1929, and 1936— indicate that there was serious need to improve the effectiveness of labor in cotton produc- tion (table 8). Man-labor requirements, at 114 to 131 hours per acre, were ex- tremely high in the older, more humid sections, and even the semiarid section, with 27 to 39 hours required, compared favorably only by contrast. Harvest labor accounted for approximately 40 to 60 per cent of all labor. Only in the semi- arid areas was there evidence of a sig- nificant trend toward decreases in total labor requirements. There was no evi- dence of any decrease in harvesting labor requirements other than in the semiarid areas, where the practice of "snapping" rather than picking the crop was largely responsible. A limited use of "strippers" was involved as well. Reductions in pre- harvest labor were also evident only in the semiarid section. This was attributed to the joint influence of a level terrain, which was more favorable to mechaniza- tion, and limited rainfall, which reduced the incidence of weeds. Holley and Arnold summarized the situation in 1938 and simultaneously underscored the strategic importance of 12 Table 8. Labor Requirements for Cotton Production by Regions 1909-1936, Selected Survey Data Region and year Yield of lint per acre Man-labor per acre Total man - labor per bale Harvest labor per hundred- weight of lint Pre- harvest Harvest Total 1 2 3 4 5 6 7 Coastal Plain : 1909 1919 1929 1936 Mississippi Delta : 1909 1919 1929 1936 Western, semiarid : 1909 pounds hours 231 209 196 239 224 227 252 302 127 163 142 176 66.2 66.8 65.4 65.0 65.3 65.6 63.3 62.1 17.2 15.6 13.1 11.7 53.9 47.7 44.3 53.5 51.5 52.0 57.6 68.7 21.4 17.9 13.9 14.9 120.1 114.5 109.7 118.5 116.8 117.6 120.9 130.8 38.6 33.5 27.0 26.6 252 263 264 237 246 247 229 210 147 97 92 72 23.3 22.8 22.6 22.4 23.0 22.9 22.8 22.7 16.8 11.0 9.8 8.5 1919 1929 1936 Source : Crop Production — Cotton. W. P. A., National Research Project Report A-7, 1938, Table 28, p. 99. Column 7: Calculated by the authors of this report. improved harvesting technology as fol- lows: A primary obstacle to the extended mechani- zation of preharvesting operations is the peak labor requirements for harvesting operations. The development of the mechanical cotton picker raises prospects for overcoming this im- pediment. It is believed that machines now under development warrant expectations of their adoption during the next decade. They are especially adaptable to use on the relatively level lands favorable to use of tractors, and they are likely to be used there first. Developments 1941-1950 Cotton producers have shared in the increased productivity of United States farmers since 1940. Both preharvest and harvesting labor requirements have been reduced. The increased labor efficiency has not been uniform, however, either among all cotton-producing sections or among all producers in any given sec- tion. The first appreciable reduction in labor requirements occurred in the semi- arid regions of Oklahoma and Texas. Holley and Arnold (1938) recorded the beginning of this reduction in preharvest requirements and the associated factors. The change-over from picking the lint from the open boll to hand "snapping" both boll and lint was largely responsible for the early reduction in harvesting labor requirements. This change, in turn, was associated with improved ginning equipment, which made it possible to remove the increased amount of dirt and trash harvested with the lint. The cotton stripper was the next logi- [13] cal development. This is a mechanical device that removes boll and lint, much as the hand laborer does in snapping. Campbell (1948) in Oklahoma, William- son and Rogers (1949) in Texas, and Sutherland and James (1948) in North Carolina have reported on the use of these machines. The Oklahoma report is based on in- terviews in 1948 with a broad cross sec- tion of cotton growers and their sup- pliers. A total of 104 tractor-mounted, factory-built strippers was reported, chiefly in western Oklahoma. It was also estimated that 50 to 100 additional locally made machines were in use. Campbell found that farmers did not consider field waste and grade loss seri- ous but did express concern about the problems caused by green bolls and leaves. Bales harvested during the season ranged from 1 to 75 and averaged 22 per machine. The author did not attempt a conclusive financial analysis but indi- cated that stripping appeared profitable in 1947. Apparently, harvesting required about one hour per acre, and stripper operators were willing to do custom work at about $1.50 per 100 pounds of seed cotton, as compared with $2.00 to $2.50 for hand snapping. The Texas report included data for 64 farms using strippers in 1948. Wil- liamson and Rogers obtained data on both dry-land and irrigated farms but found that use of the stripper on the latter was limited largely to "scrapping" after hand harvesting. Dry-land farmers, in contrast, stripped 45 per cent of their acreage. Field waste averaged 9 and 14 per cent, respectively, for storm-resistant and normal boll varieties on dry-land farms. Comparable figures for irrigated land were 6 and 13 per cent. Grade loss also was appreciable. The season's aver- age value of a bale of cotton was $127 for hand-snapped and $116 for stripped cotton. Williamson and Rogers also ana- lyzed operating costs. Their findings for total cost of harvesting and hauling to Table 9. Cost of Operating Mechanical Cotton Strippers, High Plains, Texas, 1948* Dry land Irrigated land Per acre Per bale Per acre Per bale No hand snapping : Variable cost Labor Other Fixed cost $1.55 (0.91) (0.64) 0.91 $2.46 1.46 (0.83) (0.63) 0.91 $2.37 $ 7.77 (4.54) (3.23) 4.55 $12.32 13.18 (7.50) (5.68) 8.19 $21.37 $ 2.33 (1.15) (1.18) 0.91 $ 3.10 (1.55) (1.55) 1.21 Total After hand snapping : Variable cost Labor Other Fixed cost $ 3.24 1.66 (0.91) (0.75) 0.91 $ 4.31 6.83 (3.75) (3.08) 3.75 $10.58 Total. $ 2.57 Source: Williamson, M. N., Jr., and Ralph H. Rogers. Economics of Cotton Harvesting, Texas High Plains, 1948 Season. Texas Agr. Exp. Sta. Progress Report 1200, 1949. * Does not include consideration of grade loss and field loss. [14] the gin are summarized in table 9. The cost of hand snapping plus hauling was estimated at $31.50 and $36.00 per bale, respectively, on irrigated and dry-land farms before the first frost of the season. Costs increased after the first frost. Com- parative costs for labor are important, as most of the cost of hand snapping is for labor. Grade loss was a major item in total cost of mechanically stripping cotton in North Carolina. Sutherland and James (1948), making comparisons with the value of hand-picked cotton, found lowered grade responsible for $29.90 of the total cost of $38.62 per bale. Here, as in Texas, however, hand snapping was more costly. A smaller grade loss, $24.35 per bale, was more than offset by the higher labor cost, and the total cost of hand snapping averaged $61.35 per bale. Mechanical cotton pickers of the spin- dle type were first released in 1941 when 12 were sold (Archer, 1948). In subse- quent years sales were as follows: 12 in 1942, 15 in 1943, 25 in 1944, and 75 each year from 1945 through 1947. Vol- ume production of these machines was scheduled for 1948. Mechanical pickers of this type were operating in the major cotton-producing sections by 1945, though still in small numbers. Growers in California and the Missis- sippi Delta have led in buying and using the spindle-type mechanical picker. Ven- strom (1946) reported for California: "By 1945 some twenty-odd machines were in operation in the State." Inter- views with nine operators regarding their 1945 experience led him to conclude that "operators of the larger cotton acreages in California have accepted the mechan- ical cotton picker." (This inference seems to have been somewhat optimistically premature, for later studies showed that many of the larger growers in 1948 and 1949 were still using much hand labor in the cotton harvest.) Venstrom re- ported that on a "once-over" basis for 500 acres, the mechanical picker could pick at a cost of $1.57 per 100 pounds of seed cotton, or $22.00 per 500-pound bale of lint. These figures were based on prices of cotton lint at 22 cents per pound, 10 per cent field waste, and one grade loss in quality. They compared with costs of hand picking at $2.40 to $2.70 per 100 pounds, or $32.00 to $37.00 per bale in 1945. Substantially comparable data were obtained from a limited number of machines studied three years later in the San Joaquin Valley. A sharp increase in mechanical pick- ing of California cotton was noted by Burlingame and Bailey (1950) : "Several hundred growers have bought machines and in the 1949 season probably 15 per cent of the cotton was machine-picked." The authors based their economic analy- sis on the 1948 experience of ten growers in Fresno and Kern counties. These co- operators owned 23 machines that har- vested 4,500 bales in 1948. On the aver- age, each machine picked cotton equiva- lent to 1,500 man-days of hand picking at a cost of $2.14 per 100 pounds of seed cotton. These costs, on a per-bale basis, included $16.62 for machine overhead and operation, $5.99 for field waste, and $7.37 for grade loss— a total of $29.98 as compared with a hand-picking cost of $45.00. The performance record of the machines in this study is noteworthy: "On the average each machine operated 58 working days, picked 244 bales or 3,434 hundredweight of seed cotton, and covered 313 acres (once over)." Sutherland and James (1948) in- cluded data on four mechanical pickers in their 1948 progress report from North Carolina. The season's use of the me- chanical pickers averaged 44 bales har- vested from 60 acres at a total field- operating cost of $19.69 per acre and $26.70 per bale. The authors point out: "Data on mechanical pickers are not adequate to draw conclusions but they appear to have possibilities if 100 or more bales can be picked per machine [15] Table 10. Performance and Costs for Mechanical Cotton Picker in Mississippi, 1945-1947 Item 1945 1946 1947 Number of pickers studied 27 20 26 bales Bales harvested Per picker 111.0 4.0 87.0 2.0 109.0 Per day 3.5 Selected cost items Variable cost : Labor $ 258.74 173.44 114.64 $454.17 477.27 162.35 $ 337.92 Repairs 431.52 Other 187.99 Total $ 546.82 501.08 128.46 $1,093.79 505.95 129.48 $ 957.43 Fixed cost : Depreciation, picker 516.31 Other 132.15 Total Operating cost per bale Variable cost $ 629.54 4.92 5.66 $ 635.43 12.51 7.27 $ 648.46 8.81 Fixed cost 5.96 Total $ 10.58 $ 19.78 $ 14.77 Source: Crowe, Grady B. Mechanical Cotton Picker Operation in the Yazoo-Mississippi of Agricultural Economics, Miss. Agr. Expt. Sta. Progress Report 1949, Tables 1, 2, and 8 Delta. U. S. Bureau pp. 6, 7, and 8 during the year." The findings in this report concerning harvesting costs per bale are important even though lack of data prevented a complete analysis in- cluding field waste and grade loss for the mechanical picker. Hand picking was figured at $3.00 and hand snapping at $2.00 per 100 pounds of seed cotton (and trash). Experience with mechanical pickers in Mississippi is reported by Crowe (1949a, 19496). His data covered the three sea- sons 1945-1947. It was estimated that by 1948 there were 600 to 650 of the spin- dle-type mechanical pickers in the Delta area alone. Some of the salient data con- cerning strictly field-harvesting opera- tions that resulted from Crowe's analysis are listed in table 10. The influence of increasing the num- ber of bales picked per machine on re- ducing costs per bale is emphasized in his report. Higher prices for the mechan- ical harvester have the opposite in- fluence — increasing costs per bale. An evaluation was also made of the effects of field loss and grade deterioration ac- companying machine picking. The aver- age field loss of 8 per cent in 1947 was figured at $13.00 and the loss in grade at $7.90 per bale, a total of $20.90 added to the cost of machine picking as com- pared with hand methods. This meant, at 1947 prices, a total cost of $35.67 per [16] bale, or $2.42 per hundredweight of seed cotton (table 10). Hand-picking rates, meanwhile, ranged from $2.50 to $4.50 per 100 pounds of seed cotton. Finally, the Crowe report recognizes problems concerning defoliation, weeds and grass, lack of properly equipped gins, and the need for competent mechanical opera- tion. The Situation in 1949 The present study was undertaken more than ten years after the Holley and Arnold report had focused attention on the inefficiency of labor in cotton pro- duction in the United States. Meanwhile, World War II had come and gone, re- sultant shortages of labor and changing price relationships demanded the substi- tution of capital and machines for human energy, and the postwar world need for cotton had melted away the cotton sur- pluses of the 1930's. The findings of the research studies reviewed may be summarized in the fol- lowing key points describing the situa- tion at the beginning of 1949: 1. Interest in mechanizing the cotton harvest had resulted from the relatively low effectiveness of hand labor in the cotton harvest and the consequent heavy requirements for both harvest and total labor. 2. Mechanical strippers had proved economic at postwar prices in the semi- arid areas but had not found a place in California. 3. The spindle-type mechanical cotton harvester had been used widely in the Mississippi Delta and in California. 4. Relatively heavy investments and overhead costs were involved in owning the current models of mechanical cotton pickers. 5. Volume of use in terms of total bales harvested annually was a major factor affecting harvester operating cost per bale, largely because of overhead costs. 6. Field waste of seed cotton not re- covered by the picker was considered an important cost in mechanical picking. 7. Grade deterioration was also recog- nized as a major economic cost in me- chanical picking. 8. More experience was needed before it would be possible to define precisely the range of conditions within which it is economic to own and operate a me- chanical cotton picker. It is evident that an appraisal of pro- gress in mechanizing the cotton harvest must be comprehensive, that it must con- cern itself with all economic aspects of the question. Total cost of mechanical picking includes overhead and operating expenses, field waste, reduced sale value of cotton due to quality reduction, and any added ginning, hauling, or other in- cidental expense. All these must be con- sidered in a complete evaluation. It is also evident from the literature that scale of operations, price levels for cotton and cottonseed, relative prices for labor and other operating items, yields per acre, and the probable pattern of weather dur- ing the harvest period are factors that must be recognized. So far as possible, the nature and extent of influence by these factors should be identified and measured. It is appropriate and highly important, finally, to establish certain guiding principles concerning the use of mechanical cotton pickers by farmers operating on various scales. The present study was timed to as- semble and evaluate a representative pic- ture from the experience of California farmers with the mechanical picker. It is limited to farmers owning one picker. [17] USE AND PERFORMANCE OF MECHANICAL HARVESTERS Before 1949 many growers had con- sidered machine harvesting as merely supplemental to hand picking. A number of machines had, therefore, been used only in second picking when workers were unavailable, or sometimes in fields too weedy or too low in yield to attract workers. The lack of a successful chem- ical defoliant also tended to discourage machine picking until after the first heavy frost.* Thus, there had been no widespread effort to make maximum use of harvesting machines. A considerable number of operators, however, did make reasonably full use of mechanical pickers during the 1949 season. The results of the experience are reported herein for 63 such growers interviewed in early 1950. Summary of machine use A typical grower began machine har- vesting on October 19 and finished on January 1. He operated his machine 47 working days during this 75-day period for a total of 407 machine-hours.f He covered a total of 284 acres once over, 145 acres of first picking and 139 acres Figure 3. A rear view of a mechanical cotton harvester picking cotton. of second picking. He picked a total of 229 bales, of which 182 bales were first picking and 47 were second picking. The average machine picked 3,183 hundred- weight of seed cotton, equivalent to the amount 25 hand workers could have picked in 50 working days. Mechanical pickers on the average were operated more hours (479), cov- ered more acres (317) , and picked more bales (292) in Westside than in any other area. Machines were used least in the northern area where they did the least first picking. Amount of use per machine varied considerably. Total days operated ranged from 16 to 112, machine-hours from 120 to 766, acres of picking from 80 to 535, and bales picked from 93 to 613. As might be expected, some machines oper- ated less than a full season whereas others operated at near capacity. Study of the 63 individual records revealed that 31 machines operated an average of 520 hours in 62 working days between October 11 and January 7. They picked an average of 292 bales from 356 acres, 182 acres of first picking and 174 acres of second picking. All but four of the 63 growers used their machines in first picking, and all but two growers did some second pick- ing. Eleven growers machine picked all their cotton, averaging 263 bales, of which 235 were first picking and 28 were second picking. These growers had, on the average, 158 acres of cotton, of which they second picked 110 acres. * Climate and Man. Yearbook of U. S. De- partment of Agriculture. The average date of the first killing frost in the fall is November 25 at Bakersfield, November 14 at Hanford, No- vember 15 at Coalinga, December 1 at Fresno, and November 16 at Madera. t Machine-hours exclude morning, noon, and evening service time and extended stops for adjustments or repairs but include stops for minor adjustments or repairs, for unloading the basket, and for resting the operator. [18 Summary of performance rates The number of acres the harvester picks per day depends upon the gear (speed) used and the time stopped in the field. The harvesters studied were one-row, two-gear (speed) machines de- signed to operate at 2.00 miles per hour in first gear and 2.75 miles per hour in second.* This rate would permit cover- ing 0.77 acre per hour in first gear and 1.06 acres in second on cotton in 38-inch rows if continuous operating in one di- rection were possible. Harvesters cannot maintain these rates, however, because they must stop for turning, unloading the basket, and for servicing and adjust- ment. The harvesters studied picked, on the average, 0.60 acre per machine-hour in first picking and 0.85 acre in second picking, considering time for the stops just mentioned. The quantity of cotton a harvester will pick in a given time is directly related to yield and to the quantity of open cot- ton. The harvesters studied picked, on the average, 1,021 pounds of seed cotton per machine-hour in first picking and 429 pounds in second picking (table 11) .f One machine operated successfully in cotton yielding 2.7 bales per acre; it picked 3,585 pounds of seed cotton per machine-hour in first picking. Another measure of performance is the pick per working day. The harvesters averaged 7.0 bales per workday in first picking and 2.3 bales in second picking, or 4.9 bales over the season. The machine that operated in 2.7-bale cotton averaged 17.2 bales a day. Discussion Many growers made maximum use of their machines during the 1949 season. They proceeded with machine first pick- ing, even though chemical defoliants generally were ineffective, and found * As indicated in the 1949 Owner's Manual furnished with the harvester. t Second picking was not necessarily on the same fields as first picking. Table 11. Average Performance Rates of 63 Mechanical Cotton Pickers, San Joaquin Valley, 1949 Item* San Joaquin- Valley Area Kern Eastside San Joaquin- Tran- quillity Westside Madera- Merced Seed cotton harvested per machine-hour : First picking pounds 1,021 429 783 1,182 521 903 1,016 475 792 1,040 322 856 1,062 408 826 772 584 560 Second picking All picking Bales harvested per workday : First picking bales 7.0 2.3 4.9 7.0 2.1 4.6 7.6 2.8 5.4 5.8 1.3 4.4 7.6 2.4 5.6 6.9 2.1 4.3 Second picking All picking * Other rates, "acres per machine-hour" and "bales per machine-hour," are found in table 13. [19] that results were more successful than they expected. It frequently proved pos- sible to use machines before the plants had defoliated if leaves had wilted to a dull gray-green and if a good percentage of the bolls were open.* After midseason, under pressure of a huge harvest, some machines actually picked beyond their effective capacity by operating when con- ditions were unfavorable, during nights when humidity was unduly high and days when fog or dew was excessive. Consequently, growers often found that in pressing output to the limit they re- duced output per hour and obtained lower lint grades. About half the growers, for various reasons, did not make full use of their machines. Some hesitated to use ma- chines in first picking; others, because of previous unsatisfactory experience, used machines merely to supplement hand picking. Lack of experience handi- capped those using machines for the first time, and some bought machines after the picking season had begun. Nearly all growers who used machines at less than capacity in 1949 said they would use their machines more fully in the 1950 season. This intention was not fully rea- lized because the acreage of cotton was sharply reduced in 1950; but apparently most machines operated at full capacity in 1951 when the average pick was esti- mated at 264 bales per machine (table 6, page 9). A grower who is considering convert- ing to mechanized harvesting needs to know how much a machine will pick in a normal season. The experience of 31 growers operating a full season in 1949 indicates that growers can expect to op- erate 60 working days or 500 hours in a typical season. It can safely be as- sumed that a machine can harvest com- pletely 200 acres of cotton and pick 300 bales in a season. All the cotton, as a rule, would open before first picking was fin- ished, so perhaps no more than 150 acres would require a second picking. If so, the total would be the equivalent of 350 acres once over. These estimates seem reason- able enough even though 1949 was some- what more favorable than an average season for machine harvesting. It is be- lieved that improved equipment plus greater operating skill has increased seasonal capacity per machine enough to permit such performance even in sea- sons of less favorable weather conditions. EFFECT OF MACHINE PICKING ON GIN TURNOUT Growers, ginners, and others in the cotton industry have been concerned about the effects of machine picking on gin turnout. Gin turnout is the ratio of lint to seed cotton and trash. Seed cotton usually contains moisture and varying amounts of leaves, stems, sticks, and other foreign material. Turnout is im- portant because it is an inverse expres- sion of the amount of foreign material termed "trash" in the cotton. f Low trash * Growers reported that -wilting can be in- duced by timely removal of irrigation water. t Turnout may also reflect the thoroughness with which the gin removes lint from seed, but that fact is more directly determined by inspec- tion of seed as it comes from the gin. content — high turnout — is desired be- cause the greater the amount of foreign material, the more difficult it is to gin out a high-grade sample of lint and be- cause excess foreign matter increases the cost of ginning, as ginning charges are based on the weight of seed cotton. Mechanical harvesters can reduce gin turnout by collecting with the seed cot- ton more trash than hand pickers or by adding excessive moisture from the spindle moisteners. But if technical im- provements or greater operating skill can decrease the amount of foreign material gathered with the cotton, mechanical pickers may actually improve the grade. The weight of seed cotton and lint was [20 tabulated from the gin statements for the interviewed growers and used to com- pare gin turnout of machine- and hand- picked cotton according to season aver- ages and by periods through the season. Among the 63 interviewed growers, the season average gin turnout of ma- chine-picked cotton (36.5 per cent) was less than one percentage point lower than that of hand-picked cotton (37.1 per * Gross weight of bales was used throughout in calculating gin turnout. cent,* table 12). On the average, it took 1,370 pounds of machine-picked seed cotton, as compared with 1,348 pounds of hand-picked, to make a 500-pound bale of lint cotton. Gin turnout of ma- chine-picked cotton was lower, though not to the same extent, in each of the subareas except San Joaquin-Tranquil- lity. In that area, gin turnout of machine- picked cotton (37.2 per cent) was ac- tually higher for the season than turn- out of hand-picked cotton (35.5 per Table 12. Gin Turnout by Method of Harvest, San Joaquin Valley, 1949 Four-week periods First Sept. 11- Oct. 8 Second Oct. 9- Nov. 5 Third Nov. 6- Dec. 3 Fourth Dec. 4- Dec. 31 Fifth Jan. 1- Jan. 28 Season average t San Joaquin Valley Machine picked Hand picked Difference! Subareas Kern: Machine picked Hand picked Difference! Eastside : Machine picked Hand picked Difference! San Joaquin-Tranquillity : Machine picked Hand picked Difference:}: Westside : Machine picked Hand picked Difference! Madera-Merced : Machine picked Hand picked Difference! 36.7 39.1 -2.4 32.4 37.8 -5.4 38.4 39.2 -0.8 37.3 39.3 -2.0 38.4 39.2 -0.8 per cent 37.2 36.8 34.7 32.2 37.9 36.8 35.0 30.7 -0.7 -0.3 +1.5 36.2 35.0 31.9 30.0 37.4 37.1 36.4 29.0 -1.2 -2.1 -4.5 +1.0 36.6 36.5 34.5 33.0 37.9 35.7 33.4 30.8 -1.3 +0.8 +1.1 + 2.2 37.3 37.8 36.2 33.4 38.0 37.6 35.2 31.5 -0.7 +0.2 +1.0 +1.9 37.8 37.0 35.2 31.0 37.9 37.1 36.6 31.1 -0.1 -0.1 -1.4 -0.1 37.3 36.4 34.5 33.3 38.1 37.1 35.3 27.8 -0.8 -0.7 -0.8 +5.5 36.5 37.1 -0.6 34.4 36.8 -2.4 36.2 36.9 -0.7 37.2 35.5 +1.7 36.9 37.6 -0.7 36.5 37.6 -1.1 ■ Data include all machine-picked and hand-picked cotton of 63 growers. Gin turnout is the ratio of lint to seed cotton and trash. t Season average of all cotton picked, not a simple average of the five 4-week periods. t Minus sign indicates machine below hand, plus sign, machine above hand. [21] cent) . It may be noted that San Joaquin- Tranquillity had the highest turnout from mechanical picking and the lowest from hand picking of any area. For the growers interviewed, gin turn- out of hand-picked cotton followed the usual pattern, starting high and declin- ing steadily throughout the season. In contrast, machine-picked turnout in- creased from September (36.7 per cent) to October (37.2 per cent), nearly equaled hand-picked in midseason, and was actually higher than hand-picked in late-season picking (table 12) .* In short, turnout of machine-picked cotton im- proved relative to that of hand-picked throughout the season both for the Val- ley average and for each of the sub- areas (table 12). Gin turnout of machine-picked cot- ton was remarkably good on the whole. The concern people felt when machines first came into use was apparently un- justified. Machine-picked turnout showed up less favorably early in the season, because machines collected more green leaves and other trash than hand pickers. After the cotton had defoliated, gin turn- out performance of machine-picked cot- ton was comparable with that of hand- picked. The fact that machine-picked turnout compared favorably late in the season may be explained by very poor hand picking at that time. MACHINE-PICKING COSTS The term "picking cost" as used here includes overhead and operating costs for the picker and tractor plus cost of labor. It excludes field waste and grade loss, both of which are recognized as "economic costs" in comparisons of me- chanical with hand harvesting but which are analyzed in later sections of this report. Average picking cost per machine Total machine-picking cost during the 1949 season averaged $3,355 per ma- chine for the 63 growers interviewed. The component items included an aver- age of $1,734 for overhead, $1,016 for operating cost, and $605 for labor. Overhead thus represented more than half the total cost for the Valley as a whole, and this same pattern with minor variations was repeated in the subareas (fig. 4 and table 34, page 55). Overhead costs. Overhead costs in- clude depreciation, interest on invest- ment, property taxes, and insurance on the picker and tractor. Depreciation dur- ing the life of a machine is the original * Neither snaps nor hollies were included in hand picking. cost minus the ending or salvage value. Original cost, which obviously is vital in determining depreciation, averaged $6,459 for the harvesters and $2,950 for the tractors used to harvest cotton. Length of life, the other essential factor in calculating annual depreciation, was estimated at five years for harvesters and seven years for the tractors on which they were mounted, except that machines with very high annual use depreciated more rapidly. Total depreciation aver- aged $5,490 for the harvesters, and this amount divided by 5 and corrected for machines receiving greater than annual use resulted in an average annual de- preciation of $1,112. The estimate of harvester life had to be somewhat arbitrary. The oldest ma- chines included in the survey were pur- chased in 1947, and it was not known how many years they would operate. Major modifications and improvements had already been made at the time of the interviews, and others were expected to follow. It was impossible, therefore, to determine from experience the average length of life. The five-year estimate ac- tually used was considered reasonable, however, and represented the combined [22] Cost (Dollars) SAN JOAQUIN VALLEY Acre Hour Bale Labor Operation Tractor Over- Harvester head Acre Hour Bale EASTSIDE SAN JOAQUIN-TRANQUILLITY Acre Hour Bale Acre Hour Bale WESTSIDE MADERA-MERCED ■ ■ ■MM ■ 1 Acre Hour Bale Acre Hour Bale Figure 4. Mechanical cotton picking costs in the San Joaquin Valley, 1949; costs per acre, per bale, and per hour; usual overhead, operating and labor expense. (Overhead cost of picker and tractor accounted for most of the total, with picking operations and labor the other major items.) Source of data: Tables 14 and 34. judgment of experienced machine oper- ators and dealers. The annual charge for interest on in- vestment was estimated at $149 per ma- chine, a sum much smaller than that for depreciation. Property taxes and insur- ance averaged $123 and $33, respec- tively, for each harvester. Overhead cost for the tractor was handled differently than for the har- vester in two respects : 1 ) annual repairs were included in overhead; and 2) only a part of the total overhead cost was charged to cotton harvesting. That was done because, typically, the tractor was used for other work at other times of the 23 Dollars 1500 r 1000 HARVESTER 500 E_g — ~~ ~ 1 • ■H ■1 mi H Jim Hi San Joaquin Valley Kern East- side San Joaquin- Tranquillity West- side Madera- Merced Insurance Taxes □ Interest Ijjl Depreciation Expense Items 1000r 800 600 400 200 TRACTOR San Joaquin Valley 1 Other Kern iiiiiiiiii East side \ J Repairs Expense Items San Joaquin- Tranquillity West- side Madera- Merced Depreciation Figure 5. Overhead costs for mechanical cotton picking in the San Joaquin Valley, 1949; mechanical harvesters and tractors. (Depreciation was the main item of overhead cost, and the harvester accounted for most of the total.) Source of data: Table 35. year. Depreciation still accounted for 63 per cent of the overhead cost, while in- terest on investment accounted for 12 per cent. Variations among the subareas were minor, as might be expected in view of the newness of the mechanical harvester and the close similarity in age and me- chanical condition of the machines studied. Operating costs. Expenses on the harvester averaged $869, or more than half the annual costs (fig. 6). Labor ex- penses ($605) ranked second, represent- ing about 40 per cent, and tractor oper- ating expense ($147) accounted for slightly less than 10 per cent.* * The tractor expense is somewhat mislead- ing inasmuch as repairs were included with overhead as indicated above. [24] Dollars 1000 800 600 h 400 200 HARVESTER Expense Items ■ ■ San Joaquin Valley Ker Other East- side San Joaquin Tranquillity Seasonal Repair West- side Madera- Merced Preseason Repair TRACTOR 200 100 1— - San Joaquin Valley Kern East- side San Joaquin- Tranquillity West- side Madera- Merced Other Fuel LABOR 800 600 - 400 200 San Joaquin Valley Kern East- side San Joaquin- Tranquillity West- side Madera Mercec Other Operating Figure 6. Operating expenses for mechanical cotton picking in the San Joaquin Valley, 1949; mechanical harvesters, tractors, and labor. (Major expense items were preseason repair on the harvester, operating labor, and tractor fuel.) Source of data: Table 36. The largest item in harvester expense was preseasonal repair, $505 on the av- erage, while seasonal repair required an- other $200. The two items of mainte- nance together, $705, represented 80 per cent of the operating expenses for the harvester. Preseasonal repair was diffi- cult to estimate because a number of the machines were new in 1949. The esti- mate used is based on the average re- ported by those operators who had used machines for at least one year before 1949. Costs of mounting and dismount- ing the harvester unit in order to use the tractor for other purposes ranked third among the expense items for the har- vester. Other items were relatively unim- portant in total operating costs. [25] Table 13. Average Cost of Mechanical Cotton Picking, San Joaquin Valley, 1949 Item San Joaquin Valley (63 growers) Season total Average per Acre Hour Bale Acres of picking 284 407 229 1.43 0.81 0.69 0.56 1.24 1.78 Machine-hours Bales harvested Investment Harvester $3,714 1,696 $ 13.08 5.97 $ 19.05 4.99 1.12 $ 9.12 4.17 $ 16.22 7.41 Tractor Total $ 5,410 1,417 317 $ 13.29 3.48 0.78 $ 4.26 2.14 0.36 $ 23.63 6.19 1.38 Picking Costs Overhead : Harvester Tractor Total $ 1,734 869 147 $ 6.11 3.06 0.52 $ 7.57 3.80 0.64 $ 4.44 2.21 0.37 0.03 0.03 Operating expenses : Harvester Tractor Total $ 1,016 506 84 8 7 $ 605 $ 3.58 1.78 0.30 0.03 0.02 $ 2.50 1.24 0.21 0.02 0.02 Labor: Operating* Service and repair Farm shop Compensation Total $ 2.13 $ 1.49 $ 2.64 Total picking costs $3,355 $ 11.82 $ 8.25 $ 14.65 Labor used Operating man-hours 407 101 1.43 0.36 1.79 1.00 0.25 1.25 1.78 0.44 2.22 Other Total 508 subareas: Kern, seven growers, $208; Eastside, two growers, $188; San Joaquin-Tranquillity, two growers, $247; Westside, one grower, $460; Madera-Merced, one grower, $94. [26] Labor costs. Labor for operating the harvester in the field was responsible for most of the labor expense, while labor for services and repair ranked second (fig. 6) . A few operators were paid wage bonuses, which were considerable items of expense for those reporting but which averaged only $46* for all growers inter- viewed. Because most of the repairs were made in dealers' shops, "farm labor" expense for repairing harvesters was relatively low. Cost per bale, per hour, per acre, and per 100 pounds of seed cotton The average cost of machine picking in the Valley was $14.65 per bale for the growers interviewed (fig. 4 and table 13). More than half the cost was over- head ($7.57), a third was operating ex- pense ($4.44), and the remainder was labor ($2.64). The cost varied widely * Bonuses were more commonly paid machine operators in Kern than in other areas. Almost half of the farmers interviewed in Kern paid such a bonus, averaging $208 for the season, and these farmers represented more than half of all interviewed in the Valley who reported bonus payments. among the subareas because a machine picked more bales in some areas than in others (table 14). The average cost was lowest ($11.71 per bale) in Westside where the season's pick averaged 292 bales per machine. It was highest ($20.72 per bale) in Madera-Merced where the average pick was only 158 bales. Variations in yield per acre largely caused the variation in bales picked per machine and in costs per bale. Costs per acre for picking one time over averaged $11.82 for the interviewed growers in 1949. This figure was highest in Kern ($12.62) where the smallest acreage was picked per machine and lowest in Westside ($10.80) where the largest acreage was picked. The range in cost per acre was considerably narrower than the range in costs per bale. On an hourly basis, the Valley average cost of machine picking was $8.25. The high ($10.04) was in San Tranquillity where machines the fewest hours, the low ( Westside where they were used the most. Per-hour costs also show a much rower spread than costs per bale. Joaquin- operated 7.15) in nar- Table 14. Cost of Mechanical Cotton Picking, by Subareas of the San Joaquin Valley, 1949 Item Kern Eastside San Joaquin- Tranquillity Westside Madera- Merced Acres of picking 263 414 258 309 427 245 273 316 201 317 479 292 265 387 158 Machine-hours Bales picked Cost of picking* Per acre Per hour $ 12.62 8.02 12.86 0.68 1.54 0.89 $ 11.22 8.13 14.15 0.80 1.71 1.03 $11.61 10.04 15.77 0.96 3.13 1.17 $ 10.80 7.15 11.71 0.67 1.75 0.86 $ 12.35 8.46 20.72 1.09 2.60 1.51 Per bale Per 100 pounds of seed cotton f First pick Second pick All picks * Costs per acre, per hour, and per bale are averages of all picking both first and second. These costs exclude field waste and grade loss reported elsewhere in this publication. t Includes lint, cottonseed, and trash as weighed in the trailer at the gin. [27] The average cost of machine picking per hundred pounds of seed cotton was $0.81 in first picking and $1.93 in sec- ond picking. The cost was higher in sec- ond picking because less seed cotton was picked per hour. Average costs of pick- ing in the subareas are summarized in table 14. Materials and labor used. Spindle oil was used in volume by the mechani- cal pickers. Typically, an initial pur- chase of wetting agent was made, but after this had been used, no more was bought. Some operators reported suc- cessful use of a popular brand of deter- gent instead of a special wetting agent. Fuel and oil were the principal materials required for tractor operation. The 63 growers reported an average use of 809 gallons of fuel. Most of the labor was used in operating the picker in the field. On the average, 407 of the total 517 man- hours were used in field operations. Servicing and minor field repairs amounted to another 75 hours. The labor reported does not include that used to operate trailers or trucks in hauling the seed cotton to the gin. Only a few operators indicated that added labor was used to load the trailers or to tramp the cotton; such labor was not included in this report. Discussion The average cost of $14.65 per bale for machine harvesting compares with approximately $45.00 for hand picking.* Two facts stand out regarding picking costs: 1) the dollar investment in the harvester and tractor makes high over- head costs inevitable in the form of de- preciation and interest on investment; and 2) preseasonal and seasonal repair costs for the harvester dominate operat- ing costs. * The cost of hand picking was estimated by assuming 13.5 hundredweight of seed cotton for a bale of lint and multiplying by the 1949 pick- ing cost. The latter figure averaged $3.33 for first and second hand picking according to the growers reporting. The first point emphasizes the impor- tance of making full use of the mechani- cal cotton harvester. The impression was gained during interviews that added ex- perience in using and servicing the har- vester would help to reduce breakage and wear. Many machines were delivered in 1949 and many operators gained their first experience in that year. In some in- stances, the operator had no specific training before taking the machine to the field. Further experience and definite preseason training for operators would, growers believed, help to cut costs. The quantity of cotton harvested an- nually greatly affects the cost per bale of machine picking. Cost per bale was high in Madera-Merced because fewer bales were harvested there than in the other subareas. One machine could har- vest appreciably more cotton than usu- ally was reported. Westside growers, for example, reported 292 bales harvested — 64 bales, or 28 per cent more than the Valley average. The factors responsible in 1949 for limiting the number of bales harvested per machine have been noted. Thoughtful planning can bring about more complete use. Experience in the 1951 season, when mechanical pickers in California averaged 264 bales, demon- strates that it is feasible to pick 250 or more bales per machine per year. This fact is further substantiated by the 1949 average per machine in Westside (292 bales). Added years of service life beyond the five or seven years assumed for picker units and tractors would tend to reduce harvesting costs by lowering the annual cost of depreciation. The cost of obso- lescence should be less important now that mechanical picking is established and particularly now that manufacturers have several years of field experience in testing and improving their machines. It is likely, also, that length of service life will be extended by more effective operation and maintenance of the ma- chines. [28] EFFECT OF MACHINE PICKING ON GRADE OF LINT COTTON The effect of machine picking on the quality of cotton is a matter of prime im- portance. If machine picking causes lower grades than hand picking, the cot- ton will have less value, and this reduc- tion represents an economic cost of machine harvesting. The over-all effect of machine picking on grades of lint cotton can be shown by comparing the grades of machine- picked and hand-picked bales of cotton grown, harvested, and ginned under sim- ilar conditions. All three requirements were met in this study because bales picked by both methods and coming from the same gins were analyzed to de- termine comparative grades.* Almost all cotton going through a given gin is from the immediate neighborhood where soils, weather, and other environmental condi- tions are similar. Also, all cotton was of the same Alcala variety. Most of the dif- ferences in grades, therefore, almost surely reflect variations in harvesting methods.! How machine picking can affect grades of cotton Cotton is assigned class grades accord- ing to leaf content (trash), color, and preparation of the lint. For classifica- tion; samples of lint are cut from the bale. High-grade lint has little trash, is white, and has "normal preparation." The variations in leaf (trash) content are recognized in seven grades which which range downward from Good Middling, having the least trash, to Good Ordinary, having the most. If the trash consists of grass particles, the cotton sample is further reduced one or more * Data on cotton grades supplied by inter- viewed growers were inadequate for this analy- sis because the farms were so scattered that seldom more than one or two were served by the same gin. t It was not practicable to collect data ex- clusively from fields where both machine and hand picking were done because it is uncom- mon for growers to use both methods in the same field at the same time of the season. Note: All grades analyzed in this section and throughout the report were those assigned in the U.S.D.A. Classing Office, Bakersfield, California. The official grades and standards for American Upland Cotton in effect at the time data were collected for this study were superseded on 18 August 1953. Federal Register; 15 August 1952. The new official grades and standards are as follows : White Cotton Good Middling Strict Middling Middling Strict Low Middling Low Middling Strict Good Ordinary Good Ordinary Spotted Cotton Good Middling Spotted Strict Middling Spotted Middling Spotted Strict Low Middling Spotted Low Middling Spotted The new classification has no effect on and Extra White Cotton. These colors cotton in the two top grades that have Tinged Cotton Good Middling Tinged Strict Middling Tinged Middling Tinged Strict Low Middling Tinged Low Middling Tinged Yellow Stained Cotton Good Middling Yellow Stained Strict Middling Yellow Stained Middling Yellow Stained Gray Cotton Good Middling Gray Strict Middling Gray Middling Gray Strict Low Middling Gray the applicability of findings of this study regarding White already are combined in this analysis, while there was no been eliminated — Middling Fair and Strict Good Middling. [29] grades below what otherwise would be assigned. Cotton is also classed accord- ing to color as white, spotted, tinged, yellow stained, or gray. Preparation re- fers to the arrangement of the fibers in the sample. One showing gin-cut staple, for example, may be reduced one or more grades because of this factor alone. Grades of lint cotton are affected by the quality of standing cotton, picking and handling methods, and the ginning. Immature or weather-stained seed cot- ton, with consequent discoloration from spotting, staining, or graying, will not produce good grades, however excellent the picking or ginning. Some leaf (trash) is collected along with the seed cotton in picking, whether it is done by hand or by machine. Also, discoloration, espe- cially green leaf stain, may occur regard- less of the picking method. Gins are equipped to remove much of the trash collected in picking, but they cannot re- move discoloration. Gins vary in respect to cleaning equipment, although recently most gins in California have installed lint cleaners. The skill with which they are operated also varies. Mechanical cotton pickers may reduce lint grades below those that would have been attained by hand picking in the following ways: 1. By introducing more green leaf stain in early season. 2. By introducing more trash from the dead cotton plants late in the season. 3. By gathering more foreign matter from weeds or grass. 4. By adding excess moisture (in spindle moistening) to the seed cotton, making trash removal by gin cleaners more difficult. (Whether moisture also induces graying or mildewing has not been definitely proven.) 5. By twisting or tangling lint on the spindles, thus increasing difficulty of normal gin preparation. 6. By discoloring the lint with oil or grease from the machine. On the other hand, mechanical pickers may actually improve lint grades by passing over hard-to-pick immature bolls late in the season, by adding less trash than careless hand pickers, and by mak- ing possible more timely harvesting. The grade of the crop may actually average higher if it is picked earlier by machine than later by hand. Comparative grades in the 1949-1950 season Grades were analyzed at 35 San Joaquin Valley gins, each of which ginned 500 bales or more of machine- picked cotton. These gins reported a total of 63,000 machine-picked and 238,000 hand-picked bales for the 1949- 1950 season (1949 crop). The data were also summarized to show geographical variations (fig. 7 and table 42, page 65) . Preliminary analysis showed that more than 90 per cent of the cotton picked by each method graded White. The percent- age of off-color bales was slightly higher among the machine-picked (9.0) than among the hand-picked (7.9), but this difference was not statistically signifi- cant. The evidence is not conclusive whether machine picking results in more or less discoloration than hand picking. Subsequent analyses of grades, there- fore, disregarded variation in color. Bales of machine-picked cotton con- centrated (46 per cent) in Strict Low Middling, whereas hand-picked bales were concentrated (41 per cent) in Mid- dling (table 15). Machine-picked bales included 27.1 per cent Middling or higher and 26.9 per cent Low Middling or lower. Hand-picked bales included 22.9 per cent Strict Low Middling and only 8.3 per cent Low Middling or be- low. This same pattern, with minor vari- ations, was repeated in each of the sub- areas except in Madera-Merced where machine-picked bales were concentrated in Low Middling or lower rather than in Strict Low Middling as in other areas. Average grades of both machine- and hand-picked cotton can be calculated by [30] Grades A White & EW GM SM M SLM LM SGO GO Spotted Gray Below Grade White &EW GM SM M SLM LM SGO GO Spotted Gray Below Grade White & EW GM SM M L Hand-picked Machine-picked Hand-picked Machine-picked Bales Percent Bales Bales Percent &a\es 20 20 40 40 20 20 40 1 1 ' SAN JOAQ 1 i 1 UIN VALL 1 EY 1 1 1 I KERN i i;::=:::::::::;;:;::;::::!:: ■ i;;;;::- | i i * |B f m ■ I-::;::;:;:: ;::-::;;•:::;•;:;!! 1 SAN . l:::::::::::-::::::: LLITY E C m i EASl ■ 1 ■ 1 SIDE 01 El IOAQUIN-TRANQU P :•::•::::!:!:•:::::•:■: 1 : - ::::: " : - : < :::::::::: 1 I:! " ; fe3 te:::::::::i::::::: r :::::::::::::::::::: H: 11 i WEST ■ 1 ■ 1 SIDE ■ I EH 1 ■ ■ MERCED ■ _ MADERA I:::::::::: s r- ■ |:::::::::: F ■ ■ SLM LM SGO GO Spotted Gray Below Grade 4 Uiil- 1 E I E 1 i ■ 1 ■ I 1 i | MM; ■ ■ i i l::::::::::::::i 20 20 4 4 20 20 4 Figure 7. Grades of machine- and hand-picked cotton in the San Joaquin Valley, 1949; dis- tribution of seasonal ginning. (Machine-picked bales tended to concentrate in Strict Low Middling White and hand-picked in Middling White grade. Machine-picked cotton averaged slightly less than one full grade below hand-picked cotton.) Source of data: Table 42. assigning index numbers to the various „. . 7 ,. . . , i * -n • -i i i • A complete list ol index numbers is found grades. Primary index numbers used in in table 45, page 70 . This scale is the one used this study were:f ■ by the Production and Marketing Administra- tion in its reports on cotton quality. Good Middling 105 f Each index number is multiplied by the Strict Middling 104 corresponding number of bales in a given lot of Middling 100 cotton; the totals thus obtained are summed c • t Vj- ■ j i li* anc * d^i^d D Y tne tota l number of bales to give btrict Low Middling 94 a weighted average grade index for the lot. [31] Table 1 5. Distribution of Cotton by Grades and by Method of Harvest, San Joaquin Valley, 1949* Grade f Machine-picked bales Hand- picked bales per cent Good Middling Strict Middling Middling Strict Low Middling. Low Middling Strict Good Ordinary Good Ordinary Below grade Total 0.1 4.1 22.9 46.0 21.5 4.2 0.9 0.3 100.0 2.5 25.4 40.9 22.9 6.2 1.4 0.5 0.2 100.0 * Condensed from table 41, page 64, which is based on the ginnings from 35 gins in the Valley. f Summarized by leaf or trash content only because more than 90 per cent of both machine-picked and hand-picked graded White or Extra White in color, and the small differences in color and preparation, between methods of picking, were not significant (see table 42, page 65). Low Middling 85 ing. Rain and fog contribute to colora- Strict Good Ordinary 76 tion, and the ratio of trash to seed cot- Good Ordinary 70 ton is higher because there is less seed „ , _. „ ill i . cotton to pick. Some growers have said ror the Valley as a whole, the machine- .i /. -, j .,,,,/, i • t that machine grades compared more picked bales had an average grade index x 1 1 - t i i j j j r m o i-i-i c • t tavorably with, equaled, or exceeded oi 91.8, which is between strict Low , , , , . . \, rr, , ,,.,,,. 1T Ti/r- 1 n- / ii i^x hand grades late in the season, lomake Middling and Low Middling (table 16). .. & ,, , TT i • i i ii this comparison, weekly average grade Hand-picked cotton had an average . i , _ . -, i u j_ j i.i f n n a i-i-i indexes were obtained lor each method grade index oi 97.4, which is between , . , . . . , L -. . , . a/pjji- j c* • * t Tv/r-jji- tu °* Peking at eight selected gins for the Middling and Strict Low Middling. The lrMn tan j * \. \ ao j. „ ° . t - i j- r r 1949 season (fig. 8 and table 43, page dmerence in grade index oi 5.6 repre- ^ n , A/r , . . , , , i ,. , . ? , f „ i 68) .j Machine-picked grades on the av- sents slightly less than one lull grade. -,-j , , , . , , rr,, ! r ,P -. erage did not equal hand-picked at any lhe average value oi machine- and ,. , . . , , i_ j . i i „ t t t time during the season at any ol the hand-picked cotton can also be compared . -, a . A n . , , - i * t^ eight gins. At five gins there was some in terms oi government loan value, r or . j i v«- .- i . „- . G . d , r „ , tendency toward narrower differentials the o5 gins m the Valley, the average loan value was $132.52 per bale of ma- * In estimating loan values it was assumed chine-picked and $142.84 per bale of that a11 bales av eraged l^e-inch staple length hand-picked, a difference of $10.32 a b u ecaui \ e actu A al sta P les were not re Pf * d in , , / ii 1/r \ r™ tcc these data. A very large proportion of Valley bale (table 16). The difference was an d subarea bales is IMo-inch staple, so the re- smallest m Westside ($8.06) and largest suits are not impaired by this assumption. in the Madera-Merced area ($19.75). There is no evidence whatever that machine Growers and others are familiar with picking has any effect on staple length. the downward trend in grades of cotton , t The data include all the machine- and j - .i i m . i hand-picked bales ginned the same two consecu- during the harvest season. Ihis trend , , . , , , no i . -, , i n -,-> tlve "ays m eacri wee k during an 18-week pe- ls largely attributed to weathering and riod . Wednesday and Thursday were used, increasing trash content, both of which though any other two days would have served are particularly evident in second pick- the purpose. [32] as the season progressed; otherwise, machine-picked followed the same down- ward trend as hand-picked cotton. Grades of machine-picked varied more from week to week than the hand-picked at all eight gins. Machine grades defi- nitely improved in the first week of No- vember over the last week in October at seven out of eight gins after a heavy frost had defoliated the cotton and im- proved conditions for picking. Individually, some growers had high grades of machine-picked cotton and others had low grades in all five sub- areas (table 44, page 69). That some growers in each area obtained good sea- sonal average grades indicates that suc- cessful machine picking is possible in all areas of the Valley. If grades are to be maintained, however, cotton must be grown with mechanical harvesting in mind. Rows must be properly spaced, the fields kept clean of weeds and grass, and the ground surface left smooth and free of clods. Recent research at the Shafter Station indicates that the crossrow ground profile is also important. If the ground is left higher in the stalk row than between the rows, the plant lifters on the machine have more space to feed low-growing branches into the machine. Too, dead leaves are more likely to fall away from the cotton stalk and are less likely to be picked up by the machine. Trends in machine-picked cotton grades, 1948-1950 Data are available from representative gins in the San Joaquin Valley with which to compare the average grades of machine- and hand-picked cotton in each of the three years 1948-1950 (table 17; also see table 45, page 70, for a more de- tailed distribution of bales by grade and method of picking). Apparently the dif- ference in grade narrowed during that period but not in the way most pleasing to growers. The smaller grade index dif- Table 16. Average Grade Index and Government Loan Value of Cotton by Method of Harvest, 35 Gins, San Joaquin Valley, 1949 Hand-picked Machine-picked Difference* San Joaquin Valley grade index 97.4 98.9 97.2 95.4 97.7 94.4 91.8 92.4 89.6 89.1 93.3 85.8 -5.6 -6.5 -7.6 -6.3 -4.4 -8.6 Subareas : Kern Eastside San Joaquin-Tranquillity Westside Madera-Merced San Joaquin Valley government loan value (per bale) $ 142.84 144.79 142.39 139.65 143.96 136.95 $ 132.52 133.94 127.67 127.97 135.90 117.20 - $10.32 - 10.85 - 14.72 - 11.68 - 8.06 - 19.75 Subareas : Kern Eastside San Joaquin-Tranquillity Westside Madera-Merced * Minus sign indicates machine-picked below hand-picked. [33] GIN N-l 70b^ t i i i i i i i i i i i i i i i i \y- 1 i i i i i i i i i i i i i i i 1 3 5 7 9 11 13 15 17 1 3 5 7 9 11 13 15 17 Week of Picking Season Week of Picking Season ▲ A ▲ A ▲ A A A_ Sept. October November December Jan. Sept. October November December Jan. Figure 8. Seasonal trends in grades of machine-picked and hand-picked cotton in the San Joaquin Valley, 1949; weekly average grade indexes at selected gins. (Grades from the machined cotton averaged lower than hand-picked in all periods of the season. Grades for both declined throughout with some tendency for the spread to narrow.) Source of data: Table 43. ferential of 3.1 in 1950, compared with 7.4 in 1948, resulted from a reduction in hand grades and not from any aver- age improvement in machine grades. In fact, the latter varied only one-half point during the three years — 91.3 in 1948, 91.8 in 1949, and 91.5 in 1950. Mean- while, the hand-picked declined from 98.7 to 94.6. This decline is attributed by most informed observers to progres- sively poorer hand picking owing to a larger proportion of inexperienced workers. (A similar decline has been noticed in some other cotton states.) The [34] Table 17. Grade Index and Loan Value of Cotton from Selected Gins, by Method of Harvest, San Joaquin Valley, 1 948-1 950 Item 1948 1949 1950 26 Number of gins 22 35 Number of bales : Hand-picked thousands 160 17 238 63 199 63 Machine-picked Hand-picked grade index* 98.7 91.3 97.4 91.8 94.6 91.5 Machine-picked Difference Hand-picked -7.4f -5.6 -3.1 government loan value (per bale) % $ 152.58 142.83 $ 142.84 132.52 $ 137.89 131.05 Machine-picked Difference Difference if basic loan rate were 30 cents - $9.75 - $9.25 - $10.32 - $ 9.99 - $6.84 - $6.83 * See table 43, page 68, for the indexes assigned to the individual grades of cotton from which average indexes were computed. t Minus sign indicates machine-picked below hand-picked. t Loan value per pound of Middling White, 1 1/16-inch staple, by crop years was: 1948 — 31.62 cents; 1949 — 30.98 cents ; and 1950 — 30.03 cents. average loan value of machine-picked cotton was $6.84 per bale less than that of hand-picked cotton in 1950, com- pared with $10.32 less in 1949, and $9.75 less in 1948. It is perhaps safe to say that machine picking tended to improve from 1948— 1950, even though Valley-wide averages do not show it. Some growers obtained much better grades at the end of this period than when they first started to use a machine. But the Valley average was held down, and continued to be through 1952, because of the fact that many new growers, inexperienced in mechanical harvesting, were using ma- chines for the first time and because many of these mechanical harvesting machines were used under less favorable conditions. [35 PICKING EFFICIENCY OF MECHANICAL HARVESTERS To determine the net economic advan- tage of machine harvesting over hand picking, comparative field waste must be evaluated as well as picking costs and comparative lint grades, discussed earlier. If field waste is greater in machine than in hand picking, the field value of the additional seed cotton left by the ma- chine also contributes to the total cost of mechanical harvesting. Field waste means open seed cotton left in the field. Thus, field waste, or, conversely, picking efficiency, is ex- pressed as a percentage of the total seed cotton available for picking at the time of harvesting. Actually there are two criteria for comparing the efficiency of machine harvesting. One is an absolute criterion, the answer to the question, "What percentage of all the open seed cotton does the machine pick?" The other is relative, the answer to the ques- tion, "How does the percentage of open seed cotton left by the machine compare with that left by hand pickers?" Grow- ers are interested in the answers to both. They want their machines to be as effi- cient as possible, but, in terms of eco- nomic advantage, they must compare ma- chine with hand picking, the only fea- sible alternative harvesting method. A mechanical harvester can contribute to field waste in numerous ways. Like any other mechanical device, it is lim- ited to a strictly repetitive pattern. It cannot see and therefore does not go back and pick a stray boll, once missed. The machine cannot pick cleanly if it is permitted to wander off the row. It some- times misses the lower bolls — those six inches or less from the ground; this is more noticeable when the picker drums cannot be operated close to the ground because of roughness or clods. Another characteristic of machine-picked fields is the presence of "tag" — locks or parts of locks streaming from branches of the cotton plants. Again, if the machine be- comes clogged, some cotton is soiled and must be discarded in the cleaning. In all these ways machine picking can lead to excessive field waste. Field waste is measured accurately in only one way — by hand gleaning behind the harvester, a time-consuming job, which, if it is to provide adequate data, must include a representative range of field conditions. This was beyond the scope of the present study. Very few of the growers interviewed had actually measured field waste. Consequently, this section of the analysis is based on pick- ing-efficiency studies made at the United States Cotton Field Station at Shafter (Hoover, 1949). Measurements of field waste at the Station were carefully taken under controlled conditions and repre- sent the most reliable data available. Over-all efficiency of machine harvest- ing was 96.5 per cent at the Shafter Station in the 1949 season. This means that at the end of the season, after sec- ond picking had been completed, the harvester had left in the field 3.5 per cent of the seed cotton available for picking. Efficiency was higher in 1949 than in previous seasons: 93.4 per cent in 1948 and 92.4 per cent in 1947. The Shafter experiments included no hand picking in 1949 or 1947, but the mechanization project in 1948 measured the efficiency of hand picking at 97.6 per cent. This figure appears to be about the maximum efficiency to be expected in hand picking. These efficiencies in machine and hand picking at the Shafter Station are prob- ably somewhat higher than those at- tained by the average grower. However, it is believed that they represent the ap- proximate relationship between machine- and hand-picking efficiencies that grow- ers have experienced. These efficiencies can serve to indicate the economic im- [36] portance of field loss. For the purpose of this analysis, however, it is assumed that the mechanical harvester is respon- sible for 3 per cent more field loss than hand picking. The average yield among interviewed growers (1949) was 2,171 pounds of seed cotton harvested per acre. If ma- chine harvesting was 94.6 per cent effi- cient, the natural yield (amount avail- able for picking) was 2,295 pounds per acre. Thus, machines left in the field 124 pounds of seed cotton per acre. In com- parable cotton it is estimated that hand pickers picked 2,240 pounds of seed cot- ton and left in the field 55 pounds per acre. Thus, machines left about 69 pounds more per acre than hand pickers. The value of seed cotton in the field be- fore picking was about $7.28 per 100 pounds in 1949.* The value of 69 pounds was, therefore, about $5.02, and the average value of field loss per har- vested bale was about $3.31. Discussion Although few growers had systemati- cally measured the picking efficiency of their machine harvesters, most growers said that machines did a more thorough job of "cleaning the field" in 1949 than in previous years. Very few were still concerned about field waste by the end of the 1949 season. This absence of con- cern may be partly because field waste actually was not large, partly because * Field value of seed cotton equals the value of 37.1 pounds of lint ($9.83) , plus 58.9 pounds of cotton seed ($1.40), minus (hand) harvest- ing costs ($3.25) and ginning costs ($0.70). growers had learned that machine- picked fields appear to have more waste than they do, partly because they had found that machines gather in second picking some of the waste from the first, and, finally, because they were more fully aware of the extent of waste from ordinary or poor hand picking. In general, growers expressed the opinion that field waste is relatively lower in rank-growing, high-yielding cot- ton. In such cotton a smaller percentage of the bolls are close to the ground where they are hard to reach. Another general conclusion is that competent machine operators are essential to efficient pick- ing. Some growers limited the hours per work shift to avoid overfatigue and lowered efficiency of machine operators. In actual operation, the two cautions most frequently cited by growers as es- sential to high picking efficiency were: 1) keeping the machine on the row; and 2) keeping the machine clean. Research at the Shafter Station indi- cates that row spacing of 40 inches, with cotton stalks 4 to 8 inches apart in the row, provides optimum conditions for mechanical harvesting. Other recom- mendations (Hoover, 1949) are that: The rows should be uniform in height, width, and shape. The rows should be smooth and free of clods. The crest should be at the base of the stalk. Furrows should be wide enough to per- mit steering of the picker. [37] ECONOMIC ANALYSIS OF MECHANICAL COTTON HARVESTING The analysis in preceding sections has been concerned primarily with the data and evidence as found. It is now appro- priate to see what economic generaliza- tions may be distilled from those data interpretations. There are important questions that cotton producers need answered. What is the total economic cost of mechanical harvesting? How do variations in annual use and in cotton yields affect picking cost? Under what situations is mechanical picking feasi- ble? What are the comparative costs of first and second picking? Is mechanical picking economically feasible in very light picking or in scrapping operations? Summary of economic costs Detailed data on machine-picking costs were presented in an earlier section of this report, and it was noted that the cost of picking alone does not represent the full "economic cost" of mechanical harvest as compared with hand harvest- ing. Such a comparison must also con- sider relative lint grades and field waste of the two methods. There is evidence that machine picking results in lower grades of cotton and causes somewhat more field waste than hand picking. It is appropriate, therefore, to evaluate and summarize all the economic costs of me- chanical harvesting. The analysis is based on 1949 data and prices.* The average cost of machine picking for all growers interviewed in 1949 was * Although prices, and therefore costs of picking, were relatively higher in 1950 and 1951, the 1949 price level is deemed more rep- resentative of the long-time future average. t The average pick in 1951 was estimated at 264 bales per machine. t Much of the 1950 and 1951 crops sold at prices well above loan value, but volume of sales of particular grades by specific margins is not known. § The computation of the field value of $7.28 per 100 pounds of seed cotton was shown in the footnote on page 37. L65 per bale. That cost was based on a total season pick of 229 bales (some- what less than machine capacity) in the San Joaquin Valley,f and a service life of five years for the harvester. Further study revealed that for the present anal- ysis it would be more reasonable to as- sume that a machine picks 300 bales per season and has a service life of six years. The picking cost adjusted for these as- sumptions is $12.29 per bale. Although the 1949 data showed differences in picking cost among the subareas due to variation in the season's pick, this anal- ysis assumes the same pick and the same picking cost per bale in all subareas. One measure of grade loss is the gov- ernment loan value. t The loan value of machine-picked cotton averaged $10.32 per bale lower than hand-picked cotton in 1949. Because the loan-value differ- ential varied among the subareas of the Valley, it seems reasonable to assign dif- ferent charges for grade loss in each. It is recognized that the differences found in 1949 may become smaller as the newer areas gain experience in ma- chine picking. Some differences, never- theless, are expected to persist because of cropping and cultural variations in the subareas. Machine picking in 1949 resulted in 3 per cent more field waste than hand pick- ing (see pages 36-37). In 1.5-bale cot- ton machine picking left 65 more pounds per acre of seed cotton in the field than hand picking. The economic cost thus represented amounted to $4.73 per acre, or $3.15 per harvested bale.§ No signifi- cant variations in field waste among the subareas were noted. The total economic cost of machine harvesting averages $25.76 per bale in the San Joaquin Valley, under the con- ditions assumed. The cost is highest ($35.19) in Madera-Merced where grade loss averages highest; it is lowest [38 ($23.50) in Westside where grade loss is lowest. The net economic advantage of machine picking over hand picking is $19.24 a bale in the Valley; it is highest in Westside and lowest in Madera- Merced (table 18). If grade loss continues to diminish, as the 1948-1949 trends suggest, the total economic cost of machine picking will become lower. The net economic advan- tage of machine picking, therefore, will tend to increase. Some might argue that no grade loss should be charged to mechanical harvest- ing. Machines picked some 15 per cent of the 1949 crop, hence the harvest was completed earlier than if it had been picked entirely by hand. Even with the help of machines, about 15 per cent of the crop was not picked until after De- cember 13.* Without machines, even more of the crop (perhaps 15 per cent more) would have been harvested after mid-December when the standing cotton and the ginned lint are of lower grades. * Computed from Cotton Quality Reports, Production and Marketing Administration, Bakersfield, California. f It is assumed the 900 machines each picked 229 bales, the average found among interviewed growers. California cotton ginned between mid- December and mid-January averaged Low Middling (grade index of 85.7). The average grade index for machine- picked bales for the entire season was 91.8, about three-fourths of a grade above Low Middling. The difference in loan value between a Strict Low Middling and a Low Middling bale was about $25.00. It might, therefore, be reasonable to assume that approximately 205,000 bales of cottonf were $18.75 a bale higher in value, though machine picked, than they would have been hand picked, at a later date. That is to say, a machine- picked bale in October was worth more than a hand-picked bale in December The individual grower who uses or con templates using a machine, however must consider grade loss as a cost of ma chine harvesting. Operators, on the aver age, received lower grades when they used machines. Individually, they could have used hand pickers and obtained higher grades, even though not all of them could have obtained enough hand pickers. In a sense, growers who used machines (and took lower grades) made it possible for other growers to complete their harvest with hand pickers (and get TabSe 18. Total Costs of Machine Picking Compared with Hand Picking Hand picking $ 45.00 Machine picking Item San Joaquin Valley Subareas Kern Eastside San Joaquin- Tran- quillity Westside Madera- Merced Picking cost $ 12.29 3.15 10.32 $ 12.29 3.15 10.85 $ 12.29 3.15 14.72 $ 12.29 3.15 11.68 $ 12.29 3.15 8.06 $ 12.29 Field waste 3.15 Grade loss 19.75 $ 45.00 Total economic cost* Difference in favor of machine picking . $ 25.76 $ 19.24 $ 26.29 $ 18.71 $30.16 $ 14.84 $27.12 $ 17.88 $ 23.50 $21.50 $35.19 $ 9.81 : Additional ginning costs for machine-picked cotton, averaging about 11 cents per bale, not included. [39] higher grades). In this sense, grade loss seems a real and direct loss to growers who use machines. Effect of changing prices on machine- vs. hand-harvesting costs Some farmers may question whether the cost of machine harvesting would equal or exceed the cost of hand harvest- ing if picking wage rates again become relatively low as in times past. This ques- tion is important to a farmer considering a shift from hand to machine picking. The answer turns on the probability of lower wages relative to prices and costs of operating farm machinery. As a par- tial answer, we can first examine those periods in the past when price relation- ships (cheap labor) were unfavorable to mechanization. We can then speculate about the future. Between 1935 and 1952, the year 1938 was the one in which farm wage rates were lowest relative to prices and costs of operating farm power machinery. In 1938 the index of wage rates for picking cotton was only two-thirds as high as the index for machinery prices (table 19). That means that labor was relatively cheap. On the other hand, in 1949 the index of wages was l 1 /^ times the index of machinery prices. This study has shown the cost of machine picking for that year as $2.05 per 100 pounds of Table 19. Indexes of Prices Paid by Farmers in the United States and Wage Rates for Picking Cotton in California (1910-1914=100) Year Farm machinery Farm supplies Farm wage rates Wage rate for picking 100 pounds of seed cotton* 1934 index index 144 148 150 153 158 155 153 155 164 170 174 176 182 206 240 270 275 297 308 134 134 136 143 146 142 146 156 172 192 201 204 206 222 236 246 247 264 280 99 107 114 129 130 127 129 151 197 262 318 359 387 419 442 430 432 481 ...f $0.90 0.90 1.00 0.95 0.75 0.85 0.95 1.30 1.90 2.10 2.25 2.25 2.90 3.00 3.00 3.00 3.45 3.70 3.60 123 123 136 130 102 116 130 177 259 286 307 307 396 409 409 409 470 505 491 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 Source : Bureau of Agricultural Economics. * Average rate paid up to November 1 ; includes rates paid for snapping bolls converted to seed-cotton equiva- lent. As picking wage rates were not reported before 1934, the index was converted to the 1910-1914 base by assuming that the relationship between picking rates and all United States farm wage rates was the same in 1910-1914 as in 1935-1939. f Data not available. [40] seed cotton, compared with $3.25 for hand picking. Since World War II labor generally has been scarce and wage rates high, a situation that encourages mech- anization. When 1949 machine-harvesting costs, as determined in this study, are con- verted to 1938 prices, the resulting cost of picking 100 pounds of seed cotton is: Machine-picking cost Value of grade loss Value of field waste Total ).465 .213 .065 1.743 These estimates are for a machine with full season use on 200 acres of cotton and 300 bales picked. This cost would be correspondingly higher if the machine picked less than 200 acres or 300 bales of cotton a season. The effect of season use on harvesting costs is discussed more fully on pages 42-45. The estimated cost of hand harvesting per 100 pounds of seed cotton, at 1938 prices, would be: Wages of pickers Supervisory labor Total 1.75 .08 1.83 The above comparison shows that ma- chine harvesting of cotton would be more economical than hand harvesting even at 1938 prices, a year when wage rates were very low relative to machinery prices and costs of operation. Another question is, "How representa- tive is the relationship between machine- harvesting and hand-harvesting costs as found in 1949?" Is that relationship likely to change in the future, and if so, by how much? The answer to this ques- tion might be important to the individual farmer who has less than a full season's use for a mechanical picker or whose cotton yields are below average (the above costs were estimated for a yield of 1.5 bales an acre). For this analysis, two alternative price levels were assumed. One assumes high employment and prosperous conditions. The other assumes intermediate employ- ment and less prosperous conditions. The indexes of prices received and prices paid (1910-1914-100) for each of these projected levels compared with 1949 are shown in table 20. Table 21 shows the estimated cost of picking 100 pounds of seed cotton (including a charge for grade loss and field waste in the case of machine picking) when these price Table 20. Index of Prices Received and Prices Paid (1910-1914=100) Prices received : All products Cotton, lint Prices paid : Items in production Farm machinery. . . Farm supplies Wage rates 1949 249 245 250 270 246 428 Projected High employment 215 185 215 225 210 360 Intermediate employment 150 126 155 175 175 275 Source: Bureau of Agricultural Economics. [41] Table 21. Estimated Cost of Picking 100 Pounds of Seed Cotton When 1949 Machine- and Hand-Harvesting Costs are Converted to Projected Index Levels 1949 Projections High employment Intermediate employment Machine harvest cost per 100 lbs. Full use (200 acres): 1.5-bale yield $1.86 2.75 2.05 3.13 3.25 $1.48 2.22 1.64 2.54 2.73 $1.08 1.66 1.20 1.91 1.70 .75-bale yield Half use (100 acres) : 1.5-bale yield .75-bale yield Hand harvest cost per 100 lbs. Source: Bureau of Agricultural Economics. levels and price relationships are used to convert the cost of machine harvesting and hand harvesting found in the present study. These projections indicate that farm- ers who obtain cotton yields as low as 0.75 bale per acre and who harvest only 100 acres a season would find machine- picking cost only slightly lower than hand-picking under the high employ- ment projection; under the intermediate projection, machine harvesting would cost more than hand harvesting. Farmers who usually obtain average or above- average cotton yields, even though they harvest but 100 acres of cotton a season, can expect considerably lower cost of harvesting with a machine than with hand picking. This was true in 1949 and would be true under either the high or intermediate employment conditions. It should be noted that these estimates are based on cost data in 1949 when machine harvesting, though well established, was still relatively new. Improvements in machine pickers, gins, and methods may have reduced machine-picking costs still further. An offsetting factor, of course, has been the increased prices and over- head costs of this equipment. The cost advantage of machine over hand picking, then, may have been some- what greater under 1949 price and wage conditions than it is likely to be under the prospective longer-term conditions. Nevertheless, the advantage is likely to be real and substantial, except for the farmer who obtains low yields of cotton and who makes only half-time or less than half-time use of a machine picker. Effect of acreage harvested on picking costs The machine-picking costs already presented assume a reasonably full sea- sonal use of the harvester — picking 200 acres of cotton, of which 150 acres is picked a second time. Some growers have more than 200 acres, while others have much less. This fact raises practical economic questions. What effect does the acreage harvested per season have on picking costs? Does it pay to maximize harvester use? What minimum acreage is needed to justify owning a harvester? The effect of annual use upon picking cost depends largely upon the treatment given overhead. Depreciation is the major item of overhead cost, accounting for 80 per cent of the total. In turn, deprecia- tion is related to the service life of the harvester. Service life may be the num- [42] t- 00 LO LO LO lO lO ^ m .O CO r-J CO q 00 8 © CO CT> r-4 co* id ■<** CO CO CO CO CO CO CO D C fc €0- C < **« 0) O lO 10 10 10 10 10 U» 5 q 1> 10 10 T-j 00 00 CM CM CM CM CM D u 9 Ph id eo CM iH 69- oi t>* co id id id id id id id X "0 co c s a H O «A O O *tf CM CM ts A O lO 00 q "*. *°. M H H rl H H Li 0) Pk id CO CO CM* CM* CM CM* CM* CM* CM* CM* u < 4) Q u O = +■ *«8 10 00 CM CM iH !>©•••• 10 S3 CM 00 T-i IO CO l> H IO • • • • > < CO CO t- C- t- CO q cq Li ee- s X ■B 8 ton 3 O O O O OOOO OOOO 8 8 : s O 10 t> CD i-H tH tH CO x H id 10 id" co" co" co" «o" co" u €*• 0) 5 8 *4- O O • • • >5 O O 00 c O IO eo cq 05 « 0* "> ^ # *£ 8 B 1> rH t4 to & i js ci u I I O a. S . • Q. bo •-as ° O O «-»- •J3 to 88 10 q 10 q m O CM 10 CO 10 CM O O O O O CO cq cq cq 43 a ? ° ■sis . * "8 * i-T r-T cm cm" cm" CM CO CO CO CO CO C 8 CD |gS • g 0) a& © 3 > m.2 3 CD m to 3 00 00 00 00 00 c- * * * * * * * * CO cc IO 10 ^ ^ a- b*» ► CD fl^ Ogc8 4) co0.g' gissrssi CD |*i? is: «/> "8 +■ 8 . to >, • +J • »a •■§ §g bo 3V .5 CO CM IO 00 CM lO 00 O CM lO 00 O 25 0) CO 1 j 9 m 0. •» bo X x c a, J3 5 "5 H a rH C t- CC s d 2 £ en CD O) 01 en 00 co cn lO (N O) CO NIO 00 H^HO t- CO COO CO* riddd rH u c 1 OCO c i rH O r > * a d :c CO CO «* «» ««• tt< 4 c O 'g "5 O CO bo 0" t* § s a H s 1 ■* ^ CO lOTt" O) HOIOO CO a > co co a ! S « CO* 00 00 O CO 00 ^JJ COCOCNC0 w c > Tf00 c 1 a> v ■n O ■<# OC H 5* °* o> Tfd id riddd cic ^ < «* t» te- «» • ! 1 H~ co 3 (X s s as OCO CO COOOE- O rH -^O ■«* Oit-lO ^ 2 r* 00 a n c9 0) Wt-t- CO COO CO s CO CO o~ of rH CO rHCO CO COrH CO ri rH~ CO ^ a s. § .H Xm CO » w J3 <=! eS ° I E 3 S Pi V to ^0 rt rHCO t- # co a ^ c ifjif) c ) ^3 Pi "H jg ■«* w COr- O CDrH ^ IfJrHCOO O t> Tjlf O 1 u °- z 'co "3 *"! "*. 85 6 rHrH co' dc t-^ Co'rHOd ^' Co'rH - rH rH O u d «9- «» «* 6» «» <* ► O w © -a *■ > c4 > a> in O CO J3 u. >-, > 2 2 v. coc- C coco a- O)''* CO COCOCOC5 T»< <£ Or- 1- s ° 5 CD •H 0)0) t-^ r coc O O^rHC co a Ot- t< is .g X ^ .^ HC OC tJ«'o ir: coc w . riddc CT rHC r- s ffi d '.c co M "3 u O # c COtJ c r? -S oJ i_" a3 ■9 8 3. 000 t-W *j COO) CN S 6 " ir: COO) CO lf3 c- rHIC t" +■ q ocooc <3 rH * rt — Ww eg M g CO £ lil 3 c Eh «• -2* 82£'5 .S2Wh ■So O d to j- O 0) 2 J3 g "3 H iSria ■go" 4 U K g r Oft 05 c3 O rl cf w 4S O H s ~-'bo •0 a gta ^ fl (-» OC ft"2 "3 "H ■S s rl 3 -S Table 32. Summary of Picking Costs per 100 Pounds of Seed Cotton for Mechanical Harvester Number 2, 1950 First picking . . Second picking All picking One-row machine* $1.42 4.36 1.76 Two-row machine f $1.05 4.48 1.47 * This cost covers overhead, operation of equipment, and labor, a total of $9.86 per operating hour. See tables 38 and 39, pages 61-62, for detail of items included. t This cost covers overhead, operation of equipment, and labor, a total of $12.06 per operating hour. See tables 38 and 39, pages 61-62, for detail of items included. (table 31) . The costs were $4.00 per bale cheaper in 1950 for the two-row, larger capacity machine. Picking costs per 100 pounds of seed cotton were $1.47 for the two-row ma- chine, 16 per cent lower than for the one-row. A summary of picking costs per 100 pounds of cotton by first and second pickings and for the season are summarized in table 32. APPENDIX How farmers were selected for interview From information supplied by har- vester dealers, a list was prepared of all persons known to own mechanical pick- ers in the San Joaquin Valley in 1949. About 400 farmers owned one harvester each, 75 owned two each, and 50 owned three or more. Only those farmers who owned one picker were interviewed, as it was felt that they represented the most typical situation and would be better able to provide complete information about individual machines. Owners whose op- erations were mainly custom work for others were excluded, insofar as possible, because such farmers are less typical. The harvester dealers helped edit the list to exclude such custom operators. Thus, the original list was limited to farmers with one mechanical picker used pri- marily to pick their own cotton. Owners in the revised list were then sorted into five geographical subareas (fig. 2). From the list for each area, a sample was selected for interview. Selec- tion was on the basis of every nth name; in some areas this meant every fourth name; in others, every sixth. In the Madera-Merced area, it was necessary to include all 14 farmers owning ma- chines to obtain a sufficient sample. The resulting sample included the following number of interviews by subareas: Kern 15 Eastside 16 San Joaquin-Tranquillity 9 Westside 9 Madera-Merced 14 Total 63 Since the sample was designed to rep : resent the typical situation, it does not necessarily represent the arithmetical average. However, it is believed that, from the standpoint of output per ma- chine, cost of picking per bale, and re- lated data, the typical situation would not differ greatly from the over-all aver- age. [52] ACKNOWLEDGMENTS Although many people offered helpful suggestions toward planning this study and preparing the report, only a few can be mentioned here. The authors are especially grateful for assistance provided in 1949-1950 by J. P. Fairbank, Regional Director of Agricultural Extension Service, College of Agriculture, Berkeley; Marvin Hoover, Extension Cotton Specialist, Shafter; Ray Provost, Producers Cotton Oil Company; R. V. Jensen, San Joaquin Cotton Oil Company; J. Russell Kennedy, California Cotton Cooperative Association; W. B. Lanham, Cotton Branch, Produc- tion and Marketing Administration; the Farm Advisors, Agricultural Extension Service, in Fresno, Kern, Kings, Madera, Merced, and Tulare counties; and the harvester dealer-distributors in the Valley. The authors are appreciative also of the helpful cooperation of the cotton growers who furnished records of their mechanical harvest operation in the 1949 season. LITERATURE CITED Archer, R. C. 1948. Progress in the mechanical picking of cotton. Proc, Ind. Ann. Beltwide Cotton Mech. Conf., Lubbock, Texas. Burlingame, Burt B., and Warren R. Bailey 1950. Cost of harvesting cotton with mechanical pickers — California — 1948. California Agr. Ext. Serv. (Mimeo.) Campbell, John D. 1948. Oklahoma farmers' experience with cotton strippers. Oklahoma Agr. Exp. Sta. Bui. B-324. Crowe, Grady B. 1949a. Mechanical cotton picker operation in the Yazoo-Mississippi delta. U. S. Bur. Agr. Econ. Mississippi Agr. Exp. Sta. Prog. Rept. 19496. Mechanical cotton picker operation in the Yazoo-Mississippi delta. Mississippi Agr. Exp. Sta. Bui. 465. Holley, William C, and Loyd E. Arnold 1938. Changes in technology and labor requirements in crop production — cotton. Works Prog- ress Administration, National Research Project, A-7. Hoover, Marvin 1949. Cotton mechanization. From Project Report by J. P. Fairbank and K. 0. Smith (Project 1361, Dept. of Agricultural Engineering). California Agr. Ext. Serv. (Mimeo.) Sutherland, J. Gwyn, and H. Brooks James 1948. Mechanical harvesting of cotton in North Carolina, 1947. North Carolina Agr. Exp. Sta. Agricultural Economics Information, Series 20. Venstrom, Cruz 1946. Experience in 1945 with mechanical cotton pickers in California. U. S. Bur. Agr. Econ. (Mimeo.) Williamson, M. N., Jr., Q. M. Morgan, and Ralpih H. Rogers 1951. Economics of mechanical cotton harvesting in the high plains cotton area of Texas. Texas Agr. Exp. Sta. Bui. 735. Williamson, M. N., Jr., and Ralph H. Rogers 1949. Economics of cotton harvesting, Texas high plains, 1948 season. Texas Agr. Exp. Sta. Prog. Rept. 1200. '[53] APPENDIX TABLES Table 33. Cotton Harvested by 63 Growers in the San Joaquin Valley, 1949, by Method of Harvest Number of growers Total har- vested Includes machine custom picking for other growers Area Both pickings First picking Second picking Machine- picked Hand- picked Machine- picked Hand- picked Machine- picked Hand- picked 63* 15f 16 9t 9§ 14H bales San Joaquin Valley. . . Subareas : Kern 429 443 428 350 601 357 229 258 245 201 292 158 200 185 183 149 309 199 182 201 185 183 244 116 165 144 156 53 278 170 47 57 60 18 48 42 35 41 Eastside 27 San Joaquin-Tran- quillity Westside 96 31 Madera-Merced. . 29 Note: Averages are for all growers interviewed whether or not they all used machines or hand pickers in first or second picking. * Eleven growers had no hand picking ; four had no machine first picking ; and two had no machine second picking. t Four growers had no hand picking; three had no machine first picking; and one had no machine second picking. i Two growers had no hand picking; one had no machine second picking. § Four growers had no hand picking. \ One grower had no hand picking ; one grower had no machine first picking. [54] Table 34. Average Cost of Mechanical Cotton Picking with Harvester Number 1, by Subareas, San Joaquin Valley, 1949 Kern area (15 growers) Eastside area (16 growers) Item Season total Average Season total Average per Acre per Hour per Bale per Acre per Hour per Bale Acres of picking Machine hours Bales harvested 263 414 258 1.57 0.98 0.64 0.62 1.02 1.60 309 427 245 1.38 0.79 0.72 0.57 1.26 1.74 Investment: Harvester Tractor $3,738 1,620 $5,358 1,471 335 $1,806 778 156 $ 934 499 65 11 6 $14.21 6.16 $20.37 5.59 1.27 $ 6.86 2.96 0.59 $ 3.55 1.90 0.25 0.04 0.02 $ 2.21 $ 9.03 3.91 $12.94 3.55 0.81 $ 4.36 1.88 0.38 $ 2.26 1.20 0.16 0.03 0.01 $14.49 6.28 $20.77 5.70 1.30 $ 7.00 3.02 0.60 $3,709 1,730 $5,439 1,410 306 $1,716 1,005 155 $1,160 495 89 3 7 $ 594 $12.00 5.60 $17.60 4.56 0.99 $ 8.69 4.05 $12.74 3.30 0.72 $15.14 7.06 Total Picking costs Overhead : Harvester Tractor $22.20 5.75 1.25 Total Operat'g expenses: Harvester Tractor $ 5.55 3.25 0.50 $ 3.75 1.60 0.29 0.01 0.02 $ 4.02 2.35 0.36 $ 7.00 4.10 0.63 Total Labor : Operating* Service and repair Farm shop Compensation. . . $ 3.62 1.93 0.25 0.04 0.02 $ 2.24 $ 2.71 1.16 0.21 0.01 0.02 $ 4.73 2.02 0.36 0.01 0.03 Total $ 581 $3,321 $ 1.40 $ 1.92 $ 1.40 $ 2.42 Total picking costs . $12.62 $ 8.02 $12.86 $3,470 $11.22 $ 8.13 $1,415 man-hours Labor used : Operating Other Total 414 91 505 1.57 0.35 1.92 1.00 0.22 1.22 1.60 0.35 1.95 427 110 537 1.38 0.36 1.74 1.00 0.26 1.26 1.74 0.45 2.19 Continued on next page. See end of table for footnotes.) [55] Table 34. Average Cost of Mechanical Cotton Picking with Harvester Number 1, by Subareas, San Joaquin Valley, 1949 — {Continued) San Joaquin- Tranquillity area (9 growers) Westside area (9 growers) Item Season total Average Season total Average per Acre per Hour per Bale per Acre per Hour per Bale Acres of picking Machine hours Bales harvested 273 316 201 1.16 0.74 0.86 0.64 1.36 1.57 317 479 292 1.51 0.92 0.66 0.61 1.08 1.64 Investment Harvester Tractor $3,729 1,678 $13.66 6.15 $19.81 5.21 0.88 $ 6.09 3.22 0.44 $ 3.66 1.62 0.22 t 0.02 $11.80 5.31 $17.11 4.50 0.76 $ 5.26 2.78 0.38 $ 3.16 1.41 0.19 t 0.02 $ 1.62 $18.55 8.35 $26.90 7.07 1.19 $3,672 1,653 $11.58 5.22 $16.80 4.26 0.98 $ 7.67 3.45 $11.12 2.82 0.65 $12.58 5.66 Total Picking costs Overhead : Harvester Tractor $5,407 1,421 239 $5,325 1,349 312 $18.24 4.62 1.07 Total Operating expenses Harvester Tractor $1,660 879 120 $ 999 444 61 t 6 $ 8.26 4.37 0.60 $ 4.97 2.21 0.30 t 0.03 $ 2.54 $1,661 815 173 $ 988 660 103 t 9 $ 772 $5.24 2.57 0.55 $ 3.12 2.09 0.32 t 0.03 $ 2.44 $ 3.47 1.70 0.36 $ 5.69 2.79 0.59 Total $ 2.06 1.38 0.22 t 0.02 $ 3.38 Labor : Operating* Service and repair . Farm shop Compensation. 2.26 0.35 t 0.03 Total $ 511 $ 1.86 $ 1.62 $ 2.64 Total picking costs $3,170 $11.61 $10.04 $15.77 $3,421 $10.80 $ 7.15 $11.71 man-hours Labor used : Operating Other 316 66 1.16 0.24 1.00 0.21 1.21 1.57 0.33 479 117 1.51 0.37 1.00 0.24 1.64 0.40 Total 382 1.40 1.90 596 1.88 1.24 2.04 Continued on next page. (See end of table for footnotes.) [56] Table 34. Average Cost of Mechanical Cotton Picking with Harvester Number 1, by Subareas, San Joaquin Valley, 1949 — (Continued) Item Madera-Merced area (14 growers) Season total Average per Acre per Hour per Bale Acres of picking 265 387 158 1.46 0.60 0.66 0.40 1.68 2.45 Machine hours Bales harvested Investment : Harvester Tractor Total Picking costs Overhead : Harvester Tractor Total Operating expenses : Harvester Tractor $3,712 1,779 $ 14.01 6.71 $ 9.59 4.60 $ 23.49 11.26 $ 5,491 1,410 363 $ 20.72 5.32 1.37 $ 14.19 3.64 0.94 $34.75 8.92 2.30 $ 1,773 782 126 $ 6.69 2.95 0.48 $ 4.58 2.02 0.33 $ 11.22 4.95 0.80 Total Labor : Operating* Service and repair Farm shop $ 908 466 100 20 6 $ 3.43 1.75 0.38 0.08 0.02 $ 2.35 1.20 0.26 0.05 0.02 $ 5.75 2.95 0.63 0.13 0.04 Compensation Total $ 592 $ 2.23 $ 1.53 $ 3.75 Total costs $3,273 $ 12.35 $ 8.46 $ 20.72 Labor used : Operating Other Total man-hours 387 111 1.46 0.42 1.00 0.29 2.45 0.70 498 1.88 1.29 3.15 * Includes bonuses. The following number of growers paid bonuses averaging indicated amounts by sub- areas: Kern, 7 growers, $208; Eastside, 2 growers, $188; San Joaquin-Tranquillity, 2 growers, $247; West- side, 1 grower, $460; Madera-Merced, 1 grower, $94. t Less than one-half cent. [57] Table 35. Usual Mechanical Harvester Investments and Overhead Costs in the San Joaquin Valley, 1949 Item Number of records , Acres picked*. Machine hours Bales picked . . Investment Harvester : Original cost : Less salvage value Total depreciation . . Average investment . Tractor: Original cost Less salvage value Total depreciation . . Average investment Annual overhead costs Harvester : Depreciation Interest on average invest- ment General property taxes . . . Insurance Total Tractor : Depreciation Interest on average invest- ment General property taxes . . . Insurance Repairsf Total Charged to harvesting cotton : Per cent of annual Amount San Joaquin Valley 63 284 407 229 $6,459 969 $ 5,490 3,714 2,950 442 $2,508 1,696 1,112 149 123 33 $1,417 371 68 36 15 100 $ 590 53.7 $317.00 Kern 15 263 414 258 $6,501 975 $ 5,526 3,738 2,817 422 $ 2,395 1,620 1,123 150 165 33 $ 1,471 350 65 45 14 100 $ 574 58.3 $335.00 Subareas East- side 16 309 427 245 $6,450 968 $ 5,482 3,709 3,008 451 $ 2,557 1,730 1,133 148 106 23 $ 1,410 393 69 32 11 100 $ 605 50.5 $306.00 San Joaquin- Tran- quillity 273 316 201 $ 6,485 973 $5,512 3,729 2,918 437 $2,481 1,678 1,101 149 121 50 $ 1,421 378 67 37 22 100 $ 604 39.5 $239.00 West- side 9 317 479 292 $ 6,386 958 $ 5,428 3,672 2,874 431 $ 2,443 1,653 1,093 147 72 37 $ 1,349 351 66 20 17 100 $ 554 56.3 $312.00 Madera- Merced 14 265 387 158 $ 6,455 969 $ 5,486 3,712 3,094 464 $ 2,630 1,779 1,097 149 132 32 $ 1,410 376 71 41 15 100 $ 603 60.2 $363.00 * One acre picked over one time; same acre is counted twice if picked over second time. t Included in overhead for convenience in allocating proportionate share to cotton harvesting. Table 36. Usual Mechanical Harvester Operating Expenses in the San Joaquin Valley, 1949 Item Number of records Acres picked once over*. . Machine hours Bales harvested Harvester expense : Preseason repair Seasonal repair Mount, dismount Spindle oil Grease Wetting agent Total Tractor expense :f Fuel Oil Oil filter Gear grease Total Labor expense : Operating Bonust Service and repair Farm shop Compensation insurance Total Total expenses San Joaquin Valley 63 284 407 229 505 196 79 58 6 25 $ 869 133 8 4 2 $ 147 460 46 84 8 7 $ 605 $ 1,621 Subareas Kern 15 263 414 258 $ 398 230 69 49 6 26 $ 778 143 7 4 2 $ 156 402 97 65 11 6 $ 581 $ 1,515 East- side 16 309 427 245 583 240 91 63 8 20 $ 1,005 140 9 4 2 $ 155 472 23 89 3 7 $ 594 $ 1,754 San Joaquin- Tran- quillity 273 316 201 503 145 119 92 5 15 $ 879 107 7 4 2 $ 120 388 55 61 1 6 $ 511 $ 1,510 West- side 317 479 292 $ 503 145 63 62 6 36 $ 815 159 8 4 2 $ 173 608 51 103 1 9 $ 772 $ 1,760 Madera- Merced 14 265 387 158 472 173 61 40 7 29 $ 782 113 7 4 2 $ 126 459 7 100 20 6 $ 592 $ 1,500 * One acre picked over one time; same acre is counted twice if picked over second time. f Includes only operating costs in harvesting cotton. Repair costs were included in overhead for convenience in prorating share charged to harvesting cotton. % The following number of growers paid bonuses averaging the indicated amounts, by subareas: Kern, 7 growers, $208; Eastside, 2 growers, $188; San Joaquin-Tranquillity, 2 growers, $247; Westside, 1 grower, $460; Merced-Madera, 1 grower, $94. [59] Table 37. Materials and Labor Used in Harvesting Cotton Mechanically, San Joaquin Valley, 1949 Item San Joaquin Valley Subareas Kern East- side San Joaquin- Tran- quillity West- side Madera- Merced Number of records 63 107 5 42 809 44 4 13 407 75 26 9 517 15 93 5 41 817 39 4 13 414 66 25 11 516 16 126 4 41 836 53 4 13 427 84 27 6 544 9 144 3 35 667 42 4 13 316 48 18 1 383 9 109 6 38 979 47 6 13 479 86 31 1 597 14 75 5 49 750 38 4 13 387 86 25 18 516 Harvester: Spindle oil, gallons Wetting agent, gallons Grease, pounds Tractor : Fuel, gallons Cylinder oil, quarts Oil niters Grease, pounds Labor : Operating, hours Service and repair, hours .... Mount and dismount, * hours Farm shop, hours Total labor, hours ^Average for cases reporting. [60] Table 38. Investment and Overhead Costs for Mechanical Harvester Number 2, San Joaquin Valley, 1950 Item Number of records . Acres picked*. Machine hours Bales picked . . Investment Harvester : Original cost Less salvage value Total depreciation . . Average investment . Tractor : Original cost Less salvage value Total depreciation . . Average investment . Annual overhead costs Harvester : Depreciation Interest on average investment General property taxes Insurance Total Tractor: Depreciation Interest on average investment . General property taxes Insurance Repairsf Total Charged to cotton harvesting : Per cent of annual Amount Average per grower One-row 12 100 216 88 $ 4,738 711 $4,027 2,724 1,505 226 $ 1,279 865 805 1,742 109 236 75 164 33 77 $ 1,022 182 35 (single -unit 24 machine, in- 10 cluded with 69 above) $ 320 30.9 $99.00 Two-row 6 245 347 207 $ 10,250 1,538 $8,712 5,894 $ 2,219 * One acre picked over one time; same acre is counted twice if picked over second time. f Included in overhead for convenience in allocating proportionate share to cotton harvesting. [61 Table 39. Usual Operating Expense for Mechanical Harvester Number 2, San Joaquin Valley, 1950 Item Number of records Acres picked* Machine hours Bales harvested Harvester expenses : Preseason repair Seasonal repair Mount and dismount. . . Graphite and carbon tet. Grease Wetting agent Total Tractor expense:f Fuel Oil Oil filter Gear grease Total Labor expense : Operating Bonus Service and repair Farm shop Compensation insurance Total Total expenses Average per grower One-row 12 100 216 88 284 153 45 8 3 88 $ 581 74 222 29 5 $ 354 $ 1,009 Two-row 245 347 207 $ 273 607 26 6 228 $1,140 62 199 5 4 3 3 4 $ 206 361 32 170 50 7 $ 620 $ 1,966 * One acre picked over one time; same acre is counted twice if picked over second time. f Maintenance and repair costs were estimated on an annual basis and included in overhead for convenience in prorating the share charged to harvesting cotton. [62] Table 40. Usual Materials and Labor Used in Harvesting Cotton with Mechanical Harvester Number 2, San Joaquin Valley, 1950 Item Average per machine One-row Two-row Number of records Harvester: Wetting agent, gallons Grease, pounds Tractor Fuel, gallons Cylinder oil, quarts Oil filters Grease, pounds Labor : Operating, hours Service and repair, hours . . . Mount and dismount, hours Farm shop, hours Total hours 12 14 36 20 47 359 1,104 32 27 3 4 27 216 347 96 162 29 29 48 370 557 [63] D w C o cr o Is IS g" * c.E Is c o ■3 »! O o J* £ .SO * >s ■c -a c «A u o III O giSb as ■g.' a 'CO o O « A! ffi p. .3 » «3 o "2 a> W'S. •a © •3 0) .2 2 W3. "^ T3 4) KB. J3M CO t-I CN 3 <* H 00 * 00 CO * *■< eot-co 00 CN incNCDCOOON WHOHia C-OW CO o CO 00CN • lot-" cn ooih 00 © iH CO OS tH CO CO in CO CO 1-H * rH lOO* CO CO'* *H*CO i-l CO CM iH W * CO 00 00 CO CO CO r-i **CDi-l CO iH in OS r-IWCN COCO co'io 05 CO Tf CO iH CO CO CO * MiOININOHt- CO CN 00 OS CN CN in irios OS CO t- CO CM t4 * CO CN CO t-I O g CN moot- "tf CN * CM t- iH OS CN r-l CN CO oo os w c- *h oo * w t- oooco t- CO io~coifl«* OS CO CD CO eo mcN t-cooo *CDCOCN HrtOHTf r-tOOCO tH "J t-CN 00 OS * CO i-t COCOCNCDCN m os m o m m * thooococn CN COCO ^ os'r^ rH 00CN OOOOCO O COOS iH CO t- 00 © CO CN * CN ©in COCO CN ^co'inoo'i-T OCDCD** CO*OSCN(N r-tooeo>-tcN cnco* t-m^-i' «J«CN OCOi-n CN rHCOt- 00 00 C- © OS OS 00 * tH CO OS CD OS 00 mooooeo CNOSO* r-f* WO) CO CO .-I i-HOOOS" t-mcNi ieot-*t- t- 00 O1H00 CN *H_ lOinuTtNeoeoVr moomrH OOOCOOOO CO CO ■* 00 CD CNinosoos eoeo»~r It-COiHCN co m eo CN CN t-CN* OS iHCOincD 00 iHOSOOCO CO t- OS* CO rH * * os *h * oo oo co t- m co oo in co *m*CNCD w OOrHf f CNOOOSt- cdcocDt-i cn m meo co i—l CO OS 00 OS cn com * ss CN OSCO CO iH &5 WHO x as as "§! •5 e9 OS > f8 c« as as bo (to as as > > « c -c • CO t- -^ OS N CC H NNOSOO IO • HH N C0N-* © C > © el j4 53 p, . © t- © H N C H ©COWNi- N . . OO O .©■<* i-t *a c > © M 00 H ... © s u M ■3 V (A OS T3 NHNOt-C N NCOC000 OS . . ..* i if > © Q S) S J3J4 a* . O CO t- 00 © Cs O ONWO 00 : :© © > © N«H OS © ■it c W H 0" N iTJCONCs Ifi Nt- 1 ^ CO • o o UTS OHN^IOHC CO .oooc OS H .OH© N d © k 0) OL ea 03 i £ P HOOMH^lf ^« lOOSOSC CO NO© 00 O © la o co m © t- n c ^* : d ih d c CO .OH© H C > © ^ N^H OS © t> s w ■si Nt-OlO^lOff CO iHCOCONlT ■<* •WHO © r) o ea A* 53 & OOOtOOCCHr- HOON ta ohhoc W .Ht-N © C H o © ^s H >■? la O C quin byS 03 ►^ el ol H .5 « • CO -<* t> © © If ^ ■ CO CO CO OS Ht-© t> © . O © CO © © c CO .OHO 00 H : d © •>* H 8 s* "el H o 0) ft OS CO 00 CO CO CM ^ CO HHIOW^ CO N00CO © C> © © ffi'tt O <# CO CO CO H c NOON O ONNOC OS tfi ! : © n © C c 8 H *5 M • E 3 H .5 © 00 H OS 00 C- i- Tj« N^CO OS NNt- <£ © c« AX .oococowo H^CO CO .ooo OS d : © h © t> C 8 o -o S* H t c o o U M - o "o © 00NOHNt»lf iO COt-00N^t <4< . m ■ ■ if Nt-^*H ^1 Cn © 53 o. OONOt-COHC CO CO tH CO OHOOC OS co : :d . : O ©OH© Cs c O O H s CO .so nd-P s by M co h co ■* N ^t< ir o iHlAcOtO^ H HH©N V * o OC0 00NC-NC ■* ONCOrHC 00* ONCOH t- c o S^ TH^rH 00 o 8 1 >> 4? c ** 4) Tjl \a CO OS 00 <# IT 1- iHint-^^t H H • i- .TjtlTJ© ic Cs © o 73 53 a NMCOHIOHC ^ OHHOC ^* :d : : © V c © 1 > NOON OS • ■ 1 o H o a c & u o .5 © HTfrtiOOtNNff c •OHHK U5 ■t-^H Cs M o s a eS 02 J3M o o ON00C0Ht} id OS o w o3 s 1 «■ O £ a o 3 s T3 el c8 § S o o c Eh s s a 1 S 1 S 3 c &s s "3 o "3 » o OS V 02 i- SOwftwjwO O O 02 ft CQ J 09 eiOo2ft02Hl gO02ftO2 pa c V ■ io co r^ ^ CO LO 00 ^ 00 t- t- t- CO -tf i-l 3 e : oo to co *tf CO 8" © • 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 8 w ■" — Q C 8 it • eo oo 00 CO CO CO 00 t> CO cT rH 00 CN CO CO o 5-*, • xtf ^ eo CO rH CN rH CD CD 00 t- 00 CO LO fN o wa. C . o o o o o O O CD CD CD CD CD CD CD CD CD -C n ""-"' s — ' *■ c •"■ ++ c ®— o • 00 CN H cd o 00 o CO lO IO ^. ^ CD CD CN tH *- .a* ; »d os l> 00 OS CO CO *tf LO KJIOO tH r- CO CN +■ u "8 . cd cd CD 03 as cd CD CD CD CD 00 00 00 0) o a u • io CO ^ ^ ^ rr o O CO 00 00 00 CO LO tH •■■ 2 : o CO t- c- t» CO CO CO LO 00 CO CO CN 00 CN CL V iH 1 1 1 1 | 1 1 | | 1 rH ■ to 1 "0 p c 8 Z eo (A (A £ O rH cd t> LO t~ CO CN o lO 00 ^ LO LO co tH 3 C K-a : T 5 . o o o o o O o CD CD CD CD OS 00 00 00 00 ■ *- U C n-I C ■S ® ■ CO CO CO H co CD CN © CN O CO l> CD oo o x .2 «-* . CO 8 CO CO LO 00 CO a O CD 00 rH CD CO CO . . cd CD cd CD CD CD CD CD 00 00 00 t- b- I- u V) 3" H 8 4- s-g, »- '.T. » O lO CN CO CO co o to CO CO 00 h 0) CD r- CN CN o 00 ■^ CO CD CO 3 1 1 i-i T-i rH 1 1 1 1 H 8 a *■ ^ C v O X o = *fc 8 •o © AM e3 O «-a fc • 00 t- : © o* iq d d CO d CO 00 d t> LO O CD LO oo id co oq CD rN CD d oq o> oo eo" . o o 0) o o CD CD CD CD 00 00 00 oo 00 C* i> oo 5> "8 rH rH rH rH C 8 .a J* o 5 <♦-« w O ci • *■ o -Q 8 W T) X> pd ,d J3 A d 4 •B'B'B £ 56 -t-3 +3 -^» +* -u o tH (N CO T* LO CO t- 00 »H CN CO TT IO Ci) t- 00 CD H rH rH tH tH r-l TH rH rH TtiNONCOHOHt-OOW^MlOMN dodi>6o6oc5coo6do60^oq05WO>oqNrl m to « h n d h d «i i> w od w d to d d 00)0000000)00000000000000 SI OP II ONCOHNOWt-O^rl (OiO^NNMT)iOO(fiiO«OHtM« ooooooooooooot-ooc*-t»t"" CO CO O O) CO t»(0N05WHt0NOI>C00qMIO 00 Tjl H H t> CO rl M t-' t^ tO ^ H CO ' I I I I I I I I I I I IOIOOI* CO ^ioTiitoqq»oto ' od io n " oo oo OtNOCOOOtO^COtNr-ItO ooooooooooooooo WHiHIONrtOOW^flJNNNrjCq codrlodNcoNdioco'oJd'Jd^ oooooooooooot-t-tot>i>c-t- OOtOOfCOCOCO^t-lOOOH • t>- cd-^tdt>cot>o6oed : d iH oq i> W^c4cOWNo6wif5l>CNlO^COCOCO OOOO00O000000I>t*l>C > -t-b-t» w'd'OB+S+S^+S+SOHNCO^iOtOt-OO HMW^lOIOl-COOJHrlHrlHHHHH 9 ■ b- -* Oi ^ -^ CM tH (N m o CD Oi i-H c- CD CO CO a ■ cb* io id CO t- Oi b- t> t> CD 00 00 t> m m CO CM ■c III 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0) W 3 Q C 4* M c s c o ' T3 CO lO O CD CD c- b- b- co i-i o tH o 00 CN o CO 00 u T S-3 00 CN. CM CM CN T-i 1-1 §8 r^ CD -* CO o CD m m m ** in cm x.* W CO CO m ^ i-H CO O 00 00 *tf H Oi Oi Oi Oi b- S-| . oi Oi a Oi Oi Oi Oi Oi 00 00 00 00 t> b- t- b- b- "8 *r u e Q. a m ^ ^ O CO Oi to CO t> m tH Oi 00 • Oi i ' w CD on co co CN m c- H H i-H CN m 00 CO (A V. 0) > i wa : tj« co m X JJ< . t- CD co >* "^ CO o Q CO rs] H o 00 00 00 u . Oi O^ Oi Oi Oi Oi Oi Oi 00 00 00 oo b- t- t- s s u © XI »^ u CO o O ^ Oi O CO Oi CO CD m CO CN co CO 00 9 S? ^a — 8 In O • II 00 1 1 CO CO 1 1 ^1 1 CD 1 co 1 1 1 1 co 1 CO 1 »j3 efl CJ«5 a -a a 3 (N CO CO CO CN i-H i-H o b« CD m m CO CO CN y-i CD N M 0) o, a w t/2