Vol. IVIII. Nc. 443. 



IHE AGRICULTURAL NEWS. 



126 



THE IMPROVEMENT OF THE YIELD OF 



SEA ISLAND COTTON IN THE WEST 



INDIES BY THE ISOLATION OF 



PURE STRAINS. 



In a paper contributed to the IFesi Indian Bulletin, 

 Vol. XVII, No. 3, Mr. S. C. Harland, B.Sc, Assistant for 

 ■Cotton Research on the staff of the Imperial Department of 

 Agriculture for the West Indies, points out the steps that 

 have already been taken in this direction, and describes some 

 what elaborate experiments undertaken by himself with a 

 view of increasing the yield of lint by still further isolation 

 of pure strains of cotton. Most of this paper is reproduced 

 below. 



Since the year 1903 when the tea Island cotton industry 

 was first started in the West Indies, attempts have been made 

 by the agricultural officers in the various islands to 

 obtain varieties suitable for their own special conditions, and 

 also to maintain the standard of quality of the lint at as high 

 a level as possible- The method followed by all workers 

 concerned was essentially the same in principle. Certain 

 plants were selected in the field for good vegetative characters. 

 After this the lint an:l seed characters were examined in the 

 laboratory, and finally the best were chosen for planting 

 in progeny rows. In the next year the best were again chosen, 

 and the same procedure followed, so that the final result has 

 been that in St. Vincent, St. Kitts, Monserrat, and Antigua, 

 the general appearance of the plants in field crop is now 

 remarkably uniform, and there is a notable freedom from the 

 inferior hybrids of Upland origin which were frequently seen 

 in the first year of the industry. In short, the effect of the 

 method followed has been not only to render the crop far more 

 uniform than when it was first grown, but ihe elimination of 

 inferior types has led to a marked improvement in the quality 

 of the lint — an improvement which is reflected in the high 

 prices which the spinners are willing to pay for it. Incidentally 

 each island has had its own problems to face. In St. 

 Vincent the yield has never been altogether satisfactory from 

 the point of view of the planter, and efforts have been made 

 to overcome the lowering effects of cryptogamic disease on 

 jield by the selection of higher yielding and resistant varieties. 

 In Montserrat a good deal has been done to improve the quality 

 of the cotton. Montserrat has been discriminated against in 

 the past on account of the relatively coarse nature of its 

 cotton, but Mr. Kobson has shown that cotton of very fine 

 quality can be grown in that island. In Antigua, be.sides 

 endeavouring to improve the quality of the lint, efforts have 

 been made by the Agricultural Department to produce a type 

 which will )ield well on the heavy soils of that island. 

 It will be seen that the system of selection so far 

 followed is the same in principle as the Vilmorin system of 

 pedigree selection, which is used extensively, and with great 

 successj,at the famous plant breeding station of Svalof in 

 Sweden, in connexion with the improvement of cereal crops. 



The weak points in the West Indian method have been 

 that until recently very little attempt has been made to 

 determine statistically how far the various strains were pure 

 for the character selected, and also that the frequent contami- 

 nation by cross-fertilization was not taken into account. 



The work of Dr. Lawrence Balls in Egypt in the 

 production of pure strains of Egyptian cotton has led West 

 Indian workers to realize that the accumulation of statistical 

 data in regard to the various characters studied, and also 

 ihe protection of the strains worked with from cross- 

 fertilization must form an essential part of the work of cotton 

 breeding 



At the outset we may state that the work of cotton 



improvement must always bear a dual aspect. On the one 



hand, the quality of the lint must be satisfactory to the 



i. spinner; on the other hand, the planter naturally desires to 



4 obtained the largest profit from his land. 



It will be seen that the yield of lint per acre depends on 

 a large number of factors, morphological and physiological. 

 It is by the interaction on these factors at their optimum 

 that the highest yield is built up. The failure of any one 

 factor to cooperate efficiently tends to lower the yield For 

 example, if the weather conditions are unfavourable — if there 

 is too much or too little rain — then the number of bolls per 

 plant is reduced by shedding, and the yield is immediately 

 lowered. So also, if the weight of lint per seed is zero, the 

 yield is zero. Bearing in mind that yield is, in the West 

 Indies, primarily dependent on weather conditions, it is a 

 matter of importance to make a thorough study of the mor- 

 phological and physiological characters which affect yield. 

 Weather conditions cannot be controlled, but at least an 

 endeavour can be made to produce a variety, theh ereditary 

 endowment of which will interact with environmental condi- 

 tions to the best advantage. The comparison of different 

 strains in respect of their yielding capacity must necessarily 

 prove an extremely difficult task. It is clear that the most 

 logical method of improving yield must be to determine, 

 firstly, what is the range of variation existing in Sea Island 

 cotton in respect of the morphological characters concerned; 

 secondly, to isolate strains with the highest maxima; and 

 thirdly, to produce strains po-ssessing the best combination of 

 morphological characters consistent with a high grade lint. 

 The conclusions of Balls in regard to the "compoaitio* 

 of Egyptian cotton are broadly true of Sea Island cotton 

 also, i.e. Sea Island cotton consists of an enormous mimber 

 of types which are different both physiologicilly and mor- 

 phologically. By self-fertilization and selection, strains pure 

 to any given character can l>e obtained, and it is certain that 

 the method of pedigree selection is capable of yielding rich 

 results in a comparatively short period, if the work can be 

 conducted on a large enough scale. 



After describing in detail his method of work, and the 

 results obtained from each of his numerous experiments, 

 which are very carefully and technically tabulated, Mr! 

 Harland summarizes his conclusions as follows:— 



' i . It is pointed out in this paper that the yield of 

 Sea Island cotton depends on a large number of factors both 

 morphological and physiological. Any scheme of selection 

 must aim at a type the hereditary endowment of which will 

 interact to the best advantage with environmental conditions. 



'2. The morphological characters of Sea Island cotton 

 which are concerned with yield are discussed. It is shown 

 that in respect of every one of these characters a marked 

 improvement can be effected by self-fertilization and selection. 



'3. The relative importance of lint index and lint 

 percentage, as factors in selection work, is dealt with in 

 some detail. 



'4. By following the methods of selection outlined in 

 this paper it has been possible to in?olate a strain of Sea 

 Island cotton with a weight of lint per boll 31 par cent, 

 greater than that of the ordinary type grown in the island.' 



The demand for both sisal and .sponge, the two 

 principal com modi tie.s exported from the Bahamas, has 

 been increasing steadily since the outbreak of the war. 

 The value of the former in 1917-18 was £179,195 as 

 against £47,219 in 1914-1,5, and of the latter £147,072 

 as against £9.3..395. (CotonialReports, Annual, No. 978.) 



