Two Cylinder Magazine

Editor’s Note: When the original planning to do an article on cotton harvesting machinery began, it was intended that both types of cotton harvesting machines — cotton pickers and cotton strippers — would be covered. As the article was being researched, it quickly became apparent that it was simply not possible to do justice to both types of cotton harvesters in the space available. Accordingly, the decision was made to concentrate on cotton pickers for the first article, and then cover cotton strippers in a future issue, as the two methods of cotton harvesting are completely different. Much of the history of the transition from harvesting by hand to the harvesting of cotton by machine was covered in the November-December 2006 issue of Two-Cylinder.

In the field with a home-built finger-type stripper
harvester during the harvest of 1926.

Chances are that virtually everyone reading this article has at least one high-quality denim garment in their closet. Although denim cloth can be woven from any type of cotton, some of the highest-quality denim is woven from cotton grown in the High Plains area of west Texas, Oklahoma, and parts of Kansas. American Cotton Growers (ACG), based in Littlefield, Texas, weaves as much as 38 million linear yards of high-quality denim cloth each year, and has been a major supplier to the industry for the past

31 years. Their list of customers is a virtual “Who’s Who” of the garment industry. Familiar brand names such as Lee®, Levi’s®, Wrangler®, Eddie Bauer®, and even such high-fashion designers as Calvin Klein and Tommy Hilfiger, are among the list of the many customers who purchase denim cloth from ACG. So, it’s very likely that cotton woven into denim material from the areas of the High Plains — where cotton is grown and is harvested by the stripper method — is in practically everyone’s closet. Chances approach certainty that the vast majority of the cotton used in those garments was harvested by a John Deere Cotton Stripper. The story of the transition from hand-harvesting of cotton in the High Plains to the mechanical cotton stripper method of harvest follows…

A more detailed view of a home-built
finger-type stripper harvester. 1927.

In the High Plains area of the cotton belt, where the climate was drier and yields were typically much lower, at least partial mechanization of the cotton harvest was somewhat accelerated as compared to the cotton growing regions east of the Mississippi River, primarily due to the lack of available labor. In the period just prior to World War I, the practice of “snapping” cotton (picking the whole boll from the plant instead of removing the cotton lint from the boll by hand) slowly began to become more popular because it was much faster, and that made it possible for growers to get their crop to the gin before bad weather set in. Gin owners, whose job it was to separate the seed and the trash from the lint, weren’t happy about this new practice; at first refusing to accept cotton that had been snapped rather than picked by hand, as the snapping method involved separating considerably more trash from the lint. However, since the snapping method of cotton harvest was about twice as fast as picking, and given the shortage of labor, the gin owners finally gave in and equipped their gins to handle and clean snapped cotton.

In the 1920s, the continuing shortage of affordable labor demanded an even faster method of harvest, so the “sledding” method was developed. This wasn’t a new idea, as the basic principles had been invented and patented (but never commercially developed) a half-century earlier. The first “sled” machines consisted simply of a wooden box mounted on runners, with a V-shaped slot in the front. A team of mules pulled the sled down a row of cotton, and the slot scraped the bolls off the stalk. The driver raked the bolls back into the box. As the “sleds” evolved, the V-slot was replaced by rows of fingers or tines, and the runners were sometimes replaced by wheels, but the basic principle remained unchanged. The cost of a sled was low — commercially available units could be purchased for $40 or less — and harvest productivity soared with the introduction of the sled-type cotton harvester. By 1926, a man with a sled-type harvester and two mules or horses could harvest a bale of cotton in 8-1/2 hours, whereas hand-snapped cotton required 45 man-hours per bale, and hand-picked cotton consumed 70 man-hours per bale. Even though ginning costs were higher with “sledded” cotton, the huge reduction in labor costs meant that cost savings of $12.00 or more per bale ($142 today) made the sledding method of harvest preferable whenever possible.

There was only one major drawback to the sled method of harvest; the cotton had to be completely defoliated by frost for the sled harvester to work well, and to gather the most cotton and the least amount of trash. Frost, of course, cost nothing, but usually delayed harvest until after the beginning of November, which unfortunately increased the chances of crop loss due to bad weather.

These conditions were known to the members of the Cotton Picker Committee of Deere & Company, which had been formed in 1926 to investigate the feasibility of mechanically harvesting cotton. From the beginning, the Committee had its eye on cotton strippers as a viable method of harvest. The cotton stripper, which plucked the entire boll of cotton from the stalk instead of attempting to extract the lint from the boll, was a much simpler machine than a cotton picker, so a stripper-type harvester was Deere’s first foray into the cotton harvesting arena. Fred Thomann, an engineer at Deere’s Spreader Works, was temporarily reassigned to the Wagon Works as the design engineer for the cotton stripper project. Since the primary responsibility for the cotton harvester projects had been assigned to the Spreader Works, the reasons for the Wagon Works to be involved in the cotton harvester project are unclear; possibly it was because the manager of the Wagon Works, Virgil Bozeman, was a member of the Cotton Picker Committee, and perhaps because one proposal for a stripper-type harvester involved mounting the machine on a wagon.

Experimental one-row horse-drawn cotton stripper, which appears to have had some type of cleaning device attached. 1928.

A letter from Virgil Bozeman to C.D. Wiman, dated March 24, 1927, described the plan of the Wagon Works: “to get three or four experimental cotton strippers into a rented field of wintered-over cotton near Altus, Oklahoma.” Both steel-roll and finger-type machines would be tested. If successful, the plan was that the experimental machines would be followed by a pilot run of pre-production machines in time for the 1927 harvest, with full production scheduled for the 1928 harvest season.

At that time, the principal thrust for development of a cotton stripper was directed towards a one-row machine drawn by a team of horses or mules. The cotton plant would pass between two steel rollers, studded with short pins. After the cotton bolls were stripped from the stalk, they were elevated to the collection box via flighted chain (link chain equipped with rubber paddles).

No record is available of the tests conducted in Oklahoma in the spring of 1927. An experimental stripper was tested near San Antonio, Texas, from August 31 through September 6, 1927. A delegation of Deere & Company officials observed a portion of these tests, and all concerned were of the opinion that the machine was on the right track, although improvements could be made. Harvest efficiency was in the 95-percent range, which was considered satisfactory. The principal difficulty in undefoliated cotton was that the studded rolls removed too much trash from the plant, and that the trash plugged up the elevators. Virgil Bozeman and others who observed the machine suggested revisions to the steel rolls and pins, along with the possibility of steel or fiber brushes on the rolls to reduce the amount of trash gathered. The machine was returned to the Wagon Works for rebuilding, and was sent to western Texas later that fall for further testing. C.D. Wiman requested Leonard Neighbour of the Spreader Works — who was being sent on a tour of the cotton belt to observe any and all attempts to harvest cotton mechanically — to witness the operation of the revised Deere cotton stripper. Neighbour reported that the machine worked fairly well in knee-high defoliated cotton, but no better than a wooden sled-type harvester. Since those machines could be purchased for $40 or less, a Deere machine with a projected selling price estimated to be in the range of $135 to $150 had to do better than a sled if it was going to be commercially viable.

Still, the machine had shown considerable promise, and Bozeman reported the results of the tests to the Cotton Picker Committee on December 2, 1927. Shortly thereafter, the Committee passed the following resolution: “RESOLVED; That the John Deere Wagon Works be authorized to build not over 100 Strippers for sale by December 1, 1928.” This resolution gave the Wagon Works some “wiggle room”, as the timing would allow for another year of experimentation before a production stripper would hopefully be put on the market in time for the 1928 cotton harvest.

Although the primary market and application for a cotton stripper was in the semi-arid High Plains regions of Oklahoma, New Mexico, and Texas, the concept was also tried in the Mississippi Delta with disappointing results. In the tall and heavier cotton of the Delta, the stripper tended to pull up the cotton stalks by the roots, plugging the machine; and without defoliation, the trash problem was even more pronounced than on the High Plains.

From the fall of 1927 until the fall of 1930, there is little documentation available on the cotton stripper project, but some photographic evidence exists. It is known that in 1928, Fred Thomann, who had been the lead engineer on the cotton stripper project, was transferred back to the Spreader Works to take charge of the design of a Deere cotton picker. Experimental work on the cotton stripper continued, but appears to have taken a back seat for awhile at least, as the majority of engineering time, talent, and funds would be directed towards Deere’s cotton picker project during 1928.

However, despite the emphasis on the cotton picker project, sufficient funds and engineering talent had been allocated to the Wagon Works in order to pursue further development of the cotton stripper. Experiments with the one-row horse-drawn stripper continued, and a new project was undertaken: a two-row tractor-mounted cotton stripper. Projects of this magnitude weren’t done overnight, and it was 1929 before the first machines were ready for field tests. Although the cotton stripper was primarily intended for use in the High Plains of Texas and Oklahoma, the difficulties being encountered by the design team working on the spindle-type cotton picker resulted in the cotton stripper being tested in Arkansas cotton in November 1929, where it, too, was unsuccessful in that application.

Although no written evidence is known to exist concerning the cotton stripper testing for 1929, it must have been reasonably successful, as revised tractor-mounted machines were built for the 1930 harvest season, and a production run of 50 of the No. 30 one-row horse-drawn machines were built. A letter from Leonard Neighbour to C.D. Wiman, dated December 19, 1930 summarizes the results of testing in Texas during the 1930 harvest season:

Mr. C.D. Wiman,
Deere & Company,
Moline, Illinois

Dear Sir:

Subject: Tractor Cotton Strippers

A party consisting of Mr. C.D. Wiman, Mr. C.N. Stone, Mr. B.L. Harris, Mr. V.F. Bozeman, Mr. Frank Court, Mr. Burke Healy, Mr. Pence, Mr. Essenpreis, and the writer, visited four of the tractor strippers which were in operation at the following points:

One No.10 two-row stripper on the Dixon farm near Canadian, Texas
One No. 5 one-row stripper on the Otis Wells farm at Sentinel, Oklahoma
One No. 10 two-row stripper near Memphis, Texas
One No. 10 two-row stripper near Plainview, Texas.

The Wagon Works production schedule for cotton strippers in 1930 was as follows: 5—No. 10 two-row—weight less tractor 2400 lbs., price to farmer $1450; 5—No. 5 one-row—weight less tractor 1750 lbs., price to farmer $1050; 50—No. 30 one-row horse-drawn, weight 100 lbs., price to farmer $165.

None of the tractor-operated strippers had been sold on straight sale, though it is evident that some of them can be made to stick with the farmers. This is an unusual year in western Texas, as it has been in the Mississippi Delta, due to the effect of the season as well as the economic situation on harvesting. Labor has been plentiful and cheap. Hand-picking has been as low as 75¢ per 100 pounds, as compared to a high of $3 not so long ago. Snapping costs are as low as 50¢ per 100 pounds. Cotton opened earlier than usual this year and a greater than ordinary amount of hand-picking was done. With cheap labor available, the bulk of the crop was snapped directly following the killing frost.

The inability to sell the full output of tractor-mounted strippers has been due largely to the difficulty of finding farmers with “GP” Wide-Tread Tractors, cotton to strip, and purchasing ability.

Sledding of the entire crop without intermediate hand-picking is not yet a completely accepted method of harvesting, except in years of an extremely large crop and a shortage of labor. When possible, the cotton is often picked twice before the frost on account of the fear of loss through shedding due to the high winds. It will be necessary to demonstrate in the next few years that it is economical to let the crop stay in the field until after the killing frost and then make a once-over operation. Before this method is generally accepted, it may be necessary to plant a non-shedding variety of cotton to accomplish this.

The combination of stripper and cleaner is apparently justified by the savings in ginning cost alone. In this year’s operation of a competitive machine, which stripped only and did not clean, the ginning cost ate up all the returns from the cotton seed, and this is true to almost as great an extent with hand-snapping, where it takes from 2200 to 2800 pounds of seed cotton and trash to make a 500-pound bale.

The strippers visited were doing an excellent job of cleaning, as compared with hand-snapping, and appeared to be entirely satisfactory from this standpoint. They were saving practically all of the cotton where the plants were more than four or five inches high.

It is evident that the mechanism for stripping, conveying, and cleaning the cotton is reasonably adequate and satisfactory. If cotton is to be recovered from the ground, some change must be made in the gathering snouts so that the front end of the machine can be more flexible so as to follow the ground closely. It is quite possible that sufficient change can be made to the present machine to make it useable in this respect. More adequate slip clutches and an improved elevator should be provided.

In my opinion these machines, although giving a creditable performance, are more or less in the experimental stage, and much could be learned in operating them over a full season. One or two new machines might be built next year to embody some of the more obvious changes which appear necessary from this year’s experience.

To me, the price of $1450.00 for this 2400-pound machine seems high enough to prevent the rapid immediate development of this business, although in some cases the stripper might be able to earn somewhere near this amount in a year. Mr. Court considers this machine to have a yearly capacity in the neighborhood of 300 acres, which he arrives at by operating 10 acres per day of the thirty days of the sixty-two in December and January. This seems conservative enough.

I have written a separate memorandum to Mr. Bozeman covering suggestions made to him and Mr. Court from time to time as we were operating in the field. I will be pleased to give you a copy of this if you so desire.

Yours truly,

JOHN DEERE SPREADER WORKS
L.B. Neighbour (signed)
Experimental Department

....continues