ftl f.f'T^ Issued December 1 i. 1909. U.S. DEPARTMENT OF AGRICULTURE, BUREAU OF PLANT [NDUSTKY Circular No 12 B. I . GA1 hief oi Bureau. OlilOIN OF THK HINDI cotton. (). K. COOK. WASMIHGTON : GOVERNMENT PRINTING OFFICE : 1*09 DEPOSITC I (Mr. 412 1 2 KUREAU OF PLANT INDUSTRY. Chief of Bureau, Beverly T. Galloway Assistant Chief of Bureau, Albert F. Woods. Editor, J. E. Rockwell. Chief Clerk, James K. Jones. I : I • l ORIGIN OF Till; HINDI COTTON. INTRODUCTION. Hindi is the name applied in Egypt to an undesirable type of cot- ton with a short, weak fiber, that injure- the high-grade Egyptian varieties by infesting them with hybrids. The skill and cheapness of the native Egyptian labor enable the exporters to have the cotton sorted by hand in their baling establishments, so that a high reputa- tion for uniformity has been secured in spite of the Hindi admixture. The introduction of the Egyptian cotton into the United States brings also the problem of the Hindi cotton, but without the resource of cheap labor which enables the difficulty to be surmounted in Egypt. The practicability of establishing a commercial culture of the Egyp- tian cotton in the United States depends largely upon the elimination of the Hindi contamination and other forms of diversity, so that the fiber may be produced in a satisfactory condition of uniformity. The Hindi cotton problem mighl be compared to thai of the red rice that mixes with the white and depreciates the value of the crop. In the case of the cotton there is a better prospect that adequate knowledge of the vegetative characters may enable the undesirable plant- to he removed from the fields without too seriously increasing the cost of production. DIAGNOSTIC CHARACTERS OF THE HINDI COTTON. The Hindi cotton usually appear- more vigorous and robust than the adjacent Egyptian plant- by reason of the larger number of vege- tative branches developed from the lower node- of the central -talk. The vegetative branches also take a more nearly uprighl position, rendering the plants more compact and bushj in their general shape, a- well a- more densely leafy. The leaves are much thinner in tex- ture than those of the Egyptian cotton and of a lighter and more yel- lowish green. The difference is particularly striking in Arizona, w here the Eg} ptian cotton usually i- id' a very dark grayish or bluish green. The lateral lobes appear very shorl and broad in comparison with the Egyptian cotton, or even with many of our Upland varie- [Clr. 42] 3 ORIGIN OF THE HINDI COTTON. ties. The lateral angles of the leaf are produced so little that the outer margin is left nearly straight if the middle lobe is cut off. (See fig. 1 and compare with fig. 2.) The pulvinus at the base of the leaf blade is red, as well as the adjacent part of the petiole, and es- pecially the somewhat swollen upper side of the end of the petiole, w hich may be looked upon as a part of the pulvinus. The involucral bracts are nearly orbicular, very deeply cordate at base and mar- gined with numerous long teeth. The calyx has long-pointed trian- Fig. 1. — Leaf of Jannoviteh Egyptian cotton (natural size). gular lobes. The petals are creamy white and the petal spot faint or entirely lacking. The small conic bolls have three, four, or five carpels or locks, and are of a pale-green color, with few and deeply buried oil glands. The lint is white and of very inferior quality. The seeds are longer and more angular than in the Egyp- tian cotton, and the surface is usually completely naked after the lint is removed. In rare cases there may be fuzz at the ends of the Mills, as in the Egyptian cotton, or even a larger amount. [Cir. -l-j ORIGIN OF CHE HIND] CO! ln\. SUPPOSED RELATION OF HINDI COTTON TO UNITED STATES UPLAND VARIETIES. Tl "' nature and origin of the Hindi cotton appear to have been the subject of as i ■!. popular speculation in Egypt as the red rice in the United States. The word " Hindi " is the Arabic equivalent of '""' word " fndian." Seme writers have taken this to mean that the cotton came fr om Hindustan, while others consider that the nam.' 1 ! Hindi cotton I natural size). Hindi might be applied to any foreign plant and has no particular significance as an indication of origin. A third opinion is that this cotton is either a native Egyptian variety or one that was cultivated m the country before the present commercial type. The reason given for I,M ~ idea is that this cotton is frequently found in a wild or spon- taneous condition in uncultivated or abandoned land-. [Cii b ORIGIN OF THE HINDI COTTON. The suggestions of scientific students of the Hindi cotton arc hardly more consistent. Sir George Watt's monograph of cotton connects the Hindi plant with no less than three species supposed to be native in different parts of the world, but he refers it most directly to Gossypium punctatum, and states that this species grows wild in the United State-. Some of our cult hated Upland cottons, such as the King variety, are reckoned as varieties or hybrids of Gossypium punctatum, and the Moqui cotton of the Arizona Indians is definitely referred to this species." In reality there is no wild cotton in any of the cotton-growing re- gions of the United States. In Texas and other Gulf States warm winters often allow the roots to survive and send up new shoots in the spring, but in cold years all the cotton is killed throughout the cotton belt. The only indigenous wild type of cotton known in the United States is that found in the extreme southern part of Florida and on the Florida Keys, unless we take into account the varieties cultivated by the Indians of Arizona, and these varieties have never* been planted in other part- of the United State- except in very re- cent experiments. Watt dwells in particular upon the claim that the Hindi cotton re- sembles Moqui cotton from Arizona; but when the living plants are compared, the resemblance between the Moqui and Hindi cottons ap- pears no greater than that between the Hindi and our Upland va- rieties. The Hindi cotton finds a much closer alliance with other types of cotton from southern Mexico and Central America. These types belong to the general Upland series, but they have not been known in the United States until very recently and have been planted thus far only in a few localities and only on an experimental basis. HINDI COTTON RELATED TO MEXICAN VARIETIES. The vegetative characters of the Hindi cotton show the closesl approximation to those of some of the Mexican varieties from the State of Chiapas and in particular to a type obtained by Mr. (i. N. Collins in 1906 at the town of Acala. There are the same light, yel- lowish green, broad, short-lobed, smooth, naked Leaves and the, same strongly zigzag fruiting branches which frequently branch again from the axillary hud-. A- in the Hindi cotton, the bolls are pale green, the oil glands that show a- black dots on the bolls of Egyptian cotton being buried deeply in the green tissues. The involucral bracts are rounded ami very deeply cordate at base, as in the Hindi cotton, and the margins have longer and coarser teeth, carried down nearer "Watt, si. George. The Wild and Cultivated Cotton Plants <>f the World, i... n, in,,, mil?. ],. 181. [Cir. 42] ORIGIN OF CHE HINDI COTTON. 7 to the base than in our Upland cottons. The calyx of the Hindi col ton has large triangular lobes, and these are often produced into a long, -lender tip, as in many Mexican and Central American van including thai from Acala. Many of the plants of the Acala cotton growing ai San Antonio in August, L909, were remarkably close counterparts of some of the Hindi plants of the Jannovitch row in the same Held. The chief difference lies in the greater fertility of the Mexican cottons, some of which appear worthy of cultivation in the United States, since they have lamer bolls and better lint than our United States Upland varieties. The Hindi cotton is markedly infertile or fruits very late, but this fact may be connected with its status as a reversion. Muta- tive variations, like hybrids, arc often more or less completely sterile. The Egypt ian and the I Ipland type- both have definitely specialized fruiting branches, bul the fruiting branches of the Hindi cotton -how a much greater tendency to keep an ascending position and con tinue their vegetative growth, the young flower buds being often aborted. The same tendency is often seen in aberrant plants of Egyptian cotton, including many that show Hindi characteristics. The fruiting branches of the Hindi hybrids are usually few and short and some of the Hindi like plants are completely sterile, as already stated. This is in notable contrast with the behavior of the hybrids between the Egyptian and Upland cotton, which have the fruiting branches better developed than in the pure Egyptian -locks. COTTON INDIGENOUS IN AMERICA. The resemblance between the Mexican and the Hindi cotton from Egypt may no1 appeal- to be a sufficient proof of the American origin of the Hindi cotton. It might be thought more likely that cotton had been carried from Egypt to Mexico than from Mexico to Egypt. Account must be taken of the further fact that Mexican and Central American varieties are members of a large natural group. The numerous local types are appreciably different and yet they have so many characters in common thai the whole group must be looked upon as an indigenous product instead of a recent importation. The long, narrowly attenuate lobes thai render the Hindi calyx SO widely different from the Egyptian is a feature commonly accentuated in many of the Mexican and Central American type-, though very rarely found in our United States Upland varieties. How the Hindi cotton was introduced into Egypt i- likely to remain a matter of conjecture, for the history of the Egyptian cotton itself is altogether obscure. That it came to Egypt from India is not to be considered impossible, for in India, as in Egypt, large numbers of [Cir. 42] 8 ORIGIN OF THE HINDI COTTON. varieties have been imported at different times for experimental pur- poses. Some American cottons appear to have been cultivated in India for a long time, perhaps dating back to early Portuguese intro- ductions from Brazil. All that can be said at present is that none of the cottons from India that have been grown in the United States show any close approximation to the Hindi cotton. The idea of the Hindi cotton as a wild plant in Egypt may have been strengthened, if not suggested in the first place, by the fact that Egyptian cotton stunted by dry soil or other unfavorable conditions shows a stronger resemblance to the Hindi. The first leaves of the Egyptian cotton have nearly the same shape and color as the adult leaves of the Hindi, and stunted plants continue to produce the juvenile form of leaves. The proportions of adult Hindi plants also appear to be influenced by the external conditions in different plantings of the same stock of seeds. It does not seem unreasonable to suppose that Egyptian cotton escaped from cultivation might go over more and more to the Hindi type. A further reason for considering the Hindi cotton as a collateral relative of the Egyptian, if not a truly ancestral form, may be found in the fact that many hybrids between the Egyptian cotton and United States Upland varieties show Hindi characteristics rather than those of the parental types. The fact that the affinities of the Hindi cotton have been so long misjudged would tend to show that Indian and Egyptian students of cotton have not been familiar with the Mexican and Central Ameri- can types. It is possible' that the Hindi contamination already existed in the Egyptian cotton when it was introduced into Egypt and that its existence in that country resulted from reversion rather than from local contamination. The Sea Island cotton of the United States, that has never been in Egypt, also shows sudden variations, the so-called " male stalks " or " bull cotton, " commonly reckoned as hybrids, but having a general similarity to the Hindi reversions of the Egyptian cotton and the same tendency to sterility and inferior fiber. RELATIONSHIPS OF EGYPTIAN COTTON. There are also many indigenous varieties of the general Sea Island type of cotton in the American Tropics, and often in the same locali- ties with indigenous Upland varieties, so that opportunities for crosses may have existed through long periods of time. Some of the Mexican and Central American varieties of the Upland series share the long-pointed bolls and some of the other characters of the Sea "Orton, W. A. Sea Island Cotton: Its Culture. Improvement, and Diseases. Farmers' Bulletin 302, U. S. Dept. of Agriculture, 1907, p. 20. [Cir. 42] ORIGIN OF THE HINDI COTTON. 9 [sland series, and i( is nol impossible thai a complete series of inter- mediate types maj yel be discovered in tropical America. W : 1 1 1 " — recenl assignment of the Egyptian cotton to another botan ical species (G< wypium />< ruvianum ) instead of to the Sea [sland spe- cies i <•'. barbad( ns< i should not be allowed to confuse the issue, for the two types 'I" not appear to have any essential differences to justify such a separation. The range of diversity shown by the Egyptian col ion- during the period of acclimatization leaves no doubt that thej are closely allied to the Sea [sland cotton. There are individual Egyptian plant-, with lighter color and narrower lobes than usual, that simulate the Sea [sland cotton very closely, without any serious departure from the usual Egyptian characteristics. The most pro- nounced differences that sometimes appeal' to separate the two types are the ilar!, ei- green color of the Egyptian foliage and the smaller tendency of the Egyptian cotton to produce fertile branches on the lower pari of the plant. Both these characters are known to be easily influenced by external condition- and individual selection, as in the Upland t \ pes of cotton. A planting of Sea [sland cotton at Falfurrias, Tex., in the season of 1909 showed several plants strikingly similar to Egyptian cotton. much taller and less fertile than their neighbors, and with the coarser, darker foliage and the relatively short buff lint of the Egyptian — ap- parently complete reversions from the Sea Island to the Egyptian type. Indeed, the approximation was in this instance so close as to call for repetition- of the experiment to exclude every possibility of admixture of seed. The same stock of Sea Island seed handled in the same way at New Braunfels, Tex., produced none of the Egyp- tian-like plant-, but niaii\ similar cases have occurred where diversi- ties have appeared in some places and not in others. Darker lint accompanies darker foliage among the Egyptian plant- as well as among the Sea [sland. The two series undoubtedly overlap, whether they are capable of -how Lag the same extremes or not. The question of the botanical name that should he applied to the Hindi cotton may well he left open until more definite knowledge is available regarding the botanical identity of other Mexican types. The Hindi cotton may prove to he close to the original of Todaro's C ypium mexicanum, hut may also be distinct, if Watt i- correct in referring our big-boll Upland varieties to that specie-. Todaro's 6 ypium microcarpum is another Mexican species to he considered in the identification of the Hindi cotton, tor some of the Mexican relatives of the Hindi cotton show narrowdeaved form- that may have furnished the originals of Todaro's species, though they have no apparent relation to some of the varied types that W ;i 1 1 assembles under t hi- name. [Clr. A-i\ 10 ORIGIN OF THE HINDI COTTON. POSSIBILITIES OF UPLAND ADMIXTURE IN EGYPT. That some of the so-called " Hindi contamination " in Egypt may be due to hybridization with true United States Upland cottons is not to be denied, for it is probable that many experimental plantings of Upland cotton have been made in Egypt, affording opportunities for crossing to take place. Recent reports indicate that some of the Egyptian planters are adopting the Upland cotton as a regular crop, owing to a serious decline in the }deld of the Egyptian cotton in l he last few years. Indications of a previous contamination with Upland cotton appear in the Ashmuni variety of Egyptian cotton as grown at Yuma in 1909 from newly imported seed. The Ashmuni field showed numerous Hindi plants different from those that appeared in other varieties in being distinctly hairy. In addition to the hairy Hindi plants there were several small hairy individuals that lacked other distinctive Hindi characters, such as the light-colored, short- lobed leaves, and approached in these respects some of the forms of Upland cotton. The hairy Hindi plants might also be taken to indi- cate Upland hybridization, in view of the strong tendency of the Hindi characters to come to expression in Egyptian-Upland hybrids. These hybrid reversions sometimes take on the complete Hindi form and show very few or none of the Egyptian or Upland characters. CONCLUSIONS. Experiments with Egyptian cotton in Arizona show that the so- called " Hindi " variations which appear among plants grown from seed imported from Egypt are one of the principal factors of the diversity that would diminish the commercial value of the fiber. Comparisons with other types indicate that the Hindi cotton is of American origin instead of a result of hybridization with a native Egyptian or other Old World species of cotton as various writers have assumed. On the other hand, the Hindi cotton does not prove to be identical with any of our United States Upland varieties, as supposed by Watt. It finds a much closer alliance with other types of Upland cotton indigenous in Mexico and Central America. As the Egyptian and other Sea Island types also appear to have originated in tropical America, it becomes possible to view the Hindi variants as examples of reversion to remote ancestral characters rather than as results of recent hybridization. The similarity of the Hindi foliage to that of young plants of Egyptian cotton accords with this interpretation. [Cir. 42] ORIGIN OF THE HINDI COTTON. 11 Although reversion to Hindi characters frequently occurs when the Egyptian cotton is hybridized with United State- Upland vari- eties there are also many Upland characters thai seldom or never appear among the Hindi reversions and thus enable recent contami- nation with Upland cotton to be detected. Approved : James Wn son, , v of . I iii icultun . Washington, D. ('.. October 19, H Nun. After this circular was written, the Library of the Department "f Agriculture acquired a sel of the files of the Cairo Scientific Journal, a recently ablished publication not hitherto accessible in Washington. Two papers touching upon the origin of the Hindi cotton and containing many interesting historical fads appeared in this journal in 190S, both by scientific investigators resident in Egypt. The first paper, written by Mr. \Y. Lawrence Halls for the July Dumber, inclines to the current idea thai the Hindi cotton is a native of Egypl and adjacent regions, though adducing no direct evidence. The second paper, in the November number, is by Mr. F. Fletcher, who had previously lived in India and investigated the Indian cottons. The Hindi cotton is said not to be grown in India at the present day. but Fletcher states that "it is cultivated near Bagdad undei this same title and is supposed to have been introduced there from India, as its name suggests." No consideration is given to the idea of the Hindi cotton as a native of Egypt, Watt's view of its relations i" Gossypium punctalum ami American Upland cottons being apparently accepted. The possibility of a Central African origin of the Hindi cotton is noted, on the basis of a Hindi-like herbarium specimen dating from 1863 labeled as representing a cotton introduced into Egypl from Cordofan. Fletcher .old- thai he has "received many sample- of seed from Central Africa, bul none of these have given rise to Hindi plant-." still older specimens from Upper Egypl and Abyssinia, described by early authors under the name frutescens and considered by Balls as possibly per- taining to Hindi, are shown by Fletcher to be true old World types, not related to the Hindi cotton or to the Egyptian. Halls also refers to Gossypium riti- folium as a Central African cotton with " free, naked seeds." Fletcher •' not look upon (',. vitifolium as reli 1 to the Hindi cotton, but accepts it as one of the ancestors of the Egyptian, the Sea Island as the other. Halls finds that a varietv 'if Sea Island cotton has been cultivated at Kanda. in the Menufiyeh district, for thirty years, which may explain the tendency of the Egyptian cotton to vary in the direction of the Sea Island. Fletcher also studied at Paris Lamarck's original type of vitifolium, sup- pi. -ed to come from Celebes, though the locality is doubtful. lie conclude- that an Egyptian specimen referred to Lamarck's species by Delile over a century ago was correctly identified, and gives photographs of the original specimens, which are not altogether favorable to his conclusions, it can lie seen that the involucral bracts of Lamarck's plant were of distinctly tut- ptian form, the teeth being coarse and long and extending far down toward the base of the bracts, as in the Hindi cotton. Fletcher also considers that the Delile plant agrees with a specimen of "Jumel" cotton sent from Egypl to [Clr. 42] 12 ORIGIN OF THE HINDI COTTON. Todaro about I860 with a statement that it had been introduced from Ceylon about forty years before. Historical accounts collected by Balls indicate thai the field culture of long- staple cotton in Egypt was begun by Mohammed Ali in 1*21 at the instance of Jumel, a French engineer. The superior type adopted by Jumel was not a new introduction, but a perennial " tree" cotton that was being planted as an orna- mental in gardens at Cairo, and supposed to come from India. Several direct introductions of Sea Island and Brazilian cotton appear to have been made subsequently, but without displacing the variety that had been popularized by Jumel. Balls is inclined to ascribe the brownish color of the Egyptian cotton to these Brazilian introductions, but Fletcher believes that Jumel's cotton was brown, like some of the Brazilian cottons. If the Egyptian cotton came by way of India the name Hindi that is now given to inferior plants may be only an echo of the original introduction of the Egyptian cotton itself. Any cotton brought from India might be called Hindi at first, and this name would serve in later years for the residual stock, after local varieties with special names began to be distinguished. Balls shows that there were numerous varieties of Egyptian cotton with distinctive names before the Mit AAA type was introduced in 1S82. After the use of the improved types became general the old name might still be applied to inferior variations or even to accidental hybrids. The origin of the name appears to have no bearing in this case upon the origin of the plant. Local varieties of cotton might have been taken to India from any part of tropical America, though more likely to have come from Brazil, where the Portuguese ships were accustomed to stop on their way arouud the Cape of Good Hope. [Cir. 42] O === < ^=OJ Q = h» cr 1/) O ■^ffl U- — ^^ M 00 LL CM o o> > K w — — CN cr — Z ^^^^ ^^^» , ____