IMPROVEMENT OF SUBTROPICAL FRUIT CROPS:
CITRUS

HAMILTON P. TRAUB, Senior Horticulturist
T. RALPH ROBINSON, Senior Physiologist Division of Fruit and Vegetable Crops and Diseases, Bureau of Plant Industry

[ABSTRACT]         OUTSTANDING in the story of citrus fruit improvement was the work of A. D. Shamel, of the United States Department of Agriculture, and his coworkers, in studying bud mutations. In the past 18 years, probably 10 million buds of superior strains of the Washington Navel orange, the Valencia orange, the Marsh grapefruit, the Eureka lemon, the Lisbon lemon, and miscellaneous citrus varieties have been sold to California growers alone as a result of this work. Two special strains produced fairly recently— the Robertson Navel orange and the Dawn grapefruit—are now being widely distributed and seem to have great promise. In addition, the intensive study of bud mutations, backed by careful statistics, was important in teaching growers to keep a close watch for branches mutating toward poor types, so that they could be eliminated from the orchard.

MORE than half of the 13 fruit crops known to have been cultivated longer than 4,000 years, according to the researches of DeCandolle (7), are tropical and subtropical fruits—mango, olive, fig, date, banana, jujube, and pomegranate. The citrus fruits as a group, the lychee, and the persimmon have been cultivated for thousands of years in the Orient; the avocado and papaya were important food crops in the American Tropics and subtropics long before the discovery of the New World. Other types, such as the pineapple, granadilla, cherimoya, jaboticaba, etc., are of more recent introduction, and some of these have not received the attention of the plant breeder to any appreciable extent.

Through the centuries preceding recorded history and up to recent times, progress in the improvement of most subtropical fruits was accomplished by the trial-error method, which is crude and usually expensive if measured by modern standards. With the general acceptance of the Mendelian principles of heredity—unit characters, dominance, and segregation—early in the twentieth century a starting point was provided for the development of a truly modern science of genetics.

In this article it is the purpose to consider how subtropical citrus fruit crops have been improved, are now being improved, or are likely to be improved by scientific breeding. Each of the more important crops will be considered more or less in detail. Before proceeding to these considerations, however, it is desirable to define the province of subtropical fruit culture and to take a glance at the economic importance of the subtropical fruit industry.

The region where subtropical fruits are produced, as the name indicates, is between the true Tropics, where frost never occurs, and the temperate region, where normally the temperature often falls below freezing and stays below for a considerable part of the winter season. In this intermediate region the temperature occasionally goes below freezing but not as a rule below 25° F., so that when necessary the trees can be economically protected by artificial means. Because of the influence of large bodies of water, the protection of mountain ranges, or planting where the topography gives good air drainage, this type of region may be extended as ‘islands’ considerably beyond the usual subtropical region.

The types of fruit crops grown merge into those of the true Tropics— citrus, avocado, mango, etc.—and no hard and fast division can be drawn on the basis of fruit types except that forms possessing resistance to low temperature are of major importance in the subtropics.  Diverse types are cultivated, of which the familiar citrus fruits are among the most outstanding, followed by the pineapple, fig, olive, avocado, date, persimmon, mango, papaya, guava, pomegranate, lychee, granadilla, cherimoya, loquat, jujube, and other minor types.

In the United States some of the crops, notably citrus and avocado, have become staple dessert and salad fruits. Others, notably the date and the fig, are used primarily as confections. The olive is used in preserved form or for oil. Some of these fruits were recognized as important sources of indispensable vitamins even before the true function of these chemical regulators was fully understood. Limes, for instance, have long been included by the British as a regular part of the diet of seamen as a preventive of scurvy. During the recent Ethiopian campaign, the entire Italian export crop of lemons was reserved for the army of invasion, and it is reported that deficiency diseases were at a minimum. Besides the citrus group, pineapples, papayas, dates, avocados, mangoes, and other subtropical fruits are known to be unusually high in vitamins.

Some of these fruits, for example the mango and the papaya, are extensively cultivated but primarily for local consumption. Higgins and Holt (30, p. 17) remark: “Excepting the banana, there is no fruit grown in the Hawaiian Islands that means more to the people of this Territory than the papaya, if measured in terms of the comfort and enjoyment furnished the people.” This applies to the papaya in other tropical countries as well; to the avocado in Central America and the West Indies; and to the mango in India, southeastern Asia, Malaya, Puerto Rico, and the West Indies in general.

The subtropical fruit production regions in the continental United States are indicated in figure 1. The annual farm value (1934-35) of the chief subtropical fruit crops grown in the United States for which figures are available—citrus, fig, olive, avocado, and date— amounts to approximately $140,000,000, distributed as follows:
Citrus fruits, $135,000,000; fig, $1,706,000; olive, $1,260,000; avocado, $959,000; date, $390,000.

Figure 1.—Citrus-producing areas of the United States. Solid black denotes areas of commercial production of sweet and tangerine oranges, grapefruit, or acid citrus fruits (lemons or limes). Shading denotes areas producing satsuma oranges or minor quantities of other citrus fruits. The dotted boundary lines indicate the approximate northern limits of the three subtropical crop regions: (1) Southeastern humid, (2) central irrigated, and (3) southwestern irrigated. The distribution of the total return for citrus fruits in the United States in 1934 is given in table 1.

The citrus fruits as a class are native to southeastern Asia—eastern India, Indo-China, southern China, the Philippines—and here they were first brought under cultivation. A monograph on the oranges of Wenchow, Chekiang, Nan Yen-Chih’s Chu Lu, composed in China during the period 1174 to 1189, is the earliest treatise on citrus culture extant (26). Even at this early date three horticultural groups of oranges were recognized and the total number of varieties listed was 27.

Although there are a dozen or more types of citrus fruits, only five or six are of major importance from a commercial standpoint. The most important of these grown for fruits are the sweet orange, Citrus sinensis (L.) Osbeck; the grapefruit, C. grandis (L.) Osbeck?; the acid: citrus fruits, including the lemon, C. limonia Osbeck, and the lime, C. aurantifolia (Christm.) Swingle; and the mandarin orange group, including the tangerine orange, C. nobilis var. deliciosa Swingle, and the satsuma orange, C. nobilis var. unshiu Swingle. Other types that are of importance mainly as rootstocks are the sour or Seville orange, C. aurantium L.; the rough lemon, C. limonia Osbeck; and the trifoliate orange, Poncirus trifoliata (L.) Raf. In some foreign countries the sour orange, the pummelo (C. grandis),¹ and the citron (C. medica L.) are of relatively greater commercial importance than in the United States.

From the Orient the various types and varieties spread to other parts of the world along the trade routes. The citron reached the Mediterranean region at an early date, as it is mentioned by Theophrastus. The sweet orange was apparently not introduced into Europe until the early fifteenth century. The sour orange reached Spain by way of northern Africa. The lemon and the lime were apparently introduced into Europe about the same time as the sweet orange, and several varieties are described by Ferrarius and other writers. Lemon culture first became important in Sicily, Corsica, Genoa, and other parts of southern Europe.

On his second voyage to the New World in 1493 Columbus stopped off for 2 days early in October at Gomera, Grand Canary Island, where he purchased livestock and fruit and vegetable seeds, among which were "seeds of oranges, lemons, and citrons.” He reached the island of Hispaniola on November 22, 1493, and in the course of establishing a colony he “set out orchards, planted gardens” (40). The historical records of introduction to other parts of the Americas have not been exhaustively searched, but citrus fruits were established at St. Augustine, eastern Florida, by 1579 (12) and in Peru before 1591 (82). They were introduced into southern California in 1769 by Franciscan monks at San Diego (43), and there were undoubtedly many similar introductions into Brazil, Mexico, and other regions settled by the Spanish. In Florida the sour orange, and to a lesser extent the sweet orange and the lime, escaped to the wild.

It is now known, however, that the so-called “wild lime groves” on the lower east coast keys of Florida were in fact planted out by Henry Perrine, to whom in 1838 Congress had granted a tract of land 6 miles square on Biscayne Bay for the establishment of economic tropical plants. Before his death in the Indian Key massacre of 1840 he had established a nursery of upward of 200 species and selected varieties of useful tropical plants (63).

Commercial citrus growing in the United States presents striking contrasts, due primarily to climatic conditions in the main centers of production (fig. 1). In Florida at an early date the sweet orange, the tangerine orange, and the grapefruit found a congenial home; in California the sweet orange and the lemon have proved the basis for profitable industries, with grapefruit secondary in importance. Texas and Arizona have more recently come into the picture, especially for grapefruit production.

Differences in varieties and seasons of maturity are likewise in sharp contrast. In California two varieties of oranges, the Washington Navel and the Valencia, furnish fruit maturing from November to November, a year-round shipping season. In Florida three or four sweet orange varieties, together with seedlings, are generally required to give a shipping season from October to May. In California lemons are more or less everbearing, affording a supply throughout the year. The Florida and Texas grapefruit crops mature practically during the same season (the fall and winter months), the California and Arizona crops coming in somewhat later. Similar contrasts are to be noted with the rootstocks used in these regions. These and other regional contrasts will be discussed more in detail later.

Thus it will be seen that because of differences in climate and in variety adaptation, citrus fruits, with their highly important vitamins, are available to the American consumer throughout the entire year. In general it may be said that grapefruit has been America’s chief contribution to citrus culture, its recognition in Florida as an appetizing breakfast fruit gradually changing this curiosity of the citrus family into a formidable rival of the sweet orange in the national dietary.

In the southeastern humid subtropical crops region, citrus development began in eastern Florida in the vicinity of St. Augustine and along the Indian River and in north-central Florida in the general vicinity of Palatka and Ocala as far south as Lake Monroe. On the west coast of Florida the development took place in the vicinity of Tampa Bay and southward. The outstanding pioneers in the introduction of citrus varieties during this period (1870-95) were E. H. Hart (fig. 2), H. S. Sanford, and Lyman Phelps. By the 1880’s the industry in northeastern Florida was fairly important, but in the winter of 1894— 95 it was practically wiped out by two severe freezes, and the center of the industry was moved farther south to the central ridge section and the southern coastal areas, where most of the citrus growing is now located. Today the industry is based primarily on the sweet orange, grapefruit, tangerine, and lime.

On the upper Gulf coast a citrus industry was established in the Delta district south of New Orleans, based primarily on the sweet orange. Satsuma growing along the St. Johns River and near Jacksonville began about 1900, spreading thence westward —because of the cold resistance shown by this type during the 1894-95 and 1899 freezes—to the Gulf coast region in western Florida, Alabama, Mississippi, Louisiana, and Texas.

The early plantings of sweet oranges made in various parts of Florida consisted primarily of groves established from seedlings, and it was only later that the practice of budding improved varieties was gradually adopted with the introduction of meritorious early, midseason, and late sorts, beginning in the late 1870's. Since the citrus tree is relatively long-lived, the seedling groves, producing fruit mostly midseason in maturity, are still an important factor in Florida, and they produce approximately 30 to 40 percent of the State’s total midseason sweet orange crop. However, budded varieties of recognized merit have been used almost entirely in new plantings and replacements for the last 40 years.

Figure 2.—Edmund H. Hart (1839-98), recognized as a skilled horticulturist and a pioneer in Florida citrus culture. His name is chiefly associated with the Hart’s Late or Tardiff orange, now called Valencia, which he first brought to fruiting and introduced into general use in Florida during the 1870's.

Grapefruit first attained commercial importance in the United States. This was between 1880 and 1885, when the first grapefruit were shipped from Florida to the Philadelphia and New York markets. In Florida the industry received a set-back on account of the freeze of 1894-95, but it gradually expanded again, reaching a peak by 1929.

The lemon and the lime, which were classed together in early times, were introduced into the New World by the early Spanish explorers and settlers. The everbearing and rough lemons were among the early introductions into Florida, and the latter had escaped to the wild by the time permanent settlements became common.

Prior to the great freeze of 1894-95 the lemon industry of Florida was of considerable commercial importance. During the year previous to the freeze the annual shipments amounted to 140,000 boxes of lemons. Up to the present the industry has not been rebuilt, but recently attempts have been made to reestablish it.  Limes are grown in southern Florida, chiefly on the keys, in Dade county, and in the south-central ridge section.

The mandarin oranges include the King, tangerine, and satsuma types. The tangerine was introduced into Europe from the Orient during the first half of the nineteenth century and was produced on commercial scale in Italy as early as the 1840’s.  It was introduced into Louisiana between 1840 and 1850 and later was brought to Florida, where it is grown as a fancy fruit to a greater extent than in California or Texas.  Another member of this group, the satsuma, is outstanding in being the most frost-hardy, of all the larger fruited citrus types.  It is grown primarily in the upper Gulf coast region, with the chief center of production in western Florida and southern Alabama and Mississippi.

In the southwestern irrigated subtropical crops region a citrus industry has been established in California, primarily in the southern coast and interior valley sections. In the southern coast section the industry is based almost entirely on sweet orange and lemon, and in the interior valley section on sweet orange and grapefruit. An extension into the irrigated section of Arizona occurred later, founded primarily on the grapefruit.

Citrus seeds were disseminated from other parts of Mexico to the peninsula of Baja (Lower) California probably in the early 1700’s, and later, in 1769, were introduced to Alta (Upper) California by Franciscan missionaries, who established a chain of missions for 400 miles along the coast.  According to Coit (11), the early settlers secured orange trees from the missions, and a number of small plantings were found in private gardens in the 1830’s and 1840’s in the vicinity of Los Angeles. These early plantings stimulated further interest, and in 1857, trees were planted at old San Bernardino and Highlands, and in 1865 at Crofton, and in 1871 on land now occupied by the city of Riverside.  T.A. Garey (fig. 3), of Los Angeles, the outstanding nurseryman of that time, imported large numbers of important varieties during the period 1868 to 1875. His introductions were apparently from Florida, Australia, and southern Europe, as well as from Ellwanger & Barry, of Rochester, N.Y., and Sir Thomas Rivers, of Sawbridgeworth, England. In the southern and central parts of California the industry was only of local importance until after the completion of the Southern Pacific Railroad in 1876, when the first carload of fruit was shipped to St. Louis, Mo., arriving in good condition after a month in transit.

The grapefruit industry in the Southwest—California and Arizona— began after the introduction of the Marsh variety in 1890; the plantings of other varieties previous to that time did not prove profitable.

Great success in lemon culture has been achieved in California, particularly in the southern coastal region, which produces a large proportion of high-priced summer fruit.

In the lower Rio Grande Valley of Texas, in the central irrigated subtropical crops region, an industry has been established based primarily on the grapefruit and to a lesser extent on the sweet orange.

As early as the middle of the last century scattered plantings of a few citrus trees could be found along the Texas Gulf coast, particularly from Victoria County southward to Brownsville on the Rio Grande. Experience had demonstrated by the beginning of the twentieth century that regular citrus crops could not be expected in the region above the lower Rio Grande Valley, on account of damage from low temperature (48). In the lower Rio Grande Valley—Cameron, Hidalgo, and Willacy Counties—the development of the industry may be grouped into three periods. (1) Up to 1899 citrus fruits were grown for home use and no particular attention was given to the subject of rootstocks. Seedlings and budded trees were planted. During this period, in 1869, the trifoliate orange was imported from northern China by William Saunders, of the United States Department of Agriculture, and it was used to some extent as a rootstock. (2) The resistance to frost shown by trees budded on trifoliate stock during the severe freeze of 1899 led to the extensive use of this stock during the following decade (87). Although the stock was valuable from the standpoint of frost resistance and encouraged the pioneer growers to make further attempts, it had the serious defects of dwarfing the scion variety and of being itself subject to foot rot and cotton root rot (77). (3) The observations of some of the outstanding pioneers, Charles J. Volz, Harry Banker, J. R. Robertson, F. E. Elliot, Max Melck, and A. P. Wright, beginning after 1900, made possible the growing of citrus fruits in commercial quantities.

Figure 4.—Charles J. Volz, pioneer Texas citrus grower since 1908, who demonstrated the value of sour-orange rootstock for citrus in the lower Rio Grande Valley of Texas. This demonstration proved to be a turning point in the development of the citrus industry in that section.

Charles J. Volz (fig. 4), from Indiana, settled near Mission in Hidalgo County, Tex., in 1905. He began the planting of citrus in 1908 and clearly demonstrated the superiority of the sour-orange rootstock under the lower Rio Grande Valley conditions. Harry Banker, from Oklahoma, who settled near Brownsville, in Cameron County, secured similar results with the sour-orange rootstock beginning in 1910 (87). With the solution of the rootstock problem the industry showed at first gradual and later rapid development. The citrus shipments from this section, consisting mainly of grapefruit, had reached 13 carlots in 1921 and increased to more than 5,000 carlots annually by 1931. Yields were cut down as a result of hurricane damage in 1933 and damage due to low temperature in 1933-34 and 1934-35, but the crop of 1936-37 reached a total of over 15,000 carlots. Fortunately for this new industry, the seedless type of grapefruit (Marsh variety and its pink-fleshed mutations) has been most heavily planted in Texas, and this has doubtless contributed to its favorable reception in many markets.

The sweet orange was introduced into Hawaii in 1792, and many citrus varieties have been cultivated there for over 100 years (54). The climate is well adapted to citrus culture, but for commercial production the situation is complicated by the presence of the Mediterranean fruit fly, which limits production to local needs.

While citrus fruits, especially the sweet orange, have been grown in a semiwild condition in Puerto Rico for perhaps 3 centuries, commercial planting did not begin until about 1900. The first plantings were largely of Florida orange varieties, but these were soon largely discarded in favor of grapefruit (29). The varieties most commonly grown are the Duncan and the Marsh. For budding stocks, rough lemon, sour orange, and seedling grapefruit have all been successfully used, the rough lemon being favored for rapid growth and early production and especially for the lighter grades of soil. Wind damage has caused frequent losses of trees and fruit and has directed attention to the use of such plants as bamboos and casuarinas for windbreaks.

Owing to the fact that the trees bloom intermittently, Puerto Rico is enabled to ship grapefruit over a long season, a summer crop often maturing from a previous fall bloom. Production for several years past has averaged about a million boxes, of which approximately 25 percent is marketed as canned grapefruit. Puerto Rico is credited with having initiated the canning of grapefruit “hearts", a product that has grown rapidly in favor and has made grapefruit available at all seasons and in many localities where fresh fruit seldom is offered. Orange shipments from Puerto Rico, amounting to a half million boxes in good seasons, consist largely of so-called "wild oranges", which come in a considerable measure from seedling trees cultivated as shade trees on coffee plantations.

In the Philippines citrus fruits have been grown for centuries, forms of Citrus hystrix DC. (kalpi), C. mitis Blanco (calamondin), and C. grandis (pummelo) being native to the islands. It is only in recent decades, however, that attention has been given to growing the fruit commercially. Wester (88), beginning about 1910, brought together an extensive collection of citrus varieties at Lamao, few of which proved adapted to Philippine conditions. The mandarin oranges as a class have proved best suited to commercial culture, and a local variety known as Batangas mandarin is being grown on a considerable scale chiefly for the Manila market. The Batangas, the King, and the Szinkom mandarins, several pummelo varieties— including the Siamese—and the Valencia orange constitute most of the recent plantings. Genetic studies and hybridization work have been inaugurated by Torres, and one hybrid variety, Szinbat (Szinkom X Batangas), has been introduced. It is characterized as productive, of good quality, and resistant to wind injury. Further breeding and selection work is in progress, special studies being made of polyembryony—to be discussed later—in scion and stock varieties.

The citrus breeder is concerned with two kinds of plant materials within the Rutaceae—the citrus group proper, containing the types closely related to the widely known sweet orange, and various species in genera somewhat less closely related. The first or citrus group contains all of the valuable types cultivated for their fruits or used as rootstocks, and the second is of value in some instances as stocks, as breeding material, and in furnishing a clue to the evolutionary development of the branch of the Rutaceae to which the citrus fruits belong.

The citrus group proper is characterized by great diversity in morphological characters, and this has led some systematic botanists to the multiplication of species. The classification of Swingle (68), however, is conservative, has been widely accepted, and is used in this report. The only exception made is in the case of Citrus grandis, where it has been necessary to recognize two varieties besides the type species. The horticultural differences in a number of cases are so great, as will be pointed out in the discussion of dessert quality later, that the single type designation is quite inadequate.

The following types commonly grown for their fruits or as rootstocks are in the three genera Citrus, Fortunella, and Poncirus:
          Sweet orange, Citrus sinensis (L.) Osbeck.
          Sour orange, C. aurantium L.
          King orange, C. nobilis Lour.
          Tangerine orange, C. nobilis var. deliciosa (Tenore) Swingle.
          Satsuma orange, C. nobilis var. unshiu (Mak.) Swingle.
          Shaddock, C. grandis (L.) Osbeck.
          Grapefruit, C. grandis.
          Pummelo, C. grandis.
          Citron, C. medica L.
          Lemon, C. limonia Osbeck.
          Lime, C. aurantifolia (Christm.) Swingle.
          C. ichangensis Swingle.
          Kalpi, C. hystrix DC.
          Calamondin, C. mitis Blanco.
          Kumquat, Fortunella spp.
          Trifoliate orange, Poncirus trifoliata (L.) Raf.

The citrus breeder is fortunate in possessing material that presents so many diversities—in dessert and keeping quality, season of maturity, resistance to disease, and regional adaptation. All of these will be developed in detail in the following text, but the dessert quality of the types will be discussed at this point.

Some of the fruits listed above—sweet orange, grapefruit, lemon, and lime—have become well known to northern readers and need to be only briefly differentiated; a few, however, are little known and will require more detailed descriptions. In general it may be stated that the taste qualities of mature citrus pulp and juice are dependent on various combinations of sugars, acids, glucosides, esters, and peel oil. The first two, the sugars and acids, are the basic matrix and give variations from sour through tart, sweet, and insipid, and the latter contribute bitter and aromatic principles. The bitter principle, furnished by glucosides, is apparent only if it is in solution in sufficient amount in the juice (5, 78, 79). This is normally not the case except in such types as grapefruit and lime. The aromatic quality contributed by peel oil is important in some cases.

In most commercial varieties of sweet orange the sensation of sweetness predominates, combined with a slightly perceptible tartness. In most varieties the quality contributed by esters is slight, but in such varieties as Pineapple the suggestive "pineapple" ester is outstanding.

In the mandarin orange or free-peeling group, the tangerine oranges are characterized by the pleasant "tang”, which is due to esters. The King and satsuma oranges in this same group have taste qualities similar to those of the sweet orange.

In grapefruit the bitterness of the glucoside naringin gives the sprightly taste added to the mild acidity that makes the fruit outstanding as a breakfast appetizer.

The pummelo, as distinguished from the sour shaddock, is used only as a salad fruit. The large juice sacs are separated from the locular wall tissues and are served like any other salad. The flavor in the better varieties is due to a very slight acidity and the presence of only a very little glucoside, but it is predominantly sweet, and inimitable quality is contributed by esters.

In lemon and lime, acidity is of first importance. A good acid citrus fruit, as pointed out by Traub and Robinson (80), should have from 6 to 7 percent of acid. In lime the characteristic glucoside, which has not been studied in detail, lends the “lime” taste. The peel oil of lemon and lime also gives desirable qualities.

The citron is used entirely as a preserve. The kumquat is used both in preserving and in table decorations. It is also eaten entire, out of hand. Two general types are recognized, sweet and sour, in both of which the rind has little of the pungent oil common to most citrus fruits.

In the case of hybrids, intermediates have in some instances been secured, especially in the tangelo oranges—hybrids between the grapefruit and the tangerine orange. In these hybrids, as a rule, the esters are predominant in distinguishing the flavor of the new fruits from the common sweet orange. In the Perrine lemon, a hybrid between the lemon and the lime, there is a mild suggestion of the lime glucoside.

When hybrids between citrus species first appeared it was customary to apply various compound names, such as tangelo, tangor, oranguma, limelo, lemelo, etc., to indicate the parentage. It was soon realized that this would lead to confusion from the horticultural viewpoint, since some crosses gave rise to more than one horticultural or market type. The grapefruit-tangerine cross (tangelo), for instance, gave rise to forms like the now generally known tangelo orange, and also varieties that resemble the grapefruit in structure and juice quality but with the rind, flesh, and seed color of the tangerine. A strict application of the term “tangelo” would have included both of these forms. The difficulty was overcome by the decision to place hybrids for purposes of horticultural classification with the well-known types that they most resemble, and to use the interspecies compound designations only in their scientific application. On this basis most of the tangelos already introduced, being more like the sweet orange in structure and use, were designated as a group of the sweet orange, possessing relatively high quality with special reference to a pleasing blending of esters, sugars, and acids. Such a class would naturally contain also such hybrids as the Umatilla (oranguma), a cross between the sweet orange and satsuma orange, but very similar to the other tangelo oranges, and also the Temple, apparently a naturally occurring hybrid between the tangerine and a grapefruit variety similar to the Tresca.

Representatives from a great number of related genera, Glycomis, Olaucena, Chaleas, Feronia, Feroniella, Aeglopsis, Aegle, Swinglea (Chaetospermum), Balsamocitrus, Lavanga, Hesperethusa, Triphasia, Severinia, Citropsis, Atlantia, Eremocitrus, and Microcitrus, have been introduced by the United States Department of Agriculture during the last 25 years. These are listed in table 6 in the appendix. Some of these may prove of value in citrus breeding and as rootstocks. Although certain species have entered to some extent into hybridization work, no hybrids of immediate value have been secured up to the present. The material, however, is valuable to the breeder from another standpoint, for it presents an opportunity for an evolutionary approach to the study of relationships within the group.

The early history of citrus improvement in the United States is concerned almost entirely with the introduction of varieties from other citrus-producing regions, mainly through private initiative. This period extended to the 1870’s in this country. Toward the end of the nineteenth century, in the United States, the number of varieties was increased. by the addition of those originating as chance seedlings and possibly by bud mutations accidentally propagated. Still later, improved types appeared as the result of artificial cross-pollination. Breeding work was undertaken by the Department of Agriculture in 1892, and the the State agricultural experiment stations in California in 1910, in Florida in 1924, in Alabama in 1933, and in Texas in 1934.

The breeding of citrus fruits presents two problems not met with in the case of the usual annual crops such as grains, which can be grown in great numbers on a relatively small area at small expense.

First, as a rule it takes from 6 to 10 years to fruit a seedling citrus tree unless the variety is top-worked on an older tree, in which case the time will be cut in half. The trees are expensive to produce and to test out in orchard formation on various soil types and under various climatic conditions. It is necessary, therefore, to plan breeding experiments so that only progeny are grown that promise varieties of immediate value or additions to the knowledge of citrus genetics. With such a handicap, the work has not progressed at a very rapid rate.  The other difficulty is due to the phenomenon of polyembryony—meaning “several embryos per seed” (fig, 5).  In the case of plants reproducing by seeds, each single seed as a rule gives rise to one seedling, which is the result of the union of the male gamete (reproductive cell), contributed by the pollen grain, and the female gamete (egg cell), contained in the ovary of the flower. In each of these gametes the number of chromosomes has been normally reduced by half preparatory to reproduction (the haploid number of chromosomes), and the union of the two results in a complete complement of chromosomes, called the diploid number, which is characteristic of all of the body cells of the individual plant as distinguished from the sex cells. In the case of the citrus seeds, however, a normal embryo produced by the union of the male and female gametes may be present, and in addition one or more—sometimes as high as 15—additional embryos that have arisen from projections into the embryo sac of the surrounding maternal tissue (nucellar tissue). When these projections develop into embryos they have the full chromosome complement (diploid) of the mother plant without the union of two gametes. Citrus types and varieties may vary greatly in the number of nucellar embryos produced. Seedlings that develop from nucellar embryos are called apogamic seedlings (literally, "without marriage").

Figure 5.—Nucellar embryony in citrus, showing four seedlings sprouting from single seeds (McCarty grapefruit): A, One vigorous seedling and three relatively less vigorous; B, four vigorous seedlings; C, similar to A after separation. (See also fig. 17, showing cytological details.)

Unless the parents have unlike vegetative characters, it is not possible to distinguish the sexually produced or hybrid embryo from those that arise by nucellar budding, though in the latter, of course, only the characters of the female parent will appear. This means that a great many more seedlings must be grown to the point where they can be distinguished than in the case of crops producing only normal or seminal seedlings. Citrus breeding, therefore, will continue to be even more costly than ordinary tree-fruit breeding unless a method can be worked out to achieve practical control of nucellar embryony.

Thus it is natural to expect relatively slow progress. The results from cross-pollination followed by inbreeding and selection, and from back-crossing on the parent types, will not be available in a few seasons but only after a considerable period of time. However, the earlier pioneers in this field have laid the foundation, and in the future it is probable that the rate of progress can be considerably accelerated.

Fortunately no problem is presented with reference to controlling pollination in citrus. The flowers are relatively large and the ordinary bagging technique with brown-paper bags has proved sufficient. In practice the flowers are emasculated before pollen is shed and then bagged. They are pollinated soon after opening, and the bags, which were removed for pollination, are replaced and left on until the petals have fallen and the fruit begins to grow. To protect from loss by dropping, the fruits are usually covered with cheesecloth bags.

Pollen is gathered from flowers that have been bagged when still closed and is used immediately unless flowering of the parents desired does not overlap, in which case the pollen is stored for later use. Kellerman (35) has shown that citrus pollen dried over concentrated sulphuric acid and sealed in glass vacuum tubes at about 0.5 mm pressure can be kept in a viable condition for more than 2 months.

When the seeds are removed from the harvested mature fruits secured as a result of artificial pollination, they are thoroughly washed and are planted at once in flats, for the germination percentage usually decreases if they are allowed to dry in the air. The flats containing the seeds are placed in a coldframe to prevent damage by heavy rains. When seedlings are 6 to 18 months old they are planted in nursery rows, usually 12 by 6 feet apart, and given good culture. As soon as the fruiting stage is reached, usually in 5 years, or sooner if topworked on old trees, detailed records are taken of tree and fruit characters, and all seedlings that show no immediate promise or appear to be of no value for future breeding are destroyed. These records determine the apparent worth, if any, of the plants as varieties for cultivation, and also serve as a basis for working out genetic principles.

The seedlings are tested for vigor, including disease resistance, bearing capacity, and regional and rootstock adaptation. The fruits are tested for size, shape, juice percentage, season of maturity, number of locules and seeds, rind thickness, percentage of "rag", percentage of sugars and acids, effective acidity (pH) of the juice, and the flavor of the juice. Out of a great number of seedling individuals only a very few are finally selected for introduction, and these are released only after favorable performance in preliminary fruiting tests in a number of locations.

The first consideration in citrus breeding is excellent dessert quality. What constitutes high quality has been previously discussed, and we pass to the consideration of tree and other fruit characters that the breeder has in mind when making his crosses and selections.

The tree (scion variety) should be compact in habit, but a vigorous grower and a prolific bearer. It should be resistant to the common citrus diseases and to low temperatures, and congenial with the rootstock or rootstocks used in the region. The fruit should mature at the proper season or seasons to suit market needs. There are also other characters of importance, such as thornlessness.

The fruit should have excellent dessert quality and contain few or no seeds; the shape and size should. be suited to commercial requirements and to ease in packing; shipping or keeping quality, including resistance to storage diseases, should be good; the exterior, including texture and color of rind, should be attractive; and, in particular, the vitamin C content should be up to the standard. With the rapid growth of a new industry in canning “hearts” and juice of both grapefruit and orange, special attention may need to be given to the requirements of this promising industry. For instance, it has already become evident that the pulp of Marsh grapefruit lacks the firmness necessary in a good canning grapefruit. With the growing tendency to utilize citrus fruits in juice form and in mixed drinks, the high color of the juice characteristic of the tangelo group of hybrids is proving a decided advantage.

The tree used as a rootstock should be adapted to the soil and climatic conditions, be free from or resistant to trunk and root diseases, and produce a high percentage of nucellar embryos, and it should not be so vigorous in growth as to make the fruit of the scion coarse, of poor texture, and comparatively low in total solids (sugars and acids) and therefore insipid in taste (6, 80).

As the margin between production costs and sales returns becomes narrowed with increasing supplies of citrus fruits, any adaptation that might lessen the cost of production becomes vitally important. This places emphasis on disease resistance in any breeding program, to reduce both expense for grove sanitation and losses due to infected trees and fruit. In citrus the list of such diseases is quite extensive and varies with citrus types and varieties.

The task of breeding for resistance to injury from insect and other animal pests on plants presents the major difficulty of developing a practical technique.

In Florida the chief citrus diseases are melanose, affecting tender twigs, leaves, and immature fruits, and stem-end rot, affecting mature fruit, both caused by Diaporthe (Phomopsis) citri (Fawe.) Wolf; sour orange scab (Elsinoe fawcetti Bitancourt and Jenkins), affecting leaves and fruits; key lime anthracnose (Glocosporium limetticolum Clausen) ; foot rot (Phytophthora parasitica Dastur) ; and psorosis, cause undetermined.

In the Southwest, brown rot gummosis and foot rot (Phytophthora citrophthora (Sm. and Sm.) Leonian and P. parasitica); psorosis; and shell bark (Diaporthe citri) are of major importance; and in the lower Rio Grande Valley gummosis, scaly bark, and stem-end rot (melanose).

In spite of the peculiar difficulties encountered in citrus breeding, definite achievements can be recorded, and the outlook for the future is most encouraging. The work in the past has shown that worthwhile results may be secured from appropriate crosses and that important strains and varieties may arise by bud mutation. In addition a beginning has been made in laying a foundation of genetic principles.

The work in citrus hybridization carried on by the Department of Agriculture workers since 1893 has shown that the combinations of grapefruit and tangerine and of lemon and lime give the most promising results. The first have given rise to high-quality fruits known as the tangelo oranges, and the latter to a high-quality lemon, the Perrine. The work of Frost, of the California Agricultural Experiment Station at Riverside, has shown that crosses within the mandarin orange group give worth-while results.

Apogamic seedlings from crosses have also given rise to important varieties in grapefruit and in sweet, sour, and satsuma oranges. Bud selection has given superior strains of known varieties and has served to stabilize standard varieties. Crosses with more distantly related relatives of citrus, Citrus X Poncirus, citrange X Fortunella, citrange X calamondin, etc., have proved interesting from the genetic standpoint and have also given some concrete types of possible value in horticulture—citrange, citrangequat, and citrangedin. Reference to particular hybrids in which definite objectives have been attained are found in the text and in table 6.

Thus the ground work has been laid by the earlier workers. With the wealth of breeding material and the increase of interest in this field, research in citrus breeding may be expected to show greater progress in the years to come.

If we wish to produce a variety or strain that is very similar to a standard variety, such as the Valencia, but superior in some important feature, such as eating quality, generative seedlings from either selfing or narrow crossing are very unpromising, for two reasons: (1) they are usually weak and very unproductive, and (2) they are usually very unlike the parent or parents. The best chance of success, although a very uncertain one, seems to lie in the discovery of superior bud-variation forms. Such forms may be found, of course, either as “limb sports” or as whole trees which have happened to be budded from variant branches. They may also be found as seedlings derived from nucellar embryos that happen to be formed on a variant branch; in this case the seedlings are likely to be indistinguishable from other nucellar seedlings until they produce fruit. On the basis of the time and expense required, the search for bud variations in commercial orchards seems more promising than the growing of seedlings. A good new bud-variation type may in some cases, however, be variable because of mixture with the parent variety in the same shoots to form a chimera, and the growing of seedlings may then be the best method of getting the new type separated from the parent variety. Variability may in some cases be inherent in the constitution of a new form, however, so that it will never become stable.

X-ray treatment probably can be used, as it has been with other plants, to increase greatly the frequency of new genetic variations. It seems most likely to be useful by producing variations in nucellar seedlings.

For the production of new varieties of superior quality but with flavors very unlike those of the present varieties, the great variability of hybrids offers much promise, since we can use wide crosses, which commonly give vigorous hybrids. Various hybrids of the tangelo and tangor groups, which are hybrids of mandarins with grapefruit and with sweet orange, that have been introduced by the United States Department of Agriculture, offer unique and very pleasant flavors. It is now probable, however, that the best opportunities for securing superlative quality are to be found in crossing within the mandarin species or group of species, and therefore, it is very fortunate that certain crosses between mandarin-group varieties produce vigorous hybrids.

It is advisable to make some further trial of the production of triploid hybrids by pollinating tetraploids by diploids of other species (or of other sections of the mandarin “species”). The only tetraploid form which at present seems to have much promise for such use is the tetraploid Lisbon lemon, which combines good seed production, comparatively low chromosome irregularity, and high proportion of generative progeny.

The evidence on which these suggestions are based has only partly been secured at the Citrus Experiment Station, and that on the results of selfing has very largely been obtained in Dr. H. J. Webber’s root stock experiments. The nature of the work has ranged from chromosome counts and systematic records of various tree and fruit characters, to the determination of the agreeableness of fruit flavors. Any extensive gene analysis is early, impracticable in citrus, yet certain evidence has been obtained which seems to have definite significance for general genetic theory, when interpreted on the basis of gene analysis made on more favorable organisms.

The problems that most concern the citrus breeder have been listed by Frost as the determination of—
(1) The generality of the reduction of vigor with selfing and narrow crossing. (2) The extent to which the proportion of nucellar seedlings can be predicted from counts of total embryos. (3) (a) The extent to which triploids are unproductive of fruit, as compared with diploids of the same ancestry. (b) The extent to which triploids can be produced by crossing diploids with tetraploids. (c) The frequency with which triploids occur from crosses between diploids. (4) The extent to which the prospects of high vigor, productiveness, good flavor, and other horticulturally desirable characters in crossing ean be inferred from knowledge of the species or smaller taxonomic group to which a variety belongs, on the basis of trial of other varieties of the same group. (5) The frequency with which new genetic types arise by bud variation, and the extent to which this process can be speeded up by X-ray treatment of seeds, pollen, etc. (6) The best methods of modifying physiological conditions to secure high seed production in crossing, and, if possible, high proportions of generative embryos.

With the present facilities, problems named under 1 to 4 above may be solved partially, and to a much smaller extent those under 5 and 6 in the list.

The improvement of citrus varieties will now be briefly discussed on the basis of citrus types—sweet orange, mandarin oranges, grapefruit, acid citrus fruits, and minor citrus fruits. In each case the varieties will be considered by regions. Following this, citrus bud selection and citrus rootstocks will be considered under separate headings.

Early-maturing sweet orange varieties for Florida are relatively scarce, and it is desirable to give this fact due weight in any citrus- breeding project.

The Hamlin is one of the early varieties. usually recommended. It was discovered in a grove planted in 1879 near Glenwood, which later came into possession of H. E. Hamlin. Under the best cultural conditions the acidity and sweetness are well blended, giving it excellent flavor; the rind is smooth and glossy; seeds are none too few; the season is October and November and later. Recent reports indicate that it is sensitive to overfeeding and other unfavorable growing conditions, which may lead to fruit splitting, riciness of pulp, and poor juice quality.

The Parson Brown variety was introduced by C. L. Carney, of Lake Weir, about 1878, having originated at Webster in a seedling grove owned by Parson Brown. Acidity and sweetness are fairly well blended if the fruit is picked early; seeds, 10 to 19; season, October and November.

Early sweet oranges usually have poor rind and flesh color when harvested at the beginning of the season. This is a deficiency that might be remedied by breeding methods.

The results thus far look promising. A hybrid, the Orlando tangelo orange (fig. 6), introduced by the Department in 1931 and resulting from a cross of the Bowen grapefruit pollinated by Dancy tangerine, is early in maturity and can be harvested over a long season. The variety is highly resistant to citrus scab. When first harvested in late September or early October the rind color after degreening with ethylene is a beautiful light yellow, but later in the season it takes on a natural deep reddish-orange color. It is medium in size, and the flesh color is deep orange. Its main defect is its seediness. Such a defect can apparently be overcome in future breeding work, as will be pointed out later.

A large number of midseason varieties were named and introduced beginning in the 1870’s. Among these the outstanding ones are the Homosassa and Pineapple. The Homosassa is a variety of excellent quality, with a sprightly, rich, vinous flavor; seeds, 20 to 24; season, December to February. The Pineapple is the outstanding midseason variety, having a glossy rind of deep orange color, vinous and sprightly in flavor; seeds rather large and numerous. It originated near Citra, Marion County, and received its name from the fine aroma reminiscent of the pineapple. It is today. the most important midseason sweet orange in Florida.

The navel type of sweet orange is not suited to the Florida climate. Although a number of varieties have been introduced, they have not proved successful as the Washington Navel has in California.

Two midseason tangelo oranges, of the same parentage as the Orlando variety, introduced by the Department in 1931—Minneola (December—January) and Seminole (February—April)—are outstanding from the standpoint of dessert quality, the Minneola especially having the most delicate blending of esters, sugars, and acids. These fruits have deep tangerine rind and flesh color, and their shipping quality is good. Like the Orlando tangelo orange, the Seminole (fig. 7) is highly resistant to citrus scab, the Minneola partially resistant. The fruits are somewhat seedy.

Figure 6.—Typical fruits of Orlando tangelo orange (originally introduced as the Lake variety), a hybrid between grapefruit (9 ) and tangerine (co); remarkable for earliness, maturing in October and November; highly resistant to scab. Introduced by the United States Department of Agriculture.

The sweet orange industry up to the 1870’s was based on seedlings and clones producing early and midseason fruits. An event of great importance took place when, early in the 1870’s, the late type of sweet orange, now called Valencia (fig. 8), was introduced into Florida by S. P. Parsons, a nurseryman of Long Island, N. Y., and Palatka, Fla. Parsons had secured it from Thomas Rivers in England, who had imported it from the Azores and had cataloged it under the name "Excelsior.” Parsons gave trees to E. H. Hart (fig. 2), of Federal Point, Fla.   Having lost the label; Hart distributed the trees under the name of Hart’s Late or Hart’s Tardiff. The variety was exhibited before the Florida’ Fruit Growers Association on April 25, 1877. The outstanding characteristic of this variety is its late maturity. Its season extends from March to June.

The variety was also imported into California, in a lot of citrus varieties from Thomas Rivers, by A. Chapman, of San Gabriel, Calif., between 1870 and 1872. One of these varieties, labeled as a navel, turned out to be a late-ripening nonnavel variety that fruited after the other varieties were off the market. The variety was named Valencia at the suggestion of a Spanish laborer, and Chapman sold it under the name of Valencia Late. Nurserymen in California had purchased stocks of Hart’s tardiff, and a great many trees had been set out. By the time it was discovered that the Valencia and Hart’s Tardiff were identical, the variety had attained commercial importance in California, and the name Valencia was retained.

A variety similar to the Valencia, named Lue Gim Gong, for the originator, and introduced in 1912 by the Glen St. Mary Nurseries, is described as very late in maturity. It is now generally regarded as a strain of Valencia, from which it originated as a seedling.

The production of a tangelo orange (Umatilla), which resulted from the pollination of the satsuma by the Ruby sweet orange and was introduced by the Department in 1931, indicates what may be accomplished by breeding methods in the creation of late-maturing varieties. In contrast with the parents—satsuma, early maturing, and Ruby, sweet midseason—the hybrid matures its fruit in late February, March, and April. The variety is highly resistant to citrus scab; the fruits are medium to large in size; the rind and pulp color is deep orange, and the quality is excellent. Seed content is variable.

To sum up: At the present time the sweet orange industry in Florida is based primarily on two early varieties, Hamlin and Parson Brown; on midseason seedlings and two midseason varieties, Pineapple and Homosassa; and on one late variety, Valencia.

In California one early and midseason navel and one late nonnavel variety have proved sufficient to produce an orange crop every month in the year.

Contrary to popular opinion, the navel type of sweet orange is not a modern product. It was described and pictured by John Baptisti Ferrarius in 1646 and is apparently of early origin. As early as 1820 the Bahia form of the navel orange had made its appearance in Brazil, where orange trees had been introduced by the Portuguese settlers. Its excellent qualities were soon recognized, and the variety was extensively propagated in the vicinity of Bahia. Even at this early date the variety seems to have been subject to bud mutation, and inferior types appeared that were unintentionally propagated and introduced in South Africa and Australia. Those that reached Australia also included desirable types.

The Bahia type of navel orange was first introduced into Florida in the 1830’s, but the trees were killed during the freeze of 1835. A shy-bearing form of Bahia navel was introduced into California in the early 1870’s. The strain was secured from S. B. Parsons, Flushing, N. Y., who had received it from Thomas Rivers in England. In the early 1870’s an inferior type of navel orange was imported from Australia into California, which set the precedent for, referring to inferior strains as the Australian navels to distinguish them from the superior Bahia strain.

The story of the Washington Navel orange is a dramatic illustration of the value of superior varieties of economic plants. In 1870 the citrus industry had begun in California, but there was no outstanding early and midseason variety of sweet orange generally adapted to the climate. The early mission seedlings and varieties introduced after the middle of the nineteenth century were being tested out by various growers, but there was a lack of standardization in quality. The value of alertness in using the plant material that has been produced as a result of centuries of selection is nowhere better illustrated than by the timely action of the late William Saunders (fig. 9), then superintendent of gardens and grounds of the United States Department of Agriculture, Washington, D. C.

Figure 9.—William Saunders (1822-1900), superintendent of gardens and grounds of the United States Department of Agriculture at Washington, D. C., through whose efforts the Bahia navel orange was successfully introduced from Brazil in 1870.  Three years later trees propagated by Saunders were planted in California and, under the name of Washington Navel, became the basis of an epoch-making industry.

In 1870, through the kind assistance of a missionary stationed at Bahia, Saunders imported from Brazil 12 navel orange trees in tubs. These were housed in the Department greenhouse at Washington, and propagations were made for distribution to the regions adapted to citrus culture. The first propagations were sent largely to Florida and California, but at least one of this lot is still maintained by the Department at Washington. Mr. and Mrs. Luther C. Tibbets were attracted to the settlement at Riverside, Calif., and early in 1873, before starting her journey, Mrs. Tibbets visited the Government propagation gardens at Washington, where Mr. Saunders gave her two Bahia navel trees. These were carried to California and planted beside the Tibbets’ cottage in Riverside (fig. 10). In February 1879 the fruit was awarded first prize over other navels exhibited from Orange County, and these two trees were used as the source of extensive plantings. The variety was referred to as the Washington Navel to distinguish it from the Australian importations. An attempt was made to change the name to Riverside Navel, but this proved unsuccessful. The great superiority of the Washington Navel (fig. 11) was soon recognized, since it apparently was ideally adapted to the climatic and soil conditions of California and produced a fruit of high quality with the highest market appeal. The best strain of Washington Navel, according to Shamel and associates (61), is characterized by an—
open and somewhat drooping habit of growth and dense foliage with large, oval, dark-green leaves. * * * Under normal conditions no pollen is produced by the anthers of the flowers. * * * The fruits * * * are obovoid in shape and generally of medium to large size. The rind is of medium thickness, and the texture is smooth and grained. The color of the fruit is bright orange; the rag is tender and comparatively small in quantity; the juice is abundant and of superior quality, having a pleasing and sprightly subacid flavor. The fruits are seedless, and the navel usually is small, sometimes rudimentary, with no development except in the rind.

Figure 10.—Tree of the Washington Navel orange at Riverside, Calif., the lone survivor of the original budded trees of this variety sent to California in 1873, propagated from the introduction made in 1870 from Bahia, Brazil, by William Saunders, Washington, D.C. A tablet in memory of Mrs. Eliza Tibbets, who first planted this navel orange and brought it to fruiting, stands near the tree.

By 1885 enterprising nurserymen had introduced most of the important varieties of the world, which were tested in comparison with the local seedlings of special merit. Less profitable varieties were rapidly eliminated, and by 1900 the area planted to the Washington Navel was larger than that of all other varieties in the State. It is now generally recognized that one of the outstanding events in the economic and social development of California was the introduction of this orange in 1873. During the period of more than 60 years following, a great industry has been built up from the two small trees planted by Mrs. Eliza Tibbets.

Figure 11.—Washington Navel orange fruit, typical of the variety as grown at Riverside, Calif. showing solid but juicy pulp, locules separating easily, firm, rather thick rind, small navel, and absence of seeds.

The Valencia variety of sweet orange, introduced into California and Florida between 1870 and 1872 as already detailed, is the other outstanding orange variety in the State. These two varieties are grown almost to the exclusion of others. Climatic conditions vary widely because of differences in rainfall, protection by mountain ranges, the moderating influence of the ocean, and other factors. These affect the ripening period of fruit varieties so that the same variety matures at different times in various regions. When this is coupled with “tree storage” in the case of the Valencia, which holds its fruit in good condition for several months, the combination results in a marketable fruit crop throughout the year. This tree storage is made possible by the dry summer climate with comparative freedom from fruit-destroying fungi, together with other climatic factors contributing to a long ripening season. The movement of Washington Navels begins in November and ends in May. The Valencia crop is marketed from May to November and, as a rule, overlaps by several weeks the period when navels are shipped.

Frost (21) reports that the Citrus Experiment Station of the University of California, at Riverside, has recently introduced a good early nonnavel variety, Trovita, which has pollen and a few seeds. This variety originated as one of three seedlings grown from Washington Navel seeds accidentally found. It is pointed out that this may be a promising variety in the hotter citrus districts because of its seeding tendency, since the seedless Washington Navel often fails to set fruit under these conditions. The fruit of the new variety is much like that of Washington Navel, but the navel structure is usually absent or rudimentary.

The commonly grown varieties of sweet oranges, along with other varieties of citrus, were brought into the lower Rio Grande Valley of Texas early in the twentieth century, but in general these are not as well adapted to the climate as the grapefruit. The most desirable early to midseason variety of orange is the Hamlin. The trees are prolific, regular bearers under lower Rio Grande Valley conditions, maturing their fruit in October. Although the Pineapple is the most widely planted early variety, the Weslaco substation of the Texas Agricultural Experiment Station recommends that Hamlin and Joppa should replace varieties such as Pineapple, Parson Brown, and Ruby, which are characterized in Texas by the production of undersized fruit and by erratic bearing habit (16). Even the desirable strains of Washington Navel are not satisfactory, being “rather erratic in their bearing habits” and producing juice "variable in flavor, often being quite insipid (devoid of acid)."  A recent introduction, one of several navel oranges brought by the Department from Brazil and now under test at the Weslaco substation, is very promising. It has been named Texas Navel (fig. 12) and is being tried extensively by growers. The tree is vigorous and somewhat more productive than the strains of Washington Navel under Texas conditions. The fruit is typically navel in structure, of medium size and good quality.

Figure 12.—The Texas Navel orange. A, Tree showing vigorous habit of this navel variety, an introduction from Brazil made by the United States Department of Agriculture in 1917 when the Washington Navel proved poorly adapted to lower Rio Grande Valley conditions. The tree has a height of 20 feet and a spread of 21 feet. B, Fruit of the Texas Navel orange, which resembles the standard Washington Navel in physical characteristics and quality but is more prolific under conditions in southern Texas.  Fruits average 2¾ inches in diameter.

The Valencia, which matures its fruit from December to January under lower Rio Grande Valley conditions, is the “most profitable commercial variety at the present time, since the trees are productive and fairly regular in bearing, but a large proportion of the fruit, which are of good quality, is undersized.”

The Temple orange, which belongs to the horticultural group of high-quality hybrid fruits (tangelo oranges) and is similar to the sweet orange, is promising under Texas conditions, but the rind texture is unattractive and the tree is not hardy. In this same group the Orlando, Seminole, Minneola, and Umatilla varieties are under test.

As far as the standard varieties of sweet oranges are concerned, none is entirely satisfactory under lower Rio Grande Valley conditions.

In the extreme southern tip of Louisiana on the Delta of the Mississippi, south of New Orleans, in the vicinity of Buras, a type of high quality nonnavel sweet orange of local origin has been grown for a great many years along with some navels of recent introduction. The annual output is usually about 300 cars, marketed primarily in New Orleans.

The origin of the grapefruit as a horticultural citrus type is obscure. It is apparently intermediate between the large acid shaddock and the mild salad citrus type commonly referred to as the pummelo. It was brought to Florida from the West Indies and does not correspond to any type in the Orient. It was not appreciated until it was brought to the attention of the consuming public in the 1880’s by enterprising Florida citrus growers. It represents a most important horticultural achievement, for it is now found on the breakfast table in either the fresh or the canned state not only in the United States but also in Europe, South Africa, and Australia.

The grapefruit industry in Florida was founded originally on seedling trees, which are very much like the Duncan variety in structure and quality and are now known as Florida Common. The original Duncan tree, over 100 years old in 1926, according to Hume (32), was still living in the planting known as the Snedicor Grove near Green Springs, in Pinellas County, where it originated as a seedling from a grapefruit tree grown by a Spanish settler, Don Philippe. Later reports are not available. Duncan (14, p: 136) stated in 1892 that Don Philippe brought grapefruit and orange seeds from Cuba 50 years previously and made his planting at Safety Harborin “Philippe Hammock”, and that the orange trees died out from neglect while the grapefruit trees remained in a thrifty condition. The variety was introduced and propagated by A. L. Duncan, of Dunedin, about 1892. It is a very superior variety. Although 17 or more main varieties were later introduced, Duncan remained the favorite among the seedy varieties. Duncan and other seedy varieties are now being top-worked to Marsh in some instances because of the demand for a seedless fruit.

Because of its seedlessness and other desirable qualities, the Marsh variety has been gradually replacing the seedy varieties in plantings within the last 15 years. This variety originated in the William Hancock grove at Socrum, near Lakeland, and was first cataloged by C. M. Marsh in the Catalogue of the Lakeland Nurseries for Fall 1896 and Spring 1897, with the statement that the parent tree was a seedling that had been a prolific bearer for 30 or 40 years. Propagation evidently began some years before, as the parent tree was killed in the freeze of 1894-95. A story receiving credence in recent years that the variety originated as a broken root sprout from a common seedy type tree has been definitely disproved by the testimony of members of the Hancock family and other local residents who were thoroughly familiar with the original seedling tree (56).

The Marsh variety has given rise to two pink-fleshed bud mutations. A.D. Shamel has described one of these that originated near Riverside, Calif., and was brought to his attention in July 1919 by L. V. W. Brown. The mutation that occurred in Florida, in the grove belonging to W. B. Thompson near Oneco, was discovered by S. A. Collins and introduced under the name of Thompson by Reasoner Bros. (fig. 13), of Oneco, in 1924 (65). Both of these mutations are identical with Marsh except in flesh color. The Thompson variety, although not extensively planted in Florida, has become an important variety in the lower Rio Grande Valley of Texas.

Figure 13.—Egbert N. Reasoner (1869-1926), pioneer nurseryman of Florida, active in the introduction and propagation of many subtropical fruits. He introduced Foster, Thompson, Pernambuco, and Royal grapefruit varieties and the Oneco tangerine, initiated the commercial propagation of the lychee, and was active in testing and disseminating new varieties of mangoes and avacados.

   The Foster, another pink-fleshed variety, originated as a branch mutation on a tree of the Walters grapefruit. This occurred in the Atwood Grove near Palmetto, and was first observed in the season 1906-07. It was named and introduced by Reasoner Bros. in 1914. Aside from the color, the fruit has much the same quality as the parent variety, though it is sometimes regarded as earlier maturing. The color of the pulp, as well as that of the Thompson, tends to fade as the fruit reaches full maturity.

Another seedless grapefruit, the Davis, originating from a cross between a seedling type of grapefruit and a tangerine (in the attempt by Department workers to secure a tangelo) is receiving favorable attention. In shape and size it resembles the Marsh, though it is more rounded in form, but it has the taste quality of the seedy grapefruit, with apparently less glucosidal flavor than the Marsh. Its outstanding character, however, is the fact that it has proved satisfactory for canning, the pulp remaining firm, while the Marsh tends to become soft from processing (34).

In the improvement of grapefruit there has been a steady trend toward the use of fewer varieties, and Marsh, largely because of its seedlessness, has become the standard for new plantings, gradually superseding the 17 or more other varieties—Aurantium, De Soto, Excelsior, Hall, Josselyn, Leonardy, Manville, May, Bowen, McCarty, McKinley, Inman, Pernambuco, Royal, Triumph, Walters, etc.

The Florida grapefruit varieties were introduced into California, Arizona, and the lower Rio Grande Valley of Texas.

Such varieties as Duncan, Conner, McCarty, Inman, and Walters, all seedy varieties, are grown to a limited extent for the early market, but they are at a discount after the seedless variety, Marsh, is ready for market.

The pink-fleshed varieties Foster (seedy) and Thompson (seedless) have pink flesh early in the ripening season and usually sell at a premium. The Ruby, a local mutation from Thompson, has both pink flesh and pink rind and is receiving some attention in recent plantings.

The grapefruit industry in the Southwest began after the introduction of the Marsh grapefruit in 1890; the plantings of other varieties previous to that time did not prove profitable. In the Arizona citrus districts and in the Coachella and Imperial Valleys of California, grapefruit matures fairly rapidly and is marketed in late fall and early winter; in other California grapefruit districts it does not reach full maturity until the following spring or summer.

In the mandarin orange group the tangerine and satsuma oranges are grown to supply special markets, primarily early in the season. The chief breeding problem is in connection with the production of high-quality early-maturing varieties.

The only variety of tangerine orange extensively planted in Florida is the Dancy, a prolific variety that was originated as a seedling by George L. Dancy, of Buena Vista, St. Johns County, and was introduced in 1871 or 1872. This variety is of great interest, since it is the pollen parent of most of the tangelo oranges, and it imparts to the best tangelos the deep orange rind and flesh color and the aromatic and pleasing ester qualities. In Texas, although the earlier plantings are mostly Dancy, the Clementine (Algerian) is decidedly superior to this variety, as is also the Warnuco (Ponkan) under lower Rio Grande Valley conditions.

In California the Dancy is grown to a limited extent for special markets, but apparently the small size and higher acidity developed when it is grown in this section has retarded extensive planting. Recently three mandarin varieties, developed by Frost (21) at the Citrus Experiment Station, Riverside, have been introduced for preliminary trial. They were originated as a first-generation cross between the King orange and other mandarin oranges. In the case of the Kara, the Owari satsuma was the seed parent, and in the case of the Kinnow (fig. 14) and Wilking, the King orange served as the seed parent and the Willowleaf mandarin was the pollen parent. These varieties produce fruits of good size and excellent quality, but, as pointed out by Frost, only orchard tests can determine their ultimate value in citrus culture.

Figure 14.—The Kinnow mandarin, a hybrid of King orange and Willow leaf mandarin, the King serving as the seed parent. The fruit is excellent in appearance and flavor, remaining firm when fully fipe; desirable for local consumption but shipping quality undetermined. Fruits shown are natural size, about 414 months after ripening. Introduced by Frost, of the California Agricultural Experiment Station.

The satsuma orange was first introduced into Florida by George R. Hall in 1876. In the upper Gulf coast region, from western Florida to eastern Texas, it has been grown commercially since 1910, and the industry is based primarily on the Owari satsuma. Although there are apparently two or more strains of this variety, they do not differ widely. The fruit is harvested early in the season, October and November. It is desirable to harvest as early as possible for two reasons—to secure the early market before many high-grade sweet oranges are shipped from peninsular Florida, and to remove the fruits early enough to allow for the storage of maximum food reserves in the tree before the onset of winter. In an attempt to meet this need, the Department of Agriculture imported over 50 strains of the early-maturing satsuma from Japan, and these are under test at the Gulf Coast Substation at Fairhope, Ala., in cooperation with the Alabama Agricultural Experiment Station and with growers in various parts of this section. So far the Kawano variety has shown some promise, but it does not uniformly mature earlier than the Owari. Some of the other strains may prove more regular in this respect.

A selection from extra vigorous apogamic seedlings following cross- pollination of the satsuma orange with the sweet orange is promising from a commercial standpoint. The Silverhill variety, originated in this manner, has now been tested for a number of years and appears to be a superior strain of the Owari. It has shown somewhat more resistance to cold than the other satsuma varieties, but the evidence is not conclusive.

The chief center of lemon culture is in California, and lime production is confined primarily to Florida.

The introduction of the citrus scab proved a serious drawback to lemon culture in Florida. There is also great difficulty in properly curing lemons in the humid summer climate. In 1931 the Department introduced the Perrine lemon (fig. 15), a hybrid between the Genoa lemon and the Mexican or Key lime. It is highly resistant to citrus scab and anthracnose and has given a new impetus to the industry.

The Perrine lemon is a rapid-growing, vigorous, and productive tree. The fruits are borne singly or in clusters and are of medium size, with a high acid percentage, ranging from 6.2 to 7.2 percent, ranking with standard commercial varieties in this respect (71, 80).

Figure 15.—Typical fruits of Perrine lemon, a hybrid between the lemon and the lime; remarkable for its resistance to scab and anthracnose, vigor of growth, and heavy bearing. Resembles the true lemon in shape, size, and acid content. Introduced by the United States Department of Agriculture.

   The Meyer lemon, an introduction from China named for the introducer, the late Frank N. Meyer, famous plant explorer for the Department of Agriculture, is chiefly noteworthy for its frost resistance. Its low acid content (4.3 to 4.8 percent), large size, and round shape preclude its general popularity as a commercial lemon, though it is well suited to local needs where true lemons have proved too tender.

The Key, Mexican, or West Indian lime has long been cultivated primarily on the keys, but recently the industry has declined because of hurricane damage. The Tahiti (also called Persian or Bearss) lime has recently been planted to a considerable extent in Dade County and the southern ridge section. It has high quality, and in addition the aromatic properties of the rind are highly valued by the consumer. The Lakeland and Eustis limes, hybrids of lime and kumquat, introduced by the Department, although of high juice quality, are of small size and have a thin rind, which may prove disadvantageous for storing and shipping.

In the early development of the industry a large number of European varieties of lemon were grown and also many seedlings were raised, most of which were not promising. The present plantings are practically limited to Eureka and Lisbon. The former originated from a seed planted in 1870 by C. R. Workmen in Los Angeles. The Lisbon variety originated in Europe. Shamel and coworkers (64) state that the Lisbon lemon was introduced into California in a number of importations, the first of which was made as budded trees in 1874 and 1875 from Australia. It has been pointed out that the advantages of the Eureka are its comparative freedom from thorns, its tendency toward early bearing, and its prolific bearing qualities. The tendency to bear fruit on the tips of the branches and the inclination to grow long canes with but few laterals and to drop its leaves on the long canes or branches, leaving the limbs and fruit too much exposed to the hot rays of the sun, are disadvantages.

The Lisbon has heavy foliage, which protects the fruit from sunburn, and the fruit is borne uniformly throughout the tree. This variety has a tendency to bear one large crop maturing in winter, with only a small amount of summer fruit.

   IN THE United States the pummelo, citron, kumquat, and sour orange are not grown to any appreciable extent for their fruits. Of these minor citrus fruits the pummelo is the most promising. The citron is grown mainly for exhibition purposes, although a as has been made in California with preserving it on a commercial scale. The sour orange is grown as an important marmalade fruit in some other citrus-producing countries, notably in Spain.

The pummelo in its better varieties is a most delicious salad citrus. The juice does not possess the sprightly acid and naringin (bitter) properties of the grapefruit, but when the juice sacs are served as a salad the inimitable flavor is relished by all who have had the good fortune to sample them. In times past only the sour shaddock was known in the citrus-producing districts, but since 1915 the Department of Agriculture, through its collaborator, G. Weidman Groff, in China, has introduced an important collection of pummelos. These are not as yet well known but are being distributed to those interested in the culture of the fruit. One of the outstanding varieties is named Siam. Reference to this collection will be found in table 5.

The Department has also imported a valuable collection of citron varieties, including practically all of the important ones. Corsican, the chief variety of commerce, was one of the first citrus introductions made by Fairchild, who was for many years in charge of the plant introduction work of the Department.

   Kumquats were introduced into Europe as recently as 1846 by Robert Fortune. Importations into the United States were made by Taber in 1885 and Reasoner in 1885 and 1890. The varieties Nagami (oval) and Marumi (round) were first introduced, and later the Meiwa or sweet kumquat. The Nagami predominates in all plantings made thus far. About 1910 the Department introduced the Hong Kong or wild kumquat, which produces very small round fruits and is of interest from the breeding standpoint in that it has the double haploid ac ed complement. (See the later section on cytology in this article.

The Department has introduced a superior variety of sour orange named Oklawaha. It originated as an apogamic seedling when pollen of the shaddock (sour pummelo) was used to pollinate the sour orange.  In habit the tree appears as a vigorous sour orange. The fruit is similar to the ordinary sour orange but averages larger in size, 3 to 4 inches in diameter, and it is superior to the ordinary seedling sour oranges as a marmalade variety.

With the general acceptance of De Vries’ mutation theory (1901), the improvement of varieties by the selection of bud mutations soon became a recognized mode of procedure in plant breeding. Conversely, the elimination of inferior bud mutations became of equal or even greater importance in many clones especially subject to mutations.

That the Washington Navel orange is subject to bud mutation was recognized at an early date, and by 1910 the problem was so serious that at the California State Fruit Growers Convention of that year Coit (10) proposed to "keep individual tree records for 2 years and these trees with records to be offered nurserymen for propagation as pedigreed trees." At the same meeting Coit, in discussing the application of the De Vries mutation theory to the problem, said:

If you go out into the average navel orchard to select a dozen perfect navel oranges true to old standards, you will be surprised at the amount of searching necessary. * * * It is my belief that by far the greatest part of the divergence is attributable to mutation * * *.

must be quick to see and cut out all branches sporting toward poor types. In the second place, we must be exceedingly careful in cutting budwood. * * * Select buds from those branches which produce your ideal of the navel orange.

The problem was so serious that the Department of Agriculture detailed A. D. Shamel to study it in 1909, and he began his work in cooperation with the California Citrus Experiment Station at Riverside and with citrus growers. In his first report entitled “Bud Selection in Citrus Fruits", given at the California State Fruit Growers Convention at San Bernardino, March 8, 1911, Shamel set forth the problem clearly. His work began with the Washington Navel and was extended to include other sweet oranges, the lemon and grapefruit types, as well as other crop plants.

Over a period of 27 years the work of Shamel and his coworkers (57, 58, 59, 60, 61, 62, 63, 64) has been characterized by consistent, painstaking research, which has included a study of the source of strains originating as bud mutations and unintentionally propagated by nurserymen and growers. This was followed by efforts to eliminate the inferior trees in established plantings by top-working with carefully selected buds. Further efforts were made to avoid the propagation of undesirable strains arising from bud mutations by systematic selection on the basis of individual plant performance and an intimate knowledge of the plants, Finally, a systematic search was made in cooperation with growers for superior plants originating from commercially valuable bud mutations, and these were tested in progeny plantings to single out the ones inherently superior to the parent forms for utilization by the industries concerned.

these groves consist of at least 20 strains of commercial importance with five or more others of less economic consequence. The trees of each of these strains have fruit or vegetative characteristics, or both, which serve to distinguish them from all other trees of the variety.
About 25 per cent of the total number of trees studied in the original orchards in which these investigations have been conducted were found to be of undesirable strains having consistently low yields, or bearing fruits of poor quality, or both, such as those of the Australian, Unproductive, Corrugated, Pear-Shape, Sheepnose, Flattened, Dry, and other inferior strains.

The extent of the commercial use of superior strains selected for the production of more uniformly good fruit is indicated by the sale of selected buds by the Fruit Growers Supply Co. for the period 1917-35. These data, given below, include only a part of the supplies of such buds used by nurserymen and growers—probably not more than 50 percent; but because they are conservative they will the better convey some idea of the value of this kind of work.

Two special citrus strains originating as bud mutations, the Robertson Navel orange in 1925 and the Dawn grapefruit in 1929, are now being introduced, and the indications are that they will be widely grown in certain citrus districts of California and Arizona. In the Robertson Navel orange strain the fruits are resistant to June drop on account of their very rapid early growth. They mature about 1 month earlier than those of the Washington Navel orange under comparable conditions, and the trees are more productive than the normal Washington Navel. Otherwise the mature fruits of the Robertson strain and those of the parent variety are very similar and cannot be distinguished even by those who have grown and studied them.

The Dawn is a strain of the Marsh grapefruit in which the fruits mature about 1 month earlier than those of the parent variety, and the trees are unusually productive. Otherwise the Dawn and the Marsh trees and fruits have similar characteristics. The indications are that this strain will be a particularly valuable one for growing in the desert grapefruit-growing districts.

Many other citrus bud mutations are under test in cooperation with growers in California and Arizona, and some of these promise to be of commercial importance in the near future.

In the history of subtropical fruit breeding the work of Shamel, his coworkers, and the California Agricultural Experiment Station cooperators is inspiring. It is characterized by consistent, painstaking research and unwavering purpose. In the earlier years, although the work was fully appreciated by the growers in the region, scientific workers elsewhere did not give entire credence to the remarkable evidence uncovered. As time went on, Shamel and his coworkers answered their critics by applying statistical methods to the data, which gave convincing proof of the conclusions. Later, numerous bud mutations were also discovered in other fruit types, including apple, and by the time the work was no longer challenged the results achieved stood out as monumental in this particular field of research. The practical benefits to the industry can be gauged by the millions of selected buds that have been used by the citrus growers in California and elsewhere. A summary of the bud mutations and strains isolated by Shamel and his coworkers is to be found in table 8.

In Florida a bud-selection project has been in progress since 1921. The mode of procedure differs somewhat from that followed in California and was developed cooperatively by the Florida Agricultural Experiment Station and the Department of Agriculture. At the Lake Alfred Citrus Experiment Station a progeny grove has served as the basis of variety improvement. In this progeny grove standard varieties of oranges and grapefruit are represented by rows budded from parent trees of known production, several such selected parents being included for each variety. Production records have been kept, and a detailed study of fruit characteristics has been made as the basis for reselection among the original progenies. For purposes of comparison a few off-type progenies have been included, which have served well to illustrate the importance of careful bud selection in nursery propagation. Growers and nurserymen have gradually come to realize the value of such true-to-type budwood, which has been made available through the experiment station for several seasons past at a cost not greatly in advance of common commercial rates.

Other bud mutations of value to the citrus industry are Foster, Thompson, Ruby, and Davis grapefruit (described above), Silverhill satsuma orange, and Oklawaha sour orange, the three latter being seedling or nucellar mutations.

A number of citrus types and varieties are of major importance be- cause of their value as rootstocks for the varieties grown commercially. In the early development of the citrus industry, especially in Florida, seedling citrus trees were extensively planted, but gradually this practice was discontinued, as it was realized that certain rootstocks were better adapted to the soil conditions. This advance was made by the expensive trial-and-error method, and the stage reached by the 1890's was recorded by Van Deman (81). Planned experiments with rootstocks have been relatively rare. In California, Mills (46) has reported results secured at the experiment station at Pomona, and Bonns and Mertz (3) described the experiments carried out at the Citrus Experiment Station at Riverside. The important work of Webber (85) at the citrus station is an illuminating piece of research into the seedling variation of the types commonly used as rootstocks. Unfortunately, the work of Taber (fig. 16) quoted by Hume (31, pp. 209-218) at the Glen St. Mary Nursery in Florida was discontinued too soon. Traub “and Friend (77) have reported on preliminary experiments in the lower Rio Grande Valley of Texas. Akenhead, Feilden, and Hatton (1) have recently summarized the investigation on citrus rootstocks. As indicated, the horticultural utilization of rootstock types is based at present mainly on the knowledge accumulated through many years of trial and error. The final results are very valuable, though they were secured at great economic expense. The entire field has not been exhaustively explored.

In general it may be stated that in Florida, as pointed out by Camp (4), the rootstock problem is more complex than in the lower Rio Grande Valley of Texas, in California, or in Arizona.

Three major considerations are necessary in the choice of suitable rootstock—congeniality between stock and scion, resistance to diseases, and adaptation to the soil and climatic conditions. Rootstocks have been tested over a considerable period by commercial growers, and at the present time only three are extensively used—sour orange, rough lemon, and the trifoliate orange. The Cleopatra mandarin has recently received some attention as a rootstock. In Florida, trees on grapefruit stock grow vigorously in the early stages but usually prove to be shy bearers and show decline after a number of years. Sweet orange is susceptible to foot rot, and both grapefruit and sweet orange are less cold-resistant than sour orange. The commercial lemon is susceptible to diseases, and trees on it are short-lived.  The citrange has been tried as a substitute rootstock in place of the trifoliate orange for the satsuma orange, but recent observations have shown that the citrange is not reliable in the Gulf coast region, since it is evergreen and does not go dormant sufficiently during the winter season. This has resulted in the loss of the scion variety in some seasons.

Figure 16.—George Lindley Taber (1854-1929), pioneer nurseryman of Florida, who demonstrated the limitation of the satsuma orange to trifoliate orange stock, founding a new industry; introduced Duncan grapefruit, Lue Gim Gong sweet orange, and varieties of persimmons, and cooperated with the United States Department of Agriculture in the production of early citrus hybrids.

The sour orange is a useful stock in all citrus-growing regions except South Africa, where all attempts to use it so far have failed. It is compatible with most citrus types except satsuma, kumquat, and lime. In California, lemon trees on this stock may show decline after 10 to 20 years, which has been attributed to the stock. For this reason many of the recent lemon plantings in California have been budded on the sweet orange. The sour orange is highly resistant to cold and to root and crown bark diseases, but highly susceptible to citrus scab. Roots are deeply penetrating, and the stock is well adapted to clay subsoils and wet or heavy soils. In Florida it is successful on moist hammock as well as on moist flatwoods soils and on the heavier soil types in general. In California, Arizona, and the lower Rio Grande Valley of Texas it tolerates irrigation conditions. This, together with its adaptability to southwestern conditions, has made it the rootstock almost universally used in these regions. Its main drawback is its relatively slower growth on light soils as compared with rough lemon, but on heavier soils the rate of growth is satisfactory.

The rough lemon (Citrus limonia) makes a satisfactory growth even on very light and sandy soils. This makes it valuable in certain sections of peninsular Florida. It is highly susceptible to citrus scab and susceptible to foot rot and other root and crown diseases, but this apparent handicap is minimized, for on sandy soils these diseases are less troublesome than on the heavier ones. It is extensively used as a rootstock for citrus on soil types not suitable for sour orange in Florida. It is not well adapted for use with the satsuma orange, producing coarse, raggy fruit.

The trifoliate orange was once recommended as a rootstock for citrus in the lower Rio Grande Valley of Texas (48), chiefly on account of cold resistance; but it proved susceptible to foot rot and cotton root rot and was later discarded in favor of the sour orange (77). In the upper Gulf coast region it is universally used as a root stock, since it is deciduous and goes more dormant during the winter season than evergreen citrus types. It is the hardiest of the citrus types.

Recently the Cleopatra mandarin has been recommended as a rootstock in place of rough lemon, chiefly on the basis that it is more cold-resistant than the rough lemon and tends to produce better textured fruit, holding fruit later in the season in good condition. It is immune to scab and resistant to gummosis. However, experimental work has not been carried out extensively enough to warrant unqualified recommendation.

It has already been pointed out that most citrus seeds produce more than one seedling, and that any extra seedlings not of seminal origin are produced by budding from the mother plant tissue. On the surface it would appear that this is an ideal condition from the standpoint of seedling rootstock production, since it would give a large percentage of plants like the original stock, and this is true in the main. Webber (85) and Frost (20) have shown, however, that variations may occur even among such nucellar seedlings. Frost points out the influence of mutations in this connection. The work of Webber is of special interest. His experiments, which were started in 1914, show that—
citrus seedlings of the species and varieties most commonly used as rootstocks exhibit a wide range of variation. In any lot of seedlings grown from seed of the same variety and from the same source, the great majority usually are of the same general type, but from 5 to 40 per cent of them are highly variable types which apparently differ in genetic constitution from the prevailing type and from each other.

These variants may include both apogamic and seminal seedlings. They are usually small and lacking in vigor and when used as stocks induce dwarfing of the tree. His experiments show that small seedlings and “small budlings tend to produce small, low-yielding orchard trees; and that large seedlings and large budlings tend to produce comparatively large, high-yielding orchard trees.” In order to secure the desired uniformity in orchard trees, Webber advises “a moderate culling of small seedlings at the seed bed, followed by a careful roguing and destruction of all variants and small seedlings in the nursery just prior to the budding.”

   This section, like the similar sections in connection with other sub- tropical fruit crops, is based almost entirely on replies received to the questionnaire dealing with the Cooperative Survey of Plant and Animal Improvement. The information received mainly concerns the present and future objectives of breeders.

Citrus studies are carried on at experiment stations of Montemorelos, Nuevo Leon; El Yaqui, Sonora; and Emporio Macuspana, Tabasco. Infestation of fruit by the orange maggot, Anastrepha ludens Loew, constitutes the most serious problem. Methods of attack are being studied, but without practical results thus far.

At the Instituto Agronomico do Estado de Sao Paulo in Brazil, work with citrus under the direction of C. A. Kruge, head of the genetics department, is being confined primarily to the improvement of varieties by bud selection and to the improvement of the sour orange rootstock by making extensive progeny tests. The disease resistance of rootstocks is also being studied.

In Chile no breeding work with subtropical fruits is under way at the experiment stations, but Salvador Ezquierdo at Santiago has for many years introduced varieties of citrus, avocados, etc., to test their adaptability to Chilean conditions.

At the Estacion Experimental de Concordia, Argentine Republic, Signor Ruben Bence Pieres, director of the station, is conducting experiments with citrus which concern the selection of sour orange seedlings, with a view to obtaining strong, vigorous, fast-growing plants. A naturally occurring hybrid mandarin named Malvasio, with a large fruit, fine rind color, excellent quality, and late maturity, is being tested and shows promise of commercial value. The main introduced varieties, which have been selected from a large number and are being extensively cultivated, are the Marsh and Qurian grapefruit and the Valencia and Lue (Lue Gim Gong) sweet oranges. The cultivated area of grapefruit and sweet oranges approximates 5,000 acres, with about half devoted to each type. The main problem that confronts the citrus industry is the ravages of foot rot. Work is in progress for the selection of resistant stocks, as indicated above.

At the Estacion Naranjera de Levante at Burjasot, Valencia, Spain, work has been carried on in sweet orange breeding since 1927 under the direction of Manuel Herero. The work has been confined mainly to selection from open-pollinated seedlings of Washington Navel. Two improved varieties have been selected, one round and the other oval in form. These are being cultivated to the extent of about 375 acres. Hybridization work was begun in 1932 and is being carried on to the second generation after crossing. The crosses made are those between the sweet orange and the mandarin.

A. Biraghi, pathologist, Italian Department of Agriculture, Rome, reports that the chief problem with citrus culture in Sicily is in connection with the disease mal secco, chiefly affecting lemons. Attempts are being made to breed varieties resistant to this disease. Two resistant lemon varieties of unknown origin and not desirable for quality have been found locally, and promising varieties have been imported from India (2) and the United States for use in breeding experiments.

At the Superior School of Agriculture, Laboratory of Horticulture, Athens, Greece, P. Th. Anagnostopoulos has gathered together a collection of citrus species for selection and breeding work.

In Morocco, under the leadership of F. Lacarelle, director, and Ch. Miedzyrzecki, geneticist, Experimentation Fruitiére et Maraichére, Rabat, citrus fruit improvement is being studied through several methods: (1) Bud selection, (2) hybridization to secure improved varieties and varieties resistant to disease, and (3) selection of stocks. The varieties receiving the most attention are Washington Navel, Valencia Late, and Clementine. A special study of the seed content of the Clementine led to the conclusion that seed production is largely dependent on the proximity of other varieties, especially of mandarin oranges, to furnish pollen to the flowers of the Clementine, self- fertilized flowers being almost seedless (38, 39).

At the Jewish Agency Agricultural Research Station, Rehoboth, Palestine, the experimental work was started by J. D. Oppenheim and is now being continued by Ch. Oppenheimer. The work is confined primarily to improvement by bud selection, which was begun in 1934, and by hybridization, begun in 1933. The first-generation hybrids have not reached fruiting stage. The hybrids are from crosses of the Jaffa orange with other early and late varieties, of the blood orange with the tangerine, and of the sour orange with the sweet lime. The station has gathered together a collection of about 70 citrus varieties.

Theoretical studies are under way on the inheritance of peel See in citrus and on the effect of pollination on number of seeds.

At the Imperial Horticultural Experiment Station at Okitsu, Japan, T. Tanakawa is carrying on investigations in bud variation in early- and late-maturing satsuma oranges. Sixty-one late strains and 42 early strains are being tried out. The work was started in 1925 and is still in progress. Citrus hybridization work was begun in 1909.

Tyozaburo Tanaka, professor of horticulture at the Taihoku Imperial University, Taiwan, Japan, is engaged in citrus breeding. Kimijiro Noro, pomologist at the Shizuoka (Japan) Agricultural Experiment Station, has made a collection of 80 bud mutations of the satsuma orange.

At the Department of Horticulture, Lingnan University, Canton, China, G. Weidman Groff, professor of horticulture, and Pui-Man Lei, Qu-Nin Shiu, and A. N. Benemerito, pomologists, have gathered together a comprehensive citrus collection for use in the selection of desirable types of citrus and citrus relatives, particularly from the south China area, with numerous introductions from abroad. Some of the systematic work on this collection has been carried out in collaboration with the United States Department of Agriculture.

R. D. Fordham, deputy director of gardens, United Provinces, Saharanpur, India, reports that citrus rootstock trials are under way, including Khatta (Karna lime), Citrus limonia, Sylhet, Bijori, Sadaphal, Jamberi, Bilhari, Galgal; Sweet Galgal, Turanj, and sweet lime.

B. Nazareth, superintendent, Modibag Garden, College of Agriculture, Poona, India, reports that studies of mutations occurring in the principal citrus varieties, Santra Mosambi and Ladoo, have been made, and many well-marked sporting forms have been recognized. Breeding work thus far has been confined to mass selection both for rootstocks and for scion varieties. Collections of scion and rootstock varieties are being made to initiate systematic breeding of fruit crops adapted to different soil and climatic conditions. Attempts at hybridization made in 1912, 1914, and 1918 gave negative results. Polyembryony in citrus has been studied, showing varying percentages, from none in the pummelo to 60 percent in the Ladoo orange. The Marsh (seedless) grapefruit has been introduced for culture in the Bombay region.

The Department of Agriculture, Zanzibar, East Africa, is studying local and imported varieties of citrus in a series of plots. This work is being carried out by A. Q. Findley, director.

At the Department of Agriculture and Stock, Brisbane, Queensland, bud selection of citrus has been practiced for many years and Government certified budwood is available for nurserymen. The chief sweet orange varieties grown are Washington Navel, Valencia, Jaffa, Sabina, and Joppa. The lemon varieties are Lisbon and Villa Franca. The mandarin varieties are Beauty of Glenn Retreat, Emperor of Canton, Scarlet, and Fewtrel Early. The grapefruit variety is Marsh, and Seville sour orange is grown for marmalade. Sour orange and rough lemon rootstocks are used for citrus except that mandarin is grown on mandarin.

At the Department of Agriculture, Division of Horticulture, Melbourne, Victoria, J. M. Ward, superintendent of horticulture, has carried on work in bud selection since 1931, and up to the present one variety has been propagated—a, thin-skinned early navel orange, which has not as yet been named. The department also supplies buds of selected strains of the Washington Navel and Valencia to growers and nurserymen. No work in hybridization has yet been carried out.

At the Department of Agriculture, Division of Horticulture, Wellington, New Zealand, J. A. Campbell, director, has gathered together a representative ‘collection of citrus species and varieties for intensive study.

H. Wenholz, director of plant breeding, Department of Agriculture, New South Wales, Australia, reports that citrus breeding is being carried on at the Grafton Experiment Farm and the Hawkesbury Agricultural College with the object of producing “a late hanging navel orange” and a late Valencia of fine texture. A cross has been made between Valencia and a very fine-textured seedling variety.

Figure 17.—Embryo development and “nucellar embryony” in Citrus and Poncirus, after Osawa (50): Nos. 72, 80-85, trifoliate orange; 73-16, satsuma orange; 77, Washington Navel orange. (Original magnifications changed to conform to present reduction.) 72, Two-nucleated embryo sac, X570. 73, An ovule showing embryo sac, nucellus, and inner integument, X175. 74, Details of embryo sac of no. 73, with four nuclei, X570. 75, Embryo sac with eight nuclei, X 570. 76, A mature embryo sac with egg apparatus, polar nuclei, and antipodal cells, X570. 77, The same, X 570. 80, Micropylar portion of an embryo sac, showing fertilized eggs, pollen tube, and endosperm nuclei, especially some large nucellar cells containing large nucleus and much cytoplasm, X175. 81, The same, X 265. 82, Micropylar portion of an embryo sac showing two-celled embryo and endosperm nuclei, X175. 83, An older ovule, showing embryo sac, embryo, endosperm nuclei, nucellus, and integuments, X 20. 84, Upper portion of embryo sac of no. 83 more magnified, showing six-celled embryo, endosperm nuclei, and nucellus, X 175. 85, Upper portion of an embryo sac, showing polyembryony, X 370.

A study of the chromosome numbers and chromosome behavior in citrus is basic to a consideration of citrus genetics. According to Frost (20), it may be assumed that bud-variation types originate primarily either as gene (point) mutations or as chromosomal aberrations, and an understanding of these fundamental facts may also throw some light on the great variability of of F₁ hybrids between species.

Strasburger (66) determined the haploid chromosome number as 8 in the sweet orange (Citrus sinensis), the sour or bigarade orange (C. aurantium), and the citron (C. medica). Osawa (50) reported that the haploid chromosome number in the satsuma orange (C. nobilis var. deliciosa) was probably 8 (fig. 17).

In 1924-25 both Frost (fig. 18), at the Citrus Experiment Station, Riverside, Calif., and Longley, of the Department of Agriculture, reported important work on the cytology of citrus. Frost (18) determined the chromosome numbers in two varieties of sweet orange (Citrus sinensis) and one variety of grapefruit (C. grandis), and in each case the haploid number was 9. He observed neither polyspory nor polycary.

Figure 18.—Howard B. Frost, associate plant breeder, California Agricultural Experiment Station, since 1912, has made important contributions to the cytology and genetics of citrus and is the originator of the Kara, Kinnow, and Wilking mandarins and the Trovita sweet orange.

Longley (45) made chromosome counts in 24 citrus species and Citrus relatives and verified the basic chromosome number of citrus as 9, but he also found 1 tetraploid species, the Hong Kong kumquat (Fortunella hindsii (Champ.) Swingle). He observed both polyspory and polycary in many forms. He states that “irregularities in chromosome pairing at diakinesis and in their distribution at meiosis were frequently noticed. The outcome of such irregularities was the presence of tetrads containing more than the expected four pollen grains.” In grapefruit, limes, and limequats polyspory was often observed. He points out that there may be “a relation between irregular chromosome numbers and the production of citrus with supernumerary chromosomes.” According to Longley, two factors, however, may hinder the spontaneous appearance of such polyploid forms—the possibility that only sex cells with 9 chromosomes are viable, and the infrequent use of seeds as a means of propagation. However, in the early history of citrus culture, seedlings were commonly used in planting groves, and the disappearance of polyploid forms is apparently because they are of little or no value in horticulture, as Frost points put. Longley emphasizes the possible value of the tetraploid Fortunella hindsii in hybridizing with closely related diploid forms to increase the chances of obtaining forms with unusual chromosome complexes.

Later, in 1925, Frost (19) reported on the discovery of certain "thick-leaved" apogamic seedlings of sweet orange (Citrus sinensis) and lemon (C. limonia) which proved to be tetraploids. He had observed such types to the extent of several percent of the total progeny in some cases in 12 horticultural varieties of citrus, representing 4 species, and in 2 botanical varieties, and he points out that they also may prove to be tetraploids.

As to the cause of these forms, Frost postulated the possibility of “islands” of tetraploid tissue in the parent trees, so that not all tetraploid seedlings represent distinct doubling of chromosome number. Against this view he found that the thick-leaved forms have not been found in mature trees by Shamel and his coworkers, and stated that this might be due to the slower growth of tetraploid cells which might keep them from multiplying sufficiently to dominate the apical meristem, rather than to failure of tetraploidy to originate outside the nucellus.” He further supported the hypothesis by citing the fact that in many cases a thick-leaved seedling has developed from a seed giving two or more nonhybrids, and in such cases the other apogamic seedlings from the same seed have almost always been normals.

In connection with these thick-leaved forms, Frost pointed out that they have not given promise of direct horticultural value, but they may be an aid in producing triploids or modified triploids by crossing with ordinary diploid forms, and that triploids would probably be practically seedless.

In 1926 Longley (45) reported his findings with reference to a triploid trigeneric hybrid, produced by Swingle and coworkers of the Department of Agriculture. This is a cross of the limequat (Fortunella margarita Swingle X Citrus aurantifolia) with Fortunella hindsii. Longley found 13 bivalent and a single univalent chromosome as the reduced number, and he found indications of only slight irregularity in chromosome distribution durine meiosis in the triploid plant.

These findings led him to believe that triploids may be produced by appropriate crosses, and that in the case just cited or similar crosses there is a possibility of producing a seedless kumquat. More recently Longleyψ has found a second triploid, a sister hybrid of the first triploid found. Since most triploid plants are sterile, such crosses, it is hoped, may lead to developing seedless fruits. Longley later (April 1928) found an individual of Triphasia trifolia P. Wilson having 18 chromosomes.

Figure 19.—Herbert J. Webber, pioneer worker in breeding citrus and other crops; United States Department of Agriculture, 1892-1907; Cornell University, 1907-12; California Agricultural Experiment Station, since 1913. He has carried on important work on rootstock variation as influenced by polyembryony.

In 1719 Leeuwenhoek noticed two embryos in orange seed, but it was not until 1878 that Strasburger (65) explained the true nature of the phenomenon of polyembryony in citrus as sporophytic budding from nucellar tissue. Frost (79) in 1925 found that in a minor portion of nucellar embryos—less than 1 percent—two hybrids were produced from one seed. Instances have been noted of three and even four—only two cases of the latter—apparently true hybrids produced from a single seed.† In one instance (lemon X trifoliate orange) out of 782 seeds, 16 produced 2 hybrids from 1 seed, with 1 producing 3 and one producing 4 hybrids, which is slightly more than 2 percent of “doubling.”

Webber (83) (fig. 19), Frost (20), Toxopeus (74), and Torres (73) have shown that citrus types vary widely in the percentage of nucellar embryos produced (fig. 17). In a recent study made by Torres in the Philippine Islands, based on 50-seed samples, only the pummelo type did not exhibit polyembryony (table 2).

After summing up the evidence with reference to the effect of pollination on polyembryony, Frost (20) stated that it appears very probable that citrus seeds do not develop without pollination, although seedless fruits sometimes develop without pollination even in varieties normally seedy.” This would indicate that nucellar budding, which produces apogamic embryos, is at least very largely dependent on some growth stimulus due to the fertilized egg, as suggested by Strasburger (66) and Webber (84).

Frost (20) has shown that in interspecific crosses there is a negative correlation between the total number of seedlings and the percentage of hybrid seedlings, which shows the possible effect of competition. among the embryos within the seed. During such developmental selection acting within the soma of the parent, the fertilized egg may be crowded out by the apogamic embryos, depending on the number of apogamic embryos that start and on the position and relative age and vigor of the two classes of embryos. Evidence as to selective elimination during germination was secured by noting the difference between the number of dissected embryos and the germination percentage in similar lots. During germination, survival may be determined by the “difference in size, vigor, position, morphological completeness, and susceptibility to infection.” It was also noted that albinism causes the early death of many seedlings from some parents.

The variation in the number of nucellar embryos produced within a variety, and the possible effect of environmental conditions on such variation, led Traub (76) in 1936 to offer the hypothesis, based on preliminary experiments, that the number of nucellar embryos produced might be artificially varied by difference in food supply. If the preliminary results can be firmly established by experiments now in progress, an effective method for use in breeding work would be provided.

Swingle (69) postulated that citrus varieties propagated as clones are subject to senescence with age, and he cited the supposed disappearance or reduction in size of spines in well-established clones as one of the clearest indications of such senescence. He claimed that such clones might be at least partly rejuvenated, that is, become more spiny and vigorous, for instance, when seedlings from nucellar embryos were used as a method of establishing a new clone. This supposed nutritional effect of the embryo sac on the nucellar embryo developed within he named the “new-life”’ or “neophyosis” hypothesis.

However, there are no facts to prove that citrus clones are subject to senescence, and it is questionable whether the supposed rejuvenation is explainable as a permanent genetic factor change due to a nutritional effect or may be better explained by Frost’s (20) theory of “islands” of mutating tissue in the nucellus.

Mass selection from open-pollinated seedlings was the method of citrus breeding followed by the early citrus growers. It has yielded important results over long periods, but it is not now followed by any of the United States workers in this field. The search for bud variations and the use of hybridization are so much more promising that the method will be revived only for use in special cases.

The practical bearing of a rapid mutation rate in citrus has already been discussed in detailing the improvement of varieties. The question of the cause of such variations remains to be considered. It has been pointed out that Frost had postulated islands of tetraploid nucellar tissue as a possible explanation of the variation in apogamic seedlings, and the little that is known as to the cytological basis of citrus bud mutations has been ably discussed by Frost (20). The worker in this field does not have a background of abundant experimental data and must make use largely of the principles established by workers with more facile plant material such as maize, Nicotiana, etc., in developing useful theories.

In the former case, at least, their somatic expression is doubtless favored by the presence of numerous heterozygous recessive genes. The production of recognizable bud variations, then, requires bud formation in an area of variant tissue, and may often be due to irregular tissue development in periclinal chimeras. The abundance of bud variation with some citrus forms apparently depends upon a permanent chimeral condition of the types in question.

The selection of bud mutations as a tool in citrus breeding has yielded some important results. Apogamic seedlings following interspecific crossing have given rise to the Davis grapefruit, the Silverhill satsuma orange, and the Oklawaha sour orange. In this group may be included the Everglade and Palmetto limes and Weshart and Trimble tangerines, although the supposed difference in these strains from the parental varieties has not warranted their continuance as distinct varieties. Evidence as to the behavior of the Lue orange indicates that this variety is apparently a nucellar seedling of the Valencia variety. A group of unnamed navel orange seedlings derived from seed of the Washington Navel pollinated with trifoliate orange pollen but showing no hybrid characters offer promise in securing new navel varieties adapted to Gulf coast conditions. In the same series, crosses made on the Thomson Navel gave only worthless fruits of the dry type, while a large percentage of the apogamic seedlings of the Washington Navel are vigorous growers and produce juicy fruits of more or less merit. If some of these prove to fruit satisfactorily under Florida and Gulf coast conditions, a navel variety may be found to meet the need in this section.

The work of Shamel and his coworkers in bud selection has preserved the original strains of commercial varieties and also yielded some superior new ones, as already detailed.

Chace, Church, and Denny (8, 9) studied the inheritance of fruit composition in 18 mutant strains of Washington Navel orange and several mutant strains of Eureka and Lisbon lemons isolated by Shamel, Scott, Pomeroy, and Dyer (63, 64). Chace and coworkers concluded that differences in the chemical composition of fruit exist between mutant strains, and that these are heritable. In Washington Navel strains some of these differences were closely connected with physical differences and others not. The differences generally found were in quantities of peel, oil, insoluble solids, and acids. Less variation was found in the specific gravity of the fruits and in the soluble solids and sugars of the juice. Strains of fruit with smooth skin were found to contain only small quantities of oil. In the Lisbon lemon mutants significant differences were found in specific gravity of fruit, proportion of rind, and percentage of acids; and in the Eureka lemon, in percentage of acids.

Haskins and Moore (27) observed premature flowering, albinism, fasciation, twisting, and peloric leaf formation in citrus seedlings grown from X-rayed seeds.

Selection within self-fertilized lines as a tool has little or no value in citrus breeding on account of polyembryony in most citrus types. The experience of Frost (20) at the Citrus Experiment Station has shown that there is great loss of vigor and fertility with selfing. Toxopeus (74) and others have shown that the pummelo has no nucellar embryony and is usually self-pollinated. The progeny are usually quite uniform, and in such cases it is possible to develop more highly homozygous races by selection within self-fertilized lines.

Hybridization as a means toward the improvement of citrus varieties was undertaken by the Department in 1892 and has been continued ever since, and the progenies produced are being extensively tested in cooperation with the California, Florida, Texas, and Alabama experiment stations and many citrus growers (70, 71, 84, 86). Some concrete results of definite value have been achieved as a result of this cooperation, as detailed under the improvement of citrus varieties above. Frost, in California, has carried on valuable work in this field since 1914, which has recently yielded promising citrus varieties. The work in Florida carried on by Camp and Jefferies, in Alabama by Yates, and in Texas by Yarnell and Wood was begun quite recently, and not enough time has elapsed to yield any definite results.

Breeding work by the Department was initiated by Webber and Swingle and is carried out on the cooperative testing basis with the agricultural experiment stations and citrus growers in the subtropical fruit regions. The major part of the work is concerned with interspecific hybrids in which (1) the mandarin orange type is crossed with other citrus types with the object of securing high color of rind and flesh and also the “bouquet” of the tangerine in hybrid forms; (2) the lemon and lime are crossed with each other and with other citrus types; (3) bigeneric crosses of the trifoliate orange are made with species in other genera, primarily to increase the hardiness of the offspring; and (4) various crosses are made to explore the possibilities in other directions. The practical results from this work have been previously considered in the text and are summarized in table 7 of the appendix.

Among the interspecific crosses that have given the most interesting results is the cross of the mandarin orange species on the grapefruit. In no case were hybrids produced when grapefruit pollen was transferred to the mandarin orange stigma, but the reverse operation has yielded abundant results.

The first crosses of this nature were made by Swingle in 1897 and by Webber in 1898. Webber and Swingle found wide variation in the F₁ progeny. Out of the first crosses two varieties were introduced, as already indicated, but in these the susceptibility to scab of the grapefruit parent was apparently dominant, and in addition the fruits were of such character. that the keeping and shipping quality was unsatisfactory.

During the period 1908-12 Swingle, E. M. Savage, and F. W. Savage made a second series of crosses of a similar nature. The results were similar to those already stated, except that a number of the progeny were highly resistant to citrus scab and also possessed good shipping quality. An attempt was made to proceed a step farther through selection from segregating seedlings in the F₂ generation, but on account of excessive nucellar embryony only one seminal seedling of merit was secured, which gave rise to a tangelo variety of promise, the San Jacinto. This was introduced in 1931.

A further attempt was made to secure desirable types by backcrossing the tangelo on the grapefruit, and in this case a fair number of seminal offspring were secured. In growth habit these resembled the pollen parent, the Sampson tangelo, but they were less vigorous. The fruit of two such backcrosses proved to be small, round, pink-fleshed, of low acidity and high sugar content, characters not present in either parent. One variety of the pink-fleshed backcross has been introduced as the Wekiwa.

Crosses made between the satsuma orange and the sweet orange have given rise to types somewhat similar to those secured by the tangerine-grapefruit cross with the flesh color and shape of the satsuma and the tight rind and size of the sweet orange (Ruby), but with high acidity and late maturity not found in either parent.

The third series of grapefruit-tangerine crosses were made by Traub, Robinson, and Savage, 1934-36, with the object of producing "seedless" tangelos. Marsh and its mutation, Thompson, and Davis, all “seedless” varieties, were used in place of Bowen, a seedy variety, which was chiefly utilized in previous crosses. These so-called "seedless" varieties are highly self-sterile and intersterile and produce few or no seeds even in mixed plantings.

Crosses between the tangerine and the sweet orange generally gave types similar to the latter but of small size.

Among crosses between the lime and the lemon, the Perrine lemon, a promising variety, was found to be immune to citrus scab and lime withertip. It has already been mentioned and will be discussed more fully under disease resistance later.

Citrus hybrids involving trifoliate orange (Poncirus trifoliata) as one parent have given rise to interesting results. The crosses were made by Swingle and Webber beginning in 1893 and at several times subsequently. In the first series 212 crosses were made and 13 hybrids (citranges, i. e., trifoliate X sweet orange) were secured. Most of these had the trifoliate orange as the seed parent, but in one case (Rusk citrange) the sweet orange produced the seed. There was wide variation in the characters of the hybrids. In most cases the leaves were of the trifoliate type, but unifoliate types were also secured. All were evergreen in habit; in fruit character there was also great variation in size, color, etc. In all cases the rind oil character was inherited from the trifoliate parent, and in most cases the juice character was intermediate. In none of the fruits was the juice character sufficiently like that of the sweet orange to give these fruits any prospect of commercial usefulness. Most of the progeny produced only nucellar embryos when an attempt was made to secure an F₂ generation by self-pollination, except in the case of the Sanford and Phelps citranges, which showed hybrids with segregation for leaf characters. Seedlings of these segregating varieties were distributed for trial but thus far have produced nothing of special merit.

Hybrids were later secured by crossing the citrange and the kumquat. As a result the objectionable oil content of the fruit was reduced and an acid fruit type secured which has been called the citrangequat. One of the progeny, the Thomasville, shows high resistance to citrus canker, which was inherited apparently from the kumquat ancestor. The citrange crossed with calamondin has given an acid fruit type in which the disagreeable oil is eliminated. It has been called the Glen citrangedin. This type is relatively frost-resistant. Attempts made to secure the F₂ generation have been unsuccessful, since only nucellar embryos were secured.

The lime crossed with the kumquat has given rise to small acid- fruited progeny which are immune to lime withertip and decidedly more frost-resistant than the true limes.

Various other crosses were made involving pummelo, sour orange, and other citrus species. The most interesting is the cross between the Eustis limequat, with a haploid chromosome number of 9, and the Hong Kong kumquat (Fortunella hindsii), with a haploid chromosome number of 18. This has given rise to a triploid hybrid, as reported by Longley, resembling the Hong Kong kumquat but larger and of greater vigor. It is of potential value in further crossing to secure seedless kumquat types, since many triploids are self-sterile and intersterile.

the arrangement in syngamete (zygote) of the chromatin and other hereditary substances derived from the parental gametes and the persistence of this arrangement in the cells produced by the subdivision of the syngamete.

It is assumed that the particular zygotactic arrangement taken up by the chromosomes of the parental gametes usually persists with little or no change throughout the life of the organism * * *. The fundamental idea underlying the term zygotaxis is that the architecture of the zygote with reference to its idioplastic particles, as well as its mechanisms for transmitting hereditary tendencies into expression, is determined to some extent at the moment of fusion of the two parental gametes and that this arrangement of parts is transmitted to the cells of the organism to which the zygote gives rise.

The hypothesis of zygotaxis was vigorously opposed by Hagedoorn and Hagedoorn (25) and by White (89). The former workers suggested that variable F₁ progeny in citrus crosses apparently are due to habitual self-sterility and the sexual production of seeds. White points out that the F₁ variation in citrus hybrids, in the light of the data at hand, apparently results from differences in the gametic composition of the heterozygous parents.

enough is known of the production of new characters by new combinations of genes in crossing to warn us against setting any narrow limits to the probable results of recombinations in crosses between two highly heterozygous species. * * * It is concluded, from the available evidence, that citrus forms are in general extremely heterozygous. * * * This conception seems highly significant. * * * In the evolution of heterozygosis, polyembryony probably was an important factor.

Frost suggested that lethal and sublethal effects in selfing and crossing may be the result of homozygosis of inevitably unfavorable genes and may also be in part a result of incompatible recombinations. Heterosis, however, according to Frost, is probably more often secured in crossing.

Winston, Bowman, and Bach (90) studied the resistance of citrus types and varieties to sour orange scab, Sphaceloma fawcetti, and found the following not susceptible: Kumquat; citron; Kansu orange (Citrus junos Tan.); Mexican, Woglum, and Tahiti limes; Royal and Triumph grapefruit; Cuban shaddock; Bergamot orange; most sweet oranges; and Cleopatra mandarin. The work of Peltier and Frederich (52) has shown that citrus scab cannot survive under California conditions, indicating that susceptibility under certain conditions may not be apparent.

Fulton (22) made a study of the susceptibility of citrus types and varieties to Key lime withertip or anthracnose (Gloeosporium limetticolum). He found that the West Indian (Key) lime and the Dominican (thornless) lime are highly susceptible. Other lime varieties have not given undoubted evidence of susceptibility. Hybrid West Indian limes, sweet orange, grapefruit, lemon, Citrus nobilis, and others have proved immune. Since only two types are highly susceptible, it is suggested that other closely related varieties might be substituted in culture. It is interesting to note that susceptibility seems not to be a dominant factor in F₁ hybrids.

Peltier and Frederich (51) made extensive studies of the susceptibility of Citrus and related genera to citrus canker (Pseudomonas citri) under greenhouse and Gulf coast conditions. Lee (41, 42) made similar studies in the Orient. Although there is no immediate need for varieties and stocks resistant to citrus canker, since the epidemic in the upper Gulf coast and in Florida has been practically eradicated since 1925, it is of advantage to select resistant types where possible as an insurance in case the disease should be inadvertently reintroduced. Grouped in order of susceptibility, varieties of grapefruit and pummelo (shaddock) proved extremely susceptible (except two pummelo varieties, Hirado and Siam, and the Triumph grapefruit). Lemon varieties and the trifoliate orange (Poncirus trifoliata) are only slightly less susceptible than grapefruit. Classed as moderately susceptible are the sweet oranges, sour oranges, citrons, and limes, except that the Tahiti lime is much less susceptible than the Mexican lime. The mandarin group (Citrus nobilis varieties) as a whole is only slightly attacked, as is also the calamondin. The outstanding resistant members of the edible citrus fruit group proved to be the kumquats (Fortunella spp.), except the susceptible Hong Kong kumquat (F. hindsii), resistance in kumquats amounting to practical immunity under field conditions.

The extensive citrus canker tests made by Peltier and Frederich (51) afforded opportunities for testing the numerous hybrids developed by the Department, these hybrids involving reciprocal crosses between numerous species and varieties ranging in their reaction to canker infection from extremely susceptible to very resistant. The results of these tests may be briefly summarized as follows. All the trifoliate orange first crosses proved quite susceptible to canker, like the parent, Poncirus trifoliata. In the second cross, that is, citranges crossed with other parents, those hybrids with the mandarin orange or the kumquat as one parent proved decidedly resistant; in fact, the citrangequat (Thomasville variety) proved practically immune, fully as much so as the kumquat. Likewise the limequat and the orangequat can be regarded as similar to the satsuma in resistance. The calamondin, while somewhat resistant itself, does not carry this resistance into the hybrid, with, one exception, the citrangedin (citrange X calamondin). Most of the grapefruit hybrids have proved quite susceptible, although certain of the tangerine and satsuma crosses with grapefruit (tangelos) show enough canker resistance to place them beside the mandarin oranges in their resistance to canker. Peltier and Frederich conclude: “In the search for promising canker-resistant plants the results of over 4 years’ investigations seem to point to the fact that our best plants will come from the hybrids.”

Gummosis, a disease caused by Phytophthora citrophthora, has been studied by Klotz and Fawcett (37), who tested 78 species and varieties for resistance. The sour oranges proved very resistant, while the lemons were most susceptible. Klotz (36) has shown that the resistance apparently is due to some cellular product of the host that has an inhibiting action on the fungoid enzymes.

Toxopeus (75) reported that the factor or factors for resistance to foot rot (Phytophthora parasitica) might possibly be recessive, and in that case he suggests that selection for individuals resistant to this disease be made in the second generation (F₂) after crossing.

Preliminary tests have been made indicating how capacity to produce vitamins is transmitted in citrus hybrids. The only published report (71) has to do with tests made in 1928 with the Sampson and Thornton tangelos, comparing the vitamin B content of these hybrids with that of the parental varieties, grapefruit and tangerine. These tests, made by the Bureau of Chemistry and Soils of the Department, were summarized as follows:

Both charts are in agreement in showing tangerine juice to be a better source of vitamin B than the juice of the other fruits fed, and also that tangelo juice is approximately equal in vitamin B potency to grapefruit juice. Therefore, with respect to vitamin B production the tangelo has inherited the characteristics of the grapefruit (71, footnote, p. 3).

The Thornton and Orlando tangelos (range 0.33 to 0.35) gave results approximating those of the Bowen grapefruit parent (average 0.35); the Sampson, Minneola, and Seminole (range 0.18 to 0.28) proved more like the Dancy tangerine parent (average 0.24). One variety that was grouped with the tangelos, the Umatilla, but that in fact is a hybrid between satsuma and Ruby oranges, gave a much higher content of ascorbic acid (0.40) than the true tangelos. The Clement tangelo, which has the Clementine as the pollen parent, had the highest content (0.64). The Clementine is reputed to be a natural hybrid between the sour orange and the tangerine and has a higher content of ascorbic acid (0.37) than the Dancy (0.24).
The Perrine lemon—a hybrid of lemon and lime—appears to inherit from the lemon parent rather than from the lime, having an ascorbic acid content of 0.40 as compared with 0.22 in the lime.
The limequat (Lakeland) ranks low (0.17), like the lime parent. It is interesting to note that the sweet lime (often called sweet lemon) has a fairly high ascorbic acid content (0.33), despite its almost complete lack of citric acid. Apparently there is no positive correlation between citric and ascorbic acid content.

The hybridization work at the Citrus Experiment Station at Riverside was begun by Frost (17) in 1914, and since that date pollinations were made mainly in 1915, 1916, 1928, 1929, and 1931. More work of this nature is planned for the next 2 or 3 years. The work consists of selfing and crossing and the study of genetic variation in nucellar seedlings.

In general the work is confined to crosses between species and within the mandarin orange group, Citrus nobilis. In all cases selfing of the parents is carried along as far as practicable with the work of crossing. In all, six varieties of sweet orange, three of grapefruit, and four of lemon have been used in crossing. Within the mandarin group four varieties were used in crosses. The chief results so far have been secured by selection within the first generation (F₁), but promising F₁ hybrids have been selfed, and crosses have been made between promising hybrids and standard varieties.

The F₁ hybrid combinations represented by at least one individual include the following: Lemon and mandarin (tangemon), mandarin and mandarin (of distinct botanical varieties), orange and mandarin (tangor), grapefruit and mandarin (tangelo), grapefruit and orange (orangelo), lemon and orange (oramon), and grapefruit and lemon (emelo). The most promising hybrids not yet introduced seem to be the following: One each from King mandarin X Dancy tangerine (fruit of very gt size, fair shape, and very good flavor); from Mediterranean Sweet orange X Dancy tangerine (very good fruit color and flavor, extra early); and from satsuma (Owari) X Lisbon lemon (fruit rough, very juicy, seedless, high in acid, excellent in aroma).

The most promising results have been secured by crossing within the mandarin group. Three varieties already mentioned have been introduced, and at least two or three more are under consideration as candidates for introduction. The earliest work at the Citrus Experiment Station led to the use of the King mandarin as the main seed parent in these crosses.

During the last few years certain hybrids from crosses of lemons with grapefruit and of “hedge bergamot” with grapefruit have been studied as possible stock types for lemons, On the basis of tree characters, seed production, and resistance to inoculated phytophthora gummosis, several hybrids have been selected for further trial, and a preliminary study of seedlings has been made. At present one hybrid from Imperial grapefruit X hedge bergamot and one from Eureka lemon X Imperial grapefruit seem most promising, although it is not certain whether they have enough nucellar embryony to produce adequately uniform seedlings for rootstocks.

A general survey including 3,800 trees is under way in a search for genetic differences from the seed parent. Tetraploids are listed and studied, and a special study is being made of a few other progenies showing unquestionably variant characters (dry fruit from one parent tree, nonnavel fruit and pollen production from two navel parent trees without pollen, date-ripening fruit from one parent tree). A study of possible genetic differences in fruit shape (satsuma) and pulp color (blood orange) among progeny of the same seed-parent tree has also been undertaken.

(1). Polyembryony.—Counts of total embryos and counts of generative and nucellar seedlings have been made for horticultural varieties, tetraploid and diploid nucellar progeny, and F₁ hybrids. Among about 1,200 hybrids, 10 cases of duplicate hybrids (two identical hybrids from one seed) have been found, obviously the result of embryonic fission. Nucellar embryos seem to be somewhat less abundant with tetraploids than with corresponding diploids.
(2) Triploidy and tetraploidy—About 2 percent of about 3,800 nucellar seedlings were tetraploid and have been especially studied. These are of no horticultural value. A few hybrids have recently been produced by crossing a tetraploid with diploids as pollen parents; the reverse combination failed to produce seed. About 1% percent of about 1,200 hybrids have been proved to be triploid, none tetraploid; several times as many are suggestive of triploidy, but their chromosome constitution has not been determined.
(3) Chromosome behavior.—Preliminary studies have been made on,chromo- some conjugation and segregation in diploids, triploids, and tetraploids, and on irregularities at the microsporad stage.
(4) Clonal senescence.—Studies are in progress on various juvenile characters of young seedlings and clones, especially thorniness, scarcity of flowers, and general vigor of growth, and on their decline with increasing age from seed and increasing length of shoot growth.
(5) Chimeras and bud variation—Studies of tree and fruit characters have been made on forms that seem to be chimeral, in the general variety collection of citrus.

The citrus hybridization work by Camp and Jefferies at the Florida State Citrus Experiment Station, Lake Alfred, with acid citrus fruits was begun in 1924. Numerous crosses have been made, including 18 different combinations, with the following as parents: Calamondin; Rangpur lime; Meyer, Genoa, and Villa Franca lemons; Jamaica and Key limes; and the limequat. The F₁ progency have reached the fruiting stage and are under test.

The work carried on in cooperation with the Department is concerned with bud selection and with the testing of the citrus varieties originated by the Department.

The citrus progeny testing collection, begun in 1921, contains 52 strains of oranges, tangelos, grapefruit, and tangerines selected from record trees, and these have been intensively studied. From the superior strains budwood has been distributed in quantity to nurserymen. Selections from Thompson and Marsh grapefruit and Valencia, Hamlin, Parson Brown, Temple, and Pineapple oranges are in greatest demand.

The Department hybrids have been tested out over a long period. According to the latest report, the utilization of Rusk and Morton citranges as rootstocks for the satsuma orange is promising.

The citrus work of the Alabama Agricultural Experiment Station, begun in 1933 in cooperation with the Department, is concerned with the testing out of over 50 strains of the satsuma orange, both early- and late-maturing, and with the selection in the F₂ generation of citrange, particularly for high quality fruit and cold resistance. The experiments have not been carried on for a sufficient period to yield concrete results, although several of the recently introduced satsuma orange strains appear promising for hardiness, early bearing, and fruit quality.

The citrus breeding work begun by Yarnell and Wood at the Texas Agricultural Experiment Station in 1934 comprises a study of the adaptability of citrus varieties and citrus relatives, both for fruit and rootstock purposes; mass selection of open-pollinated citrange seedlings, primarily for cold resistance; crossing and selfing in sweet orange, mandarin orange, grapefruit, and acid citrus fruits. In all cases of crossing and selfing some cytological studies of root tips, buds, and young fruits are also carried along.

The following rootstock plants introduced primarily by the Department are under test: (1) Balsamocitrus paniculata Swingle; (2) citradia; (3) Citrus aurantifolia—Key, Mexican, Rangpur, Tahiti; (4) C. aurantium—African, Bittersweet, Brazilian, Paraguay, Roubidoux, Rough Seville, standard Sour; (5) C. bergamia Risso; (6) C. ichangensis; (7) C. grandis—Triumph; (8) C. limetta Risso—sweet lime; (9) C. limonia—Keller, Meyer, Perrine, Ponderosa, Rickart, Rough; (10) C. maxima Marr.—Chinese pummelo, Cuban and Pink shaddocks; (11) C. medica—Diamante, Etrog, Indian, Italian; (12) C. mitis— calamondin; (13) C. nobilis—Cleopatra; (14) C. sinensis—Oroville, Weldon, Raymondville; (15) citrange; (16) citrangedin—Glen; (17) citrangequat—Thomasville; (18) citrumelo (citrange X pummelo); (19) limequat—Lakeland, Tavares; (20) Suen Kat orange; (21) tangelo—Altoona, Clement, Lake, Minneola, Sampson, Seminole, Thornton, Umatilla, Wekiwa, Yalaha.

The following crosses have been made: (1) Sweet orange—Ruby X Hamlin, Valencia X Hamlin, Ruby X Meyer lemon and reciprocal, Parson Brown X Meyer lemon and reciprocal, Parson Brown X Ponderosa lemon, Pineapple X Pink Marsh grapefruit, Valencia X citron; (2) mandarin orange and hybrids—Meyer lemon X Clementine tangerine, Meyer lemon X Owari satsuma, Owari X Hamlin orange, Owari X Pink Marsh grapefruit, Thornton X Clementine, Thornton X Pink Marsh; (3) grapefruit—Meyer lemon X Pink Marsh, Pineapple orange X Pink Marsh, satsuma orange X Pink Marsh, Thornton tangelo X Pink Marsh; (4) acid citrus fruits—Meyer lemon X Pink Marsh grapefruit, Meyer lemon X Mexican lime, Meyer lemon X Clementine tangerine, Meyer lemon X Eureka lemon, Meyer lemon X citron, Meyer lemon X Ruby orange, Meyer lemon X satsuma orange, Meyer lemon selfed, Parson Brown orange X Meyer lemon, Parson Brown orange X Ponderosa lemon.

The citrus varieties originated by the Department, and also introductions from abroad, have been tested out on a cooperative basis since 1930. The most promising variety in the testing collection at the Texas substation at Weslaco up to the present time is the Texas Navel, to which reference has already been made. In cooperation with the Department more than 50 strains of the satsuma orange, both early and late maturing, are under test at Texas substations at Angleton and Winter Haven, and attempts to select desirable individuals of high quality and resistant to low temperatures in second generation citrange progeny are under way at College Station.

The Hawaii Agricultural Experiment Station (28) reports that rootstock and variety tests are in progress to determine the suitability of various combinations of scion and rootstock to different sites and soils. Many recent introductions of promise are being tested in comparison with the older standard varieties.

(1)   Akenhead, D., Feilden, G., St. C., and Hatton, R. G. 1932. INVESTIGATIONS ON THE STANDARDIZATION OF CITRUS TREES BY PROPAGATION METHODS. Imp. Bur. Fruit Prod., East Malling, Kent, Tech. Commun. 3, 48 pp., illus.
(2)   Biraghi, A. 1935. RILIEVI SU ALCUNI CITRUS A FRUTTO ACIDO PRESENTI IN INDIA IN RELAZIONE ALLA RICERCA DI FORME RESISTENTI AL MAL SECCO. Boll. R. Staz. Patol. Veg. (n. s.) 15: 424-441.
(3)   Bonns, W. W., and Mertz, W. M. 1916. EXPERIMENTS WITH STOCKS For citrus. Calif. Agr. Expt. Sta. Bull. 267, pp. [275]-301, illus.
(4)   Camp, A. F. 1931. CITRUS PROPAGATION. Fla. Agr. Expt. Sta. Bull. 227, 48 pp., illus.
(5)―――  ― Traub, H. P., Gaddum, L. W., and Staul, A. L. 1932. TYPE, VARIETY, MATURITY, AND PHYSIOLOGICAL ANATOMY OF CITRUS FRUITS AS AFFECTING QUALITY OF PREPARED CITRUS JUICES, Fla. Agr. Expt. Sta. Bull. 243, 56 pp., illus.
(6)  Camp, A. F. 1934. FUNDAMENTAL PHYSIOLOGY OF FRUIT PRODUCTION. Fla. Agr. Expt. Sta. Ann. Rept. 1933-34: 64-65.
(7)  Candolle, A. L. P. P. DE 1886. ORIGIN OF CULTIVATED PLANTS. Ed. 2,468 pp. London.
(8)  Chace, E. M., and Church, C. G. 1930. INHERITANCE OF COMPOSITION OF WASHINGTON NAVEL ORANGES OF VARIOUS STRAINS PROPAGATED AS BUD VARIANTS. U. S. Dept. Agr. Tech. Bull. 163, 23 pp.
(9)―――  Church, C. G., and DENNY, F. E. 1924. INHERITANCE OF COMPOSITION IN FRUIT THROUGH VEGETATIVE PROPAGATION, BUD VARIANTS OF EUREKA AND LISBON LEMONS. U.S. Dept. Agr. Dept. Bull. 1255, 19 pp.
(10)  Coit, J. E. 1910. THE RELATION OF ASEXUAL OR BUD-MUTATION TO THE DECADENCE OF CALIFORNIA CITRUS ORCHARDS. Calif. State Fruit Growers Conv. Proc. 37: 32-39.
(11)―――  1915. CITRUS FRUITS; AN ACCOUNT OF THE CITRUS FRUIT INDUSTRY, WITH SPECIAL REFERENCE TO CALIFORNIA REQUIREMENTS AND PRACTICES AND SIMILAR CONDITIONS. 520 pp., illus., New York.
(12)  Connor, J. T., ed. and translator. 1925. COLONIAL RECORDS OF SPANISH FLORIDA: LETTERS AND REPORTS OF GOVERNORS AND SECULAR PERSONS. v. Deland, Fla. (Fla. State Hist. Soc. Pubs. no. 5, v. 1)
(13)  Daniel, E. P. 1937. ACID CONTENT OF A NUMBER OF CITRUS FRUITS. Jour. Agr. Research. (In press.)
(14)  Duncan, A. L. 1892. REPORT ON CITRUS FRUITS. Fla. State Hort. Soc. Proc. 5: 63-64.
(15)  Fawcett, H. S. 1936. CITRUS DISEASES AND THEIR CONTROL. Ed. 2, completely rewritten and enl., 656 pp., illus. New York.
(16)  Friend, W. H. 1933. CITRUS ORCHARD MANAGEMENT IN THE LOWER RIO GRANDE VALLEY. (Section on Insect Pests by 8. W. Clark.)   Tex. Agr. Expt. Sta. Cir. 67, 56 pp., illus.
(17)  Frost, H. B. 1916. CITRUS HYBRIDIZATION AT THE CITRUS EXPERIMENT STATION. Calif. Citrog. 1(10)  : 13.
(18)―――    ―1925. THE CHROMOSOMES OF CITRUS. Jour. Wash. Acad. Sci. 15: 1-3, illus.
(19)―――  1925. TETRAPLOIDY IN CITRUS. Natl. Acad. Sci. Proc. 11: 535-537, illus.
(20)―――  1926. POLYEMBRYONY, HETEROZYGOSIS AND CHIMERAS IN CITRUS. Hilgardia 1: 365-402, illus.
(21)―――  1935. FOUR NEW CITRUS VARIETIES—THE KARA, KINNOW, AND WILKING MANDARINS AND THE TROVITA ORANGE. Calif. Agr. Expt. Sta. Bull. 597, 14 pp., illus.
(22)  Fulton, H. R. 1925. RELATIVE SUSCEPTIBILITY OF CITRUS VARIETIES TO ATTACK BY GLOEOSPORIUM LIMETTICOLUM (CLAUSEN)  . Jour. Agr. Research 30: 629-635.
(23)  Gaiser, L. O. 1930. CHROMOSOME NUMBERS IN ANGIOSPERMS. 11. Bibliog. Genetica 6:171-466.
(24)―――  1930. CHROMOSOME NUMBERS IN ANGIOSPERMS. II. Genetica 12: 161-250.
(25)  Hagedoorn, A. C. F. L., and Hagedoorn, A. L. 1914. ANOTHER HYPOTHESIS TO ACCOUNT FOR DR. SWINGLE’S EXPERIMENTS with citrus. Amer. Nat. 48: 446-448.
(26)  Han Yen-Chih. 1923. HAN YEN-CHIH’s CHÜ LU. (MONOGRAPH ON THE ORANGES OF WENCHOW, CHEKIANG.)   Transl. by M. J. Hagerty, with introd. by P. Pelliot. 34 pp. Leide. (Toung Pao ser. 2, v. 22.)  
(27)  Haskins, C. P., and Moors, C. N. 1935. GROWTH MODIFICATIONS IN CITRUS SEEDLINGS GROWN FROM X-RAYED SEED. Plant Physiol. 10: 179-185, illus.
(28)  Hawaii Agricultural Experiment Station. 1936. Horticulture. Hawaii Agr. Expt. Sta. Ann. Rept. 1936: 25-33, illus.
(29)  Henricksen, H. C. 1930. CITRUS CULTURE IN PORTO RICO. Porto Rico Agr. Expt. Sta. Bull. 33, 34 pp., illus.
(30)  Higgins, J. E., and Holt, V. S. 1914. THE PAPAYA IN HAWAII. Hawaii Agr. Expt. Sta. Bull. 32, 44 pp., illus.
(31)  Hume, H. H. 1904. CITRUS FRUITS AND THEIR CULTURE. 597 pp., illus. Jacksonville, Fla.
(32)―――  1926. THE CULTIVATION OF CITRUS FRUITS. 561 pp., illus. New York.
(33)  Ishikawa, M. 1916. A LIST OF THE NUMBER OF CHROMOSOMES. Bot. Mag. [Tokyo] 30:404-448, illus.
(34)  Jefferies, J. H. 1934. CITRUS PROGENY AND BUD SELECTION. Fla. Agr. Expt. Sta. Ann. Rept. 1933-34: 82-83.
(35)  Kellerman, M. 1915. SUCCESSFUL LONG-DISTANCE SHIPMENT OF CITRUS POLLEN. Science (n. 8s.)   42: 375-877.
(36)  Klotz L. J. 1927. INHIBITION OF ENZYMATIC ACTION AS A POSSIBLE FACTOR IN THE RESISTANCE OF PLANTS TO DISEASE. Science (n. s.)   66: 631-632.
(37)―――  and Fawcett, H. S. 1930. THE RELATIVE RESISTANCE OF VARIETIES AND SPECIES OF CITRUS TO PYTHIACYSTIS GUMMOSIS AND OTHER BARK DISEASES. Jour. Agr. Research 41: 415-425.
(38)  La Carelle, F. 1936. L’ EXPERIMENTATION FRUITIÈRE ET MARAICHÈRE AU MAROC. Compt. Rend. Cong. Hort. de Casablanca, mai 1936, pp. 1-13.
(39)―――  and Mrepzyrznckt, C. 1936. CONTRIBUTION À L'ÉTUDE DU CLÉMENTINIER AU MAROC. La Terre Marocaine, January 1936, 19-25.
(40)  Las Casas, B. De. 1875. HISTORIA DE LAS INDIAs. t. 1, cap. 83.
(41)  Lee, H. A. 1918. FURTHER DATA ON THE SUSCEPTIBILITY OF RUTACKOUS PLANTS TO CITRUS-CANKER. Jour. Agr. Research 15: 661-666, illus.
(42)―――  1921. CITRUS CANKER CONTROL: A PROGRESS REPORT OF EXPERIMENTS. Philippine Jour. Sci. 19: 129-173, illus.
(43)  LeLong, B. M., and others. 1900. CULTURE OF THE CITRUS IN CALIFORNIA. 260 pp., illus. Sacramento.
(44)  Longley, A. E. 1925. POLYCARY, POLYSPORY, AND POLYPLOIDY IN CITRUS AND CITRUS RELATIVES. Jour. Wash. Acad. Sci. 15: 347-351, illus.
(45)―――  1926. TRIPLOID CITRUS. Jour. Wash. Acad. Sci. 16: 543-545, illus.
(46)  Mills, J. W. 1902. cirrus FRUIT CULTURE... Calif. Agr. Expt. Sta. Bull. 188:5-46, illus.
(47)  Nakamura, M. 1929. CYTOLOGICAL STUDIES IN THE GENUS CITRUS. I. ON THE WASE SATSUMA ORIGINATED THROUGH BUD VARIATION. Studia Citologica, Tanaka Citrus Expt. Sta. 3: 1-14, illus. [In Japanese. English summary, pp. 13-14.]
(48)  ONDERDONK, G. 1909. POMOLOGICAL POSSIBILITIES OF TEXAS. Tex. State Dept. Agr. Bull. 9, 55 pp., illus.
(49)  Oppenheimer, H. C. 1934. KULTUR-PFLANZEN: NAME, HEIMAT, MITZUNG, CHROMOSOMENZAHL, BOTANISCHE UND TECHNISCHE CHARAKTERISTICA. Tabulae Biol. Period. 3: 349-395.
(50)  Osawa, I. 1912. CYTOLOGICAL AND EXPERIMENTAL STUDIES IN CITRUS. Jour. Col. Agr. Imp. Univ. Tokyo 4: 83-116.
(51)  Peltier, B. L., and Frederich, W. J. 1924. FURTHER STUDIES ON THE RELATIVE SUSCEPTIBILITY TO CITRUS CANKER OF DIFFERENT SPECIES AND HYBRIDS OF THE GENUS CITRUS, INCLUDING THE WILD RELATIVES. Jour. Agr. Research 28: 227-239.
(52)―――  and Frederich, W. J. 1924, RELATION OF ENVIRONMENTAL FACTORS TO CITRUS SCAB CAUSED BY CLADOSPORIUM CITRI MASSEE. Jour. Agr. Research 28: 241-254, illus.
(53)  Perrine, H. E. 1885. THE TRUE STORY OF SOME EVENTFUL YEARS IN GRANDPA’S LIFE. Buffalo, N. Y.
(54)  Pope, W. T. i 1934. CITRUS CULTURE IN HAWAII Hawaii Agr. Expt. Sta. Bull. 71, 37 pp., illus.
(55)  Robinson, T. R. 1921. THE BUD-SPORT ORIGIN OF A NEW PINK-FLESHED GRAPEFRUIT IN FLORIDA. Jour. Heredity 12: 194-198, illus.
(56)―――  1934, THE ORIGIN OF THE MARSH SEEDLESS GRAPEFRUIT. Jour. Heredity 24: 437-439.
(57)  Shamel, A. D., Pomeroy, C. S., and Caryl, R. E. 1923. BUD SELECTION AS RELATED TO QUANTITY PRODUCTION IN THE WASHINGTON NAVEL ORANGE. Jour. Agr. Research 26: 319-322, illus.
(58)―――  Pomeroy, C.S., and Caryl, R. E. 1924. BUD SELECTION AS RELATED TO QUALITY OF CROP IN THE WASHINGTON NAVEL ORANGE. Jour. Agr. Research 28: 521-526, illus.
(59)―――  Pomeroy, C. S., and Caryn, R. E. 1926. BUD SELECTION IN THE WASHINGTON NAVEL ORANGE. VI. PROGENY TEST OF A DUAL LIMB VARIATION. Jour. Heredity 17: 59-65, illus.
(60)―――  Pomeroy, C. S., and Caryn, R. E. 1927. BUD SELECTION IN THE VALENCIA ORANGE: PROGENY TESTS OF LIMB variations. U.S. Dept. Agr. Bull. 1483, 38 pp., illus.
(61)―――  Pomeroy, C. S., and Caryn, R. E. 1929. BUD SELECTION IN THE WASHINGTON NAVEL. ORANGE: PROGENY TESTS OF LIMB VARIATIONS. U.S. Dept. Agr. Tech. Bull. 123, 72 pp., illus.
(62)―――  Scott, L. B., and Pomeroy, C. S. 1918. CITRUS-FRUIT IMPROVEMENT: A STUDY OF BUD VARIATION IN THE MARSH GRAPEFRUIT. U.S. Dept. Agr. Bull. 697, 112 pp., illus.
(63)―――  Scott, L. B., Pomeroy, C. S., and Dyer, C. L. 1920. CITRUS-FRUIT IMPROVEMENT: A STUDY OF BUD VARIATION IN THE EUREKA LEMON. U.S. Dept. Agr. Bull. 813, 88 pp., illus.
(64)―――  Scott, L. B., Pomeroy, C. 8., and Dyer, C. L. 1920. CITRUS-FRUIT IMPROVEMENT: A STUDY OF BUD VARIATION IN THE LISBON LEMON. U.S. Dept. Agr. Bull. 815, 70 pp., illus.
(65)  Strasburger, E. 1878. UEBER POLYEMBRYONIE. Jenaísche Ztschr. Natur. 12: 647-667, illus.
(66)  Strasburger, E. 1907. UEBER DIE INDIVIDUALITAT DER CHROMOSOMEN UND DIE PFROPFHY-BRIDEN-FRAGE. Jahrb. Wiss. Bot. 44: 482-555, illus.
(67)  Swingle, W. T. 1918. VARIATION IN FIRST GENERATION HYBRIDS (IMPERFECT DOMINANCE)  : ITS POSSIBLE EXPLANATION THROUGH ZYGOTAXIS. Conf. Internatl. Génétique, 4., Paris, Compt. Rend. pp. 381-394, illus.
(68)―――  1914. CITRUS AND RELATED GENERA. In Bailey, L. H., Standard Cyclopedia of Horticulture, 6v. New York.
(69)―――  1932. RECAPITULATION OF SEEDLING CHARACTERS BY NUCELLAR BUDS DEVELOPING IN THE EMBRYO SAC OF CITRUS. 6th Internatl. Cong. Genetics Proc., Ithaca, N. Y., 1932, v. 2, 196-197.
(70)―――  and Robinson, T. R. 1923. TWO IMPORTANT NEW TYPES OF CITROUS HYBRIDS FOR THE HOME GARDEN—CITRANGEQUATS AND LIMEQUATS. Jour. Agr. Research 23: 229-288, illus.
(71)―――  Robinson, T. R., and Savage, E. M. 1931. NEW CITRUS HYBRIDS. U.S. Dept. Agr. Cir. 181, 20 pp., illus.
(72)  TISCHLER, G. 1927-31. PFLANZLICHE CHROMOSOMEN-ZAHLEN. Tabulae Biologicae 4: 1-82, 1927; Tabulae Biologicae Periodicae 1: 109-226, 1931.
(73)  Torres, J. P. 1936. POLYEMBRYONY IN CITRUS AND STUDY OF HYBRID SEEDLINGS. Philippine Jour. Agr. 7: 37-58.
(74)  Toxopeus, H. J. 1931, ERVARINGEN EN RESULTATEN VAN HET IN 928, 929 EN 930 UITGE-VOERDE KRUISINGSWERK IN CITRUS. [EXPERIENCES AND RESULTS IN CITRUS CROSSING.] Landbouw. Tijdschr. Ver. Landb. Nederland.-Indië 6: 807-819.
(75)  ―――1936. DIE ZÜCHTUNG VON UNTERLAGEN FÜR CITRUS SINENSIS OSB. IMMUN GEGEN PHYTOPHTHORA PARASITICA, DIE URSACHE DER “GUM-DISEASE” IN JAVA. Züchter. 8: 1-10, illus.
(76)  Traub, H. P. 1936. ARTIFICIAL CONTROL OF NUCELLAR EMBRYONY IN CITRUS. Science (n.-s.)   83: 165-166.
(77)  ——— and Friend, W. H. 1930. CITRUS PRODUCTION IN THE LOWER RIO GRANDE VALLEY OF TEXAS. Tex. Agr. Expt. Sta. Bull. 419, 60 pp., illus.
(78)  ——— Gaddum, L. W., Camp, A. F., and Stahl, A. L. 1932. RELATION OF ANATOMY AND METHOD OF EXTRACTION TO QUALITY OF SATSUMA ORANGE JUICE. Science (n. s. )  76: 298-299.
(79)  ——— Gaddum, L. W., Camp, A. F., and Stahl, A. L. 1933. PHYSIOLOGICAL ANATOMY, TYPE, VARIETY, AND MATURITY OF CITRUS FRUITS AS AFFECTING QUALITY OF PREPARED JUICES. Plant Physiol. 8: 35-80, illus.
(80)  ——— and Robinson, T. R. 1935. MATURITY AND QUALITY IN ACID CITRUS FRUITS. Fla. State Hort. SoC. Proc. 48: 173-180.
(81)  Van Deman, H. E. 1891. REPORT ON THE RELATIVE MERITS OF THE VARIOUS STOCKS FOR THE ORANGE, WITH NOTES ON MALDI-GOMA AND THE MUTUAL INFLUENCE OF STOCK AND scion. U.S. Dept. Agr., Div. Pomology, Bull. 4, 25 pp.
(82)  Vega, G. DE LA. 1731. LE _COMMENTAIRE ROYAL OU L’HISTOIRE DES YNCAS ROYS DU PERU. Transl. from Spanish by I. Baudoin, C. I. Paris.
(83)  Webber, H. J. 1900. COMPLICATIONS IN CITRUS HYBRIDIZATION CAUSED BY POLYEMBRYony. Science (n. s.)   11: 308-309.
(84)  1905. NOTES ON CITRUS HYBRIDS. Amer. Breeders’ Assoc. Proc. 1: 78-86, illus.
(85)   Webber, H. J. 1932. VARIATION IN CITRUS SEEDLINGS AND THEIR RELATION TO ROOTSTOCK SELECTION. Hilgardia 7: 1-79, illus.
(86)   ——— and Swingle, W. T. 1905. NEW CITRUS CREATIONS OF THE DEPARTMENT OF AGRICULTURE. U. S. Dept. Agr. Yearbook, 1904: 221-240, illus.
(87)   Welch, J. H. 1936. THE RISE OF THE TEXAS CITRUS SUN. Tex. Farming and Citriculture 13 (3)  : 5, 20-21; (4)  : 7; (5)  : 5, 16-17, illus.
(88)   Wester, P. J. 1913. CITRICULTURE IN THE PHILIPPINES. Philippine Bur. Agr. Bull. 27, 71 pp., illus.
(89)   White, O. E. 1914. SWINGLE ON VARIATION IN F; CITRUS HYBRIDS AND THE THEORY OF ZYGOTAXIS. Amer. Nat. 48: 185-192.
(90)   Winston, J. R., Bowman, J. J., and Bach, W. J. 1925. RELATIVE SUSCEPTIBILITY OF SOME RUTACEOUS PLANTS TO ATTACK BY THE CITRUS-SCAB FUNGUS. Jour. Agr. Research 30: 1087-1093.

[An asterisk (*) designates workers all or part of whose salaries were or are paid from Federal funds]

(1) United States Department of Agriculture, Washington, D. C.:
Early workers—*H. J. Webber, *W. T. Swingle, *Eugene May, *Frank W. Savage, *R. E. Caryl, *L. B. Scott.
Present workers—*Hamilton P. Traub, senior horticulturist; *T. Ralph Robinson, senior physiologist; ,*Edward M. Savage, assistant plant breeder; *A. D. Shamel, principal physiologist; *C. S. Pomeroy, associate pomologist; *A. E. Longley, associate botanist.

(2) Alabama Agricultural Experiment Station of the Alabama Polytechnic Institute, Gulf Coast Substation, Fairhope:
Present workers—*Harold F. Yates, acting superintendent.

(3) Arizona College of Agriculture and the Agricultural Experiment Station, Tucson: Present workers—W. E. Bryan, head of Department of Plant Breeding (date breeding).

(4) California College of Agriculture and the Agricultural Experiment Station of the University of California:
Early workers—C. S. Milliken, C. L. Dyer, F. N. Harmon.
Present workers—H. J. Webber, emeritus professor of subtropical horticulture, Riverside; Ira J. Condit, associate professor of subtropical horticulture, Riverside; L. D. Batchelor, director of Citrus Experiment Station, Riverside; Robert W. Hodgson, head of Division of Subtropical Horticulture, Los Angeles; R. E. Caryl, associate in orchard management, Riverside; H. B. Frost, associate plant breeder, Riverside.

(5) College of Agriculture and Agricultural Experiment Station of the University of Florida:
Early workers—H. H. Hume, John Belling.
Present workers—H. S. Wolfe, horticulturist in charge, Subtropical Experiment Station, Homestead; A. F. Camp, horticulturist in charge, Citrus Experiment Station, Lake Alfred; J. H. Jefferies, superintendent, Citrus Experiment Station, Lake Alfred.

(6) Glen St. Mary Nurseries Co., Glen St. Mary, Fla.:
Early workers—G. L. Taber (trifoliate rootstock investigations), H. Harold Hume (persimmon investigations).

(7) University of Hawaii and the Hawaii Agricultural Experiment Station:
Early workers—*J. E. Higgins, *V. S. Holt, *J. M. Westgate.
Present workers—*J. H. Beaumont, principal horticulturist; *W. T. Pope, senior horticulturist; *R. H. Moltzan, principal scientific aide; *J. C. Thompson, principal scientific aide; *W. B. Storey, junior biological aide.

(8) Louisiana State University and Agricultural and Mechanical College and the Agricultural Experiment Station:
Present worker—J. C. Miller, horticulturist in charge of research.

(9) Puerto Rico Agricultural Experiment Station, Rio Piedras, Puerto Rico:
Early worker—O. W. Barrett.
Present worker—Julio S. Simons, agronomist for plant introduction and propagation.

(10) Federal Experiment Station, Mayaguez, Puerto Rico:
Early workers—*O. W. Barrett, *Holger Johansen, *T. B. McClelland. Present worker—*C. L. Horne, associate horticulturist.

(11) Panama Canal Zone Experiment Gardens, Summit:
Present workers—*J. E. Higgins, consultant in plant introduction and utilization; *Walter R. Lindsay, acting director.

(12) Commonwealth of the Philippines, Department of Agriculture and Commerce (citrus, avocado, pineapple, and papaya breeding):
(a) Bureau of Plant Industry, Manila:
Early workers—C. F. Baker, T. P. Reyes, P. J. Wester.
Present workers—J. P. Torres, J. de Leon, F. Galang, E. K. Morada. (b) College of Agriculture, University of the Philippines, Department of of Agronomy, Laguna:
Present workers—N. B. Mendiola, plant breeder and geneticist; J. M. Capinpin, plant breeder and cytologist; T. Mercado, assistant plant breeder.

(13) Texas Agricultural Experiment Station:
Present workers—S. H. Yarnell, chief, Division of Horticulture, College Station ; J. F. Wood, horticulturist, Weslaco; H. M. Reed, horticulturist, Angleton.

Spain:
(1) Estacion Naranjera de Levante, Burjasot, Valencia (citrus breeding): Present worker—Manuel Herrero.
Italy:
(2) Regia Stazione Sperimentale de Olivicoltura ed Oleificio, Pescara (olive breeding):
Present worker—Julio Savastano.
(3). Regia Stazione Sperimentale di Frutticoltura e de Agrumicoltura, Arcieale Catania, Sicily (breeding for resistance to mal secco in lemons).
French Morocco:
(4) Experimentation Fruitiére et Maraichére, Rabat (citrus breeding):
a Present workers—F. Lacarello, Director; Ch. Miedzyrzecki, geneticist.
Greece:
(5) Hellenic Kingdom Superior School of Agriculture, Laboratory of Horticulture, Athens (olive and fig variety studies):
Present worker—P. Th. Anagnostopolous.

Palestine:
(6) Jewish Agency, Agricultural Experiment Station, Rehoboth (citrus breeding):
Early worker—J. D. Oppenheim.
Present worker—Ch. Oppenheimer.
(7) Agricultural School, Mikweh-Israel (collection of Jaffa orange bud mutations):
Present worker—S. Yedidja.
India:
(8) Poona, Bombay Presidency; Horticultural Section (1) Ganeshkhind Fruit Experiment Station, Kirkee; (2) Modibag (Garden), College of Agriculture, Poona. (Improvement of citrus, mango, papaya, pineapple, guava, pomegranate, fig, banana, jujube, annona.)
Early workers—W. Burns, S. H. Prayag, L. B. Kulkarni, H. P. Paranjpe.
Present workers—G. S. Cheema, horticulturist and professor of horticulture; Pp. G. Dani, assistant; S. R. Gandhi, assistant; S.S. Bhat, assistant; B. Nazareth, assistant.
(9) Department of Agriculture, Bihar (mango and papaya breeding):
Present workers—R. Zarbakht-Kahn, horticulturist; R. Shah, assistant; S. Prashad, assistant.
(10) Royal Agricultural and Horticultural Society of India, Alipur, Calcutta (variety studies of tropical fruits):
Present worker—Sydney Percy Lancaster.
(11) Government Gardens, Saharanpur, United Provinces (lime, mango, and loquat improvement):
R. D. Fordham, deputy director.
Straits Settlements and Federated Malay States:
(12) Department of Agriculture, Kuala Lampur (pineapple breeding):
Present worker—W. D. P. Olds, director of agriculture.
Siam:
(13) Department of Agriculture and Fisheries, Bangkok (variety studies of tropical fruits):
Present worker—Luang Suwan, director general.
China:
(14) College of Agriculture, Lingnan Uniyersity, Canton; Department of Horticulture (variety studies of citrus, avocado, mango, papaya, pineapple, bapee lychee, diospyros, annona, artocarpus, Chinese olive (Canaria), etc.):
Present workers—G. Weidman Groff, professor of horticulture; Pui-man Lei, Iu-nin Shiu, A. N Benemerito.
Japan:
(15) Imperial Horticultural Experiment Station, Okitsu (fig, loquat, and citrus breeding):
Present worker—T. Tanikawa, acting director.
(16) Chiba Horticultural College, near Tokyo (loquat variety studies):
Present worker—Taiji Miki, professor of pomology.
(17) seen Agricultural Experiment Station (bud mutations of satsuma orange):
Present worker—Kimijiro Noro, pomologist.
(18) Taihoku Imperial University, Formosa (citrus breeding):
Present worker—Tyozaburo Tanaka, professor of citriculture.
(19) Agricultural Experiment Station, Shirin, Taihoku, Taiwau (subtropical fruit crops breeding):
Present worker—Y. Sakurai, pomologist.
Netherland East Indies:
(20) Buitenzorg Botanical Gardens (improvement of citrus and other tropical fruits): ;
Present workers—G. J. A. Lirra, horticultural adviser; H. J. Toxopeus, geneticist.

Mexico:
(32) Estacion Experimental:
(a) Acapulco, Guerrero (mango, lime, pineapple, and avocado).
(b) Oaxaca, Oaxaca (mango, lime, avocado). &
(c) Jalapa, Veracruz (papaya, mango, pineapple).
(d) Colima, Colima (lime).
(e) Leon, Guanajuato (avocado).
(f) Queretaro, Queretaro (avocado).
(g) Coalan del Rio, Morelos (papaya).
(h) Hecelchakan, Campeche (avocado).
Honduras:
(33) Lancetilla Experiment Station (United Fruit Co.), Tela (Extensive tropical fruit variety trials):
Present worker—Wilson Popenoe, agricultural director.
British Honduras:
(34) The Agricultural Office, Belize (introduction and testing of citrus varieties):
Present worker—H. P. Smart, agricultural officer.
Costa Rica:
(35) Alan Kelso, Aparto. 246, Punta Arenas. (Private work; tropical fruit crop introduction and improvement.)
Cuba:
(36) Agricultural Experiment Station, Santiago de las Vegas (variety tests and selections of citrus, avocado, mango, banana, and pineapple):
Early workers—M. Fortun, J. Ajete.
(37) Atkins Institution of the Arnold Arboretum (Harvard University), Cienfuegos (breeding of citrus and other tropical fruits):
Present worker—H. C. Gray, director.
Jamaica:
(38) Department of Agriculture, Hope, Kingston (papaya and banana breeding):
Present worker—L. N. H. Larter, Government botanist.
Trinidad, British West Indies:
(39) The Imperial College of Tropical Agriculture (banana and other tropical fruits breeding):
Present worker—E. E. Cheeseman, professor of botany.

Chile:
(40) Salvador Izquierdo, Monedo 778, Santiago. oe Z (Private work confined mostly to introduction of new varieties of subtropical fruits.)
Brazil:
(41) Instituto Agronomico do Estada de S&o Paulo, Genetics Department (citrus and banana breeding):
Present worker—C. A. Krug, head of department.
Argentina:
(42) Estacion Experimental de Concordia, Ministerio de Agricultura de la Nacion Argentina (citrus breeding):
Present worker—Ruben Bence Pieres, ing. agr. and director of the station.

Note.—The article entitled Improvement of Subtropical Fruits other than Citrus, by Hamilton P. Traub and T. Ralph Robinson, appears in the 1937 Yearbook Separate on Improvement of Subtropical Fruits.