Although a few small schismatic sects exist within the Sikh faith, the statistical data of the census bearing on them are so meager and out of date that we have not tried to map them.

Jains constitute an exceedingly old and well-established re- ligious community that at various periods has played a major political and cultural role in Indian history. (See, in particular, plate III.B.5, relating to the early spread of Jainism during the 6th to 3rd century B.C. and plate IV.4, on which are shown major Jain centers of the 8th to 12th century A.D. and events related to the decline of Jainism during that period.) Since they are largely a mercantile community, their influence is out of all proportion to their numbers. Like the Sikhs, they are found in small numbers virtually throughout India, particularly in cities and market towns, though the area over which they are found in significant strength (more than 1% of the total population) is, as our maps depict, relatively limited. Unlike the distribution pattern of Sikhs, that of the Jains is now rela- tively stable, having varied little from 1931 to 1961. Even in the areas of their greatest concentration, notably Rajasthan and Gujarat, Jains rarely exceed 5% of the population, and in only one district, Cutch (in 1931 only), did they exceed 10%.

Sectarian differences appear to be relatively more clear-cut within the Jain population than among the mass of Hindus, judging from the relatively large proportion who, as of the 1931 census, declared themselves as belonging to a particular sect. Map X.A.5 (f), is an attempt to portray the regional dis- tribution of Jain sects on the basis of the detailed, but never- theless far from complete, data provided by the 1931 census. We have no clear information on the biases the census report- ing may entail or on shifts in affiliation that may have occurred since 1931. But one point the map makes clear, which we as- sume is still true, is that over most of the areas where Jains are found in significant numbers the several sects show a high de- gree of regional intermingling. Gujarat, where the Swetambari sect is in a clear majority, is a noteworthy exception.

The distribution of Buddhists in South Asia in 1961 is dra- matically different from that in 1931. In the former year one finds that, despite India's being the land where Buddhism has its roots, there were virtually no Buddhists in what one might consider India proper. Burma, of course, was a stronghold of Buddhism then, as it has been in varying degrees for nearly two millennia, and other areas of Buddhist strength showed up in the Himalayas as extensions of the Lamaistic Buddhist region of Tibet. Outside India, Buddhism was, in 1931 and in 1961, the dominant religion of all but two provinces of Cey- lon. (For more detailed district data on Ceylon, see plate VIII.C.3, map (b), which relates communal distributions in that island to political events in the pre-independence period.) In 1961 it was also well represented in portions of Nepal, though in no census region was it the majority faith (1931 data are lacking). The striking difference, then, in the 1931 and 1961 patterns lies well within the present boundaries of India, deriving from the emergence of the so-called neo-Buddhists in and adjacent to the state of Maharashtra. The origin of this movement in 1956 is noted on the map itself. The other map notes relative to major Buddhist sects are, we feel, self- explanatory.

Plate X.A.6 portrays the distribution patterns of three non- indigenous religions of South Asia—Christianity, Zoroastrian- ism, and Judaism—as well as of two nonreligiously defined communities that are wholly or partially nonindigenous— "Europeans" (including Americans, Australians, etc.) and "Eurasians" (of varying degrees of racial admixture, includ- ing Anglo-Indians, Luso-Indians, Burghers, etc.).

Of the religiously defined groups, Christians make up by far the largest. Although they constituted only 2.2% of South Asia's total population in 1961, the proportion of Christians has more than doubled during the 20th century and apparently continues to rise. (For India alone the proportion rose from 2.4% to 2.6% in the period 1961–71.) It is in order, then, that this notable growth, like that of the Sikhs, be documented in three maps, relating to the years 1901, 1931, and 1961, rather than the usual two, 1931 and 1961, as for other faiths. On all three maps the patterns of relative Christian strength are territorially fragmented. Regions of notable strength in- clude those of the Syrian Christians of the Malabar Coast, whose origins are said to go back to the time of the Apostle Thomas, who is alleged to have died in India; areas of active Catholic proselytism by Jesuits and others, especially during the period of Portuguese prominence, in Ceylon, along the west coast of India, and in scattered parts of Madras; areas in which Protestant missionaries have concentrated their efforts, largely among depressed castes, notably in Punjab, western United Provinces and the Telugu-speaking region; and, finally, various tribal areas to which Catholic or Protestant missionary activity has been strongly directed, in Burma (especially among the Karens), the far northeast of India proper, the Chota Nag- pur, and so forth. The strength of Christians in 1931 in a num- ber of districts containing large cities or military cantonments can be explained to a large degree, of course, by the concen- tration there of the European and Eurasian populations, as is made evident from an examination of map X.A.6 (c). Com- parable data are not available for 1961.

Map X.A.6 (e), seeks to portray a generalized pattern of distribution of major Christian denominations in South Asia as of 1961. It not only differentiates among Roman Catholics, Syrian Christians, and Protestants and suggests the relative strength of each, but also indicates the areas of prominence of specific Protestant and "Syrian" groups. Of the latter it is noteworthy that a portion, the Romo-Syrians, accept the su- premacy of the pope; another group, including the Mar Thom- ites, take communion with the Protestant Church of South India; while a number of others, most notably the "Jacobites," have maintained their independence from both the Catholic and Protestant folds. The large number of modern schisms among Syrian Christians and the general state of flux within that community preclude our judging the relative sizes of the respective groups. Protestants within South Asia generally have also been experiencing considerable denominational flux in re- cent decades.

Previously, the Protestant missionary field in India had been divided by a series of "comity agreements" whereby each of a number of the leading missionary societies agreed not to pros- elytize in areas already staked out, as it were, by any of the others among them. Consequently, a patchwork pattern arose of areas in which native Indian Christians belonged, by and large, to the particular denomination that first entered the local missionary field. As early as 1908, however, there occured the first of a number of denominational mergers, initiating a trend that continues to this day. Of particular note was the establishment in 1947 of the aforementioned "Church of South India" (CSI), and in 1970 the "Church of North India" was formed to join, among other groups, not only the "United Church of Northern India," but also the Anglican "Church of India," which included those Anglican congregations outside the area that affiliated with the CSI at its inception. A "Church of Pakistan," also established in 1970, similarly united most Protestants in that country, Anglicans included.

The Zoroastrians (Parsis) and Jews in South Asia are few enough that we can plot their distribution with much greater precision than was feasible for the other religious groups con- sidered in plates X.A.3–7. Both groups, we may note, are far more influential than their small numbers suggest. Despite their long establishment on the soil of South Asia, the two groups remain concentrated in a few primarily urban local- ities. Significantly, both groups declined markedly in numbers between 1931 and 1961, the Zoroastrians, it is alleged, be- cause of an unusually low, if not negative, rate of natural in- crease and a high rate of net emigration and the Jews because of the exodus of a large proportion of their number to Israel.

Sources

General, Censuses (see note on censuses for each country in Government Documents section of main bibliography)

Ceylon (1921), (1946), (1953), (1963); India (1881), (1891), (1901), (1911), (1931), (1941), (1951), (1961); Nepal (1961); Pakistan (1951), (1961).

N.B. Indian censuses of 1881, 1891, 1911, and 1941 were used solely to obtain data on sects/denominations of various religions; Indian and Pakistani censuses of 1951 were used solely to obtain data on displaced persons.

General, Works Other Than Censuses

I. R. A. al Faruqi and D. E. Sopher (1974, listed under At- lases); H. A. Gleason, Jr. (1946); Imperial gazetteer of India (1907–9).

For Hindu Sects (plate X.A.3)

J. N. Farquhar (1915); M. Monier-Williams (1919/1971);

D. A. Pai (1928); H. H. Wilson (1958)

For Muslim Sects (plate X.A.4)

Encyclopedia of Islam (1913–36), (1960–); J. N. Hollister (1953); S. C. Misra (1964).

For Christian Denominations (plate X.A.6)

The Catholic directory of India (1969); Church of England, National Assembly (1966); Delhi Orthodox Syrian Church Society (1964); H. Emmerich (1968, listed under Atlases); J. Gausdal (1966); India News (various years); H. Jedin et al. (1970, listed under Atlases); S. Neill (1965); Pakistan Affairs (various years); W. A. Partridge (1962); H. H. Rowley (1969, listed under Atlases); W. Stewart (1964); B. Sundkler (1954); E. Tisserant (1957); World Christian Handbook (1967).

Acknowledgments (affiliations following names relate to the time when assistance was rendered)

For assistance with the map of Hindu sects we are indebted to Professor David E. Sopher and Bharat Bhatt of the Depart- ment of Geography, Syracuse University. For assistance with the map of Muslim sects we are indebted to Sir Muhammad Zafrullah Khan, President, International Court of Justice. For assistance with the map of Christian denominations we are in- debted to Professor John Brush, Department of Geography, Rutgers, the State University [of New Jersey]; Dr. H. A. Gleason, Jr., Retired; Rev. Fr. E. R. Hambye, Professor of Church History, De Nobili College, Poona; Professor Paul Hiebert, Department of Anthropology, Kansas State Univ.; Rev. Sam V. Jones, Billy Graham Evangelistic Association; William K. Mall, Executive Secretary, Pakistan Christian Council; A. D. Manuel, General Secretary, Christian Litera- ture Society, Madras; Rev. Richard W. Taylor, Christian In- stitute for the Study of Religion and Society, Bangalore; Rev. John Thetgyi, General Secretary, Burma Christian Council.

X.A.7 and 8. Tribal Population and Tribal Religions, 1931 and 1961; "Depressed"/"Scheduled" Castes, 1931 and 1961; Scheduled Castes in Area of Neo-Buddhist Conversions, 1951; Total of "Scheduled" Communities, 1961

The two plates considered in the following text relate to communities within the population of South Asia that are deemed to be particularly disadvantaged and which, both be- fore and since independence, have by various designations been officially recognized as such. While only a small fraction of the total number of low-caste and tribal persons so considered are enumerated by the 1931 and 1961 censuses as being religiously differentiated from the remainder of the South Asian societies of which they form a part, that is, as adhering to what have been collectively called "animistic" or "tribal" religions, it will be instructive to examine the patterns of distribution of those religious groups in addition to those of the larger disadvan- taged populations of which they form a part.

With respect to South Asia's "tribal" population, it must be observed at the outset that the comprehensiveness of the term varies somewhat between the census years 1931 and 1961 and from one administrative jurisdiction to another, depending mainly on the arbitrary assignment of specific ethnic groups to the tribal or nontribal category by the census commissioners of particular provinces, states, and states agencies in 1931 or by the official "schedules" drawn up for the nation's states and union territories in 1961.

For mapping the 1931 figures on tribal population, we have deviated in a number of respects from the classification of the census itself in an attempt, on the one hand, to arrive at greater internal consistency and, on the other hand, to achieve greater comparability with 1961 usage. While generally following the enumerations provided in the keys to the social maps that ap- pear in the "Report" volumes for the respective provinces, and states, throughout the map we have classified as tribal those groups that are so described in certain areas but not in others. Thus, we have added Bhils and Bhilalas in Gwalior; Santals in the plains portions of Bihar; Gonds, Agarias, Bayars, and Kols in the eastern United Provinces, and other minor groups. By and large these changes seem justified in the light of the note entitled "Primitive Tribes" following appendix II of vol. I, part 1 of the census. But even that appendix has not been fol- lowed in its entirety, for we have retained Meitheis (Mani- puris), Shans, and Ahoms as tribal, even though the census suggests that they no longer be regarded as such. In Burma we have included in the classification "tribal" all "races" except the so-called Burma Group, Indians, Chinese, Europeans, and Indo-Burmese; additionally, the term has been extended to the relatively small number of animists within the Burma Group. Throughout India both the census and this atlas have arbitrar- ily excluded Muslim ethnic groups from classification as tribal even though large segments of that religious community, par- ticularly in the northwest of the subcontinent, are organized within local societies to which the tribal designation is custom- arily applied in ethnographic literature. In light of the several rules and qualifications just outlined, we arrive, then, at an all- India total of roughly 26.1 million tribals in 1931 as against a figure of 24.6 suggested by the census itself.

The keys to the census's social maps, on which our maps (a) and (b) of plate X.A.7 were largely based, usually, but not invariably, reported data by district or by minor princely states. Where aggregation of data for different contiguous por-