Title Page
 Letter of transmittal
 Boll weevil control
 Fertilizer experiments
 The spacing test
 Topping experiment
 Blossom records
 Soil moisture studies
 Summary and general recommendations...
 Boll weevil control on Sea Island...

Group Title: New series
Title: Sea Island cotton in Florida
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00002934/00001
 Material Information
Title: Sea Island cotton in Florida report of Works Progress Administration
Series Title: New series
Physical Description: 72 p. : ill. ; 23 cm.
Language: English
Creator: Smith, George D ( George Durward ), b. 1886
United States -- Works Progress Administration
Publisher: Dept. of Agriculture
Place of Publication: Tallahassee Fla
Publication Date: 1937
Subject: Sea Island cotton -- Florida   ( lcsh )
Cotton growing -- Florida   ( lcsh )
Cotton -- Diseases and pests -- Florida   ( lcsh )
Genre: non-fiction   ( marcgt )
Statement of Responsibility: George D. Smith.
General Note: "April, 1937"--Cover.
 Record Information
Bibliographic ID: UF00002934
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltqf - AAA3247
ltuf - AKD9628
oclc - 27184200
alephbibnum - 001962951
 Related Items
Other version: Alternate version (PALMM)
PALMM Version

Table of Contents
    Title Page
        Page 1
        Page 2
        Page 3
        Page 4
    Letter of transmittal
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
    Boll weevil control
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
    Fertilizer experiments
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
    The spacing test
        Page 40
        Page 41
        Page 42
        Page 43
    Topping experiment
        Page 44
        Page 45
    Blossom records
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
    Soil moisture studies
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
    Summary and general recommendations for 1937
        Page 66
        Page 67
        Page 68
        Page 69
    Boll weevil control on Sea Island cotton
        Page 70
        Page 71
        Page 72
Full Text
Xo. 82 Xew Series
Sea Island Go
Report of Works Progress Administration

Sea Island Cotton
Report of Works Progress Administration

This bulletin is based on data secured by the Works Progress Administration and describes in detail the results of the various research experiments conducted on Sea Island Cotton in Madison County. The research work by the Works Progress Administration has been in charge of Geo. D. Smith. Entomologist, and the data presented herewith is largely a summary of the activity and research experiments of the Works Progress Administration. However, since very little scientific information has been published in recent years on the Sea Island cotton problem, the vast amount of valuable information in this report will be passed to Sea Island Cotton growers through the medium of the State Department of Agriculture.
Particular attention is called to the parts of this bulletin dealing with the "Afternoon Method of Boll Weevil Control" as this is the first time a 100 percent perfect weevil control method has been announced. Attention is also called to the results of the fertilizer test showing that soil moisture is the limiting factor, together with spacing, in Sea Island Cotton production. The correlation of soil moisture measurements with the fertilizer experiments represents a great step forward in solving some of Florida's greatest agricultural problems.
All Florida Sea Island Cotton growers are urged to read this bulletin carefully and put into practice the methods of weevil control described herein. It is generally admitted that, unless the initial weevil infestation can be eliminated, it is impossible to produce Sea Island Cotton profitably.
I took up the correspondence with the Federal Department of Agriculture in the spring of 1SK33, insisting on that Department furnishing Pure Sea Island Cotton seed to be planted in Florida. As a result of that correspondence a man was sent from Washington, D. C. to Gilchrist County to con-

fcr with Mr. W. F. Love to investigate the project. He made v. favorable report and fifty bushels of seed were furnished Mr. Love and twelve bushels were furnished the Experiment Station at Gainesvilleall that the Government had.
Since I secured the first few bushels of Sea Island seed ill l!).*i4. acreage to this important crop has jumped from about 3.000 acres in 1!)3<> to approximately 20 thousand acres in 1A3T. Florida's lost Sea Island Industry is coming back and all growers are urged to cooperate in getting the rehabilitation of the industry started off according to the scientific idea of eliminating the over-wintered weevils or else do not make an effort to grow this fine cotton.
All the data, findings and recommendations contained herein concerning Sea Island Cotton apply to Short Staple Cotton.
NATHAN MAYO, Commissioner of Agriculture.

Mr. Frank P. Ingram.
Director Florida Works Progress Administration. Jacksonville, Pla.
Sir: I am pleased to submit herewith a preliminary report on the activity of the Works Progress Administration in the revival of the Sea Island Cotton Industry for the years !!).'!." and 1936. This report has been prepared by the Entomologist in charge.
C. D. Toinlinson. District Director.
Mr. C. D. Toinlinson,
District Director, Works Progress Administration, Madison. Pla.
Sir: I have the honor to transmit herewith my preliminary report on the part played by the Works Progress Administration in the revival of the Sea Island Cotton Industry in Florida. This report covers the activity of the Works Progress Administration for the years 1 !>:$.> and 1936.
Geo. D. Smith. Entomologist in Charge.

The rehabilitation of Sea Island cotton was attempted in Madison county by the Works Progress Administration as a County Project in liKS.-) with such successful results that the Project has been made State-wide.
At the beginning of the Sea Island Cotton Rehabilitation investigation, it was realized that for the successful revival of the industry an improved method of boll weevil control would be necessary as Sea Island cotton bolls are soft and more susceptible to weevil injury than upland cotton bolls. In fact, weevil injury to Sea Island cotton is so severe it was realized from the beginning of the experiments that any method of weevil control evolved would have to be so efficient as to completely eradicate the initial weevil infestation. That the weevil problem constituted the chief menace to growing Sea Island cotton is emphasized in a circular released by the 0. S. Department of Agriculture which is quoted as follows:
"For the information of cotton growers in Florida, they are advised that many experimental plantings of Sea Island cotton have been made by the U. S. Department of Agriculture in this State during the past several years but the yields have been too small to justify any encouragement of general planting by farmers. The Sea Island is late cotton with soft bolls and the crop is especially susceptible to severe damage from weevils during the entire period of development. Unless a practical method of boll weevil control can be worked out for this special type of cotton, there is little hope of successful production under present conditions. Until more information is available, therefore, the general planting of Sea Island cotton cannot be recommended."
The situation described above clearly illustrates the problem faced by the VVPA Entomologists. It was a case of either

FIG. 1. The Works Progress Administration loading a car with Sea Island cotton at Lee, Fla. This is the first car lot of Sea Island cotton to be shipped from Lee, Fla., since the industry was destroyed by the boll weevil.
evolving an improved method of weevil control or abandon the Project. Fortunately, all previous research work on boll weevil control had ignored the thirst of the weevil when approaching the problem from a control standpoint. After a few preliminary experiments WPA Entomologists discovered that by applying poisoned-syrup mixtures in the afternoons while the cotton plants were dry and the weevils thirsty, practically all adult weevils in the fields were killed within a few hours. With this knowledge 30 farms, ranging in size from 1 to 10 acres, were used in the 1935 weevil control demonstrations and all weevils that had lived through the winter were killed by applying the poisoncd-syrup mixtures in the afternoons only. With the preliminary work of 1935 completed and a new and superior method of weevil control perfected, a total of 70 farms in Madison county were used in the 1936 weevil control demonstration. The 1935 and 1936 demonstrations in weevil control were 100 per cent successful and the 100 farms comprising about 450 acres of cotton represents the first successful experi-

merits or demonstrations in eradication of the initial weevil infestation ever accomplished. In the past, weevil control measures, such as advocated by the U. S. Department of Agriculture, had as their aim the holding down of the weevil numbers until a profitable crop of cotton could be matured.
As stated in the paragraph quoted from the U. S. Department of Agriculture circular, there was much speculation concerning probable yields of Sea Island cotton under boll weevil conditions. On account of not being able to eradicate the initial weevil infestation, yields secured hy the U. S. Department of Agriculture had been unusually low and rather discouraged the idea of attempting the revival of the Sea Island cotton industry. By using the Boll Weevil Control Method developed by the WPA Entomologists, however, yields in both 1935 and 1936, where stands were adequate, were higher than those recorded during the period prior to the coming of the weevil. In pre.-we.evit: days the average yield of Sea Island cotton iu Florida was one bale to four acres compared with yields ranging from 600 to 1400 pounds of seed cotton per acre on the WPA demonstrations. Yields on the better soils averaged slightly better than one-half bale per acre and on one demonstration 13 bales of cotton were harvested from YZ acres. In. 1.934 only 15 hales of Sea Island cotton were produced in Florida; 160 bales in 1935 and SI9 in 1936. Prospects for 1937 point to an increase to probably 5,000 bales. Incidentally, as Sea Island cotton is rehabilitated land values will rise proportionately since this type of cotton can only he grown in a few counties in North Florida and Southeastern Georgia.
At the beginning of the WPA Sea Island Cotton Rehabilitation P rogram there was considerable speculation as to whether the trade would be interested in using Sea Island cotton. Some cotton students contended that unless the cotton mills were assured of a large, permanent supply of this type of cotton the farmers would have difficulty in disposing of the small amounts produced, The fear of no market kept many farmers from planting even small demonstrations in 1935 and 1936. However, the trade purchased the 160 bales produced in 1935

FIG. 2. Map of the Sea Island cotton area of the United States showing distribution by counties. Each dot represents an average production of 500 bales. (From Orton, Bur. Plant Industry, U. S. D. A.)
and paid an average price of *>."> cents per pound for a very low-grade crop. The price received for the 1936 crop ranged from 21 to 32 cents per pound. The 1936 crop was unusually high quality with few hales grading lower than a No. 2. Concerning the outlook and price for Sea Island cotton in England we quote as follows from the November issue of the Manchester Masters Spinners Cotton Bulletin:
"The stock of Sea Island cotton returned by the Liverpool Cotton Association in August 1935, was 3,210 bales but as this included all West Indian cotton actually available, it was a

very poor guide to the quality of Sea Island cotton actually available, although it was generally known that stocks were considerably lower than those ruling for some years past.
"With improved demand, prices gradually improved, and available stocks were quickly reduced at prices commencing at 16d. per pound, which rose gradually to as high as 23cl. The latter price was obtainable owing to a scarcity of supplies at that particular moment, and buyers felt that such a price was rather above the parity of the market. At the same time the trade have been persistent buyers at from lSd. to 20d.
"Another important factor is that manufacturers have been prepared to market their goods at a reasonable price, which is largely competitive, and has therefore widened the outlet for these particular goods. This cheap marketing has to a certain extent limited the price which spinners can pay for the raw material, but at the moment lSd. to 20d., according to quality, should be obtainable, which price should be satisfactory to producers. Actually when taken into comparison with ruling prices of Egyptian Sakel cotton, present prices for Sea Island cotton look satisfactory.
"It is evident that Sea Island cotton is meeting a wider market, largely for English spinners, and a factor which has contributed to some extent to this improved demand has been the advertising campaign which has taken place for Sea Island goods. One or two large distributors have made a specialty of goods manufactured from Sea Island cotton. In addition a recognized trade mark has been established, the use of which has given reputable firms the opportunity of specially marking their goods, and will have the tendency of cutting out competition from inferior articles.
"The prospects for the present season appear hopeful. Early arrivals have been sold at satisfactory prices, and the stocks are low. Provided producers are satisfied with prices at present available, there should not be much difficulty in disposing of consignments soon after arrival, and it would appear

FIG. 3. Staples of the world's fine cottons showing comparative length of staples.
that if the cotton is not held for excessive prices, a regular market is definitely established for Sea Island cotton.
"Owing to the efforts of the Agricultural Department of the West Indies, the class of seed used in the various islands has been more regular, and has resulted in improved shipments coming forward.
"There has been a definite improvement in the quality of the cotton coming from Antigua, Anguilla and Nivis. St. Kitts cotton has varied during the last season or two, but we are glad to again report that an improvement is now noticeable in their shipments."
Soil Moisture Studies
One of the outstanding features of the research work conducted in Madison county during 1936 was a soil moisture study relative to the shedding of Sea Island cotton where different amounts of fertilizer were used. Mr. Paul W. Calhoun, Assistant Entomologist, originated this special study and the results secured are expected to be of far-reaching importance to future scientific investigations in the Cotton Belt.
The soil moisture studies proved conclusively, in so far as one year's work is concerned, that the amounts of fertilizer

that can be used under Sea Island cotton are in proportion to the moisture holding capacity of the soil type. On the light Norfolk sandy soils such as occur in the vicinity of Lee. Fla.. where the soil moisture content normally runs from ."> to (> per cent, light applications of fertilizer gave much better yields than heavy applications. This result was due to the fact that where heavy plant growth is forced by fertilizer on the light sandy soils and severe drought follows, there is insufficient moisture to prevent the plant from shedding both leaves and fruit. Where light applications of fertilizer are used and plant growth is not forced, drought, conditions do not affect the plant so adversely and better yields are secured. < )nly on soils rich in humus, and where the normal moisture content is as high as 10 or 1*2 per cent, can heavy applications of fertilizer be used under Sea Island cotton with reasonable expectation of profitable returns.
The soil moisture studies, by showing that profitable returns cannot be secured from heavy applications of fertilizer on soils of low water holding capacity should mean a yearly saving of many thousands of dollars to Sea Island cotton growers. It is the opinion of the writer that further study will show this to he equally true of many other farm crops in the Southeast where sandy types of soil prevail. The result of the one year's Study, made during a relatively dry season, proved that scientific work on fertilizers cannot be considered complete unless interpreted in the light of the availability or non-avaiiability of an ample supply of soil moisture.

Scope of the WPA Work on Sea Island Cotton
The major research projects of the Works Progress Administration pertaining to the rehabilitation of Sea island cotton in Florida are listed as follows:
1. Boll Weevil Control.
2. Fertilizer Experiments.
3. Spacing Tests.
I. Reestablishing the Sea Island Market.
3. Keeping the Sea Island Seed from becoming adulterated.
6. Cooperation with the U. S. Department of Agriculture and the Florida Experiment Station.
7. Giving advice to farmers on methods of culture and .harvesting.
8. Study of the effect of early defoliation of the cotton plants by the cotton leaf worm as it affects the numbers of weevils that enter winter quarters in the fall.
!). General recommendations for use in growing Sea Island cotton under Florida conditions.
](. Soil Moisture Studies.
CHAPTER I. Boll Weevil Control
In November. 1935, the "Afternoon Weevil Poisoning Method" was announced through the medium of the press. The new method of Weevil Control consists in mopping the cotton plants with poisoned-syrup mixture, beginning the first application when the plants reach the "pre-square" stage of develop-

merit. Four to five applications of the poisoned-syrup mixture are made depending on the date of the first application. Where cotton is planted near swamps the last mopping should he given the field about June IS and about June 12 where the cotton is planted on high, pine forest land away from swamps. The poison is applied at weekly intervals and. if three hours without rain occur after the poison is applied, maximum results are secured. Thus the New Method of Weevil Control can be successfully used in wet summer seasons.
Approaching the weevil control or eradication problem from the standpoint of the weevil's thirst is a new discovery and results secured in Madison county in 10:5") and 1936 prove that it is entirely possible to completely eradicate the initial weevil infestation by mopping the cotton plants with poisoned-syrup mixture in the afternoons while the plants are dry and the weevil is thirsty. This method of weevil eradication was used on all Sea Island cotton fields in Madison county in 1035 and 1936 and on July 1the date by which all overwintered weevils have left their hibernating quarters and entered the cotton fieldsno trace of weevil infestation could be found on any of the 30 farms in 1035 or the TO farms in 1936. The annual summer migration of the weevil, which usually occurs in Florida during the last half of July, is of no special importance as. where Sea Island cotton is planted on or before March 20, the crop is matured into the hard boll stage and practically no damage is experienced from migratory weevils. Careful check on the progress of the infestation by migratory weevils did not indicate late summer poisoning would pay in any of the early-planted fields.
The "Afternoon Method" of weevil poisoning is simple and inexpensive. The poison-syrup mixture is applied by pushing a mop through the tops of the cotton plants and not by the "churn-dasher'" method of trying to get the poison into the buds of the plants. A small mop made by tying a piece of burlap on the end of a stick about three feet long is all that is necessary to apply the poison. Children can mop as effective as adults and one man can mop at the rate of about six to eight acres of cotton a day.

lor about 15c per gallon and calcium arsenate sells for about &x/t cents per pound wholesale. On this basis the cost of one application of the poisoned-syrup mixture approximates 25 cents per acre and labor for applying the poison should not exceed about 10 cents per acre. Four to six applications of the poison are usually requireddepending on the date of the first application. Where cotton is planted near upland woods with no Spanish moss nearby, the last application of poison should be made about June 13 and about June IS near lakes and other swampy locations where moss is abundant.
One of the features of the "Afternoon Method >i Weevil Poisoning" is that only about three hours without rain arc required after the poison is applied to kill practically all weevils in the field. In other words, during a wet summer in Florida it is entirely practical to kill all weevils that have survived the winter without permitting any of them opportunity to deposit eggs for the first weevil generation. No other method of weevil control has ever offered such advantages.
Where adult weevils arc scarce or their numbers reduced to a minimum by early fall destruction of the cotton plants, it is reasonable to presume that only three moppings at weekly intervals would be sufficient to eradicate the few weevils that survived the winter.
In all experiments it was found that there was no occasion to apply the poisoned-syrup mixture other than at weekly intervals.
The following photographs are illustrative of the WPA Afternoon Method of Moll Weevil Control, as it was practiced in Madison county in 1035 and 1030.
Reestablishing The Sea Island Cotton Market
At the beginning of the WPA Sea Island Cotton Rehabilitation Program it was currently believed that textile mills would not buy Sea Island cotton unless they were assured of a rather large quantity and a jermanent supply. Cotton market-

FIG. 4. The Mopping Operation, Showing the Correct and the Incorrect Procedure.
Left, the old "churn-dasher" style of mopping. The laborer, after dipping the mop in the poisoned-syrup mixture in the container carrier in the left hand, touches the mop to the bud of each plant, from above. This is slow, tedious work. Besides, the poison is left where it is most easily washed off by rain, and most quicky dried by the sun, after which it is not very effective against the boll weevil.
Right, an improved procedure for mopping cotton. The wetted mop is simply pushed through the tops of the plants while the laborer walks full speed ahead. The tops of the plants arc pushed over, and the poisoned-syrup mixture is applied to the under side of the leaves, and to the limbs and main stem of the plants. It is much better protected from rain and sun in these locations, and therefore remains potent for a longer period of lime. In addition, only about half the labor is required for applying the poison in this manner as in the old "churn-dasher" style of mopping.
The poison (calcium arsenate) is used at the rate of from one to one and one-half pounds mixed into two quarts of water to which one gallon of syrup is added after the water and arsenate have been thoroughly stirred. The mixture is stirred briskly as the syrup is poured in and is then ready for iTsc. A gallon and one-half of the mixture is sufficient to mop one acre of cotton one time.
The cost of the poisoning operation is relatively cheap. Cane syrup, such as a number 2 or 3 grade, can be purchased

FIG. 5. Laborers Poisoning Sea Island Cotton by the WPA Afternoon Method near a mossy hammock on the farm of F. C. Glass. Madison County, Florida, under the direction of a WPA Field Assistant.
ing students were somewhat amazed, however, when the WPA succeeded in selling the 160 bales produced in 1935 for an average slightly in excess of 25c per pound. Where one buyer was induced to purchase the 1935 crop a half dozen or more buyers were interested in getting some of the 193(5 crop. The various thread mills, tire manufactories and fine yarn mills in Europe were interested in securing some of the 1936 crop. In fact, the demand is now growing faster than production and it is believed that a Sea Island cotton crop of 100,(100 bales could be sold at fancy prices.
The price of Sea Island cotton is related to upland cotton prices only in the manner upland prices affect Egyptian cotton prices. At present Egyptian Sakels represent the chief competitor of Sea Island cotton along with Pima cotton grown in Arizona. The price of Egyptian and Pima, however, is always under that of Sea Island. When Egyptian is quoted on the

Boston market at say 29 cents Sea Island is worth about 32 to .'<3 cents per pound, that is, the fancy grades.
Keeping Sea Island Seed From Becoming Contaminated
One of the big problems confronting the WPA Sea Island Cotton Program consists in keeping the present strain of Sea Island cotton from becoming mixed or contaminated with upland cotton. Grower contracts in both 193") and 193(> contained clauses to the effect that Sea Island cotton would not be planted nearer than one mile to upland cotton and where Sea Island was planted any nearer than one mile to upland cotton all Sea Island seed were to be sold to oil mills and in no way used for planting purposes. WPA field workers also "rogued" or destroyed all off-type Sea Island plants in the fields. Sea Island growers were constantly advised of the necessity of keeping the se.ed ,pure in order to continue to produce premium staple lengths. In addition, gins where Sea Island cotton was ginned, were carefully inspected and cleaned of any scattered upland seed before the Sea Island was brought to the gin.
Cooperation With The U. S. Department of Agriculture and The Florida Agricultural Experiment Station
. In all the research work and other phases of the Sea Island Cotton Rehabilitation Program, a very close cooperation has been maintained with the U. S. Department of Agriculture and the Florida Agricultural Experiment Station. Outlines for the various lines of experimentation have been submitted to the j.bove organizations before the experiments wrere undertaken and the above institutions have been informed of the progress made in all of the experiments and demonstrations.
Recommendations for Use in Growing Sea Island Cotton Under Florida Conditions
Throughout the Sea Island Rehabilitation Program efforts have been made to keep the farmers informed as to the proper planting dates, fertilization, culture and care in harvesting.

FIG. 6. Showing one type of hibernating quarters for the boll weevil commonly found in the Sea Island cotton belt.
In well drained pine woods surroundings such as arc shown above, the poisoning operation need not be continued alter June 10 or 12. Weevils emerge from such surroundings earlier in the spring than in damp mossy situations, and arc generally not as numerous.
Very small amounts of poisoned-syrup mixture arc required to mop an acre of cotton with a poor stand such as shown here. This would not be the case if the poison were applied with a power sprayer or dusting machine, which would put most of the poison on the ground instead of the cotton plants.
The WPA has definitely established the fact that planting on or before March 20 will give better results than planting at a later date. Fertilization for Sea Island cotton is discussed in another section of this report. Although only one year's work has been completed, it now appears probable that with another year's additional work on fertilizers, it will be possible to so fertilize Sea Island cotton as to lead to average production's well above one-half bale per acre even.on the poorest soil types. This information will he passed along through the medium of joint recommendations issued in cooperation with the Florida Agricultural Experiment Station.

Defoliation of the Cotton Plants by the Cotton Leaf Worm. Alabama Argillacea Hubtter As It Affects the Cotton Bell Weevil
A typical and rather severe attack of the cotton leaf worm. Alabama Argillacea IIubner in North and Central Florida in the fall of 1936 offered an excellent opportunity to study the effect which'defoliation of the cotton plants by this insect has on the probable intensity of boll weevil infestation the following year. While it is generally recognized that the cotton leaf worm is frequently a benefactor, in that it reduces weevil population in the fall, so far as the writer knows no serious study of this phase of leaf worm ecology has ever been made.
The early phases of the leaf worm attack were not studied, so it is not known which generation effected the principal part of the defoliation. However, about September 20 evidence of a widespread attack appeared in many lields of Madison county, and by (>ctober 1 practically all the fields of both upland and Sea Island cotton were either completely or almost completely defoliated. I.n most cases defoliation was complete in live days after it began.
Defoliation of the cotton plants in Madison county by the leaf worm was probably more complete than it was in most of the other counties. In many fields in this county every vestige of foliage was eaten off the plants, including that on the basal sprouts, also the tender ends of the limbs were eaten. This type >f defoliation was common in Madison county both on Sea Island and upland cotton wherever the cotton was large and succulent, such as on "new ground" patches and in bottoms. On the lighter soils, where the cotton growth was not succulent, defoliation was not. in general, complete. In such locations usually small bunches of leaves ami squares were left at the ends of the limbs, while the basal sprouts were hardly attacked at all. Occasionally a field of fairly succulent cotton was encountered in Madison county where defoliation by the leaf worm was slight. This, however, was uncommon, and n<> explanation is suggested for these exceptional cases.

In damp, mossy situations such as is shown above, the poisoning operation should he continued until June 18 or SO, since the emergence of the weevils from such hibernating quarters continues until comparatively late in the spring.
Perfect control of the weevil was obtained in this field of Sea Island cotton with the WPA afternoon poisoning method, in spite of the facl that cotton was grown near this same swamp the previous year.
In sonic instances where large succulent Sea Island cotton was completely defoliated by the leaf worm, many green half-grown bolls were in the tops of the plants. In such cases, a large number of weevils remained to feed on the bolls for about two weeks, or as long as they remained succulent. Only a few weevils remained longer than three weeks to feed on the bolls, however, since by this time most of them had begun to harden or decay. In a few of the fields of completely defoliated cotton, weevils emerged from the top bolls in large numbers after the plants were defoliated, and fed on the bolls until they hardened.
Where a good crop of succulent Sea Island bolls remained for the weevils to feed on it is impossible to say whether a marked reduction in the following season's weevil population
PIG. 7. Showing another type of hibernating quarters for boll weevil.

was effected or not. No doubt many of these weevils lived long enough to enter hibernation, but it is questionable whether any appreciable number of them will survive the winter, since green bolls are not thought to be as suitable food for the boll weevil as are squares and blossoms. If green bolls are very much inferior to squares and blossoms as weevil food, then complete defoliation of the cotton plants by the leaf worm effected a great reduction in the number of weevils to survive the winter from even those fields where many succulent bolls remained. However, if succulent Sea Island bolls arc about as suitable boll weevil food as squares and blossoms, then many weevils will survive the winter from some of the Sea Island fields where the foliage was completely consumed by the leaf worm.
However, in a majority of the cases of complete defoliation by the leaf worm, very few bolls were on the plants that were suitable for weevil feeding. In these instances, the weevils had practically all left the fields within a week. It is not thought that more than a negligible percentage of these could survive the winter and infest the following year's crop under Florida conditions.

Close watch was kept of the. large succulent cotton that had been completely defoliated to determine if a brood of weevils could emerge from squares that grew on sprouts developing after defoliation. There was no instance in Madison county where this occurred. Sprouts began to appear about two weeks after defoliation, and it was about two weeks later before squares began to appear large enough for weevils to feed on. So many other insects, such as grasshoppers, leaf hoppers, etc., fed on the squaresand foliage of the sprouts that very few such squares became large enough for weevil oviposi-tion. It is possible that some exceptions to this rule occurred in the more southern part of the belt, where sprouting and squaring proceeded at a more rapid rate. However, the principal importance of the late squares seemed to be as food for conditioning the weevils, for successful hibernation.

FIG. 8. Making sure that the hibernated weevils have been completely eliminated. WPA Field Assistants examining a field of Sea Island cotton at the end of the poisoning season. Note the dense wcods at the border of the field.
Where the leaf worm left any appreciable amount of foliage, sav as much as ten percent, an entirely different situation was observed. Most of the foliage left would be at the ends of the limbs, and some small squares would invariably be included. The weevils congregated in these locations instead of leaving the fields. Since the foliage and squares left on the plants grew very rapidly, it is doubtful if the weevils in such fields suffered for food much more than in fields where no defoliation at all tool; place. Where large numbers of weevils became adult and emerged from bolls after defoliation took placea much more common occurrence in Sea Island cotton than in upland cottonthe presence of squares left by the leaf worm in such fields undoubtedly enabled a heavy brood of the voting weevils to fatten before frost.
The result of all factors seemed to be that on fertile soil where the cotton tended to produce squares and succulent foliage, complete defoliation by the leaf worm was almost as ef-

fective in reducing the following year's boll weevil infestation as plowing under the cotton at an early date would have been. Incomplete defoliation by the leaf worm, however, did not seem to decrease the following year's weevil supply greatly, even where only a small percentage of the foliage was left on the plants.
()n poor soils complete defoliation of the plants was uncommon on both upland and Sea Island cotton. However, on the poorer soils of Florida heavy infestations of weevils in the tail are very uncommon, because of the striking determinate habit of growth of cotton on such soils. I.t is thus evident that 'hough the leaf worm did effect complete defoliation of the plants only where the cotton was most green and succulent, very little harm was done by their failure to defoliate the remainder.
In no case of incomplete defoliation during the latter part of September did a second outbreak of the leaf worm occur sufficiently severe to complete the defoliation. Whether the leaf worm was held in subjection principally by its numerous In most of the other counties inspected. Jefferson. Lafayette. Suwanee. Columbia Baker. Union. Bradford. Alachua. Dixie, Gilchrist, Marion, Sumter and Hernando, complete defoliation by the leaf worm was not as common as in Madison county, and fairly heavy infestations of weevils were more common in parts of these counties. The tendency to sprout and produce a new growth of scpiares was more pronounced in the more southern counties, possibly because of different soil .'.nd climatic conditions.
In certain sections of Union. Bradford and Alachua counties fairly frequent cases of defoliation by rust were encountered, after wdiich subsequent sprouting and square production was common. In such cases the brood of weevils that hatched from the top bolls of the Sea Island cotton found an abundance of suitable food.

FIG. 9. Sea Island cotton on the farm of Charles Ragans, near Macedonia Church, Madison county, Florida. Note open cotton near the tops of the plants.
It is impossible to say with any degree of certainty to what extent the leaf worm attack may be expected to reduce the initial boll weevil infestation in Florida in 1937; In some localities it will almost certainly reduce it a great deal. These are the localities where the leaf worm attack was so complete that the weevils left the fields shortly after defoliation. However, in the localities where defoliation was not complete, where, in other words, the weevils remained in the field until subsequent sprouting and square production took place, it is probable that something in the order of a normal infestation of weevils may be looked for in the spring of 1937";
Entomologists interested in the cotton boll weevil have considered the possibility of controlling this insect by the introduction of the leaf worm in the cotton fields at such a time and in such numbers as to cause the defoliation of the plants every year at a late enough date so as to avoid injury to the bolls but early enough to largely starve out the fall brood of weevils. This study sheds little if any light on how this might

be accomplished, but it does indicate that to be a successful weevil control measure, defoliation of the plants by the leaf worm would have to be complete in all fields that would otherwise furnish large numbers of weevils for hibernation.
After the success of the Sea Island cotton venture in 1M."> in Madison, as well as (iilchrist counties, Florida, it was immediately apparent that growers would attempt to expand the industry very rapidly. Among the problems that this forced to the front, were ones of proper fertilization and spacing of Sea island cotton under present farming conditions. The culture of upland cotton had been revolutionized since the appearance of the boll weevil and the disappearance of Sea Island cotton in Florida, and it seemed reasonable to assume that many of these changes in cultural practice would be beneficial in the case of Sea Island also, on which practically no spacing or fertilization tests had been conducted in Florida even before the advent of the weevil. Consequently a test was planned for 1 !>:{<; 10 determine, if possible, roughly what spacing and fertilization practice would prove most practicable under present conditions.
The experiments were located on the farm of C. L. Black, about four miles southeast of Lee. Florida. This district was ;< typical Florida Norfolk Sand area. The particular plot of ground selected for the experiments was Norfolk Fine Sand, deep phase, which had been in cultivation a long time, and was therefore considerably impoverished. Several small "bottom" .-pots were scattered about the experimental field, where the soil fertility was considerably higher than the average for the area as a whole. This could not be avoided, but an attempt was made to arrange the experiments so that the soil irregularities would not interfere unduly with the interpretation of tin-data.
The vegetaion was burned or otherwise cleaned off and the land in the experimental field broken somewhat shallowly in February. In the early part of March the plats were measured off. and the rows opened with a shovel plow, the fertilizer applied as indicated later, and "listed" upon. On March 1(5 the seed were planted with a Cole corn planter. Due to poor ger-
Fertilizer Experiments

FIG. 10. Sea Island cottor. on the farm of Leroy Di.xon, near Hickory Grove church, Madison county, Florida.
ruination qualities of the seed, cold weather and too deep planting, such a poor stand was secured that it was necessary to plant the entire experimental field again. This was done on April 2 and 3. A small scrape was run shallowly down the seed bed, and the seed planted directly behind the scrape with a Cole cotton planter, at a rate of a little less than a bushel per acre. An excellent stand was obtained, but it was. of course, late. The cotton was thinned, unless otherwise specified, to Hi inches in the drill, two plants per hill.
The general plan of the experiments may be outlined roughly as follows:
(!) A general fertilizer test, consisting of four acres. Comparisons of varying rates of -J-8-4, with and without additional amounts of kainit, both applied at planting time. Half of the plats received a side dressing of nitrate of soda on June 23, half did not.

(2) A spacing test of three widths of rows, and five drill spacings.
(3) A test of the value of late applications of nitrate of soda, made under several Supplementary fertilization conditions.
( I) A test of varying amounts of kainit applied at planting time, with and without a side application of nitrate of soda on, June 23.
(*>) A test of planting in checks, on a bottom spot of ground with considerable humus, two plants per hill and four plants per hill, with and without a side application of nitrate of soda on June 23.
(6) A rather extensive experiment, the topping dates being June 1">, July I. and July 15.
(1) A soil moisture study, to determine the effect of several rates of fertilization on the soil moisture supply.
The general layout of the experimental field is shown in some detail in Fig. 11.
FIG. 11. Map of Experimental Field, with General Description of Fertilization or Spacing Treatment Given Each Acre.
Acre Xo. 1 General Fertilizer Test Acre Xo. 2 General Fertilizer Test Acre Xo. :{ General Fertilizer Test Acre Xo. 4 General Fertilizer Test Acre Xo. 5 Kainit Test Acre Xo. ti 1 lanted In Checks
Acre Xo. 7 Spacing Test Acre Xo. 8 Spacing Test Acre Xo. 9 Spacing Test Acre Xo. 10 Xitrate of Soda Test Acre Xo. It Xitrate of Soda Test Acre Xo. 12 Xitrate of Soda Test
Acre No. 1. General Fertilizer Test. Varying amounts (200, -100. 600 and 800 pounds per acre) of 4-8-4 at planting

FIG. 12. Two typical Sea Island cotton fields, in which defoliation by the leaf worm was practically complete. Note that all foliage was eaten even from the tips of the limbs and the tops of the plants. Wherever defcliaticn was as complete as here shown, the boll wcevi's soon migrated and no new brood was hatched. The foliage in both fields was dense prior to the leaf worm attack, which occurred about Sept. 25.

time, with and without a side application of 75 pounds per acre of nitrate of soda on June 23.
Acres Nos. 2, 3 and 4, General Fertilizer Test. Treatment of each of these acres was identical to Acre No. 1, except that in Acres 2. 3 and 4, 50, 75, and 100 pounds per acre, respectively, of kainit, was applied, in addition to the 4-8-4, regardless of the amount of the latter, at planting time. The check rows of each acre, which received no 4-8-1, received no kainit either.
Acre No. 5, Kainit Test. Plats fertilized as follows: 0, 30, 100, and 150 pounds per acre of kainit at planting time, with, and without, a side application of 75 pounds per acre of nitrate of soda on June 23. No other fertilizer than the above used on Acre No. 5.
Acres Nos. 7, 8 and 9, Spacing Test. In each acre, five drill spacings were used, 12, .11, 16, 18 and 24 inches, two plants per hill. The row widths were 4, dj/2 and 3 feet, respectively, for the three acres. Each acre received 400 pounds per acre of 4-S-4 at planting time, with no additional applications.
Acres 10, 11 and 12, Nitrate of Soda Test. These acres received C00, 500 and 400 pounds per acre of 4-8-4 at planting time, respectively, and in addition, 75 pounds per acre of nitrate of soda on May 8. On June 23, plats on each acre received additional amounts of nitrate of soda as follows: 0, 50, 75 and 100 pounds per acre.
Acre No. 6. In this acre the cotton was planted in checks, 86 inches each way, two plants per hill and four plants per hill. All plats received 400 pounds per acre of 4-8-4 at planting time. In addition, some received a side application of 75 pounds per acre of nitrate of soda on June 23.
In attempting to interpret the picking weights of the general fertilizer test, it should be borne in mind that the test was not sufficiently extensive to warrant coming to any definite conclusion as to which fertilization rate was most satisfactory. As mentioned before, the soil was somewhat variable, and the

picking weights do not appear to be consistent in all cases. The differences in yield between the most fertile part of the field and the poorest parts were very pronounced. This, in a sense, was a detriment to the experiment as a whole, but it also served to call attention to the need of studying soil conditions, and determining what types of soil are most suitable for Sea Island cotton production. It also illustrated forcibly that a fertilization program that is profitable on one soil type might be a failure on another. Had the soil type been uniform and the results entirely consistent, they probably would have been misleading, for the probability that the farmer would have exactly the same soil type would be remote.
Although these differences in soil fertility were not apparent before the cotton was planted, it became evident as the season progressed that acres Nos. 1 and 2, of the general fertilizer test, were much more fertile than acres Nos, 3 and 4. Judging from the yields of the non-fertilized cotton, in fact, the fertility of the different acres decreased in regular order, acre No. 1 being the most fertile, and acre No. 4 being the poorest. This decrease in fertility continued into acre No. 5, which was used for a special test of kainit, with and without nitrate of soda applied June 23. All comments as to the comparative fertility of the other acres must be taken as a personal interpretation, for no system of checks was used to make a comparison as to the relative fertility of the soil possible on these acres, as was the case in acres Nos. 1 through 5.
The yields of the general fertilizer test, acres Nos. 1 through -I, are given in Table 1. For all fertilization rates, the yields given are averages of three plats.

TABLE 1. Average Yield in Pounds Seed Cotton per Acre in the General Fertilizer Test.
Average yield per acre of plats receiving amounts of fertilizer indicated below at planting time
Acre No., and Fertilization Program Non- !200 lbs. 1400 lbs. fertilized! 4-8-4 | 4-8-4 600 lbs. 4-8-4 800 lbs. | 4-8-4 |
Acre No. 1 1-8-4 alone 560 761 833 804 i 939 1
Above, plus 75 lbs. Nitrate of Soda 1 I 876 886 965 894 |
I Acre No. 2 4-8-4 phis 50 lbs. kainit per acre 50!) S06 615 823 940
Above, plus 75 lbs. Nitrate of Soda ...... i 778 753 867 856 1
i Acre No. 3 4-8-4 plus 75 lbs. kainit per acre 458 693 591 617 776
Above, plus 75 lbs. Nitrate of Soda i 671 808 810 751
i Acre No. 4 1-8-4 pius 100 lbs. kainit per acre 343 494 456 441 519
Above, plus 75 lbs. Nitrate of Soda I 638 555 613 566
The 4-8-4 and Kainit were applied at planting time, the Nitrate of Soda June 23.
It will be noted in Table 1 that the yield of the non-fertilized rows decreased from acre to acre. This may be taken as a fair measure of the comparative fertility of the soil in these acres. There also seems to be a general tendency for the yield of the fertilized plots to decrease also from acre to acre, in spite of the fact that each succeeding acre was fertilized heavier than the one preceeding it. The most that can be said from a general inspection of this table is that the addition of kainit. in the amounts indicated, did not overcome the decreasing fertility of the soil. However, the addition of 75 pounds of nitrate of soda per acre on June 23, in most instances, increased the yield appreciably.
The data of Table 1 are shown in Table 2 in a somewhatj different arrangement. Here, acres 1 and 2 are grouped to-

gether, and compared to acres 3 and 4, which are grouped. The former two acres were fairly fertile, while the latter were rather poor. This gives a good comparison of the results of fertilizer applied to fertile soil as compared to poor soil. The amounts of fertilizer shown at the top of the table represents the average amount applied at planting time, which included all the 4-8-4 and kainit applied.
TABLE 2. Comparison of the Yields of the Two Most Fertile Acres with the Two Poorest Acres of the General Fertilizer Test.
Acre No., and Fertilization Program Average yield per acre of piats receiving approximately the amount of fertilizer indicated below at planting time
Non-fertilized plats 225 lbs. 425 lbs. 625 lbs. 825 lbs.
ACRES 1, PLUS 2, Fertile Soil 4-8-4 plus kainit Above, plus 75 lbs. Nitrate of Soda 535 784 724 814 040
-- 827 820 916 875 !
ACRES 3, PLUS 4. Poor Soil 4-8-4 plus kainit Above, plus 75 lbs. Nitrate of Soda Non-fertilized plats 285 lbs. 485 lbs. 685 lbs. 885 lbs. i
401 504 524 520 | I 043 1
655 682 712 : 659 j
NOTE: In the above table, all plats used in arriving at an average yield received within 25 pounds of the gross amount indicated None of the plats whose yields are given under the non-fertilized column received any fertilizer at all, not even Nitrate of Soda.
From Table 2 it is apparent that the yield on the more fertile acres was much higher than on the less fertile ones in spite of the fact that the latter received larger amounts of fertilizer. It is also apparent that on the more fertile soil, the higher rates

FIG. 13. Average Yields of the Two Most Fertile Acres, (left), and the Two Poorest Acres, (right), of the General Fertilizer Test
ef fertilization gave much greater increases in yield above the non-fertilized plats than was the case on the less fertile soil.
This fact is even more apparent from Fig. 13. In this graph, the yield per acre is plotted against the total poundage of fertilizer used, including the 4-8-4 and kainit applied at planting time and the nitrate of soda applied later. As in Table 2 the poundage of fertilizer indicated is an approximation, being the mean of two rates. Since in this graph the total poundage of fertilizer used is plotted against the yield, the slope of the curve is a good indication of the rate of increase in yield effected by the fertilizer. It must be borne in mind though that the nitrate of soda is the more expensive ingredient, and unless it produced a substantial increase in- yield, its use was not justified from an economic standpoint.
A comparison of the left and right hand graphs of Fig. 13 indicates plainly that the yield was much higher on the more fertile acres. In the case of acres Nos. 1 and 2, the trend of the graph continues sharply upward, with the exception of a break in the line at the 225 pound rate. The increase here seems to be about a half pound of seed cotton for each pound
535 5 535 5
Arere(e pound! fertiliier Applied per acre
290 TEo SB 53o
ATcrtge pounds fertilizer applied per ecre

FIG. 14. Unfertilized cotton, left, and fertilized cotton, right, of the General Fertilizer Test. The wide space separates aero No. 1 from acre No. 2.
of fertilizer applied, with the exception of the 225 pound rate, where the increase seems to have been about a pound of seed cotton per pound of fertilizer. Nitrate of soda appeared to be practically valueless on these two acres. The nitrate of soda plats produced the highest yields in most instances, it is true, but apparently this same increase would have been gotten from a somewhat larger amount of 4-8-4 and kainit, costing as little or perhaps less than the 75 pounds of nitrate of soda.
On acres Nos. :i and -I the situation appears to be different. The heavier fertilizations of l-S-l and kainit gave no better yields than the 285 pound rate, where no side dressing of nitrate of soda was used. The 485 and 685 pound applications gave a lower yield than the 2S5 pound application for some reason, where no nitrate of soda was used. The nitrate of soda corrected this failure to a large extent but even with its use the higher rates of fertilization did not give profitable increases in yield over the 285 pound application of 4-8-4 and kainit at

planting time, and the 75 pound per acre application of nitrate of soda on June 23.
Tests of Kainit and Nitrate of Soda on Poor Sandy Soil
Acre No. 5 was devoted ot a test of kainit. nitrate of soda, and combinations of the two. In each instance the kainit was applied at planting time, and the nitrate of soda, at the rate of ".'"> pounds per acre, was side dressed to the cotton on June .23.
The soil of this acre appeared to be as poor as any on the twelve acres. The arrangement of the test was somewhat different from general fertilizer test, in that the non-fertilized cotton, instead of being planted the entire length of the test acre, was divided into three blocks, and placed at random over the test acre, as were the plats of each of the fertilization treatments.
Table 3 gives the average yield of the different treatments, and Fig. 1G shows the yield per acre plotted against the total poundage of fertilizer used for each treatment.
TABLE 3. Average Yield in Pounds Seed Cotton per Acre on Poor Sandy Soil Fertilized with Kainit, with and without Nitrate of Scda, or not Fertilized, as indicated.
Average yield per acre of plots receiving no fertilizer, kainit, nitrate of soda, or kainit and nitrate of soda as indicated
1 i i r Pounds kainit applied per 1 1 I acre
i i 1 Noiie 1 50 1 1 100 150 1 1
1 I Kainit as indicated, no Nitrate of Soda 2S4 1 405 540 572 1
75 pounds Nitrate of Soda per acre, plus Kainit as indicated 1 1 4S5 | 543 I COS 725
The Kainit was applied at planting time, the Nitrate of Soda side dressed June 23.

FIG. 15. General Fertilizer Test, left, and Spacing Test, right, at the WPA Experimental Farm.
The concurrence of the two lines in Fig. 1(6, and the similarity of their trends, indicate that kainit was as efficient as nitrate of soda in increasing the yield where only one of these ingredients was used. If this were true for large amounts of kainit as well as small amounts, it would not pay to use nitrate of soda under the conditions of this experiment, hut very likely this would not be the case. The yield curve of the non-nitrated plots broke appreciably at the heaviest fertilization rate, that is. the 150 pound application produced but little more cotton than the loo pound application. It thus appears probable that higher applications of kainit, without the side dressing of nitrate of soda, would have increased the yield but little, but the experiment was not extensive enough to state this as a fact.
On the other hand, the yield curve of the plats receiving the side dressing of nitrate of soda did not break at any point; that is, for each additional amount of kainit applied, in combination with the 75 pound per acre side dressing of nitrate of soda, a substantial increase in yields was obtained. This suggests that, in all probability, still larger amounts of kainit would have paid in this series.

u o j
*} o +> P o o
xi B 3
,No nitrate of\ Vsoda applied i
/75# per acre N.S.) ____
\applied June 23
. gio IcJo H5 2U0 260
Total pounds fertilizer applied per acre
FIG. 16. Yields on poor sandy soil where only kainit was applied at planting, with and without a side application of nitrate of soda June 23.
The greatest percentage increase in yield clue to fertilization was obtained in this acre, the increase being about 1.1 a pounds of seed cotton per pound of fertilizer used. This was a surprise, for the cotton on this acre was small and somewhat lacking in color.
The Nitrate of Soda Experiment
Nitrate of soda was used, to some extent, on all plats except those devoted to the spacing test. In all cases discussed so far however, only one rate of application was used, 75 pounds per acre, and that applied on June 23, a later date than is usually recommended. In the series to be described here, however, several rates of application were made on June 23.
These experiments were conducted on acres Nos. 10. 11, and 12. Reference to Fig. 11 will show that acre No. 10 re-

ceived GOO pounds of 4-S-4, acre No. 11. f>00 pounds, and acre No. 12, 400 pounds, respectively, at planting time. Each acre also received a 75 pound per acre application of nitrate of soda on May 8. On June 23, three plats of each acre received 100 pounds per acre of nitrate of soda, three plats 75 pounds, three plats 50 pounds, while three plats received no nitrate of soda at all. In all cases, it will he noted, the cotton was rather heavily fertilized, the lightest fertilization rate being 400 pounds per acre of 4-8-4 at planting time and 75 pounds per acre of nitrate of soda on May 8.
Table 4 gives the average yield in pounds of seed cotton per acre for each treatment on each acre of the nitrate of soda test, also the average yield for each rate of nitrate of soda application for all three acres. In the latter case, each yield given is an average of nine plats, differing but little in fertilization, that difference being in the amount of 4-8-4 applied at planting time. Fig. 17 shows this in graphic form.
TABLE 4. Average Yield per Acre in Nitrate of Soda Experiment in Plats Fertilized as indicated
Acre No., and Fertilization Rate Prior to June 23 Average yield per acre of plats receiving side dressing of Nitrate of Soda on June 23 in per acre amounts indicated
1 None j 50 lbs. 75 lbs. 1 100 lbs. j
ACRE No. 10 600 lbs. 4-8-4 at planting, 75 lbs. Nitrate of Soda. May 8 691 7.(4 768 764
ACRE No. 11 500 lbs. 4-8-4 at planting, 75 lbs. Nitrate of Soda, May 8 545 377 500 662
ACRE No. 12 400 lbs. 4-8-4 at planting, 75 lbs. Nitrate of Soda, May 8 000 540 678 725
Average, Acres 10, 11 & 12 Average 500 lbs. 4-8-4, 75 lbs Nitrate of Soda 612 553 649 717

Pounds nitrate of soda applied June 23
FIG. 17. Average yields of plats in nitrate of soda experiment receiving amounts of nitrate of soda indicated on June 23.
Evidently the last application of nitrate of soda did not have a very pronounced effect on the yield, though the trend of the graph is definitely upward. Most likely, the fertilizer applied previous to June 23 was sufficient to produce near maximum yields on the rather poor soil in part of the field. No unfertilized plats were left, so it is not possible to say how much the fertilization program as a whole increased the yield. The application of nitrate of soda on May S caused the plants to grow very rapidly. From the appearance of the cotton on these plats, one would have expected larger yields than were secured.
The Spacing Test
As is shown in Fig. 11, the spacing test was conducted on acres Nos. 7, S and 0. Three widths of rows were used, 3, 3J4 and 4 feet. Five drill spacings were used on each acre, 12, 14,

1G, 18 and 21 inches, two plants per hill in all cases. Four hundred pounds of 4-8-1 fertilizer were applied uniformly on each acre at planting time, and no subsequent applications were made.
The soil in this part of the field appeared to be medium fertilenot as fertile as acres Nos. 1 and 2. and not so poor as acres Nos. 3, 4 and ">. Each drill spacing was triplicated on each acre, the plots being scattered at random, just as were the fertilized plots on the other acres.
The yields of the different spaced plats are given in Table 5 and arc shown graphically in Fig. 18.
TABLE 5. Average Yields per Acre in the Spacing Test.
i Acre No., fertilization rate and width of row Average yield per acre of plats spaced in the drill as indicated
12 in. 14 in. 16 in. 18 in. l 21 in.
ACRE No. 7 J00 lbs. 4-8-4 4 ft. Rows 675 730 730 771 638
ACRE No. 8 400 lbs. 4-8-4 zy2 ft. Rows 74S 780 742 799 SOS
ACRE No. 9 400 lbs. 4-8-4 3 ft. Rows 5(50 773 736 769 820
I Average, all plats spaced in drill as indicated fie, 4 761 736 755 7r,5
In this experiment, there appeared to be no definite trend either in favor of wide or narrow spacing in the drill in any one of the acres. In acre 7, the intermediate spacings gave best results, both narrow and wide spacing in the drill giving low yields. In acre 8, the trend is a little in favor of wide spacing in the drill. In acre i), the trend is somewhat more in favor of wide spacing in the drill, though the graph is irregular. It is perhaps significant that in the four foot rows the

FIG. 18. Average yields per acre in the spacing test.
Spaaing In drill, inches
24" spacing yielded lowest, while in the three foot rows the 24" spacing gave the highest yield, the 12" spacing in the drill giving a very low yield. This would seem to indicate that under the conditions of the experiment. 21" in the drill is too wide tor four foot rows, and that 12" in the drill is too close for three foot rows. Everything considered, the 1 !. l(i, and IS inch spacings were the most consistent performers, since these gave good yields for all widths of rows tried.
It is quite possible that on a different soil type different results would have been secured. I.n interpreting the vields of the spacing test, as well as the fertilizer test, it should be borne in mind that they were conducted on fairly poor sandy soil for the most part, and that the season was a dry one.
Planting In Checks
Many old time growers of Sea Island cotton insist that it should be planted in checks, about three feet each way. Therefore, one acre. No. (i, was planted in this way. Most of this acre was in a slight bottom, therefore the soil was more fertile there than the average of the experimental field as a whole.

Apparently, it was about equal in fertility to acre No. 1. Four hundred pounds of 4-8-4 fertilizer were distributed evenly over this acre at planting time. Half the acre was thinned to two plants per hill, '}(!" in the drill, and the other half was thinned to four plants per hill, 36 inches in the drill. The rows were three feet wide in both instances. Twelve rows of each part received 7~> pounds per acre of nitrate of soda on June
The cotton on this acre was greener and healthier looking than in any other part of the experimental field, and appeared to hold up better late in the season, especially the part thinned to two plants per hill. However, it started blooming later than the other plats, and would therefore have suffered heavier weevil damage had no control measures been employed on any of the plats. It is not possible to say. of course, whether the size and greeness of the cotton of this acre were due to the wide spacing of the cotton or to the more fertile soil on this part of the field, but it was probably due in part to both causes.
The yields of the four plats of this acre are given in Table G.
TABLE 6. Yield of plats in the test of planting in checks, 36" each way.
Description of plat Yield in pounds seed cotton per acre
Two plants per hill, without Nitrate of Soda, June 23 1015
Two plants per hill, with 75 lbs. per acre of N.S. June 23 1042
Four plants per hill, without
Nitrate of Soda. June 23_|_S74_
Four plants per hill, with 75
lbs per acre of N.S. June 23 I_55S_
It would appear from Table (i that planting two plants per hill in this manner was considerably superior to planting four plants per hill. However, there was only one plat of each treatment, so such a conclusion would hardly be justified. From the fact that the nitrate of soda rows of the four plants per hill plat produced less cotton than the non-nitrated part, it seems safe to conclude that that half of the field was less fertile than the half that was planted two plants per hill.

Topping Experiment
Four rows on every plat of the spacing test and the nitrate of soda test were topped. Two were topped on June IS, the border row of each plat and the one adjacent to it. The third row of each plat was topped on July 1. and the fourth row on July 15. The topping consisted in breaking about four inches off the tops of the plants. Blossom counts and picking weights were kept on only three of the topped rows of each plat, no records being kept on the border rows.
The yields of the topped plants are given in Table. 7,

TABLE 7. Yields in the Topping Experiment
Acre Xo., and General Average yield per acre of all Description of Flats rows lopped on dates indicated
Not i Topped Topped jTopped June 151 July 1 Topped July 15
Acre No. 7. Spacing test. Four foot rows, drill spacing from 12 to 24 in. Average of 15 plats 709 682 757 749
Acre Xo. 8. Spacing test. Row width, V/2 feet. Drill spacing from 12 to 24 in. Average of 15 plats 775 748 757 820
1 ' Acre No. 9. Spacing test. Row width, 3 feet. Drill spacing from 12 to 24 in. Average of 15 plats 719 697 702 642
1 i Acre No. 10. Nitrate test. Average of 12 plats 739 585 627 621
Acre No. 11. Nitrate test. Average of 12 plats 1 1 521 i 513 519 539
1 i Acre No. 12. Nitrate test. 1 1 Average of 12 plats 1 638 \ 612 618 i 595
! 1 1 Average of all plats topped on date indicated. 81 plats in all 6S3 639 I 1 1 663 1 j 661 1 1
Apparently topping had but little effect on the yield either way. The average of all plats indicates that topping decreased ihe yield somewhat. In this case, the rows topped June 15 yielded lowest, and the non-topped rows yielded highest, but the difference was only about 7% of the total yield. In view of the fact that these figures were taken from an average of 81 plats for each topping date, this difference in yield is perhaps significant.
The fact that lopping would decrease the yield in an average such as this, however, would not prove that it would not increase it under some conditions represented in the experiment. Study of the picking weights in some detail, however, failed to

disclose any consistent advantage or disadvantage of the topping in any phase of the experiment. In some instances the topped rows yielded appreciably more than the untopped rows, but in these cases chance was apparently the biggest factor.
Study of the blossom records seemed to indicate that in some instances the topping stimulated blossoming early in the season somewhat, and depressed it late in the season. When an average of many plats was taken, however, this effect appeared to be so slight as to lose significance.
Rates of Blossoming
Daily blossom records were'kept of every row on which picking weights were secured during practically all the blossoming season, except on Sundays, and a few holidays. These vacancies were filled in by simple interpolation. A report on all the analyses made of the blossom records would be voluminous and much of it would be useless. Only those parts of this study are given that appear to be significant and consistent, and these data are given only in graphic form.
The total number of blossoms that occurred during the season did not forecast the \rield of the plats very well. This will be referred to later.
The rate of blossoming of the fertilized cotton differed very strikingly from that of the non-fertilized cotton. The differences in the rates of blossoming of the different fertilization rates, however, in most cases were not striking. To determine this, the number of blossoms that occured in four day intervals was determined for a large number of plats, and graphs made of all that appeared to be outstanding or of peculiar significance. Figures 19, 20 and 21 show the "rate of blossoming'" graphs for acre No. 1 of the general fertilizer test, acre No. 4 of the general fertilizer test, and acre No. 5, the kainit test, for all rates of application of fertilizer used on these acres. As has

F-IG. 19. Two views of the experimental farm.
Sea Island Cotton planted in checks at the WPA Experimental Farm.
Laborers hoeing Sea Island cotton in the Xitrate of Soda Test, WPA Experimental Farm.

been already pointed out, acre No. 1 was considerably more fertile than either acre No. 4 or acre No. 5.
The most outstanding feature of all graphs is the striking difference in the rate of blossoming of the fertilized plats as compared to the non-fertilized plats. In the non-fertilized plats, the blossoms are seen to appear late in the season, and over a iong period of time, while the blossoms of the fertilized plats appear early in the season, and over a short period of time.
This is a very significant difference from the standpoint of the susceptibility of the cotton to the ravages of the boll weevil. Assuming that most of the early blossoms "stuck," it appears that the fertilized cotton set a large percentage of its crop something like two weeks earlier than the unfertilized cotton. In addition, the fertilized cotton practically ceased to blossom approximately two weeks before the unfertilized cotton practically ceased to blossom approximately two weeks before the unfertilized cotton did. Roth these phenomena are important. The earlier the cotton sets a crop of bolls, the less chance the weevil has to infest them. This is a well appreciated fact, in both upland and Sea Island cotton culture. On the other hand, the importance of the early cessation of blossoming, from the standpoint of boll weevil control, is not so well appreciated. However, especially in the case of Sea Island cot-Ion, if blossoming and the formation of young bolls continue iatc enough in the fall, the top bolls invariably become infested by the weevil, unless a vigorous and expensive poisoning program is instituted to prevent it. Where this occurs, adult weevils often emerge from the top bolls in extremely large numbers, just at the right time to feed on the squares and tender foliage of the top of the plant when it "leafs out" in the fall. On the other hand, if the cotton sets a heavy crop early, and ceases boll production early, as a general thing very few weevils develop in the field in the fall, since most of the squares are so riddled with feeding punctures at this time of year that few of them develop adult weevils.
It appears from both Figures 19 and 20 that differing the rate of fertilization did not change the rate of blossoming curve

FIG. 20. Average Rate of Blossoming for Different Rates of Fertilization on Acre No. 1, General Fertilizer Test.
Two hundred pounds por aoro at planting
of 4-8-4 / \^ ng tine t/ \'v
/No nitrate of. ooda applied '
/75# per acre N.S.\ ^applied J""
TI vs i 2s 2? 5 4* 5"
J u 1 y
A u j.
Four hundred Vounds per acre of 4-8-4 /C\
at planting tine ^
1V0 nitrate oft laoda applied I
5 .---*!. VttppllH J..n 23'-----
'applied June 23'
is s*o s1fn Tz 552irr-jiir
June July Aug.
Six hundrod pounda per aora at planting
of 4-8-4 /-V
ng time V coca appnea
\ \/75# per acre H.S.%_____
___\. ^applied June 23 '
No nitrate of aoda applied
~i& S'O 3 7 }l V S 1'9 2l 27 S'l 4 8" July
af U la
S, it600-S ^400
.2 g
Eight hundred pounda per acre of 4-8-4 at planting tine
.No nitrate of. eoda appllad '
/75# por acre N.S. vappllad June 23 '
2'6 3'0 3" ? 11 i'6 1'9 2'3 21 June July Aug.
!'7 314
a great deal. In all cases the lowest rate of fertilization appeared to produce a blossom curve with about as good characteristics, from every viewpoint, as the higher rates. Often the addition of nitrate of soda on June 2:? appeared to cause the blossom curve to "break" quicker. In many of these plats, the addition of nitrate of soda increased the vield. as has been

pointed out already. Whether a larger percentage of the blossoms "stuck," or whether the bolls produced were larger, was not determined.
In Fig. 21 several types of blossoming curves are shown. The rate of fertilization in most of the plats of this series of graphs was much lower than that of the plats of Figs. 9 and 20. In most of the graphs of Fig. 21, the shape of the curve is seen to be very much like the blossoming curve of the non-fertilized cotton, except for the total number of blossoms represented. J.n the highest rate of fertilization, however, which was 150 pounds of kainit and 75 pounds of nitrate of soda per acre, the graph assumes the same general shape of those of Figs. 19 and 20, though it does not decline as rapidly.
Percentage of Blossoms Producing Open Bolls
Earlier it was stated that there appeared to be no very definite relationship between the total number of blossoms produced during the season and the amount of cotton produced. Since there was a definite relationship between the amount of fertilizer applied and the amount of cottorr produced in many instances, an attempt was made to determine why the blossom counts did not give a reliable forecast of the yield of the plats. Consequently, for every plat of the experimental area a calculation was made of the amount of cotton that each would have produced had every blossom produced a mature boll weighing 1-130 of a pound. The largest bolls no doubt were some heavier than this, while it. is likely that a few of the lightest weighed less. The bolls of acres Nos. 1 and 2 were appreciably heavier than the bolls of acres Nos. 3, -I and 5, and probably of some of the other plats in the experimental field. The average weight of the bolls of each plat was not determined however, so the assumption of a weight was the best that coidd be done. The theoretical production of each plat, calculated in this manner, was divided into the actual production. This gave the percentage of the blossoms that produced open bolls of 1-130 pound each.
The results of most of the calculations are shown in graphic form in Figs. 22 and 23.

FIG. 21. Average Rate of Blossoming for Different Rates of Fertilization on Acre No. 4, General Fertilizer Test.
O 3
Two hundrod pounds per acre of 4-8-4 and 100 p.p kainit at plant-In g til
or // -K
soda applied /7S# per acre U.S. ) '.pplied June .23
i's i5
25 2'7 Sri""*' T
Four hundred pound per acre of 4-8-4 and 100 p.p.a.of ,
kainit at plant- S
ing tlmo
(Wo nitrate of\ _
ioda applied '
"IT; To"
rs is
~S 2'7 5"1 4
S 800 i. u
3 i-
2^400 I ?
Qt- 200
Six hundred pounds per acre of 4-8-4 and 100 p.p.a. or kainit at plant-
ingtfe ~ SZFXVti*
2 'e "s'o
/Ko nitrate ofi_
vaoda applied '
t1S per acre K.S.i ''applied June 23 '
0. e.
9 >j
d -a n
5 !>
~. o
CO *
~2 800 COO 400 200
Eight hundred pounds per acre of 4-6-4 and 100 p.p.a. of kainit at plants in6 tine
US 19 July
Z^S S'l *
(So nitrate of\_ V-li-.i aoplled '
'soia appl
f7*#. Pr ao; (applied Jul
>re N.S. 1V.0 23
The calculations of the percentage of blossoms producing open bolls of the general fertilizer test are shown in Fig. 22. The blossoms of acres Nos. 1 and 2 are seen to have been much more productive of cotton than the blossoms of acres Nos. 3 and -1. Although the blossoms of practically all the plats receiving the side application of nitrate of soda on June 23 in this test were more productive than those of the plats not receiving

FIG. 22. Average Rate of Blossoming for Different Rates of Fertilization in Kainit Test, Acre No. 5.
?800-'600 400 200
No kainit Applied
(No nitrate or\ _
'aoda applied ~~
<78# par acre V.S.\____
appllad June 23 '
S.P. Not fertlllted
TS 2"? 3l
?800 iX.600
5 &
Fifty pounds per acre of kainit at planting time
Sic nitrate of* lOda applied '
s: hs-)----
X\ (IpfiKS S!
y---H.,-p.-'-^*-<^s. N. F.--Not fertilized
216 TS T June
ii V6 i9 2s wn 2 T
July Aug.
g 800
2 4C0 u o
- 200
One hundred pounda per acre of kainit at planting tine
/No nitrate at\ lsoda applied '
7S$ par acre N.S. applied June 23
>. ?. Not fertlllied
Hi 2"7 31
26 3 0 June
i's is
I eoc
o >>
ii m
o -u
S i;
One hundred ar.d fifty pounds of kainit '
per acre at plant- f ing titna
/No nltrete of \ _
viod applied ,75# per acre N.S.\ 'applied June 23 '---
Not fertilized
4 8 Aug.
the side application, the difference is much more pronounced in acres Nos. .'{ and 1 than in acres Nos. 1 and 2.
Probably the most significant faclor about these calculations in the general fertilizer test is the fact that in the case of acres Nos. 1 and 2, the blossoms of the higher rates of fertilization were more productive than the blossoms of the low

FIG. 23. Per cent of blossoms producing open bolls in each acre of the general fertilizer test.
o60 c
5 40-
Pi ?*
a 20 3
Varying asounU of 4-8-4
ftt -r tint
Percentages represented by graphs 65 68 64 67
26o <3o S5o ado
Poinds 4-8-4 applied per acre
Acre So. 2
Varying asounta of 4-8-4 plus kainit per acre at planting tin*
Percentages represented by graphs 69 65 69 67
51 62 55 63 63
2T0 400 ioo eHo
Pounds 4-8-4 applied par acre
Acre No. 5
1 60-d
r. 40-
SO-li 20-* 10-0
Vat/.:.," AJtOUlUf of 4-8-4 plu
"S* kainit per acre at planting title
Piroentages represented by graphs 64 71 67 68
63 S3 4T 45
200 435 60*3 860
Pound 4-8-4 applied pr acre
2 60 to
5 40
tv t>
S 20
-----._ no y.s.
Varying aeounts of 4-8-4 plus iO0f par aora of kainit par acr* at planting tltna
Percentagee represented by graphs 66 48 64 38
' 10 41 44 40 39 33
~Tot5 480 600 555
Pounds 4-8-4 applied par aore
Solid lines (N'o N.S.), no nitrate of soda applied. Dotted lines (N. S.), "5 lbs. per acre of nitrate of soda applied June 23.
rates of fertilization, in general, while in acres Nos. 3 and 4. the blossoms of the low rates of fertilization were the most efficient in producing cotton. In the latter acres, there appeared to have been a definite failure of the plant to mature the bolls set in the early part of the growing season. The side dressing of nitrate of soda seemed to correct this to some extent, but not altogether.
In Figure 'i'.i. the nitrate of soda application again appears to have been beneficial, for the efficiency of the blossoming of the plats of acre No. 5 was definitely higher in the plats rcceiv-

ing the side application than in those not receiving it.
Since the fertility of the soil in this acre was more comparable to that of acre No. 4 than to any of the acres of the general fertilizer test, it is fair to draw a comparison between those two, to judge the efficiency of the kainit applications in producing cotton from a given number of blossoms. In this respect, the kainit and nitrate of soda applications of acre No. "> appear to be about on par with most of the 4-8-4 and nitrate of soda applications of acre No. 4.
Study of the graph of the spacing test in Fig. 23 indicates that the blossoms of the wider spaced cotton were somewhat more efficient than the blossoms of the close spacings. The difference is so slight, however, that it is probably of no significance.
The graph of the nitrate of soda test in Fig. 23 corresponds almost exactly to the production curve of this test, as shown in Fig. 1(5. The heavier applications of nitrate of soda on June 23 seem to have increased the efficiency of the blossoms to some extent.

The graph of the topping experiment shows a negligible difference in the efficiency of the blossoms of the topped versus the non-topped rows. The differences indicated could hardly be considered significant.
In some of the instances noted above, the difference probably was more in the size of the bolls than in the number of blossoms producing open bolls. More than likely, the bolls of the plats receiving the side application of nitrate of soda on June 23 were heavier, in general, than the bolls of the corresponding plats that received no side application. However, where a large difference in the efficiency of the blossoms was found, as in acres Nos. 3 and 4, there is little doubt that much of it was due to heavier shedding in the non-nitrated plats.

CHAPTER VI. Soil Moisture Studies
Although cotton is ordinarily considered a dry weather plant, it. like other plants, draws considerable moisture from the soil. A marked shortage of soil moisture results in a stunting of growth, excessive shedding, and low production. So far as the writer knows, however, no study was ever made of the changes taking place in soil moisture under lightly fertilized cotton as compared to such changes on the heavily fertilized cotton prior to the very brief one made in connection with the WPA Sea Island cotton fertilizer and spacing experiment of 1936. This study was purely exploratory, and limited in extent, but the results were sufficiently striking to warrant their full discussion.
The primary object of the study -was to determine, first, whether the soil moisture was depleted very much more rapidly under heavily fertilized Sea Island cotton than under lightly fertilized Sea Island cotton; second, if this was found to be the case, is there danger of such serious depletion of soil moisture before the maturity of the crop as to markedly impair the efficiency of the fertilizer application.
An outline of the plan of the brief study contemplated was presented to the officials of the Florida Agricultural Experiment Station, with the request that they cooperate to the extent of determining the moisture content of the samples taken. This they consented to do. All moisture determinations were made by weighing before and after drying with heat, in the conventional manner.
Soil samples were taken at irregular intervals at ], 2 and 3 foot depths under the following conditions:
(!) In a roadway, kept free of vegetation. Corn was on one side, and cotton on the other, the samples being taken about eight feet from either.
(2) Under Sea Island cotton not fertilized, about 50 feet from the roadway.

(o) Under Sea Island cotton fertilized with 400 pounds of 4-8-4 and 100 pounds of kainit per acre at planting time, about 10 feet from the roadway and non-fertilized cotton.
(4) Under Sea island cotton fertilized with 800 pounds of 4-8-4 and 100 pounds of kainit per acre at planting time, about 10 feet from the roadway. 40 feet from (3) and 80 feet from (2).
The sampling, in fact, was done in one of the plats of acre No. 4 receiving 800 pounds of 4-8-4. one receiving 400 pounds of 4-8-4, at planting time, the non-fertilized rows of this acre, and in the middle of a roadway along one side of it.
Table 8 gives the percentage of soil moisture found in all the samplings. Fig. 2

TABLE 8, Comparisons of Soil Moisture Percentages at One, Two, and Three Foot Depths Under Bare Ground, Non-Fertilized Sea Island Cotton, and Sea Island Cotton Heavily Fertilized
Conditions under which samples were taken Date of sampling, nd percentage of moisture found
General data on plats sampled Depth sample was taken March 25 April 9 May 10 June 12 July 9 July 21 July 30
ABare ground, 1 foot 4.9'J 5.00 | 4i | 5,51 | 4,88 j 7,45 4,76
kept free of 2 feet 4,50 5,48 1 3,96 | 4,83 4,74 5,38 4,79
vegetation 3 feet 4.45 5,15 | 3,82 4,52 4,39 4,04 4,42
BNon-fertilized lfoot 5.55 5.69 5.87 I 3,88 3,96 6,30 3.77
Sea Island 2 feet 4,63 5,23 | 3,84 | 3,20 j 3,20 2,73 3,74
Cotton 3 feet 4J6 4,82 4,20 4,57 3,70 3,53 3,92
GSea Island Cotton fertilized with 400 lbs. 4-8-4 1 foot 5.03 5.58 5.29 4,70 2,10 5,41 1.73
and 1(1(1 His, hinit 2 feet 4,73 5,07 1 4,42 i 4,38 \ 1,88 1,67 1.48
at planting time 3 feet 41 5,11 I 4,37 | 4,49 j 3,04 1,74 2.30
DSea Island Cotton fertilized with 800 lbs. 4-8-4 lfoot 5,12 ;ui 4.90 5,10 2,31 5,51 1,80
and 100 lbs, kainit 2 feet 4,42 5,03 I 4,08 4,07 1,98 1,79 1,52
at planting time 3 feet 4.49 4,90 3,85 4,45 3,10 1,64 2,02
E- lfoot 5,10 5,58 5,10 4,96 2,21 5,46 1,77
Avcragc of 2 feet 4,58 5,00 1 4,25 1 4,23 1,93 1,73 1,50.
C and D 3 feet 4.(15 5,01 4,11 4,47 3.07 1,69 2,10

FIG. 24. Per cent of blossoms producing open bolls in kainit test, spacing test, nitrate of soda test, and topping test
o D
f so K
4i 20-
i io>
Kainit Test, aore NO* 5
Percentages represented 50 41 48 49
40 36 41 41
50 100 Tso
Pounds kainit appllad par acre
1 S.30H
a, 10
Spaaing Tost, Aoros N0. 7. 8 k 8, averaged.
Peroentaron represented hy erapha 66 60 59 63 61
12 M IS 18 W
Spacing in drill, inches
g 20 o
Nitrata of Soda Teat, Acres Nos. 10, 11 4 12, averaged
Percentages represented Vy graph 54 49 54 60
J 20-
55 76 155
Pounda N.S. applied per aora June 23
Topping Teat, Aoras H0b. ?; 8, 9, 10, 11 & 12, averaged
Percentages represented by graph 57 66 53 67
Not topped
6116 Ml 7-06^
Date cotton xaa topped
Fig. 25A shows the soil moisture percentages for the 1 foot depth. They are seen to he very irregular. This can be accounted for largely by the fact that this near the surface, the moisture content of the soil is influenced considerably by even light rains. In several instances, samples were taken only a few days following a substantial rain. However, after the middle of June there was a pronounced tendency for the soil to be drier under the fertilized cotton than under the non-fertilized cotton. In the latter part of July the soil moisture at the one foot depth was depleted with great rapidity. After being pretty

thoroughly wetted by a series of light rains occurring from the 15th to the 20th of July, the soil moisture percentage dropped from an average of 5.4.6% to 1.77* in nine days on the fertilized plats at the 1 foot depth, while in the same period the drop on the non-fertilized plat was only from G.30% to 3.77%. The fertilized cotton was wilting badly within a week after the cessation of the rains.
No samples were taken after the G.GO rainfall that occurred on July 31 and August 1. This heavy rainfall thoroughly wetted the soil at least five feet deep in all parts of the experimental field, and a series of rains following for some time at almost daily intervals made it appear useless to pursue the study further during that particular season.
Fig. 25B shows the soil moisture percentages for the two foot depth. Here it is seen that during April, May, and part of June, very little difference was found in the soil moisture content under the cotton as compared to that under the bare ground. The upturn on June 12 was probably due to the .80 inch rainfall on June 7. After June 12 however, the soil moisture percentage fell rapidly under both the non-fertilized and the fertilized cotton, but the drop was considerably more rapid under the fertilized cotton. The lowest percentage recorded for the non-fertilized cotton at the two foot depth was 2.73%, on July 21, while the average percentage for the fertilized plats was lower than 2.0% from July 0 until the very heavy rainfall of July 31. When it is considered that the lowest percentage recorded in the experiment wras 1.48%, it is apparent that after Jul)- 9 very little moisture was available to the plants at the two foot depth under the fertilized cotton, while there was considerable available under the non-fertilized cotton.
Fig. 25C shows the moisture percentages found at the. three foot depth. Again, the moisture content of the soil under the cotton differed but little from that under bare ground until, after June 12. The drop was very little at this depth after June 12 under the non-fertilized cotton, whereas it was almost as

marked under the fertilized cotton at the three foot depth as at the two foot depth. This leaves no room to doubt that the heavy fertilization caused the cotton to consume a much greater quantity of moisture than it would have if it had not been heavily fertilized.
Whether or not this extra water consumption by the plants resulted in poor utilization of the fertilizer applied is, of course, another question. However, this appeared to be the case. After the moisture content got as low as 2.5% the soil was dry and powdery, and difficult to bring to the surface with a post hold digger. Certainly the cotton plants were unable to extract much water from such soil. But was more water actually needed to produce a satisfactory crop of matured bolls?
We can only attempt to answer this question conjectur-ally, for data were not taken to demonstrate conclusively that more moisture would have substantially increased the production of cotton, but there was considerable evidence to indicate that it would have. During the latter part of June much of the cotton on the experimental farm began to wilt during the hot part of the day. The .93 inch rain on July 3 relieved this situation temporarily, but it was again wilting badly before the rains in the middle of July began. These rains, of course, stopped the wilting, but it was again in evidence before the G.60 inch rainfall of July 31.
Wilting was much more pronounced on the poor acres that bad been heavily fertilized than in any other parts of the field. Meanwhile, the non-fertilized cotton did not wilt at all. The cotton on acres Nos. 1, 2 and 6, where the growth was very heavy, wilted very little. These acres were more fertile than the average of the field as a whole, and very probably the soil there had more water holding capacity than that of most other parts of the field, although no measurements were made to determine this with certainty. Conversely, the worst wilting occured on acres Nos. 10, 11 and 12, which received and application of nitrate of soda on May S. Probably the earlier vigorous growth of the cotton, due to the nitration, caused a severe

FIG. 25. Percentage of Soil Moisture by Weight at One, Two, and Three Foot Depths Under Bare Ground Non-Fertilized Sea Island Cotton, and Heavily Fertilized Sea Island Cotton on Various Dates During the Growing Season.
" *3 -J
gSoil samples taken at two foor depth ^V
--^-X-sr. / \.
Legond for air graphs:
3.0.--Surface kept free of vogotation N'.F. Onder non-fertilised Sea leland cotton
K.F. Under heavily fer-tUUed Sea leland cotton
I Aug.
C Soil samples taken at three foot depth
Hoh.l April I Say
June I July P Aug.
moisture shortage even earlier than was experienced where the samples were taken.
Acre No. 4, where the samples were taken, was particularly poor, as has already been pointed out. The per acre yield of

the non-fertilized rows of this acre was 343 pounds of seed cotton, while the yields of the plats receiving 500 and 900 pounds of fertilizer per acre (kainit plus 4-S-4) were 456 and 519 pounds per acre respectively. Fig. 21 indicates that blossoming on this acre progressed fairly satisfactorily for all plats, but Fig. 23 shows that, on the non-nitrated plats of this acre, cither a small percentage of the blossoms produced open bolls, or the bolls produced were very small. The former is thought to be a prominent factor in this instance. Whatever the cause, the net result was that the fertilizer applications, especially at the high rates, were very inefficiently utilized by the plants on this acre, insofar as converting fertilizer into cotton is concerned. This appeared to be true wherever the cotton wilted badly during the dry periods.
On the other hand, a very efficient utilization of the fertilizer characterized acre No. 5, as can be seen in Table 3 and Fig. 16. Here the cotton was small, and never wilted. No samples were taken to show that the moisture content under this cotton was higher than in acre No. 4, where the samples were taken, but if the size of the plants can be taken as an indication of the rate of moisture consumption, the moisture content of the soil in this acre must have been almost as high as that under the non-fertilized cotton of acre No. 4.
The season was a dry one. What probably would have happened under different rainfall conditions?
If April and May had been wet, quite obviously most of the fertilizer applied would have been leached out. Since but little moisture was consumed by the cotton prior to June 12, even where heavily fertilized, it is quite plain that much of the rainfall of this period percolates through the surface soil to the lower depths, whether it occurs as a few comparativelv hard rains, or as a series of light ones, extended over the entire two months. This is because normally the soil is fully charged with water in the spring unless a winter crop has been grown, and therefore all additional water added must pass on through. This of course would be true only in the case of readily leachable soils, such as deep phase sands.

Had a few more good rains occurred during the latter part of June and the middle of July, in connection with the comparatively dry April and May experienced, probably much higher yields would have been obtained on some of the heavily fertilized plats. However, ideal seasons are rare. Cotton growers considered the 1936 season an unusually favorable one for cotton in North Florida. Dry periods lasting two or more weeks are not at all uncommon, and the longest one occurring at the WPA Experimental Farm in 1936, after the plants began consuming moisture rapidly, was from June 18 to July 3, 15 days. The next longest was from July 3 to July 15, 12 days. It is true, however, that the weather was extremely hot and somewhat windy during these periods, which was conductive to high evaporation of water by the plants.
The above considerations seem to make two suggestions justified, in the case of Sea Island cotton planted on sandy soil having low water holding capacity. First, apply very little soluble fertilizer at planting time, so as to avoid the danger of losing much plant food from the time it is planted until the plants get large enough to consume moisture rapidly. Apparently this would be about June 1, when it would be time to make a second application. Second, do not apply very large amounts of fertilizer, particularly nitrogen, on such soils. These moisture studies apparently bear out the indication of the yields of the test plots that on poor soil little will be gained hy applying more than about two or three hundred pounds per acre of 4-8-4 or 5-7-5 grade base fertilizer in connection with about 150 or 200 pounds of 0-9-12 or 0-9-15, or possibly 75 pounds per acre of nitrate of soda or its equivalent as a side dressing later in the season.

PIG, 26,
County, during March, April, May, June, July and August, 1336,*
I minimum temperatures and inches precipitation for day and" month indicated
llontli March I April May June July August
1 1 M 83-66 82-36 88-67 92-71 84 ,50
2 M 79-64 84-60 89-62 94-73 90 ,60
I U 68-41 88-62 89-61 96-72 J 1 91
4 77-50 i 66-37 88-64 90-69 94-70 93 ,60
5 73-59 81-52 86-39 95-72 ,20,' 92-68 91
6 76-55 81-65 86-39 92-70 96-71 91 ,14
7 73-52 83-50 87-56 89-69 ,8 98-74 93 ,05
8 79-4!) 63-48 85-59 85-67 94-68 94-73
1) 67-56 84-57 ,38 90-59 83-65 94-67 91-73
10 64-55 76-64 94-64 | 87-H8 ,1( 97-118 89-72
11 79-52 74-54 92-65 92-68 98-71 88-72-2 ,38
12 | 69-51) 78-49 M 92-69 95-69 ,05j 84-72 ,02
13 | 69-38 83-52 91-66 ,?5j 93-71 92-72 90-72
14 1 66-52 1 87-61 92-66 87-71 85-69 93-72
IS 82-50 87-112 82-66 90-69 92-71 ,23|
16 811-58 85-59 87-63 93-68 90-71 ,24, 92
17 m 75-33 89-61 98-74 90-72 87-72
18 M 71-46 88-61 ,50 94-71 1,31) 86-71 ,34 91-73
19 1 74-41 1 67-42 90-63 93-72 89-73 90-72
20 | 81-15 mi 88-64 86-68 88-69 ,13 90-72
21 73-42 81-50 86-68 90-68 87-69 90-72
22 80-46 79-56 83-67 1 88-69 91-73 ,05| 92-74
23 81-53 79-59 83-65 90-70 92-69 92-72
o c

34 74-58 63-58 .56 86-60 83-72 91-72 93-70
ti 82-66 70-59 83-60 86-63 94-72 93-72
N EM) 73-39 87-60 88-62 97-73 95-72
2? 85-30 85-36 84-69 90-02 93-74 96-73
28 81-35 86-59 76-66 JO' 91-119 96-73 96-72
2! 1 M 85-60 81-67 94-7;i 95 .01 92-71
30 "62 84-59 65-66 92-7,1 90 90-73
Jl 84-68 85-66 83 6,60, 88-67
'Temperatures ure LIS, Weather Bureau observations at Madison, 12 miles from ilic Experimental Farm, while precipitation recordings were lab at the Experimental Farm These were not commenced until the early part of April, Several rains occurred during March and April which were not recorded,
Rain Fall Records Xot Taken in March,

Summary and General Recommendations for 1937
On account of being more susceptible to damage by the boll weevil, the Sea Island cotton industry was wiped out almost immediately with the arrival of the weevil in II) IT. Production of this staple dropped from over 90.000 bales in 1917 to 11 bales in 1924. due largely to the ravages of this insect. The permanent rehabilitation of the Sea Island industry is more dependent on the satisfactory solution of the boll weevil problem than on any other one factor.
During both 1935 and 19.16 the application of suitable poisoncd-syrup mixtures, according to an improved technique described in this report, adequately protected Sea Island cotton grown under a wide range of conditions in Madison county from weevil damage, without a single failure. Thirty fields of Sea Island cotton were poisoned in Madison county by this method in 1935, and seventy fields in 1936practically the entire acreage of Sea Island cotton in the county each year. In most instances either upland or Sea Island cotton had been grown in the immediate vicinity the previous year.
Poisoncd-syrup mixtures were found to be more effective when applied in the afternoon, when the plants are dry and the weevils arc thirsty. By pushing the wetted mop through the tops of the small plants while walking full speed, the liquid poison is applied to the under side of the bud leaves, as well as on the limbs and main stem of the plants, where it is shaded from the sun and protected from rain to a large extent. This method of applying the poison is faster and more effective than the old style, or "churn-dasher," method of mopping, where the tops of the bud leaves are wetted with the mixture by dropping the mop on the top of each plant.
The present strains of Sea Island cotton, properly fertilized and cultivated, were found to produce a satisfactory crop of

cotton' "sufficiently"' early to "escape damage from" trie midsummer migrations, provided the cotton is planted early, and" all the hibernated weevils arc killed by poison" applications as they enter the field during May and June.
Defoliation of the Plants by the Cotton Leaf Worm
The complete defoliation of the cotton plants in early fall by the cotton leaf worm appeared to markedly reduce the number of boll weevils left in the fields in late October and November. Partial defoliation by the leaf worm appeared to be less effective in this respect, even where the percentage of foliage left on the plants was comparatively small.
In 193G complete defoliation by the leaf worm was much more common in rank cotton, which otherwise would have supplied large numbers of weevils for winter quarters, than in cotton grown on poor soils which would not have furnished many weevils anyway.' In many localities, therefore, the leaf worm attack of 1036 greatly reduced the number of weevils which would start the 1937 infestation.
Fertilization and Spacing
On fertile soil areas, relatively large amounts of 4-8-4 or 4^-8-4 in combination with kainit, applied at planting time, produced substantial increases in yield over non-fertilized or lightly fertilized plats. Late nitrate of soda applications on the better grades of soils gave but slight increases in yields.
On the poor soil areas, heavy applications of 4-8-4 and kainit applied at planting time gave little or no increase in yield over the lightly fertilized plats. Late applications of nitrate of soda gave substantial increases in yield in the poor soil areas) where the previous fertilizer applications had not been too heavy.
On poor sandy soil, 150 pounds per acre af kainit at planting time, followed by 75 pounds per acre of nitrate of soda on June 23, gave excellent yields, considering the very low yields of the non-fertilized plats on the same area.

Medium spacing, around 16 or 18 inches in the drill, two plants per hill, in 3y2 foot rows, appeared most satisfactory for fairly poor soil, all things considered. On the better grades of soil, 4 or 4^2 foot rows appeared to be somewhat preferable. High yields were obtained from cotton planted in checks, 36 inches each way. in a bottom area, where the soil was more fertile than most of the other parts of the experimental field, but the bolls matured somewhat late.
A study of the blossoming rates of the different treatments indicated that fertilized cotton set its crop and practically ceased to blossom considerably earlier than non-fertilized cotton. This is an important factor from the standpoint of escaping heavy damage from migratory weevils.
On poor soil, heavily fertilized, a smaller percentage of the blossoms appeared to produce open bolls of normal weight than on like soil moderately fertilized. On fertile soil, the reverse appeared to be true. A late side dressing of nitrate of soda increased the cotton production per blossom in all instances observed save one, and here the decrease was too small to be significant.
Topping the plants appeared to decrease the yield f lightly.
Soil Moisture Studies
Soil moisture studies showed conclusively that heavily fertilized Sea Island cotton draws water from the soil at a much more rapid rate than non-fertilized Sea Island cotton. On poor sandy soil, which will store but small amounts of soil moisture as compared to loams, clays, and other fertile types, heavily fertilized Sea Island cotton will consume all the available moisture in the soil very quickly, after it is large, but still green and thrifty, and will wilt and shed badly if a dry period that is at all prolonged occurs at this critical stage of its growth.

General Recommendations
Purity of Seed: Exercise all care possible to prevent contamination of planting seed either in the field or at the gin.
Time and manner of planting: Plant as early as weather conditions make it possible to secure a good stand. In Florida Sea Island cotton should be planted by March 20 if possible. Do not cover seed more than an inch.
Fertilization: On fertile soils, capable of storing a good moisture supply, fairly large amounts of a balanced fertilizer. r>uch as a 1-8-4. .VT-"), or 4-10-C, may be used with profit. From four to six hundred pounds per acre at planting are recommended under these conditions. A side application of 0-0-12 or O-O-l"). applied about May 20, will usually be beneficial on such soils.
On poor sandy soil, a fertilizer high in potash should be used, such as a 4-10-6. This should be applied in moderation, say. 200 pounds per acre at planting. A side dressing of 150 or 200 pounds per acre of 0-0-12 or 0-9-15 should be made about May 20, in order to sustain the cotton in its active fruiting stage. If the above formulae are not available, a side dressing of 60 or 75 pounds per acre of nitrate of soda, or its equivalent, will usually be found profitable. ()n poor soils, however, it will perhaps be safer to apply nitrate of soda, or other quick acting nitrates, somewhat later in the season, to avoid forcing early plant growth too much, otherwise the cotton might be severely damaged by a dry period in July.
Boll Weevil Control
When the cotton plants reach the prc-square stage, apply a poisoned-syrup mixture, made as follows: Add \ l/> pounds of calcium arsenate to gallon of water, and stir well. Add to this, while stirring. 1 gallon of low grade cane syrupnot black strap molasses.
To apply the above mixture, make a "mop" by tying a piece of coarse cloth, such as burlap, about the size of a handkerchief.

to the end of a stick about 3 feet long. Wrap the cloth firmly around the end of the stick, with about 4 inches of the roll projecting beyond the end of the handle. Slit this projecting roll with a knife several times, and the mop is ready to use. Dip the mop in a pail of the poisoned-syrup mixture, carried in one hand, touch it to the side of the pail in withdrawing, in order to drain off some of the excess mixture, then, while walking steadily forward, hold the mop so that the wet portion drags through the tops of the small plants. Point the handle slgihtly forward, and do not oscillate it up and down. A few drops left on the under side of the bud leaves, and on the limbs and main stems of the plants, is all that is necessary. Rcmoisten the mop about every 30 feet. Apply about 1>4 gallons of the above mixture per acre on each application.
Apply this mixture in the afternoon, or dry part of the day. If three or four hours without rain elapse, wait one week and then make another application. Continue at weekly intervals until June 10 in pine woods areas, or June 20 in swampy areas. This will require four or five applications.
If a hard shower occurs within three hours after applying the poison, repeat the application as soon as the plants are thoroughly dry.
Boll Weevil Control on Sea Island
By GEORGE D. SMITH, WPA Entomologist
(May, 1938)
Sea Lsland cotton growers in weevil infested areas cannot expect to produce profitable yields unless an aggressive weevil control program is conducted. Weevils that live through the winter and deposit eggs to start the first weevil generation must be killed early by applying poisoncd-syrup mixture at weekly intervalsbeginning the first application the week before squares are developed on the plants. Poison applications

must be continued to the middle of June in the Florida Belt to get the late-emerging weevils. The following suggestions will prove helpful:
1. Apply I he poisoned-syrup mixture in the afternoons only while the cotton plants are dry and the weevil is thirsty. On mornings when the dew is light, poisoning can be started as early as 11 A.M.. but not before that hour.
2. .The poison should remain on the plants at least four hours without rain. Where the poison is washed off the plants by rain within two or three hours after its application it should be applied again as soon as the plants are thoroughly dry.
3. Poison applications should be made at weekly intervals to about June 12 to 15 on fields well away from woodland areas and to June IS to 20 on fields near swamps.
1. In mixing the poison the grower should keep in mind the following: fa) Properly mixed syrup mixture should not be too thin or watery as it will run off the plants rapidly and not absorb much moisture overnight. Tu the case of heavy or thick-bodied syrups, or molasses, mix a pound of poison in three quarts of water and stir thoroughly. Next add one gallon of syrup and stir. The mixture is now ready for use. (b) Where light-bodied table syrups are used use only two quarts of water to one pound of poison and one gallon of syrup. Experiments have proven that the weevil is fond of sweets but they arc not attracted to them, and that the weevil prefers a good grade to a low grade of syrup. This is the reason for using sweet syrup and black-strap molasses mixed in equal parts,
5. Mops arc made by tying a piece of burlap or fertilizer sack around the end of a stick about four feet long. A small persimmon sprout about one-fourth inch in diameter makes an excellent mop handle, Cut the fertilizer sack into strips about 6 inches wide and 10 inches long and wrap around the end of the mop handle allowing the sack to extend over the end of the handle about four inches. Next slit the part of the

sack extending over the handle several times with a knife blade. Always make small mops as only one drop of the poison mixture placed immediately under the bud of the plant is sufficient to kill all weevils. Mops wear out after two or three applications and should be replaced with new ones.
All mops, buckets and other materials used in the poisoning operation should be stored where livestock cannot get to them.
Four to six applications of the poison mixture will be necessary to kill all the overwintered weevils. The last two applications of the poison are about as important as the first applications. Don't stop poisoning because there is no evidence of weevils.
<>. In case a weevil escapes the poison operation and a few weevil-puncTured squares are found, mark the location and pick up and destroy all the punctured squares. Likewise, pick off any punctured squares observed on the plants. Next mop the plants around the infested spot for about 50 feet in all directions to kill the old weevil.
1. Thus far midsummer applications of poison for controlling the weevil have not proven successfulneither dusting or mopping. Under certain conditions some profit might result from midsummer poisoning but experimental results to date do not warrant the recommendation.
Mop should strike plant just beneath the bud. Don't try to mop budpush mop through top of plant and walk briskly. Only one drop of mixture needed to kill the weevil. Don't try to whitewash plant with poison mixture.
Cotton plant in pre-square stage is just right for beginning first poison application. Push mop through top of plant don't try to mop bud.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs