The Tribes and Castes of the Central Provinces of India

The territory controlled by the Chief Commissioner of the Central Provinces and Berār has an area of 131,000 square miles and a population of 16,000,000 persons. Situated in the centre of the Indian Peninsula, between latitudes 17°47′ and 24°27′ north, and longitudes 76° and 84° east, it occupies about 7.3 per cent of the total area of British India. It adjoins the Central India States and the United Provinces to the north, Bombay to the west, Hyderābād State and the Madras Presidency to the south, and the Province of Bihār and Orissa to the east. The Province was constituted as a separate administrative unit in 1861 from territories taken from the Peshwa [6]in 1818 and the Marātha State of Nāgpur, which had lapsed from failure of heirs in 1853. Berār, which for a considerable previous period had been held on a lease or assignment from the Nizām of Hyderābād, was incorporated for administrative purposes with the Central Provinces in 1903. In 1905 the bulk of the District of Sambalpur, with five Feudatory States inhabited by an Uriya-speaking population, were transferred to Bengal and afterwards to the new Province of Bihār and Orissa, while five Feudatory States of Chota Nāgpur were received from Bengal. The former territory had been for some years included in the scope of the Ethnographic Survey, and is shown coloured in the annexed map of linguistic and racial divisions.

The main portion of the Province may be divided, from north-west to south-east, into three tracts of upland, alternating with two of plain country. In the north-west the Districts of Sangor and Damoh lie on the Vindhyan or Mālwa plateau, the southern face of which rises almost sheer from the valley of the Nerbudda. The general elevation of this plateau varies from 1500 to 2000 feet. The highest part is that immediately overhanging the Nerbudda, and the general slope is to the north, the rivers of this area being tributaries of the Jumna and Ganges. The surface of the country is undulating and broken by frequent low hills covered with a growth of poor and stunted forest. The second division consists of the long and narrow valley of the Nerbudda, walled in by the Vindhyan and Satpūra hills to the north and south, and extending for a length of about 200 miles from Jubbulpore to Handia, with an average width of twenty miles. The valley is situated to the south of the river, and is formed of deep alluvial deposits of extreme richness, excellently suited to the growth of wheat. South of the valley the Satpūra range or third division stretches across the Province, from Amarkantak in the east (the sacred source of the Nerbudda) to Asīrgarh in the Nimār District in the west, where its two parallel ridges bound the narrow valley of the Tapti river. The greater part consists of an elevated plateau, in some parts merely a rugged mass of hills hurled together by volcanic action, in others a succession of bare stony ridges and narrow fertile valleys, in which the soil has [7]been deposited by drainage. The general elevation of the plateau is 2000 feet, but several of the peaks rise to 3500, and a few to more than 4000 feet. The Satpūras form the most important watershed of the Province, and in addition to the Nerbudda and Tapti, the Wardha and Wainganga rivers rise in these hills. To the east a belt of hill country continues from the Satpūras to the wild and rugged highlands of the Chota Nāgpur plateau, on which are situated the five States recently annexed to the Province. Extending along the southern and eastern faces of the Satpūra range lies the fourth geographical division, to the west the plain of Berār and Nāgpur, watered by the Purna, Wardha and Wainganga rivers, and further east the Chhattīsgarh plain, which forms the upper basin of the Mahānadi. The Berār and Nāgpur plain contains towards the west the shallow black soil in which autumn crops, like cotton and the large millet juāri, which do not require excessive moisture, can be successfully cultivated. This area is the great cotton-growing tract of the Province, and at present the most wealthy. The valleys of the Wainganga and Mahānadi further east receive a heavier rainfall and are mainly cropped with rice. Many small irrigation tanks for rice have been built by the people themselves, and large tank and canal works are now being undertaken by Government to protect the tract from the uncertainty of the rainfall. South of the plain lies another expanse of hill and plateau comprised in the zarmīndāri estates of Chānda and the Chhattīsgarh Division and the Bastar and Kanker Feudatory States. This vast area, covering about 24,000 square miles, the greater part of which consists of dense forests traversed by precipitous mountains and ravines, which formerly rendered it impervious to Hindu invasion or immigration, producing only on isolated stretches of culturable land the poorer raincrops, and sparsely peopled by primitive Gonds and other forest tribes, was probably, until a comparatively short time ago, the wildest and least-known part of the whole Indian peninsula. It is now being rapidly opened up by railways and good roads.

Up to a few centuries ago the Central Provinces remained outside the sphere of Hindu and Muhammadan conquest. To the people of northern India it was known as Gondwāna, [8]an unexplored country of inaccessible mountains and impenetrable forests, inhabited by the savage tribes of Gonds from whom it took its name. Hindu kingdoms were, it is true, established over a large part of its territory in the first centuries of our era, but these were not accompanied by the settlement and opening out of the country, and were subsequently subverted by the Dravidian Gonds, who perhaps invaded the country in large numbers from the south between the ninth and twelfth centuries. Hindu immigration and colonisation from the surrounding provinces occurred at a later period, largely under the encouragement and auspices of Gond kings. The consequence is that the existing population is very diverse, and is made up of elements belonging to many parts of India. The people of the northern Districts came from Bundelkhand and the Gangetic plain, and here are found the principal castes of the United Provinces and the Punjab. The western end of the Nerbudda valley and Betūl were colonised from Mālwa and Central India. Berār and the Nāgpur plain fell to the Marāthas, and one of the most important Marātha States, the Bhonsla kingdom, had its capital at Nāgpur. Cultivators from western India came and settled on the land, and the existing population are of the same castes as the Marātha country or Bombay. But prior to the Marātha conquest Berār and the Nimar District of the Central Provinces had been included in the Mughal empire, and traces of Mughal rule remain in a substantial Muhammadan element in the population. To the south the Chānda District runs down to the Godāvari river, and the southern tracts of Chānda and Bastar State are largely occupied by Telugu immigrants from Madras. To the east of the Nāgpur plain the large landlocked area of Chhattīsgarh in the upper basin of the Mahānadi was colonised at an early period by Hindus from the east of the United Provinces and Oudh, probably coming through Jubbulpore. A dynasty of the Haihaivansi Rājpūt clan established itself at Ratanpur, and owing to the inaccessible nature of the country, protected as it is on all sides by a natural rampart of hill and forest, was able to pursue a tranquil existence untroubled by the wars and political vicissitudes of northern India. The population of Chhattīsgarh thus constitutes to [10]some extent a distinct social organism, which retained until quite recently many remnants of primitive custom. The middle basin of the Mahānadi to the east of Chhattīsgarh, comprising the Sambalpur District and adjoining States, was peopled by Uriyas from Orissa, and though this area has now been restored to its parent province, notices of its principal castes have been included in these volumes. Finally, the population contains a large element of the primitive or non-Aryan tribes, rich in variety, who have retired before the pressure of Hindu cultivators to its extensive hills and forests. The people of the Central Provinces may therefore not unjustly be considered as a microcosm of a great part of India, and conclusions drawn from a consideration of their caste rules and status may claim with considerable probability of success to be applicable to those of the Hindus generally. For the same reason the standard ethnological works of other Provinces necessarily rank as the best authorities on the castes of the Central Provinces, and this fact may explain and excuse the copious resort which has been made to them in these volumes.

The word ‘Caste,’ Dr. Wilson states,1 is not of Indian origin, but is derived from the Portuguese casta, signifying race, mould or quality. The Indian word for caste is jāt or jāti, which has the original meaning of birth or production of a child, and hence denotes good birth or lineage, respectability and rank. Jātha means well-born. Thus jāt now signifies a caste, as every Hindu is born into a caste, and his caste determines his social position through life.

The two main ideas denoted by a caste are a community or persons following a common occupation, and a community whose members marry only among themselves. A third distinctive feature is that the members of a caste do not as a rule eat with outsiders with the exception of other Hindu castes of a much higher social position than their own. None of these will, however, serve as a definition of a caste. In a number of castes the majority of members have abandoned their traditional occupation and taken to others. Less than a fifth of the Brāhmans of the Central Provinces are performing any priestly or religious functions, and [11]the remaining four-fifths are landholders or engaged in Government service as magistrates, clerks of public offices, constables and orderlies, or in railway service in different grades, or in the professions as barristers and pleaders, doctors, engineers and so on. The Rājpūts and Marāthas were originally soldiers, but only an infinitely small proportion belong to the Indian Army, and the remainder are ruling chiefs, landholders, cultivators, labourers or in the various grades of Government service and the police. Of the Telis or oil-pressers only 9 per cent are engaged in their traditional occupation, and the remainder are landholders, cultivators and shopkeepers. Of the Ahīrs or graziers only 20 per cent tend and breed cattle. Only 12 per cent of the Chamārs are supported by the tanning industry, and so on. The Bahnas or cotton-cleaners have entirely lost their occupation, as cotton is now cleaned in factories; they are cartmen or cultivators, but retain their caste name and organisation. Since the introduction of machine-made cloth has reduced the profits of hand-loom weaving, large numbers of the weaving castes have been reduced to manual labour as a means of subsistence. The abandonment of the traditional occupation has become a most marked feature of Hindu society as a result of the equal opportunity and freedom in the choice of occupations afforded by the British Government, coupled with the rapid progress of industry and the spread of education. So far it has had no very markedly disintegrating effect on the caste system, and the status of a caste is still mainly fixed by its traditional occupation; but signs are not wanting of a coming change. Again, several castes have the same traditional occupation; about forty of the castes of the Central Provinces are classified as agriculturists, eleven as weavers, seven as fishermen, and so on. Distinctions of occupation therefore are not a sufficient basis for a classification of castes. Nor can a caste be simply defined as a body of persons who marry only among themselves, or, as it is termed, an endogamous group; for almost every important caste is divided into a number of subcastes which do not marry and frequently do not eat with each other. But it is a distinctive and peculiar feature of caste as a social institution that it splits up the people into a multitude of these [12]divisions and bars their intermarriage; and the real unit of the system and the basis of the fabric of Indian society is this endogamous group or subcaste.

The subcastes, however, connote no real difference of status or occupation. They are little known except within the caste itself, and they consist of groups within the caste which marry among themselves, and attend the communal feasts held on the occasions of marriages, funerals and meetings of the caste panchāyat or committee for the judgment of offences against the caste rules and their expiation by a penalty feast; to these feasts all male adults of the community, within a certain area, are invited. In the Central Provinces the 250 groups which have been classified as castes contain perhaps 2000 subcastes. Except in some cases other Hindus do not know a man’s subcaste, though they always know his caste; among the ignorant lower castes men may often be found who do not know whether their caste contains any subcastes or whether they themselves belong to one. That is, they will eat and marry with all the members of their caste within a circle of villages, but know nothing about the caste outside those villages, or even whether it exists elsewhere. One subdivision of a caste may look down upon another on the ground of some difference of occupation, of origin, or of abstaining from or partaking of some article of food, but these distinctions are usually confined to their internal relations and seldom recognised by outsiders. For social purposes the caste consisting of a number of these endogamous groups generally occupies the same position, determined roughly according to the respectability of its traditional occupation or extraction.

No adequate definition of caste can thus be obtained from community of occupation or intermarriage; nor would it be accurate to say that every one must know his own caste and that all the different names returned at the census may be taken as distinct. In the Central Provinces about 900 caste-names were returned at the census of 1901, and these were reduced in classification to about 250 proper castes.

In some cases synonyms are commonly used. The caste of pān or betel-vine growers and sellers is known indifferently as Barai, Pansāri or Tamboli. The great caste [13]of Ahīrs or herdsmen has several synonyms—as Gaoli in the Northern Districts, Rawat or Gahra in Chhattīsgarh, Gaur among the Uriyas, and Golkar among Telugus. Lohārs are also called Khāti and Kammāri; Masons are called Larhia, Rāj and Beldār. The more distinctly occupational castes usually have different names in different parts of the country, as Dhobi, Wārthi, Baretha, Chakla and Parit for washermen; Basor, Burud, Kandra and Dhulia for bamboo-workers, and so on. Such names may show that the subdivisions to which they are applied have immigrated from different parts of India, but the distinction is generally not now maintained, and many persons will return one or other of them indifferently. No object is gained, therefore, by distinguishing them in classification, as they correspond to no differences of status or occupation, and at most denote groups which do not intermarry, and which may therefore more properly be considered as subcastes.

Titles or names of offices are also not infrequently given as caste names. Members of the lowest or impure castes employed in the office of Kotwār or village watchmen prefer to call themselves by this name, as they thus obtain a certain rise in status, or at least they think so. In some localities the Kotwārs or village watchmen have begun to marry among themselves and try to form a separate caste. Chamārs (tanners) or Mahars (weavers) employed as grooms will call themselves Sais and consider themselves superior to the rest of their caste. The Thethwār Rāwats or Ahīrs will not clean household cooking-vessels, and therefore look down on the rest of the caste and prefer to call themselves by this designation, as ‘Theth’ means ‘exact’ or ‘pure,’ and Thethwār is one who has not degenerated from the ancestral calling. Sālewārs are a subcaste of Koshtis (weavers), who work only in silk and hence consider themselves as superior to the other Koshtis and a separate caste. The Rāthor subcaste of Telis in Mandla have abandoned the hereditary occupation of oil-pressing and become landed proprietors. They now wish to drop their own caste and to be known only as Rāthor, the name of one of the leading Rājpūt clans, in the hope that in time it will be forgotten that they ever were Telis, and they will be admitted into the community of Rājpūts. It [14]occurred to them that the census would be a good opportunity of advancing a step towards the desired end, and accordingly they telegraphed to the Commissioner of Jubbulpore before the enumeration, and petitioned the Chief Commissioner after it had been taken, to the effect that they might be recorded and classified only as Rāthor and not as Teli; this method of obtaining recognition of their claims being, as remarked by Sir Bampfylde Fuller, a great deal cheaper than being weighed against gold. On the other hand, a common occupation may sometimes amalgamate castes originally distinct into one. The sweeper’s calling is well-defined and under the generific term of Mehtar are included members of two or three distinct castes, as Dom, Bhangi and Chuhra; the word Mehtar means a prince or headman, and it is believed that its application to the sweeper by the other servants is ironical. It has now, however, been generally adopted as a caste name. Similarly, Darzi, a tailor, was held by Sir D. Ibbetson to be simply the name of a profession and not that of a caste; but it is certainly a true caste in the Central Provinces, though probably of comparatively late origin. A change of occupation may transfer a whole body of persons from one caste to another. A large section of the Banjāra caste of carriers, who have taken to cultivation, have become included in the Kunbi caste in Berār and are known as Wanjāri Kunbi. Another subcaste of the Kunbis called Mānwa is derived from the Māna tribe. Telis or oilmen, who have taken to vending liquor, now form a subcaste of the Kalār caste called Teli-Kalār; those who have become shopkeepers are called Teli-Bania and may in time become an inferior section of the Bania caste. Other similar subcastes are the Ahīr-Sunars or herdsmen-goldsmiths, the Kāyasth-Darzis or tailors, the Kori-Chamārs or weaver-tanners, the Gondi Lohārs and Barhais, being Gonds who have become carpenters and blacksmiths and been admitted to these castes; the Mahār Mhālis or barbers, and so on.

It would appear, then, that no precise definition of a caste can well be formulated to meet all difficulties. In classification, each doubtful case must be taken by itself, and it must be determined, on the information available, whether any body of persons, consisting of one or more endogamous [15]groups, and distinguished by one or more separate names, can be recognised as holding, either on account of its traditional occupation or descent, such a distinctive position in the social system, that it should be classified as a caste. But not even the condition of endogamy can be accepted as of universal application; for Vidūrs, who are considered to be descended from Brāhman fathers and women of other castes, will, though marrying among themselves, still receive the offspring of such mixed alliances into the community; in the case of Gosains and Bairāgis, who, from being religious orders, have become castes, admission is obtained by initiation as well as by birth, and the same is the case with several other orders; some of the lower castes will freely admit outsiders; and in parts of Chhattīsgarh social ties are of the laxest description, and the intermarriage of Gonds, Chamārs and other low castes are by no means infrequent. But notwithstanding these instances, the principle of the restriction of marriage to members of the caste is so nearly universal as to be capable of being adopted as a definition.

The well-known traditional theory of caste is that the Aryans were divided from the beginning of time into four castes: Brāhmans or priests, Kshatriyas or warriors, Vaishyas or merchants and cultivators, and Sūdras or menials and labourers, all of whom had a divine origin, being born from the body of Brahma—the Brāhmans from his mouth, the Kshatriyas from his arms, the Vaishyas from his thighs, and the Sudras from his feet. Intermarriage between the four castes was not at first entirely prohibited, and a man of any of the three higher ones, provided that for his first wife he took a woman of his own caste, could subsequently marry others of the divisions beneath his own. In this manner the other castes originated. Thus the Kaivarttas or Kewats were the offspring of a Kshatriya father and Vaishya mother, and so on. Mixed marriages in the opposite direction, of a woman of a higher caste with a man of a lower one, were reprobated as strongly as possible, and the offspring of these were relegated to the lowest position in society; thus the Chandāls, or descendants of a Sūdra father and Brāhman mother, were of all men the most base. It has been recognised that this genealogy, though in substance the [16]formation of a number of new castes through mixed descent may have been correct, is, as regards the details, an attempt made by a priestly law-giver to account, on the lines of orthodox tradition, for a state of society which had ceased to correspond to them.

In the ethnographic description of the people of the Punjab, which forms the Caste chapter of Sir Denzil Ibbetson’s Census Report of 1881, it was pointed out that occupation was the chief basis of the division of castes, and there is no doubt that this is true. Every separate occupation has produced a distinct caste, and the status of the caste depends now mainly or almost entirely on its occupation. The fact that there may be several castes practising such important callings as agriculture or weaving does not invalidate this in any way, and instances of the manner in which such castes have been developed will be given subsequently. If a caste changes its occupation it may, in the course of time, alter its status in a corresponding degree. The important Kāyasth and Gurao castes furnish instances of this. Castes, in fact, tend to rise or fall in social position with the acquisition of land or other forms of wealth or dignity much in the same manner as individuals do nowadays in European countries. Hitherto in India it has not been the individual who has undergone the process; he inherits the social position of the caste in which he is born, and, as a rule, retains it through life without the power of altering it. It is the caste, as a whole, or at least one of its important sections or subcastes, which gradually rises or falls in social position, and the process may extend over generations or even centuries.

In the Brief Sketch of the Caste System of the North-Western Provinces and Oudh, Mr. J.C. Nesfield puts forward the view that the whole basis of the caste system is the division of occupations, and that the social gradation of castes corresponds precisely to the different periods of civilisation during which their traditional occupations originated. Thus the lowest castes are those allied to the primitive occupation of hunting, Pāsi, Bhar, Bahelia, because the pursuit of wild animals was the earliest stage in the development of human industry. Next above these come [17]the fishing castes, fishing being considered somewhat superior to hunting, because water is a more sacred element among Hindus than land, and there is less apparent cruelty in the capturing of fish than the slaughtering of animals; these are the Kahārs, Kewats, Dhīmars and others. Above these come the pastoral castes—Ghosi, Gadaria, Gūjar and Ahīr; and above them the agricultural castes, following the order in which these occupations were adopted during the progress of civilisation. At the top of the system stands the Rājpūt or Chhatri, the warrior, whose duty is to protect all the lower castes, and the Brāhman, who is their priest and spiritual guide. Similarly, the artisan castes are divided into two main groups; the lower one consists of those whose occupations preceded the age of metallurgy, as the Chamārs and Mochis or tanners, Koris or weavers, the Telis or oil-pressers, Kalārs or liquor-distillers, Kumhārs or potters, and Lunias or salt-makers. The higher group includes those castes whose occupations were coeval with the age of metallurgy, that is, those who work in stone, wood and metals, and who make clothing and ornaments, as the Barhai or worker in wood, the Lohār or worker in iron, the Kasera and Thathera, brass-workers, and the Sunār or worker in the precious metals, ranking precisely in this order of precedence, the Sunār being the highest. The theory is still further developed among the trading castes, who are arranged in a similar manner, beginning from the Banjāra or forest trader, the Kunjra or greengrocer, and the Bharbhūnja or grain-parcher, up to the classes of Banias and Khatris or shopkeepers and bankers.

It can hardly be supposed that the Hindus either consciously or unconsciously arranged their gradation of society in a scientific order of precedence in the manner described. The main divisions of social precedence are correctly stated by Mr. Nesfield, but it will be suggested in this essay that they arose naturally from the divisions of the principal social organism of India, the village community. Nevertheless Mr. Nesfield’s book will always rank as a most interesting and original contribution to the literature of the subject, and his work did much to stimulate inquiry into the origin of the caste system.

In his Introduction to the Tribes and Castes of Bengal [18]Sir Herbert Risley laid stress on the racial basis of caste, showing that difference of race and difference of colour were the foundation of the Indian caste system or division of the people into endogamous units. There seems reason to suppose that the contact of the Aryans with the indigenous people of India was, to a large extent, responsible for the growth of the caste system, and the main racial divisions may perhaps even now be recognised, though their racial basis has, to a great extent, vanished. But when we come to individual castes and subcastes, the scrutiny of their origin, which has been made in the individual articles, appears to indicate that caste distinctions cannot, as a rule, be based on supposed difference of race. Nevertheless Sir H. Risley’s Castes and Tribes of Bengal and Peoples of India will, no doubt, always be considered as standard authorities, while as Census Commissioner for India and Director of Ethnography he probably did more to foster this branch of research in India generally than any other man has ever done.

M. Emile Senart, in his work Les Castes dans l’Inde, gives an admirable sketch of the features marking the entry of the Aryans into India and their acquisition of the country, from which the following account is largely taken. The institution of caste as it is understood at present did not exist among the Aryans of the Vedic period, on their first entry into India. The word varna, literally ‘colour,’ which is afterwards used in speaking of the four castes, distinguishes in the Vedas two classes only: there are the Arya Varna and the Dasa Varna—the Aryan race and the race of enemies. In other passages the Dasyus are spoken of as black, and Indra is praised for protecting the Aryan colour. In later literature the black race, Krishna Varna, are opposed to the Brāhmans, and the same word is used of the distinction between Aryas and Sūdras. The word varna was thus used, in the first place, not of four castes, but of two hostile races, one white and the other black. It is said that Indra divided the fields among his white-coloured people after destroying the Dasyus, by whom may be understood the indigenous barbarian races.2 The word Dasyu, which frequently recurs in the Vedas, [19]probably refers to the people of foreign countries or provinces like the Goim or Gentiles of the Hebrews. The Dasyus were not altogether barbarians, for they had cities and other institutions showing a partial civilisation, though the Aryas, lately from more bracing climes than those which they inhabited, proved too strong for them.3 To the Aryans the word Dasyu had the meaning of one who not only did not perform religious rites, but attempted to harass their performers. Another verse says, “Distinguish, O Indra, between the Aryas and those who are Dasyus: punishing those who perform no religious rites; compel them to submit to the sacrifices; be thou the powerful, the encourager of the sacrificer.”4

Rakshasa was another designation given to the tribes with whom the Aryans were in hostility. Its meaning is strong, gigantic or powerful, and among the modern Hindus it is a word for a devil or demon. In the Satapatha Brāhmana of the white Yajur-Veda the Rakshasas are represented as ‘prohibiters,’ that is ‘prohibiters of the sacrifice.’5 Similarly, at a later period, Manu describes Aryavarrta, or the abode of the Aryas, as the country between the eastern and western oceans, and between the Himalayas and the Vindhyas, that is Hindustān, the Deccan being not then recognised as an abode of the Aryans. And he thus speaks of the country: “From a Brāhman born in Aryavarrta let all men on earth learn their several usages.” “That land on which the black antelope naturally grazes, is held fit for the performance of sacrifices; but the land of Mlechchhas (foreigners) is beyond it.” “Let the three first classes (Brāhmans, Kshatriyas and Vaishyas) invariably dwell in the above-mentioned countries; but a Sūdra distressed for subsistence may sojourn wherever he chooses.”6

Another passage states: “If some pious king belonging to the Kshatriya or some other caste should defeat the Mlechchhas7 and establish a settlement of the four castes in their territories, and accept the Mlechchhas thus defeated as Chandālas (the most impure caste in ancient Hindu society) [20]as is the case in Aryavarrta, then that country also becomes fit for sacrifice. For no land is impure of itself. A land becomes so only by contact.” This passage is quoted by a Hindu writer with the same reference to the Code of Manu as the preceding one, but it is not found there and appears to be a gloss by a later writer, explaining how the country south of the Vindhyas, which is excluded by Manu, should be rendered fit for Aryan settlement.8 Similarly in a reference in the Brāhmanas to the migration of the Aryans eastward from the Punjab it is stated that Agni the fire-god flashed forth from the mouth of a priest invoking him at a sacrifice and burnt across all the five rivers, and as far as he burnt Brāhmans could live. Agni, as the god of fire by which the offerings were consumed, was addressed as follows: “We kindle thee at the sacrifice, O wise Agni, the sacrificer, the luminous, the mighty.”9 The sacrifices referred to were, in the early period, of domestic animals, the horse, ox or goat, the flesh of which was partaken of by the worshippers, and the sacred Soma-liquor, which was drunk by them; the prohibition or discouragement of animal sacrifices for the higher castes gradually came about at a later time, and was probably to a large extent due to the influence of Buddhism.

The early sacrifice was in the nature of a communal sacred meal at which the worshippers partook of the animal or liquor offered to the god. The Dasyus or indigenous Indian races could not worship the Aryan gods nor join in the sacrifices offered to them, which constituted the act of worship. They were a hostile race, but the hostility was felt and expressed on religious rather than racial grounds, as the latter term is understood at present.

M. Senart points out that the division of the four castes appearing in post-Vedic literature, does not proceed on equal lines. There were two groups, one composed of the three higher castes, and the other of the Sūdras or lowest. The higher castes constituted a fraternity into which admission was obtained only by a religious ceremony of initiation and investment with the sacred thread. The Sūdras were excluded and could take no part in sacrifices. The punishment [21]for the commission of the gravest offences by a Brāhman was that he became a Sūdra, that is to say an outcast. The killing of a Sūdra was an offence no more severe than that of killing certain animals. A Sūdra was prohibited by the severest penalties from approaching within a certain distance of a member of any of the higher castes. In the Sutras10 it is declared11 that the Sūdra has not the right (Adhikāra) of sacrifice enjoyed by the Brāhman, Kshatriya and Vaishya. He was not to be invested with the sacred thread, nor permitted, like them, to hear, commit to memory, or recite Vedic texts. For listening to these texts he ought to have his ears shut up with melted lead or lac by way of punishment; for pronouncing them, his tongue cut out; and for committing them to memory, his body cut in two.12 The Veda was never to be read in the presence of a Sūdra; and no sacrifice was to be performed for him.13 The Sūdras, it is stated in the Harivansha, are sprung from vacuity, and are destitute of ceremonies, and so are not entitled to the rites of initiation. Just as upon the friction of wood, the cloud of smoke which issues from the fire and spreads around is of no service in the sacrificial rite, so too the Sūdras spread over the earth are unserviceable, owing to their birth, to their want of initiatory rites, and the ceremonies ordained by the Vedas.14 Again it is ordained that silence is to be observed by parties of the three sacrificial classes when a Sūdra enters to remove their natural defilements, and thus the servile position of the Sūdra is recognised.15 Here it appears that the Sūdra is identified with the sweeper or scavenger, the most debased and impure of modern Hindu castes.16 In the Dharmashāstras or law-books it is laid down that a person taking a Sūdra’s food for a month becomes a Sudra and after death becomes a dog. Issue begotten after eating a Sūdra’s food is of the Sūdra caste. A person who dies with Sūdra’s food in his stomach becomes a village pig, or is reborn in a Sūdra’s family.17 An Arya who had sexual intimacy with a [22]Sūdra woman was to be banished; but a Sūdra having intimacy with an Arya was to be killed. If a Sūdra reproached a dutiful Arya, or put himself on equality with him on a road, on a couch or on a seat, he was to be beaten with a stick.18 A Brāhman might without hesitation take the property of a Sūdra; he, the Sūdra, had indeed nothing of his own; his master might, doubtless, take his property.19 According to the Mahābhārata the Sūdras are appointed servants to the Brāhmans, Kshatriyas and Vaishyas.20 A Brāhman woman having connection with a Sūdra was to be devoured by dogs, but one having connection with a Kshatriya or Vaishya was merely to have her head shaved and be carried round on an ass.21 When a Brāhman received a gift from another Brāhman he had to acknowledge it in a loud voice; from a Rājanya or Kshatriya, in a gentle voice; from a Vaishya, in a whisper; and from a Sūdra, in his own mind. To a Brāhman he commenced his thanks with the sacred syllable Om; to a king he gave thanks without the sacred Om; to a Vaishya he whispered his thanks; to a Sūdra he said nothing, but thought in his own mind, svasti, or ‘This is good.’22 It would thus seem clear that the Sūdras were distinct from the Aryas and were a separate and inferior race, consisting of the indigenous people of India. In the Atharva-Veda the Sudra is recognised as distinct from the Arya, and also the Dasa from the Arya, as in the Rig-Veda.23 Dr. Wilson remarks, “The aboriginal inhabitants, again, who conformed to the Brāhmanic law, received certain privileges, and were constituted as a fourth caste under the name of Sūdras, whereas all the rest who kept aloof were called Dasyus, whatever their language might be.”24 The Sūdras, though treated by Manu and Hindu legislation in general as a component, if enslaved, part of the Indian community, not entitled to the second or sacramental birth, are not even once mentioned in the older parts of the Vedas. They are first locally brought to notice in the Mahābhārata, along with [23]the Abhīras, dwelling on the banks of the Indus. There are distinct classical notices of the Sūdras in this very locality and its neighbourhood. “In historical times,” says Lassen, “their name reappears in that of the town Sudros on the lower Indus, and, what is especially worthy of notice, in that of the people Sudroi, among the Northern Arachosians.”25

“Thus their existence as a distinct nation is established in the neighbourhood of the Indus, that is to say in the region in which, in the oldest time, the Aryan Indians dwelt. The Aryans probably conquered these indigenous inhabitants first; and when the others in the interior of the country were subsequently subdued and enslaved, the name Sūdra was extended to the whole servile caste. There seems to have been some hesitation in the Aryan community about the actual religious position to be given to the Sūdras. In the time of the liturgical Brāhmanas of the Vedas, they were sometimes admitted to take part in the Aryan sacrifices. Not long afterwards, when the conquests of the Aryans were greatly extended, and they formed a settled state of society among the affluents of the Jumna and Ganges, the Sūdras were degraded to the humiliating and painful position which they occupy in Manu. There is no mention of any of the Sankara or mixed castes in the Vedas.”26

From the above evidence it seems clear that the Sūdras were really the indigenous inhabitants of India, who were subdued by the Aryans as they gradually penetrated into India. When the conquering race began to settle in the land, the indigenous tribes, or such of them as did not retire before the invaders into the still unconquered interior, became a class of menials and labourers, as the Amalekites were to the children of Israel. The Sūdras were the same people as the Dasyus of the hymns, after they had begun to live in villages with the Aryans, and had to be admitted, though in the most humiliating fashion, into the Aryan polity. But the hostility between the Aryas and the Dasyus or Sūdras, though in reality racial, was felt and expressed on religious grounds, and probably the Aryans had no real [24]idea of what is now understood by difference of race or deterioration of type from mixture of races. The Sūdras were despised and hated as worshippers of a hostile god. They could not join in the sacrifices by which the Aryans renewed and cemented their kinship with their god and with each other; hence they were outlaws towards whom no social obligations existed. It would have been quite right and proper that they should be utterly destroyed, precisely as the Israelites thought that Jehovah had commanded them to destroy the Canaanites. But they were too numerous, and hence they were regarded as impure and made to live apart, so that they should not pollute the places of sacrifice, which among the Aryans included their dwelling-houses. It does not seem to have been the case that the Aryans had any regard for the preservation of the purity of their blood or colour. From an early period men of the three higher castes might take a Sūdra woman in marriage, and the ultimate result has been an almost complete fusion between the two races in the bulk of the population over the greater part of the country. Nevertheless the status of the Sūdra still remains attached to the large community of the impure castes formed from the indigenous tribes, who have settled in Hindu villages and entered the caste system. These are relegated to the most degrading and menial occupations, and their touch is regarded as conveying defilement like that of the Sūdras.27 The status of the Sūdras was not always considered so low, and they were sometimes held to rank above the mixed castes. And in modern times in Bengal Sūdra is quite a respectable term applied to certain artisan castes which there have a fairly good position. But neither were the indigenous tribes always reduced to the impure status. Their fortunes varied, and those who resisted subjection were probably sometimes accepted as allies. For instance, some of the most prominent of the Rājpūt clans are held to have been derived from the aboriginal28 tribes. On the Aryan expedition to southern India, which is preserved in the legend of Rāma, as related in the Rāmāyana, it is stated that Rāma was assisted by [25]Hanumān with his army of apes. The reference is generally held to be to the fact that the Aryans had as auxiliaries some of the forest tribes, and these were consequently allies, and highly thought of, as shown by the legend and by their identification with the mighty god Hanumān. And at the present time the forest tribes who live separately from the Hindus in the jungle tracts are, as a rule, not regarded as impure. But this does not impair the identification of the Sūdras with those tribes who were reduced to subjection and serfdom in the Hindu villages, as shown by the evidence here given. The view has also been held that the Sūdras might have been a servile class already subject to the Aryans, who entered India with them. And in the old Pārsi or Persian community four classes existed, the Athornan or priest, the Rathestan or warrior, the Vasteriox or husbandman, and the Hutox or craftsman.29 The second and third of these names closely resemble those of the corresponding Hindu classical castes, the Rājanya or Kshatriya and the Vaishya, while Athornan, the name for a priest, is the same as Atharvan, the Hindu name for a Brāhman versed in the Atharva-Veda. Possibly then Hutox may be connected with Sūdra, as h frequently changes into s. But on the other hand the facts that the Sūdras are not mentioned in the Vedas, and that they succeeded to the position of the Dasyus, the black hostile Indians, as well as the important place they fill in the later literature, seem to indicate clearly that they mainly consisted of the indigenous subject tribes. Whether the Aryans applied a name already existing in a servile class among themselves to the indigenous population whom they subdued, may be an uncertain point.

In the Vedas, moreover, M. Senart shows that the three higher castes are not definitely distinguished; but there are three classes—the priests, the chiefs and the people, among whom the Aryans were comprised. The people are spoken of in the plural as the clans who followed the chiefs to battle. The word used is Visha. One verse speaks of the Vishas (clans) bowing before the chief (Rājan), who was preceded by a priest (Brāhman). Another verse says: “Favour the prayer [26](Brahma), favour the service; kill the Rakshasas, drive away the evil; favour the power (khatra) and favour the manly strength; favour the cow (dherm, the representative of property) and favour the people (or house, visha).”30

Similarly Wilson states that in the time of the Vedas, visha (related to vesha, a house or district) signified the people in general; and Vaishya, its adjective, was afterwards applied to a householder, or that appertaining to an individual of the common people. The Latin vicus and the Greek οἶκος are the correspondents of vesha.31 The conclusion to be drawn is that the Aryans in the Vedas, like other early communities, were divided by rank or occupation into three classes—priests, nobles and the body of the people. The Vishas or clans afterwards became the Vaishyas or third classical caste. Before they entered India the Aryans were a migratory pastoral people, their domestic animals being the horse, cow, and perhaps the sheep and goat. The horse and cow were especially venerated, and hence were probably their chief means of support. The Vaishyas must therefore have been herdsmen and shepherds, and when they entered India and took to agriculture, the Vaishyas must have become cultivators. The word Vaishya signifies a man who occupies the soil, an agriculturist, or merchant.32 The word Vasteriox used by the ancestors of the Pārsis, which appears to correspond to Vaishya, also signifies a husbandman, as already seen. Dr. Max Müller states: “The three occupations of the Aryas in India were fighting, cultivating the soil and worshipping the gods. Those who fought the battles of the people would naturally acquire influence and rank, and their leaders appear in the Veda as Rājas or kings. Those who did not share in the fighting would occupy a more humble position; they were called Vish, Vaishyas or householders, and would no doubt have to contribute towards the maintenance of the armies.33 According to Manu, God ordained the tending of cattle, giving alms, sacrifice, study, trade, usury, and also agriculture for [27]a Vaishya.”34 The Sūtras state that agriculture, the keeping of cattle, and engaging in merchandise, as well as learning the Vedas, sacrificing for himself and giving alms, are the duties of a Vaishya.35 In the Mahābhārata it is laid down that the Vaishyas should devote themselves to agriculture, the keeping of cattle and liberality.36 In the same work the god Vayu says to Bhishma: “And it was Brahma’s ordinance that the Vaishya should sustain the three castes (Brāhman, Kshatriya and Vaishya) with money and corn; and that the Sūdra should serve them.”37

In a list of classes or occupations given in the White Yajur-Veda, and apparently referring to a comparatively advanced state of Hindu society, tillage is laid down as the calling of the Vaishya, and he is distinguished from the Vāni or merchant, whose occupation is trade or weighing.38 Manu states that a Brāhman should swear by truth; a Kshatriya by his steed and his weapons; a Vaishya by his cows, his seed and his gold; and a Sūdra by all wicked deeds.39 Yellow is the colour of the Vaishya, and it must apparently be taken from the yellow corn, and the yellow colour of ghī or butter, the principal product of the sacred cow; yellow is also the colour of the sacred metal gold, but there can scarcely have been sufficient gold in the hands of the body of the people in those early times to enable it to be especially associated with them. The Vaishyas were thus, as is shown by the above evidence, the main body of the people referred to in the Vedic hymns. When these settled down into villages the Vaishyas became the householders and cultivators, among whom the village lands were divided; the Sūdras or indigenous tribes, who also lived in the villages or in hamlets adjoining them, were labourers and given all the most disagreeable tasks in the village community, as is the case with the impure castes at present.

The demonstration of the real position of the Vaishyas [28]is important, because the Hindus themselves no longer recognise this. The name Vaishya is now frequently restricted to the Bania caste of bankers, shopkeepers and moneylenders, and hence the Banias are often supposed to be the descendants and only modern representatives of the original Vaishyas. Evidence has been given in the article on Bania to show that the existing Bania caste is mainly derived from the Rājpūts. The name Bāni, a merchant or trader, is found at an early period, but whether it denoted a regular Bania caste may be considered as uncertain. In any case it seems clear that this comparatively small caste, chiefly coming from Rājputāna, cannot represent the Vaishyas, who were the main body or people of the invading Aryans. At that time the Vaishyas cannot possibly have been traders, because they alone provided the means of subsistence of the community, and if they produced nothing, there could be no material for trade. The Vaishyas must, therefore, as already seen, have been shepherds and cultivators, since in early times wealth consisted almost solely of corn and cattle. At a later period, with the increased religious veneration for all kinds of life, agriculture apparently fell into some kind of disrepute as involving the sacrifice of insect life, and there was a tendency to emphasise trade as the Vaishya’s occupation in view of its greater respectability. It is considered very derogatory for a Brāhman or Rājpūt to touch the plough with his own hands, and the act has hitherto involved a loss of status: these castes, however, did not object to hold land, but, on the contrary, ardently desired to do so like all other Hindus. Ploughing was probably despised as a form of manual labour, and hence an undignified action for a member of the aristocracy, just as a squire or gentleman farmer in England might consider it beneath his dignity to drive the plough himself. No doubt also, as the fusion of races proceeded, and bodies of the indigenous tribes who were cultivators adopted Hinduism, the status of a cultivator sank to some extent, and his Vaishyan ancestry was forgotten. But though the Vaishya himself has practically disappeared, his status as a cultivator and member of the village community appears to remain in that of the modern cultivating castes, as will be shown subsequently. [29]

The settlement of the Aryans in India was in villages and not in towns, and the Hindus have ever since remained a rural people. In 1911 less than a tenth of the population of India was urban, and nearly three-quarters of the total were directly supported by agriculture. Apparently, therefore, the basis or embryo of the gradation of Hindu society or the caste system should be sought in the village. Two main divisions of the village community may be recognised in the Vaishyas or cultivators and the Sūdras or impure serfs and labourers. The exact position held by the Kshatriyas and the constitution of their class are not quite clear, but there is no doubt that the Brāhmans and Kshatriyas formed the early aristocracy, ranking above the cultivators, and a few other castes have since attained to this position. From early times, as is shown by an ordinance of Manu, men of the higher castes or classes were permitted, after taking a woman of their own class for the first wife, to have second and subsequent wives from any of the classes beneath them. This custom appears to have been largely prevalent. No definite rule prescribed that the children of such unions should necessarily be illegitimate, and in many cases no doubt seems to exist that, if not they themselves, their descendants at any rate ultimately became full members of the caste of the first ancestor. According to Manu, if the child of a Brāhman by a Sūdra woman intermarried with Brāhmans and his descendants after him, their progeny in the seventh generation would become full Brāhmans; and the same was the case with the child of a Kshatriya or a Vaishya with a Sūdra woman. A commentator remarks that the descendants of a Brāhman by a Kshatriya woman could attain Brāhmanhood in the third generation, and those by a Vaishya woman in the fifth.40 Such children also could inherit. According to the Mahābhārata, if a Brāhman had four wives of different castes, the son by a Brāhman wife took four shares, that by a Kshatriya wife three, by a Vaishya wife two, and by a Sūdra wife one share.41 Manu gives a slightly different distribution, but also permits to the son by a Sūdra wife a share of the [30]inheritance.42 Thus the fact is clear that the son of a Brāhman even by a Sūdra woman had a certain status of legitimacy in his father’s caste, as he could marry in it, and must therefore have been permitted to partake of the sacrificial food at marriage;43 and he could also inherit a small share of the property.

The detailed rules prescribed for the status of legitimacy and inheritance show that recognised unions of this kind between men of a higher class and women of a lower one were at one time fairly frequent, though they were afterwards prohibited. And they must necessarily have led to much mixture of blood in the different castes. A trace of them seems to survive in the practice of hypergamy, still widely prevalent in northern India, by which men of the higher subcastes of a caste will take daughters in marriage from lower ones but will not give their daughters in return. This custom prevails largely among the higher castes of the Punjab, as the Rājpūts and Khatris, and among the Brāhmans of Bengal.44 Only a few cases are found in the Central Provinces, among Brāhmans, Sunārs and other castes. Occasionally intermarriage between two castes takes place on a hypergamous basis; thus Rājpūts are said to take daughters from the highest clans of the cultivating caste of Dāngis. More commonly families of the lower subcastes or clans in the same caste consider the marriage of their daughters into a higher group a great honour and will give large sums of money for a bridegroom. Until quite recently a Rājpūt was bound to marry his daughters into a clan of equal or higher rank than his own, in order to maintain the position of his family. It is not easy to see why so much importance should be attached to the marriage of a daughter, since she passed into another clan and family, to whom her offspring would belong. On the other hand, a son might take a wife from a lower group without loss of status, though his children would be the future representatives of the family. Another point, [31]possibly connected with hypergamy, is that a peculiar relation exists between a man and the family into which his daughter has married. Sometimes he will accept no food or even water in his son-in-law’s village. The word sāla, signifying wife’s brother, when addressed to a man, is also a common and extremely offensive term of abuse. The meaning is now perhaps supposed to be that one has violated the sister of the person spoken to, but this can hardly have been the original significance as sasur or father-in-law is also considered in a minor degree an opprobrious term of address.

But though among the four classical castes it was possible for the descendants of mixed unions between fathers of higher and mothers of lower caste to be admitted into their father’s caste, this would not have been the general rule. Such connections were very frequent and the Hindu classics account through them for the multiplication of castes. Long lists are given of new castes formed by the children of mixed marriages. The details of these genealogies seem to be destitute of any probability, and perhaps, therefore, instances of them are unnecessary. Matches between a man of higher and a woman of lower caste were called anuloma, or ‘with the hair’ or ‘grain,’ and were regarded as suitable and becoming. Those between a man of lower and a woman of higher caste were, on the other hand, known as pratiloma or ‘against the hair,’ and were considered as disgraceful and almost incestuous. The offspring of such unions are held to have constituted the lowest and most impure castes of scavengers, dog-eaters and so on. This doctrine is to be accounted for by the necessity of safeguarding the morality of women in a state of society where kinship is reckoned solely by male descent. The blood of the tribe and clan, and hence the right to membership and participation in the communal sacrifices, is then communicated to the child through the father; hence if the women are unchaste, children may be born into the family who have no such rights, and the whole basis of society is destroyed. For the same reason, since the tribal blood and life is communicated through males, the birth and standing of the mother are of little importance, and children are, as has been seen, easily admitted to their father’s rank. But already in Manu’s [32]time the later and present view that both the father and mother must be of full status in the clan, tribe or caste in order to produce a legitimate child, has begun to prevail, and the children of all mixed marriages are relegated to a lower group. The offspring of these mixed unions did probably give rise to a class of different status in the village community. The lower-caste mother would usually have been taken into the father’s house and her children would be brought up in it. Thus they would eat the food of the household, even if they did not participate in the sacrificial feasts; and a class of this kind would be very useful for the performance of menial duties in and about the household, such as personal service, bringing water, and so on, for which the Sūdras, owing to their impurity, would be unsuitable. In the above manner a new grade of village menial might have arisen and have gradually been extended to the other village industries, so that a third group would be formed in the village community ranking between the cultivators and labourers. This gradation of the village community may perhaps still be discerned in the main social distinctions of the different Hindu castes at present. And an attempt will now be made to demonstrate this hypothesis in connection with a brief survey of the castes of the Province.

An examination of the social status of the castes of the Central Provinces, which, as already seen, are representative of a great part of India, shows that they fall into five principal groups. The highest consists of those castes who now claim to be directly descended from the Brāhmans, Kshatriyas or Vaishyas, the three higher of the four classical castes. The second comprises what are generally known as pure or good castes. The principal mark of their caste status is that a Brāhman will take water to drink from them, and perform ceremonies in their houses. They may be classified in three divisions: the higher agricultural castes, higher artisan castes, and serving castes from whom a Brāhman will take water. The third group contains those castes from whose hands a Brāhman will not take water; but their touch does not convey impurity and they are permitted to enter Hindu temples. They consist mainly of certain cultivating castes of low status, some of them recently [33]derived from the indigenous tribes, other functional castes formed from the forest tribes, and a number of professional and menial castes, whose occupations are mainly pursued in villages, so that they formerly obtained their subsistence from grain-payments or annual allowances of grain from the cultivators at seedtime and harvest. The group includes also some castes of village priests and mendicant religious orders, who beg from the cultivators. In the fourth group are placed the non-Aryan or indigenous tribes. Most of these cannot properly be said to form part of the Hindu social system at all, but for practical purposes they are admitted and are considered to rank below all castes except those who cannot be touched. The lowest group consists of the impure castes whose touch is considered to defile the higher castes. Within each group there are minor differences of status some of which will be noticed, but the broad divisions may be considered as representing approximately the facts. The rule about Brāhmans taking water from the good agricultural and artisan castes obtains, for instance, only in northern India. Marātha Brāhmans will not take water from any but other Brāhmans, and in Chhattīsgarh Brāhmans and other high castes will take water only from the hands of a Rāwat (grazier), and from no other caste. But nevertheless the Kunbis, the great cultivating caste of the Marātha country, though Brāhmans do not take water from them, are on the same level as the Kurmis, the cultivating caste of Hindustān, and in tracts where they meet Kunbis and Kurmis are often considered to be the same caste. The evidence of the statements made as to the origin of different castes in the following account will be found in the articles on them in the body of the work.

The castes of the first group are noted below:

The Brāhmans are, as they have always been, the highest caste. The Rājpūts are the representatives of the ancient Kshatriyas or second caste, though the existing Rājpūt clans are probably derived from the Hun, Gūjar and other invaders of the period before and shortly after the commencement of the Christian era, and in some cases from the indigenous or [34]non-Aryan tribes. It does not seem possible to assert in the case of a single one of the present Rājpūt clans that any substantial evidence is forthcoming in favour of their descent from the Aryan Kshatriyas, and as regards most of the clans there are strong arguments against such a hypothesis. Nevertheless the Rājpūts have succeeded to the status of the Kshatriyas, and an alternative name for them, Chhatri, is a corruption of the latter word. They are commonly identified with the second of the four classical castes, but a Hindu law-book gives Rājapūtra as the offspring of a Kshatriya father and a mother of the Karan or writer caste.45 This genealogy is absurd, but may imply the opinion that the Rājpūts were not the same as the Aryan Kshatriyas. The Khatris are an important mercantile caste of the Punjab, who in the opinion of most authorities are derived from the Rājpūts. The name is probably a corruption of Kshatri or Kshatriya. The Banias are the great mercantile, banking and shopkeeping caste among the Hindus and a large proportion of the trade in grain and ghī (preserved butter) is in their hands, while they are also the chief moneylenders. Most of the important Bania subcastes belonged originally to Rājputāna and Central India, which are also the homes of the Rājpūts, and reasons have been given in the article on Bania for holding that they are derived from the Rājpūts. They, however, are now commonly called Vaishyas by the Hindus, as, I think, under the mistaken impression that they are descended from the original Vaishyas. The Bhāts are the bards, heralds and genealogists of India and include groups of very varying status. The Bhāts who act as genealogists of the cultivating and other castes and accept cooked food from their clients may perhaps be held to rank with or even below them. But the high-class Bhāts are undoubtedly derived from Brāhmans and Rājpūts, and rank just below those castes. The bard or herald had a sacred character, and his person was inviolable like that of the herald elsewhere, and this has given a special status to the whole caste.46 The Kāyasths are the writer caste of Hindustān, and the Karans and Prabhus are [35]the corresponding castes of Orissa and Bombay. The position of the Kāyasths has greatly risen during the last century on account of their own ability and industry and the advantages they have obtained through their high level of education. The original Kāyasths may have been village accountants and hence have occupied a lower position, perhaps below the cultivators. They are an instance of a caste whose social position has greatly improved on account of the wealth and importance of its members. At present the Kāyasths may be said to rank next to Brāhmans and Rājpūts. The origin of the Prabhus and Karans is uncertain, but their recent social history appears to resemble that of the Kāyasths. The Guraos are another caste whose position has greatly improved. They were priests of the village temples of Siva, and accepted the offerings of food which Brāhmans could not take. But they also supplied leaf-plates for festivals, and were village musicians and trumpeters in the Marātha armies, and hence probably ranked below the cultivators and were supported by contributions of grain from them. Their social position has been raised by their sacred character as priests of the god Siva and they are now sometimes called Shaiva Brāhmans. But a distinct recollection of their former status exists.

Thus all the castes of the first group are derived from the representatives of the Brāhmans and Kshatriyas, the two highest of the four classical castes, except the Guraos, who have risen in status owing to special circumstances. The origin of the Kāyasths is discussed in the article on that caste. Members of the above castes usually wear the sacred thread which is the mark of the Dwija or twice-born, the old Brāhmans, Kshatriyas and Vaishyas. The thread is not worn generally by the castes of the second group, but the more wealthy and prominent sections of them frequently assume it.

The second group of good castes from whom a Brāhman can take water falls into three sections as already explained: the higher agricultural castes, the higher artisans, and the serving or menial castes from whom a Brāhman takes water from motives of convenience. These last do not properly [36]belong to the second group but to the next lower one of village menials. The higher agricultural castes or those of the first section are noted below:

In this division the Kurmis and Kunbis are the typical agricultural castes of Hindustān or the plains of northern India, and the Bombay or Marātha Deccan. Both are very numerous and appear to be purely occupational bodies. The name Kurmi perhaps signifies a cultivator or worker. Kunbi may mean a householder. In both castes, groups of diverse origin seem to have been amalgamated owing to their common calling. Thus the Kunbis include a subcaste derived from the Banjāra (carriers), another from the Dhangars or shepherds, and a third from the Mānas, a primitive tribe. In Bombay it is considered that the majority of the Kunbi caste are sprung from the non-Aryan or indigenous tribes, and this may be the reason why Marātha Brāhmans do not take water from them. But they have now become one caste with a status equal to that of the other good cultivating castes. In many tracts of Berār and elsewhere practically all the cultivators of the village belong to the Kunbi caste, and there is every reason to suppose that this was once the general rule and that the Kunbis or ‘householders’ are simply the cultivators of the Marātha country who lived in village communities. Similarly Sir H. Risley considered that some Kurmis of Bihār were of the Aryan type, while others of Chota Nāgpur are derived from the indigenous tribes. The Chasas are the cultivating caste of Orissa and are a similar occupational group. The word Chasa has the generic meaning of a cultivator, and the caste are said by Sir H. Risley to be for the most part of non-Aryan origin, the loose organisation of the caste system among the Uriyas making it possible on the one hand for outsiders to be admitted into the caste, and on the other for wealthy Chasas, who gave up ploughing with their own hands and assumed the respectable title of Mahanti, to raise themselves to [37]membership among the lower classes of Kāyasths. The Koltas are another Uriya caste, probably an offshoot of the Chasas, whose name may be derived from the kulthi47 pulse, a favourite crop in that locality.

Similarly the Vellālas are the great cultivating caste of the Tamil country, to whom by general consent the first place in social esteem among the Tamil Sūdra castes is awarded. In the Madras Census Report of 1901 Mr. Francis gives an interesting description of the structure of the caste and its numerous territorial, occupational and other subdivisions. He shows also how groups from lower castes continually succeed in obtaining admission into the Vellāla community in the following passage: “Instances of members of other castes who have assumed the name and position of Vellālas are the Vettuva Vellālas, who are only Puluvāns; the Illam Vellālas, who are Panikkāns; the Karaiturai (lord of the shore) Vellālas, who are Karaiyāns; the Karukamattai (palmyra leaf-stem) Vellālas, who are Balijas; the Guha (Rama’s boatmen) Vellālas, who are Sembadavāns; and the Irkuli Vellālas, who are Vannāns. The children of dancing-girls also often call themselves Mudali, and claim in time to be Vellālas, and even Paraiyāns assume the title of Pillai and trust to its eventually enabling them to pass themselves off as members of the caste.”

This is an excellent instance of the good status attaching to the chief cultivating caste of the locality and of the manner in which other groups, when they obtain possession of the land, strive to get themselves enrolled in it.

The Jāts are the representative cultivating caste of the Punjab. They are probably the descendants of one of the Scythian invading hordes who entered India shortly before and after the commencement of the Christian era. The Scythians, as they were called by Herodotus, appear to have belonged to the Mongolian racial family, as also did the white Huns who came subsequently. The Gūjar and Ahīr castes, as well as the Jāts, and also the bulk of the existing Rājpūt clans, are believed to be descended from these invaders; and since their residence in India has been comparatively short in comparison with their Aryan predecessors, [38]they have undergone much less fusion with the general population, and retain a lighter complexion and better features, as is quite perceptible to the ordinary observer in the case of the Jāts and Rājpūts. The Jāts have a somewhat higher status than other agricultural castes, because in the Punjab they were once dominant, and one or two ruling chiefs belonged to the caste.48 The bulk of the Sikhs were also Jāts. But in the Central Provinces, where they are not large landholders, and have no traditions of former dominance, there is little distinction between them and the Kurmis. The Gūjars for long remained a pastoral freebooting tribe, and their community was naturally recruited from all classes of vagabonds and outlaws, and hence the caste is now of a mixed character, and their physical type is not noticeably distinct from that of other Hindus. Sir G. Campbell derived the Gūjars from the Khazars, a tribe of the same race as the white Huns and Bulgars who from an early period had been settled in the neighbourhood of the Caspian. They are believed to have entered India during the fifth or sixth century. Several clans of Rājpūts, as well as considerable sections of the Ahīr and Kunbi castes were, in his opinion, derived from the Gūjars. In the Central Provinces the Gūjars have now settled down into respectable cultivators. The Ahīrs or cowherds and graziers probably take their name from the Abhīras, another of the Scythian tribes. But they have now become a purely occupational caste, largely recruited from the indigenous Gonds and Kawars, to whom the business of tending cattle in the jungles is habitually entrusted. In the Central Provinces Ahīrs live in small forest villages with Gonds, and are sometimes scarcely considered as Hindus. On this account they have a character for bucolic stupidity, as the proverb has it: ‘When he is asleep he is an Ahīr and when he is awake he is a fool.’ But the Ahīr caste generally has a good status on account of its connection with the sacred cow and also with the god Krishna, the divine cowherd.

The Marāthas are the military caste of the Marātha country, formed into a caste from the cultivators, shepherds and herdsmen, who took service under Sivaji and subsequent [39]Marātha leaders. The higher clans may have been constituted from the aristocracy of the Deccan states, which was probably of Rājpūt descent. They have now become a single caste, ranking somewhat higher than the Kunbis, from whom the bulk of them originated, on account of their former military and dominant position. Their status was much the same as that of the Jāts in the Punjab. But the ordinary Marāthas are mainly engaged in the subordinate Government and private service, and there is very little distinction between them and the Kunbis. The Khandaits or swordsmen (from khanda, a sword) are an Uriya caste, which originated in military service, and the members of which belonged for the most part to the non-Aryan Bhuiya tribe. They were a sort of rabble, half military and half police, Sir H. Risley states, who formed the levies of the Uriya zamīndārs. They have obtained grants of land, and their status has improved. “In the social system of Orissa the Sreshta (good) Khandaits rank next to the Rājputs, who are comparatively few in number, and have not that intimate connection with the land which has helped to raise the Khandaits to their present position.”49 The small Rautia landholding caste of Chota Nāgpur, mainly derived from the Kol tribe, was formed from military service, and obtained a higher status with the possession of the land exactly like the Khandaits.

Several Rājput clans, as the Panwārs of the Wainganga Valley, the Rāghuvansis, the Jādums derived from the Yādava clan, and the Daharias of Chhattīsgarh, have formed distinct castes, marrying among themselves. A proper Rājpūt should not marry in his own clan. These groups have probably in the past taken wives from the surrounding population, and they can no longer be held to belong to the Rājpūt caste proper, but rank as ordinary agricultural castes. Other agricultural castes have probably been formed through mixed descent from Rājpūts and the indigenous races. The Agharias of Sambalpur say they are sprung from a clan of Rājpūts near Agra, who refused to bend their heads before the king of Delhi. He summoned all the Agharias to appear before him, and fixed a sword across the door at the [40]height of a man’s neck. As the Agharias would not bend their heads they were as a natural consequence all decapitated as they passed through the door. Only one escaped, who had bribed a Chamār to go instead of him. He and his village fled from Agra and came to Chhattīsgarh, where they founded the Agharia caste. And, in memory of this, when an Agharia makes a libation to his ancestors, he first pours a little water on the ground in honour of the dead Chamār. Such stories may be purely imaginary, or may contain some substratum of truth, as that the ancestors of the caste were Rājpūts, who took wives from Chamārs and other low castes. The Kīrārs are another caste with more or less mixed descent from Rājpūts. They are also called Dhākar, and this means one of illegitimate birth. The Bhilālas are a caste formed of the offspring of mixed alliances between Rājpūts and Bhīls. In many cases in Nimār Rājpūt immigrants appear to have married the daughters of Bhīl chieftains and landholders, and succeeded to their estates. Thus the Bhilālas include a number of landed proprietors, and the caste ranks as a good agricultural caste, from whom Brāhmans will take water. Among the other indigenous tribes, several of which have in the Central Provinces retained the possession of large areas of land and great estates in the wilder forest tracts, a subcaste has been formed of the landholding members of the tribe. Such are the Rāj-Gonds among the Gonds, the Binjhāls among Baigas, and the Tawar subtribe of the Kawar tribe of Bilāspur, to which all the zamīndārs50 belong. These last now claim to be Tomara Rājpūts, on the basis of the similarity of the name. These groups rank with the good agricultural castes, and Brāhmans sometimes consent to take water from them. The Dāngis of Saugor appear to be the descendants of a set of freebooters in the Vindhyan hills, much like the Gūjars in northern India. The legend of their origin is given in Sir B. Robertson’s Census Report of 1891: “The chief of Garhpahra or old Saugor detained the palanquins of twenty-two married women and kept them as his wives. The issue of the illicit intercourse were named Dāngis, and there are thus twenty-two subdivisions of these people. There are also three other subdivisions who claim [41]descent from pure Rājpūts, and who will take daughters in marriage from the remaining twenty-two, but will not give their daughters to them.” Thus the Dāngis appear to have been a mixed group, recruiting their band from all classes of the population, with some Rajpūts as leaders. The name probably means hillman, from dāng, a hill. Khet men bāmi, gaon men Dāngi or ‘A Dāngi in the village is like the hole of a snake in one’s field,’ is a proverb showing the estimation in which they were formerly held. They obtained estates in Saugor and a Dāngi dynasty formerly governed part of the District, and they are now highly respectable cultivators. The Mīnas or Deswālis belonged to the predatory Mīna tribe of Rājputāna, but a section of them have obtained possession of the land in Hoshangābād and rank as a good agricultural caste. The Lodhas of the United Provinces are placed lowest among the agricultural castes by Mr. Nesfield, who describes them as little better than a forest tribe. The name is perhaps derived from the bark of the lodh tree, which was collected by the Lodhas of northern India and sold for use as a dyeing agent. In the Central Provinces the name has been changed to Lodhi, and they are said to have been brought into the District by a Rāja of the Gond-Rājpūt dynasty of Mandla in the seventeenth century, and given large grants of waste land in the interior in order that they might clear it of forest. They have thus become landholders, and rank with the higher agricultural castes. They are addressed as Thākur, a title applied to Rājpūts, and Lodhi landowners usually wear the sacred thread.

The above details have been given to show how the different agricultural castes originated. Though their origin is so diverse they have, to a great extent, the same status, and it seems clear that this status is dependent on their possession of the land. In the tracts where they reside they are commonly village proprietors and superior tenants. Those who rank a little higher than the others, as the Jāts, Marāthas, Dāngis and Lodhis, include in their body some ruling chiefs or large landed proprietors, and as a rule were formerly dominant in the territory in which they are found. In primitive agricultural communities the land is the principal, if not almost the sole, source of wealth. Trade in the [42]modern sense scarcely exists, and what interchange of commodities there is affects, as a rule, only a trifling fraction of the population. India’s foreign trade is mainly the growth of the last century, and the great bulk of the exports are of agricultural produce, yet in proportion to the population the trading community is still extremely small. It thus seems quite impossible that the Aryans could have been a community of priests, rulers and traders, because such a community would not have had means of subsistence. And if the whole production and control of the wealth and food of the community had been in the hands of the Sūdras, they could not have been kept permanently in their subject, degraded position. The flocks and herds and the land, which constituted the wealth of early India, must thus have been in the possession of the Vaishyas; and grounds of general probability, as well as the direct evidence already produced, make it clear that they were the herdsmen and cultivators, and the Sūdras the labourers. The status of the modern cultivators seems to correspond to that of the Vaishyas, that is, of the main body of the Aryan people, who were pure and permitted to join in sacrifices. The status, however, no longer attaches to origin, but to the possession of the land; it is that of a constituent member of the village community, corresponding to a citizen of the city states of Greece and Italy. The original Vaishyas have long disappeared; the Brāhmans themselves say that there are no Kshatriyas and no Vaishyas left, and this seems to be quite correct. But the modern good cultivating castes retain the status of the Vaishyas as the Rājpūts retain that of the Kshatriyas. The case of the Jāts and Gūjars supports this view. These two castes are almost certainly derived from Scythian nomad tribes, who entered India long after the Vedic Aryans. And there is good reason to suppose that a substantial proportion, if not the majority, of the existing Rājpūt clans were the leaders or aristocracy of the Jāts and Gūjars. Thus it is found that in the case of these later tribes the main body were shepherds and cultivators, and their descendants have the status of good cultivating castes at present, while the leaders became the Rājpūts, who have the status of the Kshatriyas; and it therefore seems a reasonable [43]inference that the same had previously been the case with the Aryans themselves. It has been seen that the word Visha or Vaishya signified one of the people or a householder. The name Kunbi appears to have the same sense, its older form being kutumbika, which is a householder or one who has a family,51 a pater familias.

It has been seen also that Visha in the plural signified clans. The clan was the small body which lived together, and in the patriarchal stage was connected by a tie of kinship held to be derived from a common ancestor. Thus it is likely that the clans settled down in villages, the cultivators of one village being of the same exogamous clan. The existing system of exogamy affords evidence in favour of this view, as will be seen. All the families of the clan had cultivating rights in the land, and were members of the village community; and there were no other members, unless possibly a Kshatriya headman or leader. The Sūdras were their labourers and serfs, with no right to hold land, and a third intermediate class of village menials gradually grew up.

The law of Mirāsi tenures in Madras is perhaps a survival of the social system of the early village community. Under it only a few of the higher castes were allowed to hold land, and the monopoly was preserved by the rule that the right of taking up waste lands belonged primarily to the cultivators of the adjacent holdings; no one else could acquire land unless he first bought them out. The pariahs or impure castes were not allowed to hold land at all. This rule was pointed out by Mr. Slocock, and it is also noticed by Sir Henry Maine: “There are in Central and Southern India certain villages to which a class of persons is hereditarily attached, in such a manner that they form no part of the natural and organic aggregate to which the bulk of the villagers belong. These persons are looked upon as essentially impure; they never enter the village, or only enter reserved portions of it; and their touch is avoided as contaminating. Yet they bear extremely plain marks of their origin. Though they are not included in the village, they are an appendage solidly connected with it; they have [44]definite village duties, one of which is the settlement of boundaries, on which their authority is allowed to be conclusive. They evidently represent a population of alien blood whose lands have been occupied by the colonists or invaders forming the community.”52 Elsewhere, Sir Henry Maine points out that in many cases the outsiders were probably admitted to the possession of land, but on an inferior tenure to the primary holders or freemen who formed the cultivating body of the village; and suggests that this may have been the ground for the original distinction between occupancy and non-occupancy tenants. The following extract from a description of the Marātha villages by Grant Duff53 may be subjoined to this passage: “The inhabitants are principally cultivators, and are now either Mirāsidars or Ooprees. These names serve to distinguish the tenure by which they hold their lands. The Oopree is a mere tenant-at-will, but the Mirāsidar is a hereditary occupant whom the Government cannot displace so long as he pays the assessment on his field. With various privileges and distinctions in his village of minor consequence, the Mirāsidar has the important power of selling or transferring his right of occupancy at pleasure. It is a current opinion in the Marātha country that all the lands were originally of this description.”

As regards the internal relations of clans and village groups, Sir H. Maine states: “The men who composed the primitive communities believed themselves to be kinsmen in the most literal sense of the word; and, surprising as it may seem, there are a multitude of indications that in one stage of thought they must have regarded themselves as equals. When these primitive bodies first make their appearance as landowners, as claiming an exclusive enjoyment in a definite area of land, not only do their shares of the soil appear to have been originally equal, but a number of contrivances survive for preserving the equality, of which the most frequent is the periodical redistribution of the tribal domain.”54 Similarly Professor Hearn states: “The settlement of Europe was made by clans. Each clan occupied a certain territory[45]—much, I suppose, as an Australian squatter takes up new country. The land thus occupied was distributed by metes and bounds to each branch of the clan; the remainder, if any, continuing the property of the clan.”55 And again: “In those cases where the land had been acquired by conquest there were generally some remains of the conquered population who retained more or less interest in the lands that had once been their own. But as between the conquerors themselves it was the clansmen, and the clansmen only, who were entitled to derive any advantage from the land that the clan had acquired. The outsiders, the men who lived with the clan but were not of the clan, were no part of the folk, and had no share in the folkland. No services rendered, no participation in the common danger, no endurance of the burden and heat of the day, could create in an outsider any colour of right. Nothing short of admission to the clan, and of initiation in its worship, could enable him to demand as of right the grass of a single cow or the wood for a single fire.”56

Thus it appears that the cultivating community of each village constituted an exogamous clan, the members of which believed themselves to be kinsmen. When some caste or tribe occupied a fresh area of land they were distributed by clans in villages, over the area, all the cultivators of a village being of one caste or tribe, as is still the case with the Kunbis in Berār. Sometimes several alien castes or groups became amalgamated into a single caste, such as the Kurmis and Kunbis; in others they either remained as a separate caste or became one. When the non-Aryan tribes retained possession of the land, there is every reason to suppose that they also were admitted into Hinduism, and either constituted a fresh caste with the cultivating status, or were absorbed into an existing one with a change of name. Individual ownership of land was probably unknown. The patel or village headman, on whom proprietary right was conferred by the British Government, certainly did not possess it previously. He was simply the spokesman and representative [46]of the village community in its dealings with the central or ruling authority. But it seems scarcely likely either that the village community considered itself to own the land. Cases in which the community as a corporate body has exercised any function of ownership other than that of occupying and cultivating the soil, if recorded at all, must be extremely rare, and I do not know that any instance is given by Sir Henry Maine. A tutelary village god is to be found as a rule in every Hindu village. In the Central Provinces the most common is Khermāta, that is the goddess of the village itself or the village lands. She is a form of Devi, the general earth-goddess. When a village is founded the first thing to be done is to install the village god. Thus the soil of the village is venerated as a goddess, and it seems doubtful whether the village community considered itself the owner. In the Marātha Districts, Hanumān or Mahābīr, the monkey god, is the tutelary deity of the village. His position seems to rest on the belief of the villagers that the monkeys were the lords and owners of the soil before their own arrival. For the worship of these and the other village gods there is usually a village priest, known as Bhumka, Bhumia, Baiga or Jhānkar, who is taken from the non-Aryan tribes. The reason for his appointment seems to be that the Hindus still look on themselves to some extent as strangers and interlopers in relation to the gods of the earth and the village, and consider it necessary to approach these through the medium of one of their predecessors. The words Bhumka and Bhumia both mean lord of the soil, or belonging to the soil. As already seen, the authority of some menial official belonging to the indigenous tribes is accepted as final in cases of disputed boundaries, the idea being apparently that as his ancestors first occupied the village, he has inherited from them the knowledge of its true extent and limits. All these points appear to tell strongly against the view that the Hindu village community considered itself to own the village land as we understand the phrase. They seem to have looked on the land as a god, and often their own tutelary deity and protector. What they held themselves to possess was a right of occupancy, in virtue of prescriptive settlement, not subject to removal or [47]disturbance, and transmitted by inheritance to persons born into the membership of the village community. Under the Muhammadans the idea that the state ultimately owned the land may have been held, but prior to them the existence of such a belief is doubtful. The Hindu king did not take rent for land, but a share of the produce for the support of his establishments. The Rājpūt princes did not call themselves after the name of their country, but of its capital town, as if their own property consisted only in the town, as Jodhpur, Jaipur and Udaipur, instead of Mārwār, Dhūndhār and Mewār. Just as the village has a priest of the non-Aryan tribes for propitiating the local gods, so the Rājpūt chief at his accession was often inducted to the royal cushion by a Bhīl or Mīna, and received the badge of investiture as if he had to obtain his title from these tribes. Indeed the right of the village community to the land was held sometimes superior to that of the state. Sir J. Malcolm relates that he was very anxious to get the village of Bassi in Indore State repopulated when it had lain waste for thirty-six years. He had arranged with the Bhīl headman of a neighbouring village to bring it under cultivation on a favourable lease. The plan had other advantages, and Holkar’s minister was most anxious to put it into execution, but said that this could not be done until every possible effort had been made to discover whether any descendant of the former patel or of any watandār or hereditary cultivator of Bassi was still in existence; for if such were found, he said, “even we Marāthas, bad as we are, cannot do anything which interferes with their rights.” None such being found at the time, the village was settled as proposed by Malcolm; but some time afterwards, a boy was discovered who was descended from the old patel’s family, and he was invited to resume the office of headman of the village of his forefathers, which even the Bhīl, who had been nominated to it, was forward to resign to the rightful inheritor.57 Similarly the Marātha princes, Sindhia, Holkar and others, are recorded to have set more store by the headship of the insignificant Deccan villages, which were the hereditary offices of their families, than by the great principalities which they had carved out for themselves with the [48]sword. The former defined and justified their position in the world as the living link and representative of the continuous family comprising all their ancestors and all their descendants; the latter was at first regarded merely as a transient, secular possession, and a source of wealth and profit. This powerful hereditary right probably rested on a religious basis. The village community was considered to be bound up with its village god in one joint life, and hence no one but they could in theory have the right to cultivate the lands of that village. The very origin and nature of this right precluded any question of transfer or alienation. The only lands in which any ownership, corresponding to our conception of the term, was held to exist, were perhaps those granted free of revenue for the maintenance of temples, which were held to be the property of the god. In Rome and other Greek and Latin cities the idea of private or family ownership of land also developed from a religious sentiment. It was customary to bury the dead in the fields which they had held, and here the belief was that their spirits remained and protected the interests of the family. Periodical sacrifices were made to them and they participated in all the family ceremonies. Hence the land in which the tombs of ancestors were situated was held to belong to the family, and could not be separated from it.58 Gradually, as the veneration for the spirits of ancestors decayed, the land came to be regarded as the private property of the family, and when this idea had been realised it was made alienable, though not with the same freedom as personal property. But the word pecunia for money, from pecus a flock, like the Hindi dhan, which means wealth and also flocks of goats and sheep, and feudal from the Gaelic fiu, cattle, point to conditions of society in which land was not considered a form of private property or wealth. M. Fustel de Coulanges notices other primitive races who did not recognise property in land: “The Tartars understand the term property as applying to cattle, but not as applying to land. According to some authors, among the ancient Germans there was no ownership of land; every year each member of the tribe received a holding to cultivate, and the holding was changed in the following year. The German [49]owned the crop; he did not own the soil. The same was the case among a part of the Semitic race and certain of the Slav peoples.”59 In large areas of the Nigeria Protectorate at present, land has no exchangeable value at all; but by the native system of taxation a portion of the produce is taken in consideration of the right of use.60 In ancient Arabia ‘Baal’ meant the lord of some place or district, that is, a local deity, and hence came to mean a god. Land naturally moist was considered as irrigated by a god and the special place or habitation of the god. To the numerous Canaanite Baalims, or local deities, the Israelites ascribed all the natural gifts of the land, the corn, the wine, and the oil, the wool and the flax, the vines and fig trees. Pasture land was common property, but a man acquired rights in the soil by building a house, or, by ‘quickening’ a waste place, that is, bringing it under cultivation.61 The Israelites thought that they derived their title to the land of Canaan from Jehovah, having received it as a gift from Him. The association of rights over the land with cultivation and building, pointed out by Professor Robertson Smith, may perhaps explain the right over the village lands which was held to appertain to the village community. They had quickened the land and built houses on it, establishing the local village deity on their village sites, and it was probably thought that their life was bound up with that of the village god, and only they had a right to cultivate his land. This would explain the great respect shown by the Marāthas for hereditary title to land, as seen above; a feeling which must certainly have been based on some religious belief, and not on any moral idea of equity or justice; no such deep moral principle was possible in the Hindu community at the period in question. The Hindu religious conception of rights to land was thus poles apart from the secular English law of proprietary and transferable right, and if the native feeling could have been, understood by the early British administrators the latter would perhaps have been introduced only in a much modified form. [50]

The suggested conclusion from the above argument is that the main body of the Aryan immigrants, that is the Vaishyas, settled down in villages by exogamous clans or septs. The cultivators of each village believed themselves to be kinsmen descended from a common ancestor, and also to be akin to the god of the village lands from which they drew their sustenance. Hence their order had an equal right to cultivate the village land and their children to inherit it, though they did not conceive of the idea of ownership of land in the sense in which we understand this phrase.

The original status of the Vaishya, or a full member of the Aryan community who could join in sacrifices and employ Brāhmans to perform them, was gradually transferred to the cultivating member of the village communities. In process of time, as land was the chief source of wealth, and was also regarded as sacred, the old status became attached to castes or groups of persons who obtained or held land irrespective of their origin, and these are what are now called the good cultivating castes. They have now practically the same status, though, as has been seen, they were originally of most diverse origin, including bands of robbers and freebooters, cattle-lifters, non-Aryan tribes, and sections of any castes which managed to get possession of an appreciable quantity of land.

The second division of the group of pure or good castes, or those from whom a Brāhman can take water, comprises the higher artisan castes:

The most important of these are the Sunār or goldsmith; the Kasār or worker in brass and bell-metal; the Tamera or coppersmith; the Barhai or carpenter; and the Halwai and Bharbhūnja or confectioner and grain-parcher. The Sānsia or stone-mason of the Uriya country may perhaps also be included. These industries represent a higher degree of civilisation than the village trades, and the workers may probably have been formed into castes at a later period, when the practice of the handicrafts was no longer despised. The metal-working castes are now [51]usually urban, and on the average their members are as well-to-do as the cultivators. The Sunārs especially include a number of wealthy men, and their importance is increased by their association with the sacred metal, gold; in some localities they now claim to be Brāhmans and refuse to take food from Brāhmans.62 The more ambitious members abjure all flesh-food and liquor and wear the sacred thread. But in Bombay the Sunār was in former times one of the village menial castes, and here, before and during the time of the Peshwas, Sunārs were not allowed to wear the sacred thread, and they were forbidden to hold their marriages in public, as it was considered unlucky to see a Sunār bridegroom. Sunār bridegrooms were not allowed to see the state umbrella or to ride in a palanquin, and had to be married at night and in secluded places, being subject to restrictions and annoyances from which even Mahārs were free. Thus the goldsmith’s status appears to vary greatly according as his trade is a village or urban industry. Copper is also a sacred metal, and the Tameras rank next to the Sunārs among the artisan castes, with the Kasārs or brass-workers a little below them; both these castes sometimes wearing the sacred thread. These classes of artisans generally live in towns. The Barhai or carpenter is sometimes a village menial, but most carpenters live in towns, the wooden implements of agriculture being made either by the blacksmith or by the cultivators themselves. Where the Barhai is a village menial he is practically on an equality with the Lohār or blacksmith; but the better-class carpenters, who generally live in towns, rank higher. The Sānsia or stone-mason of the Uriya country works, as a rule, only in stone, and in past times therefore his principal employment must have been to build temples. He could not thus be a village menial, and his status would be somewhat improved by the sanctity of his calling. The Halwai and Bharbhūnja or confectioner and grain-parcher are castes of comparatively low origin, especially the latter; but they have to be given the status of ceremonial purity in order that all Hindus may be able to take sweets and parched grain from their hands. Their position resembles that of the barber [52]and waterman, the pure village menials, which will be discussed later. In Bengal certain castes, such as the Tānti or weaver of fine muslin, the Teli or oil-presser, and the Kumhār or potter, rank with the ceremonially pure castes. Their callings have there become important urban industries. Thus the Tāntis made the world-renowned fine muslins of Dacca; and the Jagannāthia Kumhārs of Orissa provide the earthen vessels used for the distribution of rice to all pilgrims at the temple of Jagannāth. These castes and certain others have a much higher rank than that of the corresponding castes in northern and Central India, and the special reasons indicated seem to account for this. Generally the artisan castes ranking on the same or a higher level than the cultivators are urban and not rural. They were not placed in a position of inferiority to the cultivators by accepting contributions of grain and gifts from them, and this perhaps accounts for their higher position. One special caste may be noticed here, the Vidūrs, who are the descendants of Brāhman fathers by women of other castes. These, being of mixed origin, formerly had a very low rank, and worked as village accountants and patwāris. Owing to their connection with Brāhmans, however, they are a well-educated caste, and since education has become the door to all grades of advancement in the public service, the Vidūrs have taken advantage of it, and many of them are clerks of offices or hold higher posts under Government. Their social status has correspondingly improved; they dress and behave like Brāhmans, and in some localities it is said that even Marātha Brāhmans will take water to drink from Vidūrs, though they will not take it from the cultivating castes. There are also several menial or serving castes from whom a Brāhman can take water, forming the third class of this group, but their real rank is much below that of the cultivators, and they will be treated in the next group.

The third main division consists of those castes from whom a Brāhman cannot take water, though they are not regarded as impure and are permitted to enter Hindu temples. The typical castes of this group appear to be the village artisans and menials and the village priests. The annexed list shows the principal of these. [53]

Village menials.

Castes of village watchmen.

Village priests and mendicants.

Others.

Other village traders and artisans.

Banjāra—Carrier.

The essential fact which formerly governed the status of this group of castes appears to be that they performed various services for the cultivators according to their different vocations, and were supported by contributions of grain made to [54]them by the cultivators, and by presents given to them at seed-time and harvest. They were the clients of the cultivators and the latter were their patrons and supporters, and hence ranked above them. This condition of things survives only in the case of a few castes, but prior to the introduction of a metal currency must apparently have been the method of remuneration of all the village industries. The Lohār or blacksmith makes and mends the iron implements of agriculture, such as the ploughshare, axe, sickle and goad. For this he is paid in Saugor a yearly contribution of 20 lbs. of grain per plough of land held by each cultivator, together with a handful of grain at sowing-time and a sheaf at harvest from both the autumn and spring crops. In Wardha he gets 50 lbs. of grain per plough of four bullocks or 40 acres. For new implements he must either be paid separately or at least supplied with the iron and charcoal. In Districts where the Barhai or carpenter is a village servant he is paid the same as the Lohār and has practically an equal status. The village barber receives in Saugor 20 lbs. of grain annually from each adult male in the family, or 22½ lbs. per plough of land besides the seasonal presents. In return for this he shaves each cultivator over the head and face about once a fortnight. The Dhobi or washerman gets half the annual contribution of the blacksmith and carpenter, with the same presents, and in return for this he washes the clothes of the family two or three times a month. When he brings the clothes home he also receives a meal or a wheaten cake, and well-to-do families give him their old clothes as a present. The Dhīmar or waterman brings water to the house morning and evening, and fills the earthen water-pots placed on a wooden stand or earthen platform outside it. When the cultivators have marriages he performs the same duties for the whole wedding party, and receives a present of money and clothes according to the means of the family, and his food every day while the wedding is in progress. He supplies water for drinking to the reapers, receiving three sheaves a day as payment, and takes sweet potatoes and boiled plums to the field and sells them. The Kumhār or potter is not now paid regularly by dues from the cultivators like other village menials, as the ordinary system of sale has been found to be more convenient [55]in his case. But he sometimes takes for use the soiled grass from the stalls of the cattle and gives pots free to the cultivator in exchange. On Akti day, at the beginning of the agricultural year, the village Kumhār in Saugor presents five pots with covers on them to each cultivator and is given 2½ lbs. of grain. He presents the bride with seven new pots at a wedding, and these are filled with water and used in the ceremony, being considered to represent the seven seas. At a funeral he must supply thirteen vessels which are known as ghāts, and must replace the household earthen vessels, which are rendered impure on the occurrence of a death in the house, and are all broken and thrown away. In the Punjab and Marātha country the Kumhār was formerly an ordinary village menial.

The office of village watchman is an important one, and is usually held by a member of the indigenous tribes. These formerly were the chief criminals, and the village watchman, in return for his pay, was expected to detect the crimes of his tribesmen and to make good any losses of property caused by them. The sections of the tribes who held this office have developed into special castes, as the Khangārs, Chadārs and Chauhāns of Chhattīsgarh. These last are probably of mixed descent from Rājpūts and the higher castes of cultivators with the indigenous tribes. The Dahāits were a caste of gatekeepers and orderlies of native rulers who have now become village watchmen. The Pankas are a section of the impure Gānda caste who have embraced the doctrines of the Kabīrpanthi sect and formed a separate caste. They are now usually employed as village watchmen and are not regarded as impure. Similarly those members of the Mahār servile caste who are village watchmen tend to marry among themselves and form a superior group to the others. The village watchman now receives a remuneration fixed by Government and is practically a rural policeman, but in former times he was a village menial and was maintained by the cultivators in the same manner as the others.

The village priests are another class of this group. The regular village priest and astrologer, the Joshi or Parsai, is a Brāhman, but the occupation has developed a [56]separate caste. The Joshi officiates at weddings in the village, selects auspicious names for children according to the constellations under which they were born, and points out the auspicious moment or mahūrat for weddings, name-giving and other ceremonies, and for the commencement of such agricultural operations as sowing, reaping, and threshing. He is also sometimes in charge of the village temple. He is supported by contributions of grain from the villagers and often has a plot of land rent-free from the proprietor. The social position of the Joshis is not very good, and, though Brāhmans, they are considered to rank somewhat below the cultivating castes. The Gurao is another village priest, whose fortune has been quite different. The caste acted as priests of the temples of Siva and were also musicians and supplied leaf-plates. They were village menials of the Marātha villages. But owing to the sanctity of their calling, and the fact that they have become literate and taken service under Government, the Guraos now rank above the cultivators and are called Shaiva Brāhmans. The Gondhalis are the village priests of Devi, the earth-goddess, who is also frequently the tutelary goddess of the village. They play the kettle-drum and perform dances in her honour, and were formerly classed as one of the village menials of Marātha villages, though they now work for hire. The Gārpagāri, or hail-averter, is a regular village menial, his duty being to avert hail-storms from the crops, like the χαλαζοφύλαξ in ancient Greece. The Gārpagāris will accept cooked food from Kunbis and celebrate their weddings with those of the Kunbis. The Jogis, Mānbhaos, Sātanis, and others, are wandering religious mendicants, who act as priests and spiritual preceptors to the lower classes of Hindus.

With the village priests may be mentioned the Māli or gardener. The Mālis now grow vegetables with irrigation or ordinary crops, but this was not apparently their original vocation. The name is derived from māla, a garland, and it would appear that the Māli was first employed to grow flowers for the garlands with which the gods and also their worshippers were adorned at religious ceremonies. Flowers were held sacred and were an essential adjunct to worship in India as in Greece and Rome. The sacred flowers of [57]India are the lotus, the marigold and the champak63 and from their use in religious worship is derived the custom of adorning the guests with garlands at all social functions, just as in Rome and Greece they wore crowns on their heads. It seems not unlikely that this was the purpose for which cultivated flowers were first grown, at any rate in India. The Māli was thus a kind of assistant in the religious life of the village, and he is still sometimes placed in charge of the village shrines and is employed as temple-servant in Jain temples. He would therefore have been supported by contributions from the cultivators like the other village menials and have ranked below them, though on account of the purity and sanctity of his occupation Brāhmans would take water from him. The Māli has now become an ordinary cultivator, but his status is still noticeably below that of the good cultivating castes and this seems to be the explanation. With the Māli may be classed the Barai, the grower and seller of the pān or betel-vine leaf. This leaf, growing on a kind of creeper, like the vine, in irrigated gardens roofed with thatch for protection from the sun, is very highly prized by the Hindus. It is offered with areca-nut, cloves, cardamom and lime rolled up in a quid to the guests at all social functions. It is endowed by them with great virtues, being supposed to prevent heartburn, indigestion, and other stomachic and intestinal disorders, and to preserve the teeth, while taken with musk, saffron and almonds, the betel-leaf is held to be a strong aphrodisiac. The juice of the leaf stains the teeth and mouth red, and the effect, though repulsive to Europeans, is an indispensable adjunct to a woman’s beauty in Hindu eyes. This staining of the mouth red with betel-leaf is also said to distinguish a man from a dog. The idea that betel preserves the teeth seems to be unfounded. The teeth of Hindus appear to be far less liable to decay than those of Europeans, but this is thought to be because they generally restrict themselves to a vegetable diet and always rinse out their mouths with water after taking food. The betel-leaf is considered sacred; a silver ornament is made in its shape and it is often invoked in spells and [58]magic. The original vine is held to have grown from a finger-joint of Bāsuki, the Queen of the Serpents, and the cobra is worshipped as the tutelary deity of the pān-garden, which this snake is accustomed to frequent, attracted by the moist coolness and darkness. The position of the Barai is the same as that of the Māli; his is really a low caste, sometimes coupled with the contemned Telis or oil-pressers, but he is considered ceremonially pure because the betel-leaf, offered to gods and eaten by Brāhmans and all Hindus, is taken from him. The Barai or Tamboli was formerly a village menial in the Marātha villages.

The castes following other village trades mainly fall into this group, though they may not now be village menials. Such are the Kalār or liquor-vendor and Teli or oil-presser, who sell their goods for cash, and having learnt to reckon and keep accounts, have prospered in their dealings with the cultivators ignorant of this accomplishment. Formerly it is probable that the village Teli had the right of pressing all the oil grown in the village, and retaining a certain share for his remuneration. The liquor-vendor can scarcely have been a village menial, but since Manu’s time his trade has been regarded as a very impure one, and has ranked with that of the Teli. Both these castes have now become prosperous, and include a number of landowners, and their status is gradually improving. The Darzi or tailor is not usually attached to the village community; sewn clothes have hitherto scarcely been worn among the rural population, and the weaver provides the cloths which they drape on the body and round the head.64 The contempt with which the tailor is visited in English proverbial lore for working at a woman’s occupation attaches in a precisely similar manner in India to the weaver.65 But in Gujarāt the Darzi is found living in villages and here he is also a village menial. The Kachera or maker of the glass bangles which every Hindu married woman wears as a sign of her estate, ranks with the village artisans; his is probably an urban trade, but he has never become [59]prosperous or important. The Banjāras or grain-carriers were originally Rājpūts, but owing to the mixed character of the caste and the fact that they obtained their support from the cultivators, they have come to rank below these latter. The Wanjāri cultivators of Berār have now discarded their Banjāra ancestry and claim to be Kunbis. The Nat or rope-dancer and acrobat may formerly have had functions in the village in connection with the crops. In Kumaon66 a Nat still slides down a long rope from the summit of a cliff to the base as a rite for ensuring the success of the crops on the occasion of a festival of Siva. Formerly if the Nat or Bādi fell to the ground in his course, he was immediately despatched with a sword by the surrounding spectators, but this is now prohibited. The rope on which he slid down the cliff is cut up and distributed among the inhabitants of the village, who hang the pieces as charms on the eaves of their houses. The hair of the Nat is also taken and preserved as possessing similar virtues. Each District in Kumaon has its hereditary Nat or Bādi, who is supported by annual contributions of grain from the inhabitants. Similarly in the Central Provinces it is not uncommon to find a deified Nat, called Nat Bāba or Father Nat, as a village god. A Natni, or Nat woman, is sometimes worshipped; and when two sharp peaks of hills are situated close to each other, it is related that there was once a Natni, very skilful on the tight-rope, who performed before the king; and he promised her that if she would stretch a rope from the peak of one hill to that of the other, and walk across it, he would marry her and make her wealthy. Accordingly the rope was stretched, but the queen from jealousy went and cut it nearly through in the night, and when the Natni started to walk, the rope broke, and she fell down and was killed. Having regard to the Kumaon rite, it may be surmised that these legends commemorate the death of a Natni or acrobat during the performance of some feat of dancing or sliding on a rope for the magical benefit of the crops. And it seems possible that acrobatic performances may have had their origin in this manner. The point bearing on the present argument [60]is, however, that the Nat performed special functions for the success of the village crops, and on this account was supported by contributions from the villagers, and ranked with the village menials.

Some of the castes already mentioned, and one or two others having the same status, work as household servants as well as village menials. The Dhīmar is most commonly employed as an indoor servant in Hindu households, and is permitted to knead flour in water and make it into a cake, which the Brāhman then takes and puts on the girdle with his own hands. He can boil water and pour pulse into the cooking-pot from above, so long as he does not touch the vessel after the food has been placed in it. He will take any remains of food left in the cooking-pot, as this is not considered to be polluted, food only becoming polluted when the hand touches it on the dish after having touched the mouth. When this happens, all the food on the dish becomes jūtha or leavings of food, and as a general rule no caste except the sweepers will eat these leavings of food of another caste or of another person of their own. Only a wife, whose meal follows her husband’s, will eat his leavings. As a servant, the Dhīmar is very familiar with his master; he may enter any part of the house, including the cooking-place and the women’s rooms, and he addresses his mistress as ‘Mother.’ When he lights his master’s pipe he takes the first pull himself, to show that it has not been tampered with, and then presents it to him with his left hand placed under his right elbow in token of respect. Maid-servants frequently belong also to the Dhīmar caste, and it often happens that the master of the household has illicit intercourse with them. Hence there is a proverb: ‘The king’s son draws water and the water-bearer’s son sits on the throne,’—similar intrigues on the part of high-born women with their servants being not unknown. The Kahār or palanquin-bearer was probably the same caste as the Dhīmar. Landowners would maintain a gang of Kahārs to carry them on journeys, allotting to such men plots of land rent-free. Our use of the word ‘bearer’ in the sense of a body-servant has developed from the palanquin-bearer who became a personal attendant on his master. Well-to-do families often have a Nai or barber [61]as a hereditary family servant, the office descending in the barber’s family. Such a man arranges the marriages of the children and takes a considerable part in conducting them, and acts as escort to the women of the family when they go on a journey. Among his daily duties are to rub his master’s body with oil, massage his limbs, prepare his bed, tell him stories to send him to sleep, and so on. The barber’s wife attends on women in childbirth after the days of pollution are over, and rubs oil on the bodies of her clients, pares their nails and paints their feet with red dye at marriages and on other festival occasions. The Bāri or maker of leaf-plates is another household servant. Plates made of large leaves fastened together with little wooden pins and strips of fibre are commonly used by the Hindus for eating food, as are little leaf-cups for drinking; glazed earthenware has hitherto not been commonly manufactured, and that with a rougher surface becomes ceremonially impure by contact with any strange person or thing. Metal vessels and plates are the only alternative to those made of leaves, and there are frequently not enough of them to go round for a party. The Bāris also work as personal servants, hand round water, and light and carry torches at entertainments and on journeys. Their women are maids to high-caste Hindu ladies, and as they are always about the zenana are liable to lose their virtue.

The castes of village and household menials form a large group between the cultivators on the one hand and the impure and servile labourers on the other. Their status is not exactly the same. On the one hand, the Nai or barber, the Kahār and Dhīmar or watermen, the household servants, the Bāri, Ahīr, and others, some of the village priests and the gardening castes, are considered ceremonially pure and Brāhmans will take water from them. But this is a matter of convenience, as, if they were not so held pure, they would be quite useless in the household. Several of these castes, as the Dhīmars, Bāris and others, are derived from the primitive tribes. Sir H. Risley considered the Bāris of Bengal as probably an offshoot from the Bhuiya or Mūsahar tribe: “He still associates with the Bhuiyas at times, and if the demand for leaf-plates and cups is greater than he [62]can cope with himself, he gets them secretly made up by his ruder kinsfolk and passes them off as his own production. Instances of this sort, in which a non-Aryan or mixed group is promoted on grounds of necessity or convenience to a higher status than their antecedents would entitle them to claim, are not unknown in other castes, and must have occurred frequently in outlying parts of the country, where the Aryan settlements were scanty and imperfectly supplied with the social apparatus demanded by the theory of ceremonial purity. Thus the undoubtedly non-Aryan Bhuiyas have in parts of Chota Nāgpur been recognised as Jal-Acharani (able to give water to the higher castes) and it may be conjectured that the Kahārs themselves only attained this privilege in virtue of their employment as palanquin-bearers.”67 The fact that Brāhmans will take water from these castes does not in any way place them on a level with the cultivators; they remain menial servants, ranking, if anything, below such castes as Lohār, Teli and Kalār, from whom Brāhmans will not take water; but these latter are, as corporate bodies, more important and prosperous than the household menial castes, because their occupation confers a greater dignity and independence.

On the other hand, one or two of the village menials, such as the Dhobi or washerman, are considered to some extent impure. This is due to specially degrading incidents attaching to their occupation, as in the case of the Dhobi, the washing of the clothes of women in childbirth.68 And the Sungaria subcaste of Kumhārs, who keep pigs, are not touched, because the impurity of the animal is necessarily communicated to its owner’s house and person. Still, in the village society there is little real difference between the position of these castes and those of the other village menials.

The status of the village menial castes appears to be fixed by their dependent position on the cultivators. The latter are their patrons and superiors, to whom they look for a livelihood. Before the introduction of a currency in the rural tracts (an event of the last fifty to a hundred years) the village artisans and menials were supported by [63]contributions of grain from the cultivators. They still all receive presents, consisting of a sowing-basketful of grain at seed-time and one or two sheaves at harvest. The former is known as Bīj phūtni, or ‘The breaking of the seed,’ and the latter as Khanvār, or ‘That which is left’ Sometimes, after threshing, the menials are each given as much grain as will fill a winnowing-fan. When the peasant has harvested his grain, all come and beg from him. The Dhīmar brings some water-nut, the Kāchhi or market-gardener some chillies, the Barai betel-leaf, the Teli oil and tobacco, the Kalār liquor (if he drinks it), the Bania some sugar, and all receive grain in excess of the value of their gifts. The Joshi or village priest, the Nat or acrobat, the Gosain or religious mendicant and the Fakīr or Muhammadan beggar solicit alms. On that day the cultivator is said to be like a little king in his fields, and the village menials constitute his court. In purely agricultural communities grain is the principal source of wealth, and though the average Hindu villager may appear to us to be typical of poverty rather than wealth, such standards are purely relative. The cultivator was thus the patron and supporter of the village artisans and menials, and his social position was naturally superior to theirs. Among the Hindus it is considered derogatory to accept a gift from another person, the recipient being thereby placed in a position of inferiority to the donor. Some exception to this rule is made in the case of Brāhmans, though even with them it partly applies. Generally the acceptance of a gift of any value among Hindus is looked upon in the same manner as the taking of money in England, being held to indicate that the recipient is in an inferior social position to the giver. And the existence of this feeling seems to afford strong support to the reason suggested here for the relative status of the cultivating and village menial castes.

The group of village menial and artisan castes comes between the good cultivating castes who hold the status of the Vaishyas or body of the Aryans, and the impure castes, the subjected aborigines. The most reasonable theory of their status seems to be that it originated in mixed descent. As has already been seen, it was the common practice of [64]members of the higher classes to take lower-caste women either as wives or concubines, and a large mixed class would naturally result. Such children, born and brought up in the households of their fathers, would not be full members of the family, but would not be regarded as impure. They would naturally be put to the performance of the menial household duties, for which the servile castes were rendered unsuitable through their impure status. This would correspond with the tradition of the large number of castes originating in mixed descent, which is given in the Hindu sacred books. It has been seen that where menial castes are employed in the household, classes of mixed descent do as a matter of fact arise. And there are traces of a relationship between the cultivators and the menial castes, which would be best explained by such an origin. At a betrothal in the great Kunbi cultivating caste of the Marāthas, the services of the barber and washerman must be requisitioned. The barber washes the feet of the boy and girl and places vermilion on the foreheads of the guests; the washerman spreads a sheet on the ground on which the boy and girl sit. At the end of the ceremony the barber and washerman take the bride and bridegroom on their shoulders and dance to music in the marriage-shed, for which they receive small presents. After a death has occurred at a Kunbi’s house, the impurity is not removed until the barber and washerman have eaten in it. At a Kunbi’s wedding the Gurao or village priest brings the leafy branches of five trees and deposits them at Māroti’s69 temple, whence they are removed by the parents of the bride. Before a wedding, again, a Kunbi bride must go to the potter’s house and be seated on his wheel, while it is turned round seven times for good luck. Similarly at a wedding among the Hindustāni cultivating castes the bride visits the potter’s house and is seated on his wheel; and the washerman’s wife applies vermilion to her forehead. The barber’s wife puts red paint on her feet, the gardener’s wife presents her with a garland of flowers and the carpenter’s wife gives her a new wooden doll. At the wedding feast the barber, the washerman and the Bāri or personal servant also eat with the guests, though sitting [65]apart from them. Sometimes members of the menial and serving castes are invited to the funeral feast as if they belonged to the dead man’s caste. In Madras the barber and his wife, and the washerman and his wife, are known as the son and daughter of the village. And among the families of ruling Rājpūt chiefs, when a daughter of the house is married, it was customary to send with her a number of handmaidens taken from the menial and serving castes. These became the concubines of the bridegroom and it seems clear that their progeny would be employed in similar capacities about the household and would follow the castes of their mothers. The Tamera caste of coppersmiths trace their origin from the girls so sent with the bride of Dharam-Pāl, the Haihaya Rājpūt Rāja of Ratanpur, through the progeny of these girls by the Raja.

Many other castes belong to the group of those from whom a Brāhman cannot take water, but who are not impure. Among these are several of the lower cultivating castes, some of them growers of special products, as the Kāchhis and Mowārs or market-gardeners, the Dāngris or melon-growers, and the Kohlis and Bhoyars who plant sugarcane. These subsidiary kinds of agriculture were looked down upon by the cultivators proper; they were probably carried out on the beds and banks of streams and other areas not included in the regular holdings of the village, and were taken up by labourers and other landless persons. The callings of these are allied to, or developed from, that of the Māli or gardener, and they rank on a level with him, or perhaps a little below, as no element of sanctity attaches to their products. Certain castes which were formerly labourers, but have now sometimes obtained possession of the land, are also in this group, such as the Rajbhars, Kīrs, Mānas, and various Madras castes of cultivators. Probably these were once not allowed to hold land, but were afterwards admitted to do so. The distinction between their position and that of the hereditary cultivators of the village community was perhaps the original basis of the different kinds of tenant-right recognised by our revenue law, though these now, of course, depend solely on length of tenure and other incidents, and make [66]no distinction of castes. The shepherd castes who tend sheep and goats (the Gadarias, Dhangars and Kuramwārs) also fall into this group. Little sanctity attached to these animals as compared with the cow, and the business of rearing them would be left to the labouring castes and non-Aryan tribes. The names of all three castes denote their functional origin, Gadaria being from gādar, a sheep, Dhangar from dhan or small-stock, the word signifying a flock of sheep or goats and also wealth; and Kuramwār from kurri, the Telugu word for sheep. Others belonging to this group are the digging and earth-working castes, the Beldārs, Murhas, Nunias and so on, practically all derived from the indigenous tribes, who wander about seeking employment from the cultivators in the construction and repair of field embankments and excavation of wells and tanks; and various fishing and boating castes, as the Injhwārs, Naodas, Murhas and Kewats, who rank as equal to the Dhīmars, though they may not be employed in household or village service. Such castes, almost entirely derived from the non-Aryan tribes, may have come gradually into existence as the wants of society developed and new functions were specialised; they would naturally be given the social status already attaching to the village menial castes.

The fourth group in the scheme of precedence comprises the non-Aryan or indigenous tribes, who are really outside the caste system when this is considered as the social organisation of the Hindus, so long at least as they continue to worship their own tribal deities, and show no respect for Brāhmans nor for the cow. These tribes have, however, entered the Hindu polity in various positions. The leaders of some of them who were dominant in the early period were admitted to the Kshatriya or Rājpūt caste, and the origin of a few of the Rājput clans can be traced to the old Bhar and other tribes. Again, the aristocratic or landholding sections of several existing tribes are at present, as has been seen, permitted to rank with the good Hindu cultivating castes. In a few cases, as the Andhs, Halbas and Mānas, the tribe as a whole has become a Hindu caste, when it retained possession of the land in the centre [67]of a Hindu population. These have now the same or a slightly higher position than the village menial castes. On the other hand, those tribes which were subjugated and permitted to live with a servile status in the Hindu villages have developed into the existing impure castes of labourers, weavers, tanners and others, who form the lowest social group. The tribes which still retain their distinctive existence were not enslaved in this manner, but lived apart in their own villages in the forest tracts and kept possession of the land. This seems to be the reason why they rank somewhat higher than the impure castes, even though they may utterly defile themselves according to Hindu ideas by eating cow’s flesh. Some tribes, such as the Gonds, Binjhwārs and Kawars, counted amongst them the owners of large estates or even kingdoms, and consequently had many Hindu cultivators for their subjects. And, as the Hindus themselves say, they could not regard the Gonds as impure when they had a Gond king. Nevertheless, the Gond labourers in Hindu villages in the plains are more despised than the Gonds who live in their own villages in the hill country. And the conversion of the tribes as a whole to Hinduism goes steadily forward. At each census the question arises which of them should be classed as Hindus, and which as Animists or worshippers of their own tribal gods, and though the classification is necessarily very arbitrary, the process can be clearly observed. Thus the Andhs, Kolis, Rautias and Halbas are now all Hindus, and the same remark applies to the Kols, Bhīls and Korkus in several Districts. By strict abstention from beef, the adoption of Hindu rites, and to some extent of child-marriage, they get admission to the third group of castes from whom a Brāhman cannot take water. It will be desirable here to digress from the main argument by noticing briefly the origin and affinities of the principal forest tribes of the Central Provinces.

These tribes are divided into two families, the Munda or Kolarian, named after the Kol tribe, and the Dravidian, of which the former are generally held to be the older and more primitive. The word Kol is probably the Santāli hār, a man. “This word is used under various forms, such [68]as hār, hāra, ho and koro by most Munda tribes in order to denote themselves. The change of r to l is familiar and presents no difficulty.”70 The word is also found in the alternative name Ho for the Kol tribe, and in the names of the cognate Korwa and Korku tribes. The word Munda is a Sanskrit derivative meaning a head, and, as stated by Sir H. Risley, is the common term employed by the Kols for the headman of a village, whence it has been adopted as an honorific title for the tribe. In Chota Nāgpur those Kols who have partly adopted Hinduism and become to some degree civilised are called Munda, while the name Ho or Larka (fighting) Kol is reserved for the wilder section of the tribe.

The principal tribes of the Munda or Kolarian family in the Central Provinces are shown below:

One large group includes the Kol, Munda or Ho tribe itself and the Bhumij and Santāls, who appear to be local branches of the Kols called by separate names by the Hindus. The Kharias seem to be the earliest Kol settlers in Chota Nāgpur, who were subjugated by the later comers. The name Kol, as already seen, is probably a form of the Santali hār, a man. Similarly the name of the Korku tribe is simply a corruption of Koraku, young men, and that of the Korwa tribe is from the same root. The dialects of the Korku and Korwa tribes closely approximate to Mundāri. Hence it would seem that they were originally one tribe with the Kols, but have been separated for so long a period that their direct connection can no longer be proved. The disintegrating causes which have split up what was originally one into a number of distinct tribes, are probably no more than distance and settlement in different parts of the country, leading to cessation of intermarriage and social intercourse. The tribes have then obtained some variation in the original names or been given separate territorial or occupational designations by the Hindus, [69]and their former identity has gradually been forgotten. Both the Korwas of the Chota Nāgpur plateau and the Korkus of the Satpūra hills were known as Muāsi, a term having the meaning of robber or raider. The Korwas have also a subtribe called Korāku, and Mr. Crooke thinks that they were originally the same tribe. Sir G. Grierson states that the Korwa dialect is closely allied to Kharia. Similarly the resemblance of the name raises a presumption that the great Koli tribe of Gujarāt and western India may be a branch of the Kols who penetrated to the western coast along the Satpūlra and Central India hill ranges. The Kolis and Bhīls are tribes of the same country and are commonly spoken of together. Both have entirely lost their own language and cannot therefore be classified definitely either as Kolarian or Dravidian, but there is a probability that they are of the Kolarian family. The Nāhals, another tribe of the western Satpūra range, are an offshoot of the Korkus. They are coupled with the Bhīls and Kolis in old Hindu accounts.

The Savars, Sawaras or Saonrs are also a widely distributed tribe, being found as far west as Bundelkhand and east in Orissa and Ganjām. In the Central Provinces they have lost their own language and speak Hindi or Uriya, but in Madras they still retain their original speech, which is classified by Sir G. Grierson with Gadba as a Munda or Kolarian dialect. The name occurs in Vedic literature, and the tribe is probably of great antiquity. In the classical stories of their origin the first ancestor of the Savars is sometimes described as a Bhīl. The wide extension of the Savar tribe east and west is favourable to the hypothesis of the identity of the Kols and Kolis, who have a somewhat similar distribution. The Gadbas of Ganjām, and the Māl or Māle Pahāria tribe of Chota Nāgpur seem to be offshoots of the Savars. The Khairwārs or Kharwārs are an important tribe of Mīrzāpur and Chota Nāgpur. There is some reason for supposing that they are an occupational offshoot of the Kols and Cheros, who have become a distinct group through taking to the manufacture of edible catechu from the wood of the khair tree.71

Another great branch of the Kolarian family is that [70]represented by the Bhuiya and Baiga tribes and their offshoots, the Bhunjias, Bhainas and Binjhwārs. The Kolarian origin of the Bhuiyas has been discussed in the article on that tribe, and it has also been suggested that the Baiga tribe of the Central Provinces are an offshoot of the Bhuiyas. These tribes have all abandoned their own languages and adopted the local Aryan vernaculars. The name Bhuiya is a Sanskrit derivative from bhu, earth, and signifies ‘belonging to the soil.’ Bhumij, applied to a branch of the Kol tribe, has the same origin. Baiga is used in the sense of a village priest or a sorcerer in Chota Nāgpur, and the office is commonly held by members of the Bhuiya tribe in that locality, as being the oldest residents. Thus the section of the tribe in the Central Provinces appears to have adopted, or been given, the name of the office. The Bharias or Bharia-Bhumias of Jubbulpore seem to belong to the great Bhar tribe, once dominant over large areas of the United Provinces. They also hold the office of village priest, which is there known as Bhumia, and in some tracts are scarcely distinguished from the Baigas. Again, in Sambalpur the Bhuiyas are known as Bhumia Kol, and are commonly regarded as a branch of the Kol tribe. Thus it would seem that two separate settlements of the Kolarian races may have occurred; the earlier one would be represented by the Bhars, Bhuiyas, Baigas and kindred tribes who have entirely lost their own languages and identity, and have names given to them by the Hindus; and a later one of the Kols or Mundas and their related tribes, whose languages and tribal religion and organisation, though in a decaying state, can be fully recognised and recorded. And the Dravidian immigration would be subsequent to both of them. To judge from the cases in which the fissure or subdivision of single tribes into two or more distinct ones can still be observed, it seems quite a plausible hypothesis that the original immigrants may have consisted only of a single tribe on each occasion, and that the formation of new ones may have occurred after settlement. But the evidence does not warrant any definite assertion.

The principal Dravidian tribes are the Gonds, Khonds and Oraons. The Gonds were once dominant over the greater part of the Central Provinces, which was called Gondwana [71]after them. The above three names have in each case been given to the tribes by the Hindus. The following tribes are found in the Province:

Gond, Oraon or Kurukh, Khond, Kolām, Parja, Kamār. Tribal Castes: Bhatra, Halba, Dhoba. Doubtful: Kawar, Dhanwār.

The Gonds and Khonds call themselves Koi or Koitur, a word which seems to mean man or hillman. The Oraon tribe call themselves Kurukh, which has also been supposed to be connected with the Kolarian horo, man. The name Oraon, given to them by the Hindus, may mean farmservant, while Dhangar, an alternative name for the tribe, has certainly this signification.

There seems good reason to suppose that the Gonds and Khonds were originally one tribe divided through migration.72 The Kolāms are a small tribe of the Wardha Valley, whose dialect resembles those of the Gonds and Khonds. They may have split off from the parent tribe in southern India and come northwards separately. The Parjas appear to represent the earliest Gond settlers in Bastar, who were subjugated by later Gond and Rāj-Gond immigrants. The Halbas and Bhatras are mixed tribes or tribal castes, descended from the unions of Gonds and Hindus.

The Munda languages have been shown by Sir G. Grierson to have originated from the same source as those spoken in the Indo-Pacific islands and the Malay Peninsula. “The Mundas, the Mon-Khmer, the wild tribes of the Malay Peninsula and the Nicobarese all use forms of speech which can be traced back to a common source though they mutually differ widely from each other.”73 It would appear, therefore, that the Mundas, the oldest known inhabitants of India, perhaps came originally from the south-east, the islands of the Indian Archipelago and the Malay Peninsula, unless India was their original home and these countries were colonised from it.

Sir Edward Gait states: “Geologists tell us that the Indian Peninsula was formerly cut off from the north of Asia by sea, while a land connection existed on the one side with Madagascar and on the other with the Malay Archipelago; and though there is nothing to show that India was then [72]inhabited, we know that it was so in palaeolithic times, when communication was probably still easier with the countries to the north-east and south-west than with those beyond the Himalayas.”74 In the south of India, however, no traces of Munda languages remain at present, and it seems therefore necessary to conclude that the Mundas of the Central Provinces and Chota Nāgpur have been separated from the tribes of Malaysia who speak cognate languages for an indefinitely long period; or else that they did not come through southern India to these countries but by way of Assam and Bengal or by sea through Orissa. There is good reason to believe from the names of places and from local tradition that the Munda tribes were once spread over Bihār and parts of the Ganges Valley; and if the Kolis are an offshoot of the Kols, as is supposed, they also penetrated across Central India to the sea in Gujarāt and the hills of the western Ghāts. The presumption is that the advance of the Aryans or Hindus drove the Mundas from the open country to the seclusion of the hills and forests. The Munda and Dravidian languages are shown by Sir G. Grierson to be distinct groups without any real connection.

Though the physical characteristics of the two sets of tribes display no marked points of difference, the opinion has been generally held by ethnologists who know them that they represent two distinct waves of immigration, and the absence of connection between their languages bears out this view. It has always been supposed that the Mundas were in the country of Chota Nāgpur and the Central Provinces first, and that the Dravidians, the Gonds, Khonds and Oraons came afterwards. The grounds for this view are the more advanced culture of the Dravidians; the fact that where the two sets of tribes are in contact those of the Munda group have been ousted from the more open and fertile country, of which, according to tradition, they were formerly in possession; and the practice of the Gonds and other Dravidian tribes of employing the Baigas, Bhuiyas and other Munda tribes for their village priests, which is an acknowledgment that the latter as the earlier residents have a more familiar acquaintance with the local deities, and can solicit their favour and [73]protection with more prospect of success. Such a belief is the more easily understood when it is remembered that these deities are not infrequently either the human ancestors of the earliest residents or the local animals and plants from which they supposed themselves to be descended.

The Dravidian languages, Gondi, Kurukh and Khond, are of one family with Tamil, Telugu, Malayalam and Canarese, and their home is the south of India. The word Dravida comes from an older form Damila or Dramila, and was used in ancient Pali and Jain literature as a name for the people of the Tamil country.75 Afterwards it came to signify generally the people of southern India as opposed to Gaur or northern India.

As stated by Sir Edward Gait there is at present no evidence to show that the Dravidians came to southern India from any other part of the world, and for anything that is known to the contrary the languages may have originated there. The existence of the small Brahui tribe in Baluchistān who speak a Dravidian language but have no physical resemblance to other Dravidian races cannot be satisfactorily explained, but, as he points out, this is no reason for holding that the whole body of speakers of Dravidian languages entered India from the north-west, and, with the exception of this small group of Brahuis, penetrated to the south and settled there without leaving any traces of their passage.

The Dravidian languages occupy a large area in Madras, Mysore and Hyderābād, and they extend north into the Central Provinces and Chota Nāgpur where they die out, practically not being found west and north of this tract. As the languages are more highly developed and the culture of their speakers is far more advanced in the south, it is justifiable to suppose, pending evidence to the contrary, that the south is their home and that they have spread thence as far north as the Central Provinces. The Gonds and Oraons, too, have stories to the effect that they came from the south. The belief has hitherto been, at least in the Central Provinces, that both the Gonds and Baigas have been settled in this territory for an indefinite period, that is, from prior to any [74]Aryan or Hindu immigration. Mr. H.A. Crump, C.S., has however pointed out that if this was the case the Munda or Kolarian tribes, which have lost their own languages, should have adopted Dravidian and not Hindu forms of speech. As already seen, numerous Kolarian tribes, as the Binjhwār, Bhaina, Bhuiya, Baiga, Bhumij, Chero, Khairwār and the Kols themselves in the Central Provinces have entirely lost their own languages, as well as the Bhīls and Kolis, if these are held to be Kolarian tribes. None of them have adopted a Dravidian language, but all speak corrupt forms of the ancient Aryan vernaculars derived from Sanskrit. The fact seems to indicate that at the time when they abandoned their own languages these tribes were in contact with Hindus, and were not surrounded by Gonds, as several of them are at present. The history of the Central Provinces affords considerable support to the view that the Gond immigration occurred at a comparatively late period, perhaps in the ninth or tenth century, or even later, after a considerable part of the Province had been governed for some centuries by Rājpūt dynasties.76 The Gonds and Oraons still have well-defined legends about their immigration, which would scarcely be the case if it had occurred twenty centuries or more ago.

Any further evidence or argument as to the date of the Dravidian immigration would be of considerable interest.

The fifth or lowest group in the scheme of precedence is that of the impure castes who cannot be touched. If a high-caste Hindu touches one of them he should bathe and have his clothes washed. These castes are not usually allowed to live inside a Hindu village, but have a hamlet to themselves adjoining it. The village barber will not shave them, nor the washerman wash their clothes. They usually have a separate well assigned to them from which to draw water, and if the village has only one well, one side of it is allotted to them and the Hindus take water from the other side. Formerly they were subjected to more humiliating restrictions. In Bombay a Mahār might not spit on the ground lest a Hindu should be polluted by touching it with his foot, but had to hang an earthen pot round his neck to hold his spittle. He was [75]made to drag a thorny branch with him to brush out his footsteps, and when a Brāhman came by had to lie at a distance on his face lest his shadow might fall on the Brāhman.77 Even if the shadow of a Mahār or Māng fell on a Brāhman he was polluted and dare not taste food and water until he had bathed and washed the impurity away. In Madras a Paraiyān or Pariah pollutes a high-caste Hindu by approaching within a distance of 64 feet of him.78 The debased and servile position of the impure castes corresponds to that which, as already seen, attached to the Sūdras of the classical period. The castes usually regarded as impure are the tanners, bamboo-workers, sweepers, hunters and fowlers, gipsies and vagrants, village musicians and village weavers. These castes, the Chamārs, Basors, Mahārs, Koris, Gāndas and others are usually also employed as agricultural and casual labourers. Formerly, as already seen, they were not allowed to hold land. There is no reason to doubt that the status of impurity, like that of the Sūdra, was originally the mark of a subjugated and inferior race, and was practically equivalent to slavery. This was the position of the indigenous Indians who were subjugated by the Aryan invaders and remained in the country occupied by them. Though they were of different races, and the distinction was marked and brought home to themselves by the contrast in the colour of their skins, it seems probable that the real basis for their antagonism was not social so much as religious. The Indians were hated and despised by the immigrants as the worshippers of a hostile god. They could not join in the sacrifices by which the Aryans held communion with their gods, and the sacrifice itself could not even be held, in theory at least, except in those parts of India which were thoroughly subdued and held to have become the dwelling-place of the Aryan gods. The proper course prescribed by religion towards the indigenous residents was to exterminate them, as the Israelites should have exterminated the inhabitants of Canaan. But as this could not be done, because their numbers were too great or the conquerors not sufficiently ruthless, they were reduced to the servile condition [76]of impurity and made the serfs of their masters like the Amalekites and the plebeians and helots.

If the whole of India had been thoroughly subjugated and settled like the Punjab and Hindustān, it may be supposed that the same status of impurity would have been imposed upon all the indigenous races; but this was very far from being the case. In central and southern India the Aryans or subsequent immigrants from Central Asia came at first at any rate only in small parties, and though they may have established territorial states, did not regularly occupy the land nor reduce the indigenous population to a condition of servitude. Thus large bodies of these must have retained a free position, and on their acceptance of the new religion and the development of the caste system, became enrolled in it with a caste status on the basis of their occupation. Their leaders were sometimes admitted to rank as Kshatriyas or Rājpūts, as has been stated.

Subsequently, as the racial distinction disappeared, the impure status came to attach to certain despised occupations and to customs abhorrent to Hinduism, such as that of eating beef. But, as already seen, the tribes which have continued to live apart from the Hindus are not usually regarded as impure, though they may eat beef and even skin animals. The Dhīmars, who keep pigs, still have a higher status than the impure castes because they are employed as water-bearers and household servants. It is at least doubtful whether at the time when the stigma of impurity was first attached to the Sūdras the Hindus themselves did not sacrifice cows and eat beef.79 The castes noted below are usually regarded as impure in the Central Provinces.

The Dhobi (washerman) and Kumhār (potter) are sometimes included among the impure castes, but, as already noted, their status is higher than that of the castes in this list.

Audhelia: Labouring caste of mixed descent who keep pigs.

Balāhi: Weavers and village messengers and watchmen.

Basor: Bamboo basket-makers and village musicians.

Chamār: Tanners and labourers.

Gānda: Weavers and village musicians. [77]

Ghasia: Grass-cutters, labourers and sweepers.

Kaikāri: Vagrant basket-makers.

Kanjar, Beria, Sānsia: Gipsies and thieves.

Katia: Cotton-spinners.

Kori: Weavers and labourers.

Mādgi: Telugu tanners and hide-curriers.

Mahār: Weavers and labourers.

Māla: Telugu weavers and labourers.

Māng: Broom- and mat-makers and village musicians. They also castrate cattle.

Mehtar: Sweepers and scavengers.

Certain occupations, those of skinning cattle and curing hides, weaving the coarse country cloth worn by the villagers, making baskets from the rind of the bamboo, playing on drums and tom-toms, and scavenging generally are relegated to the lowest and impure castes. The hides of domestic animals are exceedingly impure; a Hindu is defiled even by touching their dead bodies and far more so by removing the skins. Drums and tom-toms made from the hides of animals are also impure. But in the case of weaving and basket-making the calling itself entails no defilement, and it would appear simply that they were despised by the cultivators, and as a considerable number of workers were required to satisfy the demand for baskets and cloth, were adopted by the servile and labouring castes. Basket- and mat-making are callings naturally suited to the primitive tribes who would obtain the bamboos from the forests, but weaving would not be associated with them unless cloth was first woven of tree-cotton. The weavers of the finer cotton and silk cloths, who live in towns, rank much higher than the village weavers, as in the case of the Koshtis and Tāntis, the latter of whom made the famous fine cotton cloth, known as abrawān, or ‘running water,’ which was supplied to the imperial Zenāna at Delhi. On one occasion a daughter of Aurāngzeb was reproached on entering the room for her immodest attire and excused herself by the plea that she had on seven folds of cloth over her body.80 In Bengal Brāhmans will take water from Tāntis, and it seems clear that their higher status is a consequence of the lucrative and important nature of their occupation.

The Katias are a caste of cotton-spinners, the name being derived from kātna, to cut or spin. But hand-spinning is now practically an extinct industry and the [78]Katias have taken to weaving or ordinary manual labour for a subsistence. The Kanjars and Berias are the gipsy castes of India. They are accustomed to wander about carrying their grass-matting huts with them. Many of them live by petty thieving and cheating. Their women practise palmistry and retail charms for the cure of sickness and for exorcising evil spirits, and love-philtres. They do cupping and tattooing and also make reed mats, cane baskets, palm-leaf mats and fans, ropes from grass- and tree-fibre, brushes for the cotton-loom, string-net purses and balls, and so on; and the women commonly dance and act as prostitutes. There is good reason for thinking that the Kanjars are the parents of the European gipsies, while the Thugs who formerly infested the high-roads of India, murdering solitary travellers and small parties by strangulation, may also have been largely derived from this caste.81

It can only be definitely shown in a few instances that the existing impure occupational castes were directly derived from the indigenous tribes. The Chamār and Kori, and the Chuhra and Bhangi, or sweepers and scavengers of the Punjab and United Provinces, are now purely occupational castes and their original tribal affinities have entirely disappeared. The Chamārs and Mehtars or sweepers are in some places of a superior physical type, of comparatively good stature and light complexion;82 this may perhaps be due to a large admixture of Hindu blood through their women, during a social contact with the Hindus extending over many centuries, and also to the fact that they eat flesh when they can obtain it, including carrion. Such types are, however, exceptional among the impure castes, and there is no reason to doubt their general origin from the non-Aryan tribes, which in a few instances can be directly traced. Thus it seems likely that the Kanjars, Berias, Sānsias and other gipsy groups, as well as the Mirāsis, the vagrant bards and genealogists of the lower classes of Hindus, are derived from the Dom caste or tribe of Bengal, who are largely employed as sweepers and scavengers as well as on ordinary labour. The evidence for the origin [79]of the above groups from the Doms is given in the article on Kanjar. Sir H.M. Elliot considered the Doms to be one of the original tribes of India. Again, there is no doubt that the impure Gānda caste, who are weavers, labourers and village musicians in the Uriya country and Chhattīsgarh Districts of the Central Provinces, are derived from the Pān tribe of Chota Nāgpur. The Pāns or Pābs are a regular forest tribe, and are sometimes called Gānda, while the Gāndas may be alternatively known as Pān. But the section of the tribe who live among the Hindus and are regarded as impure have now become a distinct caste with a separate name. The Bhuiya tribe were once the rulers of Chota Nāgpur; they still install the Rāja of Keonjhar, and have a traditional relation to other ruling families. But in parts of Chota Nāgpur and southern Bihār the Bhuiyas living in Hindu villages have become a separate impure caste with the opprobrious designation of Mūsahar or rat-eater. The great Mahār caste of the Marātha country or Bombay are weavers and labourers, and formerly cured hides, like the Chamārs and Koris of northern India. They are regarded as impure and were the serfs or villeins of the Kunbis, attached to the land. An alternative name for them is Dher, and this is supposed to be a corruption of Dharada a hillman, a name applied in Manu to all the indigenous races of India. Though the connection cannot be traced in all cases, there is thus no reason to doubt that the existing impure castes represent the subjected or enslaved section of the primitive non-Aryan tribes.

It has been seen that the old Aryan polity comprised four classes: the Brāhmans and Kshatriyas or priestly and military aristocracy; the Vaishyas or body of the Aryans, who were ceremonially pure and could join in sacrifices; and the Sūdras or servile and impure class of labourers. The Vaishyas became cultivators and herdsmen, and their status of ceremonial purity was gradually transferred to the cultivating members of the village community, because land was the main source of wealth. Between the last two there arose another class of village menials and craftsmen, originating principally from the offspring of fathers of the Aryan classes and Sūdra women, to whom [80]was left the practice of the village industries, despised by the cultivators. In spite of the almost complete fusion of races which the intercourse of centuries has effected, and the multiplication and rearrangement of castes produced by the diversity of occupation and other social factors, the divisions of the village community can still be recognised in the existing social gradation.

It has been seen also that occupation is the real basis of the division and social precedence of castes in India, as in all communities which have made any substantial progress in civilisation and social development. Distinctions of race, religion and family gradually disappear, and are merged in the gradation according to wealth or profession. The enormous majority of castes are occupational and their social position depends on their caste calling. Thus in the case of an important industry like weaving, there are separate castes who weave the finer kinds of cloth, as the Tāntis and Koshtis, while one subcaste of Koshtis, the Sālewārs, are distinguished as silk-weavers, and a separate caste of Patwas embroider silk and braid on cloth; other castes, as the Mahārs, Gāndas and Koris, weave coarse cloth, and a distinct caste of Katias existed for the spinning of thread, and the Muhammadan caste of Bahnas for cleaning cotton. The workers in each kind of metal have formed a separate caste, as the Lohārs or blacksmiths, the Kasārs or brass-workers, the Tameras or coppersmiths, and the Sunārs or gold- and silversmiths, while the Audhia subcaste of Sunārs83 and the Bharewas, an inferior branch of the Kasārs, work in bell-metal. Each of these castes makes ornaments of its own metal, while the Kachera caste84 make glass bangles, and the Lakheras make bangles from lac and clay. In the case of agriculture, as has been seen, there is usually a functional cultivating caste for each main tract of country, as the Jāts in the Punjab, the Kurmis in Hindustān, the Kunbis in the Deccan, the Chasas in Orissa, the Kāpus in the Telugu country and the Vellālas in the Tamil country. Except the Jāts, who were perhaps originally a racial caste, the above castes appear to include a number of heterogeneous [81]groups which have been welded into a single body through the acquisition of land and the status which it confers. Various other cultivating castes also exist, whose origin can be traced to different sources; on obtaining possession of the land they have acquired the cultivating status, but retained their separate caste organisation and name. Other agricultural castes have been formed for the growing of special products. Thus the Mālis are gardeners, and within the caste there exist such separate groups as the Phūlmālis who grow flowers, the Jire Mālis cumin and the Halde Mālis turmeric.85 Hindus generally object to cultivate sān-hemp,86 and some special castes have been formed from those who grew it and thus underwent some loss of status; such are the Lorhas and Kumrawats and Pathinas, and the Santora subcaste of Kurmis. The āl87 or Indian madder-dye is another plant to which objection is felt, and the Alia subcastes of Kāchhis and Banias consist of those who grow and sell it. The Dāngris and Kāchhis are growers of melons and other vegetables on the sandy stretches in the beds of rivers and the alluvial land on their borders which is submerged in the monsoon floods. The Barais are the growers and sellers of the betel-vine.

Several castes have been formed from military service, as the Marāthas, Khandaits, Rautias, Taonlas and Pāiks. All of these, except the Marāthas, are mainly derived from the non-Aryan tribes; since they have abandoned military service and taken as a rule to agriculture, their rank depends roughly on their position as regards the land. Thus the Marāthas and Khandaits became landowners, receiving grants of property as a reward for, or on condition of, military service like the old feudal tenures; they rank with, but somewhat above, the cultivating castes. The same is the case, though to a less degree, with the Rautias of Chota Nāgpur, a military caste mainly formed from the Kol tribe. On the other hand, the Pāiks or foot-soldiers and Taonlas have not become landholders and rank below the cultivating [82]castes. The Hatkars are a caste formed from Dhangars or shepherds who entered the Marātha armies. They are now called Bangi Dhangars or shepherds with the spears, and rank a little above other Dhangars.

The great majority of castes have been formed from occupation, but other sources of origin can be traced. Several castes are of mixed descent, as the Vidūrs, the descendants of Brāhman fathers and mothers of other castes; the Bhilālas, by Rajpūt fathers and Bhīl mothers; the Chauhāns, Audhelias, Khangārs and Dhākars of Bastar, probably by Hindu fathers and women of various indigenous tribes; the Kirārs of mixed Rajpūt descent, and others. These also now generally take rank according to their occupation and position in the world. The Vidūrs served as village accountants and ranked below the cultivators, but since they are well educated and have done well in Government service their status is rapidly improving. The Bhilālas are landholders and rank as a good cultivating caste. The Chauhāns and Khangārs are village watchmen and rank as menials below the cultivators, the Dhākars are farmservants and labourers with a similar position, while the Audhelias are labourers who keep pigs and are hence regarded as impure. The Halbas or ‘ploughmen’ are another mixed caste, probably the descendants of house-servants of the Uriya Rājas, who, like the Khandaits, formed a sort of militia for the maintenance of the chiefs authority. They are now mainly farmservants, as the name denotes, but where they hold land, as in Bastar, they rank higher, almost as a good cultivating caste.

Again, very occasionally a caste may be formed from a religious sect or order. The Bishnois were originally a Vaishnava sect, worshipping Vishnu as an unseen god, and refusing to employ Brāhmans. They have now become cultivators, and though they retain their sectarian beliefs, and have no Brāhman priests, are generally regarded as a Hindu cultivating caste. The Pankas are members of the impure Gānda caste who adhered to the Kabīrpanthi sect. They are now a separate caste and are usually employed as village watchmen, ranking with menials above the Gāndas and other similar castes. The Lingāyats are a large sect of [83]southern India, devoted to the worship of Siva and called after the lingam or phallic emblem which they wear. They have their own priests, denying the authority of Brāhmans, but the tendency now is for members of those castes which have become Lingāyats to marry among themselves and retain their relative social status, thus forming a sort of inner microcosm of Hinduism.

Occupation is the real determining factor of social status in India as in all other societies of at all advanced organisation. But though in reality the status of occupations and of castes depends roughly on the degree to which they are lucrative and respectable, this is not ostensibly the case, but their precedence, as already seen, is held to be regulated by the degree of ceremonial purity or impurity attaching to them. The Hindus have retained, in form at any rate, the religious constitution which is common or universal in primitive societies. The majority of castes are provided with a legend devised by the Brāhmans to show that their first ancestor was especially created by a god to follow their caste calling, or at least that this was assigned to him by a god. The ancestors of the bearer-caste of Kahārs were created by Siva or Mahādeo from the dust to carry his consort Parvati in a litter when she was tired; the first Māng was made by Mahādeo from his own sweat to castrate the divine bull Nandi when he was fractious, and his descendants have ever since followed the same calling, the impiety of mutilating the sacred bull in such a manner being thus excused by the divine sanction accorded to it. The first Māli or gardener gave a garland to Krishna. The first Chamār or tanner made sandals for Siva from a piece of his own skin; the ancestor of the Kāyasth or writer caste, Chitragupta, keeps the record of men’s actions by which they are judged in the infernal regions after death; and so on.

All important castes are divided into a number of subordinate groups or subcastes, which as a rule marry and take food within their own circle only. Certain differences of status frequently exist among the subcastes of the occupational or social type, but these are usually too minute to be recognised by outsiders. The most common type of [84]subcaste is the local, named after the tract of country in which the members reside or whence they are supposed to have come. Thus the name Kanaujia from the town of Kanauj on the Ganges, famous in ancient Indian history, is borne by subcastes of many castes which have immigrated from northern India. Jaiswār, from the old town of Jais in the Rai Bareli District, is almost equally common. Pardeshi or foreign, and Pūrabia or eastern, are also subcaste names for groups coming from northern India or Oudh. Mahobia is a common name derived from the town of Mahoba in Central India, as are Bundeli from Bundelkhand, Narwaria from Narwar and Mārwāri from Mārwār in Rājputāna. Groups belonging to Berār are called Berāri, Warade or Baone; those from Gujarāt are called Lād, the classical term for Gujarāt, or Gujarāti, and other names are Deccani from the Deccan, Nimāri of Nimar, Havelia, the name of the wheat-growing tracts of Jubbulpore and Damoh; Chhattīsgarhia, Kosaria, Ratanpuria (from the old town of Ratanpur in Bilāspur), and Raipuria (from Raipur town), all names for residents in Chhattīsgarh; and so on. Brāhmans are divided into ten main divisions, named after different tracts in the north and south of India where they reside;88 and these are further subdivided, as the Mahārāshtra Brāhmans of the Marātha country of Bombay into the subcastes of Deshasth (belonging to the country) applied to those of the Poona country above the western Ghāts; Karhāra or those of the Satara District, from Karhar town; and Konkonasth or those of the Concan, the Bombay coast; similarly the Kanaujia division of the Pānch-Gaur or northern Brāhmans has as subdivisions the Kanaujia proper, the Jijhotia from Jajhoti, the old name of the Lalitpur and Saugor tract, which is part of Bundelkhand; the Sarwaria or those dwelling round the river Sarju in the United Provinces; the Mathuria from Muttra; and the Prayāgwāls or those of Allahabad (Prayāg), who act as guides and priests to pilgrims who come to bathe in the Ganges at the sacred city. The creation of new local subcastes seems to arise in two ways: when different groups of a caste settle in different tracts of country and are prevented from attending [85]the caste feasts and assemblies, the practice of intermarriage and taking food together gradually ceases, they form separate endogamous groups and for purposes of distinction are named after the territory in which they reside; this is what has happened in the case of Brāhmans and many other castes; and, secondly, when a fresh body of a caste arrives and settles in a tract where some of its members already reside, they do not amalgamate with the latter group, but form a fresh one and are named after the territory from which they have come, as in the case of such names as Pardeshi, Pūrabia, Gangapāri (‘from the other side of the Ganges’), and similar ones already cited. In former times, when the difficulties of communication were great, these local subcastes readily multiplied; thus the Kanaujia Brāhmans of Chhattīsgarh are looked down upon by those of Saugor and Damoh, as Chhattīsgarh has been for centuries a backward tract cut off from the rest of India, and they may be suspected of having intermarried with the local people or otherwise derogated from the standard of strict Hinduism. Similarly the Kanaujia Brāhmans of Bengal are split into several local subcastes named after tracts in Bengal, who marry among themselves and neither with other Kanaujias of Bengal nor with those of northern India. Since the opening of railways people can travel long distances to marriage and other ceremonies, and the tendency to form new subcastes is somewhat checked; a native gentleman said to me, when speaking of his people, that when a few families of Khedāwāl Brāhmans from Gujarāt first settled in Damoh they had the greatest difficulty in arranging their marriages; they could not marry with their caste-fellows in Gujarāt because their sons and daughters could not establish themselves, that is, could not prove their identity as Khedāwāl Brāhmans; but since the railway has been opened intermarriage takes place freely with other Khedāwāls in Gujarāt and Benāres. Proposals are on foot to authorise the intermarriage of the three great subcastes of Marātha Brāhmans: Deshasth, Konkonasth and Karhāra. As a rule, there is no difference of status between the different local subcastes, and a man’s subcaste is often not known except to his own caste-fellows. But occasionally a [86]certain derogatory sense may be conveyed; in several castes of the Central Provinces there is a subcaste called Jharia or jungly, a term applied to the oldest residents, who are considered to have lapsed in a comparatively new and barbarous country from the orthodox practices of Hinduism. The subcaste called Deshi, or ‘belonging to the country,’ sometimes has the same signification. The large majority of subcastes are of the local or territorial type.

Many subcastes are also formed from slight differences of occupation, which are not of sufficient importance to create new castes. Some instances of subcastes formed from growing special plants or crops have been given. Audhia Sunārs (goldsmiths) work in brass and bell-metal, which is less respectable than the sacred metal, gold. The Ekbeile Telis harness one bullock only to the oil-press and the Dobeile two bullocks. As it is thought sinful to use the sacred ox in this manner and to cover his eyes as the Telis do, it may be slightly more sinful to use two bullocks than one. The Udia Ghasias (grass-cutters) cure raw hides and do scavengers’ work, and are hence looked down upon by the others; the Dingkuchia Ghasias castrate cattle and horses, and the Dolboha carry dhoolies and palanquins. The Māngya Chamārs are beggars and rank below all other subcastes, from whom they will accept cooked food. Frequently, however, subcastes are formed from a slight distinction of occupation, which connotes no real difference in social status. The Hathgarhia Kumhārs (potters) are those who used to fashion the clay with their own hands, and the Chakarias those who turned it on a wheel. And though the practice of hand pottery is now abandoned, the divisions remain. The Shikāri or sportsmen Pardhis (hunters) are those who use firearms, though far from being sportsmen in our sense of the term; the Phanse Pardhis hunt with traps and snares; the Chitewale use a tame leopard to run down deer, and the Gayake stalk their prey behind a bullock. Among the subcastes of Dhīmars (fishermen and watermen) are the Singaria, who cultivate the singāra or water-nut in tanks, the Tānkiwālas or sharpeners of grindstones, the Jhīngars or prawn-catchers, the Bansias and Saraias or anglers (from bansi or sarai, a bamboo fishing-rod), the Kasdhonias who wash the [87]sands of the sacred rivers to find the coins thrown or dropped into them by pious pilgrims, and the Sonjharas who wash the sands of auriferous streams for their particles of gold.89 The Gāriwān Dangris have adopted the comparatively novel occupation of driving carts (gāri) for a livelihood, and the Pānibhar are water-carriers, while the ordinary occupation of the Dāngris is to grow melons in river-beds. It is unnecessary to multiply instances; here, as in the case of territorial subcastes, the practice of subdivision appears to have been extended from motives of convenience, and the slight difference of occupation is adopted as a distinguishing badge.

Subcastes are also occasionally formed from differences of social practice which produce some slight gain or loss of status. Thus the Biyāhut or ‘Married’ Kalārs prohibit the remarriage of widows, saying that a woman is married once for all, and hence rank a little higher than the others. The Dosar Banias, on the other hand, are said to take their name from dūsra, second, because they allow a widow to marry a second time and are hence looked upon by the others as a second-class lot. The Khedāwāl Brāhmans are divided into the ‘outer’ and ‘inner’: the inner subdivision being said to exist of those who accepted presents from the Rāja of Kaira and remained in his town, while the outer refused the presents, quitted the town and dwelt outside. The latter rank a little higher than the former. The Suvarha Dhīmars keep pigs and the Gadhewāle donkeys, and are considered to partake of the impure nature of these animals. The Gobardhua Chamārs wash out and eat the undigested grain from the droppings of cattle on the threshing-floors. The Chungia group of the Satnāmi Chamārs are those who smoke the chongi or leaf-pipe, though smoking is prohibited to the Satnāmis. The Nāgle or ‘naked’ Khonds have only a negligible amount of clothing and are looked down upon by the others. The Makaria Kamārs eat monkeys and are similarly despised.

Subcastes are also formed from mixed descent. The Dauwa Ahīrs are held to be the offspring of Ahīr women who were employed as wet-nurses in the houses of Bundela Rājpūts and bore children to their masters. The Halbas and [88]Rautias are divided into subcastes known as Puraīt or ‘pure,’ and Surāit or of ‘mixed’ descent. Many castes have a subcaste to which the progeny of illicit unions is relegated, such as the Dogle Kāyasths, and the Lahuri Sen subcaste of Barais, Banias and other castes. Illegitimate children in the Kasār (brass-worker) caste form a subcaste known as Tākle or ‘thrown out,’ Vidur or ‘illegitimate,’ or Laondi Bachcha, the issue of a kept wife. In Berār the Mahādeo Kolis, called after the Mahādeo or Pachmarhi hills, are divided into the Khas, or ‘pure,’ and the Akarāmāse or ‘mixed’; this latter word means gold or silver composed of eleven parts pure metal and one part alloy. Many subcastes of Bania have subcastes known as Bīsa or Dasa, that is ‘Twenty’ or ‘Ten’ groups, the former being of pure descent or twenty-carat, as it were, and the latter the offspring of remarried widows or other illicit unions. In the course of some generations such mixed groups frequently regain full status in the caste.

Subcastes are also formed from members of other castes who have taken to the occupation of the caste in question and become amalgamated with it; thus the Korchamārs are Koris (weavers) adopted into the Chamār (tanner) caste; Khatri Chhīpas are Khatris who have become dyers and printers; the small Dāngri caste has subcastes called Teli, Kalār and Kunbi, apparently consisting of members of those castes who have become Dāngris; the Bāman Darzis or tailors will not take food from any one except Brāhmans and may perhaps be derived from them, and the Kaith Darzis may be Kāyasths; and so on.

Occasionally subcastes may be formed from differences of religious belief or sectarian practice. In northern India even such leading Hindu castes as Rājpūts and Jāts have large Muhammadan branches, who as a rule do not intermarry with Hindus. The ordinary Hindu sects seldom, however, operate as a bar to marriage, Hinduism being tolerant of all forms of religious belief. Those Chamārs of Chhattīsgarh who have embraced the doctrines of the Satnāmi reforming sect form a separate endogamous subcaste, and sometimes the members of the Kabīrpanthi sect within a caste marry among themselves.

Statistics of the subcastes are not available, but their [89]numbers are very extensive in proportion to the population, and even in the same subcaste the members living within a comparatively small local area often marry among themselves and attend exclusively at their own caste feasts, though in the case of educated and well-to-do Hindus the construction of railways has modified this rule and connections are kept up between distant groups of relatives. Clearly therefore differences of occupation or social status are not primarily responsible for the subcastes, because in the majority of cases no such differences really exist. I think the real reason for their multiplication was the necessity that the members of a subcaste should attend at the caste feasts on the occasion of marriages, deaths and readmission of offenders, these feasts being of the nature of a sacrificial or religious meal. The grounds for this view will be given subsequently.

The caste or subcaste forms the outer circle within which a man must marry. Inside it are a set of further subdivisions which prohibit the marriage of persons related through males. These are called exogamous groups or clans, and their name among the higher castes is gotra. The theory is that all persons belonging to the same gotra are descended from the same male ancestor, and so related. The relationship in the gotra now only goes by the father’s side; when a woman marries she is taken into the clan of her husband and her children belong to it. Marriage is not allowed within the clan and in the course of a few generations the marriage of persons related through males or agnates is prohibited within a very wide circle. But on the mother’s side the gotra does not serve as a bar to marriage and the union of first cousins would be possible, other than the children of two brothers. According to Hindu law, intermarriage is prohibited within four degrees between persons related through females. But generally the children of first cousins are allowed to marry, when related partly through females. And several castes allow the intermarriage of first cousins, that of a brother’s daughter to a sister’s son and in a less degree of a brother’s son to a sister’s daughter being specially favoured. One or two Madras castes allow a man to marry his niece, and the small Dhoba caste of Mandla permit the union of children of the same mother but different fathers. [90]

Sir Herbert Risley classed the names of exogamous divisions as eponymous, territorial or local, titular and totemistic. In the body of this work the word clan is usually applied only to the large exogamous groups of the Rājpūts and one or two other military castes. The small local or titular groups of ordinary Hindu castes are called ‘section,’ and the totemic groups of the primitive tribes ‘sept.’ But perhaps it is simpler to use the word ‘clan’ throughout according to the practice of Sir J.G. Frazer. The vernacular designations of the clans or sections are gotra, which originally meant a stall or cow-pen; khero, a village; dih, a village site; baink, a title; mul or mur, literally a root, hence an origin; and kul or kuri, a family. The sections called eponymous are named after Rishis or saints mentioned in the Vedas and other scriptures and are found among the Brāhmans and a few of the higher castes, such as Vasishta, Garga, Bhāradwāj, Vishvamitra, Kashyap and so on. A few Rājpūt clans are named after kings or heroes, as the Rāghuvansis from king Rāghu of Ajodhia and the Tilokchandi Bais from a famous king of that name. The titular class of names comprise names of offices supposed to have been held by the founder of the clan, or titles and names referring to a personal defect or quality, and nicknames. Instances of the former are Kotwār (village watchman), Chaudhri, Meher or Māhto (caste headman), Bhagat (saint), Thākuria and Rawat (lord or prince), Vaidya (physician); and of titular names and nicknames: Kuldip (lamp of the family), Mohjaria (one with a burnt mouth), Jāchak (beggar), Garkata (cut-throat), Bhātpagar (one serving on a pittance of boiled rice), Kangāli (poor), Chīkat (dirty), Petdukh (stomach-ache), Ghunnere (worm-eater) and so on. A special class of names are those of offices held at the caste feasts; thus the clans of the Chitrakathi caste are the Atak or Mānkari, who furnish the headman of the caste panchāyat or committee; the Bhojin who serve the food at marriages and other ceremonies; the Kākra who arrange for the lighting; the Gothārya who keep the provisions, and the Ghorerao (ghora, a horse) who have the duty of looking after the horses and bullock-carts of the caste-men who assemble. Similarly the five principal clans of the small Turi caste are [91]named after the five sons of Singhbonga or the sun: the eldest son was called Mailuar and his descendants are the leaders or headmen of the caste; the descendants of the second son, Chardhagia, purify and readmit offenders to caste intercourse; those of the third son, Suremār, conduct the ceremonial shaving of such offenders, and those of the fourth son bring water for the ceremony and are called Tirkuār. The youngest brother, Hasdagia, is said to have committed some caste offence, and the four other brothers took the parts which are still played by their descendants in his ceremony of purification. In many cases exogamous clans are named after other castes or subcastes. Many low castes have adopted the names of the Rājpūt clans, either from simple vanity as people may take an aristocratic surname, or because they were in the service of Rājpūts, and have adopted the names of their masters or are partly descended from them. Other names of castes found among exogamous groups probably indicate that an ancestor belonging to that caste was taken into the one in which the group is found. The Bhaina tribe have clans named after the Dhobi, Ahīr, Gond, Māli and Panka castes. The members of such clans pay respect to any man belonging to the caste after which they are named and avoid picking a quarrel with him; they also worship the family gods of the caste.

Territorial names are very common, and are taken from that of some town or village in which the ancestor of the clan or the members of the clan themselves resided.90 The names are frequently distorted, and it seems probable that the majority of the large number of clan names for which no meaning can be discovered were those of villages. These unknown names are probably more numerous than the total of all those classes of names to which a meaning can be assigned.

The last class of exogamous divisions are those called totemistic, when the clan is named after a plant or animal or other natural object. These are almost universal among the non-Aryan or primitive tribes, but occur also in most Hindu castes, including some of the highest. The commonest totem names are those of the prominent animals, including several which are held sacred by the Hindus, as bāgh or [92]nāhar, the tiger; bachās, the calf; morkuria, the peacock; kachhwāha or limuān, the tortoise; nāgas, the cobra; hasti, the elephant; bandar, the monkey; bhainsa, the buffalo; richharia, the bear; kuliha, the jackal; kukura, the dog; karsayāl, the deer; heran, the black-buck, and so on. The utmost variety of names is found, and numerous trees, as well as rice, kodon and other crops, salt, sandalwood, cucumber, pepper, and some household implements, such as the pestle and rolling-slab, serve as names of clans. Names which may be held to have a totemistic origin occur even in the highest castes. Thus among the names of eponymous Rishis or saints, Bhāradwāj means a lark, Kaushik may be from the kūsha grass, Agastya from the agasti flower, Kashyap from kachhap, a tortoise; Taittiri from tītar, a partridge, and so on. Similarly the origin of other Rishis is attributed to animals, as Rishishringa to an antelope, Mandavya to a frog, and Kanāda to an owl.91 An inferior Rājpūt clan, Meshbansi, signifies descendants of the sheep, while the name of the Baghel clan is derived from the tiger (bāgh), that of the Kachhwāha clan perhaps from kachhap, a tortoise, of the Haihaivansi from the horse, of the Nāgvansi from the cobra, and of the Tomara clan from tomar, a club. The Karan or writer caste of Orissa, similarly, have clans derived from the cobra, tortoise and calf, and most of the cultivating and other middle castes have clans with totemistic names. The usual characteristics of totemism, in its later and more common form at any rate, are that members of a clan regard themselves as related to, or descended from, the animal or tree from which the clan takes its name, and abstain from killing or eating it. This was perhaps not the original relation of the clan to its clan totem in the hunting stage, but it is the one commonly found in India, where the settled agricultural stage has long been reached. The Bhaina tribe have among their totems the cobra, tiger, leopard, vulture, hawk, monkey, wild dog, quail, black ant, and so on. Members of a clan will not injure the animal after which it is named, and if they see the corpse of the animal or hear of its death they throw away an earthen cooking-pot, and bathe and shave themselves as for one of the family. At a wedding the bride’s [93]father makes an image in clay of the bird or animal of the groom’s sept and places it beside the marriage-post. The bridegroom worships the image, lighting a sacrificial fire before it, and offers to it the vermilion which he afterwards smears on the forehead of the bride. Women are often tattooed with representations of their totem animal, and men swear by it as their most sacred oath. A similar respect is paid to the inanimate objects after which certain septs are named. Thus members of the Gawad or cowdung clan will not burn cowdung cakes for fuel; and those of the Mircha clan do not use chillies. One clan is named after the sun, and when an eclipse occurs they perform the same formal rites of mourning as others do on the death of their totem animal. The Bāghani clan of Majhwārs, named after the tiger, think that a tiger will not attack any member of their clan unless he has committed an offence entailing temporary excommunication from caste. Until this offence has been expiated his relationship with the tiger as head of the clan is in abeyance, and the tiger will eat him as he would any other stranger. If a tiger meets a member of the clan who is free from sin, he will run away. Members of the Khoba or peg clan will not make a peg nor drive one into the ground. Those of the Dūmar or fig-tree clan say that their first ancestor was born under this tree. They consider the tree to be sacred and never eat its fruit, and worship it once a year. Sometimes the members of the clan do not revere the object after which it is named but some other important animal or plant. Thus the Markām clan of Gonds, named after the mango-tree, venerate the tortoise and do not kill it. The Kathotia clan of Kols is named after kathota, a bowl, but they revere the tiger. Bāgheshwar Deo, the tiger-god, resides on a little platform in their verandas. They may not join in a tiger-beat nor sit up for a tiger over a kill. In the latter case they think that the tiger would not come and would be deprived of his food, and all the members of their family would get ill. The Katharia clan take their name from kathri, a mattress. A member of this sept must never have a mattress in his house, nor wear clothes sewn in crosspieces as mattresses are sewn. The name of the Mudia or Mudmudia clan is said to mean shaven head, but they [94]apparently revere the white kumhra or gourd, perhaps because it has some resemblance to a shaven head. They give a white gourd to a woman on the day after she has borne a child, and her family then do not eat this vegetable for three years. The Kumraya sept revere the brown kumhra or gourd. They grow this vegetable on the thatch of their house-roof and from the time of planting it till the fruits have been plucked they do not touch it, though of course they afterwards eat the fruits. The Bhuwar sept are named after bhu or bhumi, the earth. They must always sleep on the earth and not on cots. The Nūn (salt) and Dhān (rice) clans of Oraons cannot dispense with eating their totems or titular ancestors. But the Dhān Oraons content themselves with refusing to consume the scum which thickens on the surface of the boiled rice, and the Nūn sept will not lick a plate in which salt and water have been mixed. At the weddings of the Vulture clan of the small Bhona caste one member of the clan kills a small chicken by biting off the head and then eats it in imitation of a vulture. Definite instances of the sacrificial eating of the totem animal have not been found, but it is said that the tiger and snake clans of the Bhatra tribe formerly ate their totems at a sacrificial meal. The Gonds also worship the cobra as a household god, and once a year they eat the flesh of the snake and think that by doing so they will be immune from snake-bite throughout the year. On the festival of Nāg-Panchmi the Mahārs make an image of a snake with flour and sugar and eat it. It is reported that the Singrore Dhīmars who work on rivers and tanks must eat the flesh of a crocodile at their weddings, while the Sonjharas who wash the sands of rivers for gold should catch a live crocodile for the occasion of the wedding and afterwards put it back into the river. These latter customs may probably have fallen into abeyance owing to the difficulty of catching a crocodile, and in any case the animals are tribal gods rather than totems.

Exogamy and totemism are found not only in India, but are the characteristics of primitive social groups over the greater part of the world. Totemism establishes a relation of kinship between persons belonging to one clan who are not related by blood, and exogamy prescribes that [95]the persons held to be so related shall not intermarry. Further, when terms of relationship come into existence it is found that they are applied not to members of one family, but to all the persons of the clan who might have stood in each particular relationship to the person addressing them. Thus a man will address as mother not only his own mother, but all the women of his clan who might have stood to him in the relation of mother. Similarly he will address all the old men and women as grandfather or grandmother or aunt, and the boys and girls of his own generation as brother and sister, and so on. With the development of the recognition of the consanguineous family, the use of terms of relationship tends to be restricted to persons who have actual kinship; thus a boy will address only his father’s brothers as father, and his cousins as brothers and sisters; but sufficient traces of the older system of clan kinship remain to attest its former existence. But it seems also clear that some, at least, of the terms of relationship were first used between persons really related; thus the word for mother must have been taught by mothers to their own babies beginning to speak, as it is a paramount necessity for a small child to have a name by which to call its mother when it is wholly dependent on her; if the period of infancy is got over without the use of this term of address there is no reason why it should be introduced in later life, when in the primitive clan the child quickly ceased to be dependent on its mother or to retain any strong affection for her. Similarly, as shown by Sir J.G. Frazer in Totemism and Exogamy, there is often a special name for the mother’s brother when other uncles or aunts are addressed simply as father or mother. This name must therefore have been brought into existence to distinguish the mother’s brother at the time when, under the system of female descent, he stood in the relation of a protector and parent to the child. Where the names for grandfather and grandmother are a form of duplication of those for father and mother as in English, they would appear to imply a definite recognition of the idea of family descent. The majority of the special names for other relatives, such as fraternal and maternal uncles and aunts, must also have been devised to designate those relatives in particular, and hence there is a [96]probability that the terms for father and brother and sister, which on a priori grounds may be considered doubtful, were also first applied to real or putative fathers and brothers and sisters. But, as already seen, under the classificatory system of relationship these same terms are addressed to members of the same clan who might by age and sex have stood in such a relationship to the person addressing them, but are not actually akin to him at all. And hence it seems a valid and necessary conclusion that at the time when the family terms of relationship came into existence, the clan sentiment of kinship was stronger than the family sentiment; that is, a boy was taught or made to feel that all the women of the clan of about the same age as his mother were as nearly akin to him as his own mother, and that he should regard them all in the same relation. And similarly he looked on all the men of the clan of an age enabling them to be his fathers in the same light as his own father, and all the children of or about his own age as his brothers and sisters. The above seems a necessary conclusion from the existence of the classificatory system of relationship, which is very widely spread among savages, and if admitted, it follows that the sentiment of kinship within the clan was already established when the family terms of relationship were devised, and therefore that the clan was prior to the family as a social unit. This conclusion is fortified by the rule of exogamy which prohibits marriage between persons of the same clan between whom no blood-relationship can be traced, and therefore shows that some kind of kinship was believed to exist between them, independent of and stronger than the link of consanguinity. Further, Mr. Hartland shows in Primitive Paternity92 that during the period of female descent when physical paternity has been recognised, but the father and mother belong to different clans, the children, being of the mother’s clan, will avenge a blood-feud of their clan upon their own father; and this custom seems to show clearly that the sentiment of clan-kinship was prior to and stronger than that of family kinship.

The same argument seems to demonstrate that the idea of kinship within the clan was prior to the idea of descent [97]from a common ancestor, whether an animal or plant, a god, hero or nicknamed ancestor. Because it is obvious that a set of persons otherwise unconnected could not suddenly and without reason have believed themselves to be descended from a common ancestor and hence related. If a number of persons not demonstrably connected by blood believe themselves to be akin simply on account of their descent from a common ancestor, it can only be because they are an expanded family, either actually or by fiction, which really had or might have had a common ancestor. That is, the clan tracing its descent from a common ancestor, if this was the primary type of clan, must have been subsequent to the family as a social institution. But as already seen the sentiment of kinship within the clan was prior to that within the family, and therefore the genesis of the clan from an expanded family is an impossible hypothesis; and it follows that the members of the clan must first have believed themselves to be bound together by some tie equivalent to or stronger than that of consanguineous kinship, and afterwards, when the primary belief was falling into abeyance, that of descent from a common ancestor came into existence to account for the clan sentiment of kinship already existing. If then the first form of association of human beings was in small groups, which led a migratory life and subsisted mainly by hunting and the consumption of fruits and roots, as the Australian natives still do, the sentiment of kinship must first have arisen, as stated by Mr. M’Lennan, in that small body which lived and hunted together, and was due simply to the fact that they were so associated, that they obtained food for each other, and on occasion protected and preserved each other’s lives.93 These small bodies of persons were the first social units, and according to our knowledge of the savage peoples who are nearest to the original migratory and hunting condition of life, without settled habitations, domestic animals or cultivated plants, they first called themselves after some animal or plant, usually, as Sir J.G. Frazer has shown in Totemism and Exogamy,94 after some edible animal or plant. The most probable theory of totemism on a priori grounds [98]seems therefore to be that the original small bodies who lived and hunted together, or totem-clans, called themselves after the edible animal or plant from which they principally derived their sustenance, or that which gave them life. While the real tie which connected them was that of living together, they did not realise this, and supposed themselves to be akin because they commonly ate this animal or plant together. This theory of totemism was first promulgated by Professor Robertson Smith and, though much disputed, appears to me to be the most probable. It has also been advocated by Dr. A.C. Haddon, F.R.S.95 The Gaelic names for family, teadhloch and cuedichc or coedichc, mean, the first, ‘having a common residence,’ the second, ‘those who eat together.’96 The detailed accounts of the totems of the Australian, Red Indian and African tribes, now brought together by Sir J.G. Frazer in Totemism and Exogamy, show a considerable amount of evidence that the early totems were not only as a rule edible animals, but the animals eaten by the totem-clans which bore their names.97 But after the domestication of animals and the culture of plants had been attained to, the totems ceased to be the chief means of subsistence. Hence the original tie of kinship was supplanted by another and wider one in the tribe, and though the totem-clans remained and continued to fulfil an important purpose, they were no longer the chief social group. And in many cases, as man had also by now begun to speculate on his origin, the totems came to be regarded as ancestors, and the totem-clans, retaining their sentiment of kinship, accounted for it by supposing themselves to be descended from a common ancestor. They thus also came to base the belief in clan-kinship on the tie of consanguinity recognised in the family, which had by now come into existence. This late and secondary form of totemism is that which obtains in India, where the migratory and hunting [99]stage has long been passed. The Indian evidence is, however, of great value because we find here in the same community, occasionally in the same caste, exogamous clans which trace their descent sometimes from animals and plants, or totems, and sometimes from gods, heroes, or titular ancestors, while many of the clans are named after villages or have names to which no meaning can be attached. As has been seen, there is good reason to suppose that all these forms of the exogamous clan are developed from the earliest form of the totem-clan; and since this later type of clan has developed from the totem-clan in India, it is a legitimate deduction that wherever elsewhere exogamous clans are found tracing their descent from a common ancestor or with unintelligible names, probably derived from places, they were probably also evolved from the totem-clan. This type of clan is shown in Professor Hearn’s Aryan Household to have been the common unit of society over much of Europe, where no traces of the existence of totemism are established.98 And from the Indian analogy it is therefore legitimate to presume that the totem-clan may have been the original unit of society among several European races as well as in America, Africa, Australia and India. Similar exogamous clans exist in China, and many of them have the names of plants and animals.99

In order to render clear the manner in which the clan named after a totem animal (or, less frequently, a plant) came to hold its members akin both to each other and their totem animals, an attempt may be made to indicate, however briefly and imperfectly, some features of primitive man’s conception of nature and life. Apparently when they began dimly to observe and form conscious mental impressions of the world around them, our first ancestors made some cardinal, though natural and inevitable, mistakes. In the first place they thought that the whole of nature was animate, and that every animal, plant, or natural object which they saw around them, was alive and self-conscious like themselves. They had, of course, no words or ideas connoting life or consciousness, or distinguishing animals, vegetables or lifeless [100]objects, and they were naturally quite incapable of distinguishing them. They merely thought that everything they saw was like themselves, would feel hurt and resentment if injured, and would know what was done to it, and by whom; whenever they saw the movement of an animal, plant, or other object, they thought it was volitional and self-conscious like their own movements. If they saw a tree waving in the wind, having no idea or conception of the wind, they thought the tree was moving its branches about of its own accord; if a stone fell, they, knowing nothing of the force of gravity, thought the stone projected itself from one place to another because it wished to do so. This is exactly the point of view taken by children when they first begin to observe. They also think that everything they see is alive like themselves, and that animals exercise volition and have a self-conscious intelligence like their own. But they quickly learn their mistakes and adopt the point of view of their elders because they are taught. Primitive man had no one to teach him, and as he did not co-ordinate or test his observations, the traces of this first conception of the natural world remain clearly indicated by a vast assortment of primitive customs and beliefs to the present day. All the most prominent natural objects, the sun and moon, the sky, the sea, high mountains, rivers and springs, the earth, the fire, became objects of veneration and were worshipped as gods, and this could not possibly have happened unless they had been believed to have life. Stone images and idols are considered as living gods. In India girls are married to flowers, trees, arrows, swords, and so on. A bachelor is married to a ring or a plant before wedding a widow, and the first ceremony is considered as his true marriage. The Saligrām, or ammonite stone, is held to represent the god Vishnu, perhaps because it was thought to be a thunderbolt and to have fallen from heaven. Its marriage is celebrated with the tulsi or basil-plant, which is considered the consort of Vishnu. Trees are held to be animate and possessed by spirits, and before a man climbs a tree he begs its pardon for the injury he is about to inflict on it. When a tank is dug, its marriage is celebrated. To the ancient Roman his hearth was a god; the walls and doors and threshold of his house were gods; [101]the boundaries of his field were also gods.100 It is precisely the same with the modern Hindu; he also venerates the threshold of his house, the cooking-hearth, the grinding-mill, and the boundaries of his field. The Jains still think that all animals, plants and inanimate objects have souls or spirits like human beings. The belief in a soul or spirit is naturally not primitive, as man could not at first conceive of anything he did not see or hear, but plants and inanimate objects could not subsequently have been credited with the possession of souls or spirits unless they had previously been thought to be alive. “The Fijians consider that if an animal or a plant dies its soul immediately goes to Bolotoo; if a stone or any other substance is broken, immortality is equally its reward; nay, artificial bodies have equal good luck with men and hogs and yams. If an axe or a chisel is worn out or broken up, away flies its soul for the service of the gods. If a house is taken down or any way destroyed, its immortal part will find a situation on the plains of Bolotoo. The Finns believed that all inanimate objects had their haltia or soul.”101 The Malays think that animals, vegetables and minerals, as well as human beings, have souls.102 The Kawar tribe are reported to believe that all articles of furniture and property have souls or spirits, and if any such is stolen the spirit will punish the thief. Theft is consequently almost unknown among them. All the fables about animals and plants speaking and exercising volition; the practice of ordeals, resting on the belief that the sacred living elements, fire and water, will of themselves discriminate between the innocent and guilty; the propitiatory offerings to the sea and to rivers, such incidents as Xerxes binding the sea with fetters, Ajax defying the lightning, Aaron’s rod that budded, the superstitions of sailors about ships: all result from the same primitive belief. Many other instances of self-conscious life and volition being attributed to animals, plants and natural objects are given by Lord Avebury in Origin of Civilisation, by Dr. Westermarck in The Origin and Development [102]of the Moral Ideas,103 and by Sir J.G. Frazer in The Golden Bough104

Thus primitive man had no conception of inanimate matter, and it seems probable that he did not either realise the idea of death. Though it may be doubtful whether any race exists at present which does not understand that death is the cessation of life in the body, indications remain that this view was not primary and may not have been acquired for some time. The Gonds apparently once thought that people would not die unless they were killed by magic, and similar beliefs are held by the Australian and African savages. Several customs also point to the belief in the survival of some degree of life in the body after death, apart from the idea of the soul.

Primitive man further thought that life, instead of being concentrated in certain organs, was distributed equally over the whole of the body. This mistake appears also to have been natural and inevitable when it is remembered that he had no name for the body, the different limbs and the internal organs, and no conception of their existence and distribution, nor of the functions which they severally performed. He perceived that sensation extended over all parts of the body, and that when any part was hurt or wounded the blood flowed and life gradually declined in vigour and ebbed away. For this reason the blood was subsequently often identified with the life. During the progress of culture many divergent views have been held about the source and location of life and mental and physical qualities, and the correct one that life is centred in the heart and brain, and that the brain is the seat of intelligence and mental qualities has only recently been arrived at. We still talk about people being hard-hearted, kind-hearted and heartless, and about a man’s heart being in the right place, as if we supposed that the qualities of kindness and courage were located in the heart, and determined by the physical constitution and location of the heart. The reason for this is perhaps that the soul was held to be the source of mental qualities, and to be somewhere in the [103]centre of the body, and hence the heart came to be identified with it. As shown by Sir J.G. Frazer in The Golden Bough many peoples or races have thought that the life and qualities were centred in the whole head, not merely in the brain. And this is the reason why Hindus will not appear abroad with the head bare, why it is a deadly insult to knock off a man’s turban, and why turbans or other head-gear were often exchanged as a solemn pledge of friendship. The superstition against walking under a ladder may have originally been based on some idea of its being derogatory or dangerous to the head, though not, of course, from the fear of being struck by a falling brick. Similarly, as shown in the article on Nai, the belief that the bodily strength and vigour were located in the hair, and to a less extent in the nails and teeth, has had a world-wide prevalence. But this cannot have been primary, because the hair had first to be conceived of apart from the rest of the body, and a separate name devised for it, before the belief that the hair was the source of strength could gradually come into existence. The evolution of these ideas may have extended over thousands of years. The expression ‘white-livered,’ again, seems to indicate that the quality of courage was once held to be located in the liver, and the belief that the liver was the seat of life was perhaps held by the Gonds. But the primary idea seems necessarily to have been that the life was equally distributed all over the body. And since, as will be seen subsequently, the savage was incapable of conceiving the abstract idea of life, he thought of it in a concrete form as part of the substance of the flesh and blood.

And since primitive man had no conception of inanimate matter it followed that when any part of the body was severed from the whole, he did not think of the separate fraction as merely lifeless matter, but as still a part of the body to which it had originally belonged and retaining a share of its life. For according to his view of the world and of animate nature, which has been explained above, he could not think of it as anything else. Thus the clippings of hair, nails, teeth, the spittle and any other similar products all in his view remained part of the body from which they had been severed and retained part of its life. In the case of the elements, earth, [104]fire and water, which he considered as living beings and subsequently worshipped as gods, this view was correct. Fractional portions of earth, fire and water, when severed from the remainder, retained their original nature and constitution, and afforded some support to his generally erroneous belief. And since he had observed that an injury done to any part of the body was an injury to the whole, it followed that if one got possession of any part of the body, such as the severed hair, teeth or nails, one could through them injure that body of which they still formed a part. It is for this reason that savages think that if an enemy can obtain possession of any waste product of the body, such as the severed hair or nails, that he can injure the owner through them. Similarly the Hindus thought that the clippings of the hair or nails, if buried in fertile ground, would grow into a plant, through the life which they retained, and as this plant waxed in size it would absorb more and more of the original owner’s life, which would consequently wane and decline. The worship of relics, such as the bones or hair of saints, is based on the same belief that they retain a part of the divine life and virtue of him to whom they once belonged.

It is probable that qualities were first conceived of by being observed in animals or natural objects. Prior to the introduction of personal names, the individuality of human beings could neither be clearly realised nor remembered after they were dead. But man must have perceived at an early period that certain animals were stronger or swifter than he was, or more cunning, and since the same quality was reproduced in every animal of the species, it could easily become permanently associated with the animal. But there were no names for qualities, nor any independent conception of them apart from the animal or animals in which they were observed. Supposing that strength and swiftness were mainly associated with the horse, as was often the case, then they would be necessarily conceived of as a part or essence of the horse and his life, not in the way we think of them, as qualities appertaining to the horse on account of the strength of his muscles and the conformation of his limbs. When names were devised for these qualities, they would be something equivalent to horsey or horse-like. The association of [105]qualities with animals is still shown in such words as asinine, owlish, foxy, leonine, mulish, dogged, tigerish, and so on; but since the inferiority of animals to man has long been recognised, most of the animal adjectives have a derogatory sense.105 It was far otherwise with primitive man, who first recognised the existence of the qualities most necessary to him, as strength, courage, swiftness, sagacity, cunning and endurance, as being displayed by certain animals in a greater degree than he possessed them himself. Birds he admired and venerated as being able to rise and fly in the air, which he could not do; fish for swimming and remaining under water when he could not; while at the same time he had not as yet perceived that the intelligence of animals was in any way inferior to his own, and he credited many of them with the power of speech. Thus certain animals were venerated on account of the qualities associated with them, and out of them in the course of time anthropomorphic gods personifying the qualities were evolved. The Australian aborigines of the kangaroo totem, when they wish to multiply the number of kangaroos, go to a certain place where two special blocks of stone project immediately one above the other from the hillside. One is supposed to represent an ‘old man’ kangaroo and the other a female. The stones are rubbed and then painted with alternate red and white stripes, the red stripes representing the red fur of the kangaroos, and the white ones its bones. After doing this some of them open veins in their arms and allow the blood to spurt over the stones. The other men sing chants referring to the increase in the numbers of the kangaroos, and they suppose that this ceremony will actually result in producing an increased number of kangaroos and hence an additional supply of food.106 Here the inference seems to be that the stones represent the centre or focus of the life of kangaroos, and when they are quickened by the painting, and the supply of blood, they will manifest their creative activity and increase the kangaroos. If we suppose that some similar stone existed on the Acropolis and was considered by the owl clan as the centre [106]of the life of the owls which frequented the hill, then when the art of sculpture had made some progress, and the superiority of the human form and intellect began to be apprehended, if a sculptor carved the stone into the semblance of a human being, the goddess Athena would be born.

It has been seen that primitive man considered the life and qualities to be distributed equally over the body in a physical sense, so that they formed part of the substance and flesh. The same view extended even to instrumental qualities or functions, since his mental powers and vision were necessarily limited by his language. Language must apparently have begun by pointing at animals or plants and making some sound, probably at first an imitation of the cry or other characteristic of the animal, which came to connote it. We have to suppose that language was at the commencement a help in the struggle for life, because otherwise men, as yet barely emerged from the animal stage, would never have made the painful mental efforts necessary to devise and remember the words. Words which would be distinctly advantageous in the struggle would be names for the animals and plants which they ate, and for the animals which ate them. By saying the name and pointing in any direction, the presence of such animals or plants in the vicinity would be intimated more quickly and more accurately than by signs or actions. Such names were then, it may be supposed, the first words. Animals or plants of which they made no use nor from which they apprehended any danger, would for long be simply disregarded, as nothing was to be gained by inventing names for them. The first words were all nouns and the names of visible objects, and this state of things probably continued for a long period and was the cause of many erroneous primitive conceptions and ideas. Some traces of the earliest form of language can still be discerned. Thus of Santāli Sir G. Grierson states: “Every word can perform the function of a verb, and every verbal form can, according to circumstances, be considered as a noun, an adjective or a verb. It is often simply a matter of convenience which word is considered as a noun and which as an adjective ... Strictly speaking, in Santāli there is no real verb as distinct from the other classes of words. Every [107]independent word can perform the function of a verb, and every verbal form can in its turn be used as a noun or adjective.”107 And of the Dravidian languages he says: “The genitive of ordinary nouns is in reality an adjective, and the difference between nouns and adjectives is of no great importance ... Many cases are both nouns and verbs. Nouns of agency are very commonly used as verbs.”108 Thus if it be admitted that nouns preceded verbs as parts of speech, which will hardly be disputed, these passages show how the semi-abstract adjectives and verbs were gradually formed from the names of concrete nouns. Of the language of the now extinct Tasmanian aborigines it is stated: “Their speech was so imperfectly constituted that there was no settled order or arrangement of words in the sentence, the sense being eked out by face, manner and gesture, so that they could scarcely converse in the dark, and all intercourse had to cease with nightfall. Abstract forms scarcely existed, and while every gum-tree or wattle-tree had its name, there was no word for ‘tree’ in general, nor for qualities such as hard, soft, hot, cold, etc. Anything hard was ‘like a stone,’ anything round ‘like the moon,’ and so on, the speaker suiting the action to the word, and supplementing the meaning to be understood by some gesture.”109 Here the original concrete form of language can be clearly discerned. They had a sufficiency of names for all the objects which were of use to them, and apparently verbal ideas were largely conveyed by gesture. Captain Forsyth states110 that though the Korkus very seldom wash themselves, there exist in their language eight words for washing, one for washing the face, another for the hands and others for different parts of the body. Thus we see that the verbal idea of washing was originally conceived not generally, but separately with reference to each concrete object or noun, for which a name existed and to which water was applied.

The primitive languages consisted only of nouns or the names of visible objects, possibly with the subsequent addition [108]of a few names for such conceptions as the wind and the voice, which could be heard, but not seen. There were no abstract nor semi-abstract terms nor parts of speech. The resulting inability to realise any abstract conception and the tendency to make everything concrete is a principal and salient characteristic of ethnology and primitive religion.111 All actions are judged by their concrete aspect or effects and not by the motives which prompted them, nor the results which they produce. For a Hindu to let a cow die with a rope round its neck is a grave caste offence, apparently because an indignity is thus offered to the sacred animal, but it is no offence to let a cow starve to death. A girl may be married to inanimate objects as already seen, or to an old man or a relative without any intention that she shall live with him as a wife, but simply so that she may be married before reaching puberty. If she goes through the ceremony of marriage she is held to be married. Yet the motive for infant-marriage is held to be that a girl should begin to bear children as soon as she is physically capable of doing so, and such a marriage is useless from this point of view. Some castes who cannot afford to burn a corpse hold a lighted brand to it or kindle a little fire on the grave and consider this equivalent to cremation. Promises are considered as concrete; among some Hindus promises are tied up in knots of cloth, and when they are discharged the knots are untied. Mr. S.C. Roy says of the Oraons: “Contracts are even to this day generally not written but acted. Thus a lease of land is made by the lessor handing over a clod of earth (which symbolises land) to the lessee; a contract of sale of cattle is entered into by handing over to the buyer a few blades of grass (which symbolise so many heads of cattle); a contract of payment of bride-price is made by the bridegroom’s father or other relative handing over a number of baris or small cakes of pulse (which symbolise so many rupees) to the bride’s father or other relative; and a contract of service is made by the mistress of the house anointing the head of the intended servant with oil, and making a present of a few pice, and entertaining him to a feast, thus signifying that he would receive food, lodging and some [109]pay.”112 Thus an abstract agreement is not considered sufficient for a contract; in each case it must be ratified by a concrete act.

The divisions of time are considered in a concrete sense. The fortnight or Nakshatra is presided over by its constellation, and this is held to be a nymph or goddess, who controls events during its course. Similarly, as shown in The Golden Bough,113 many kinds of new enterprises should be begun in the fortnight of the waxing moon, not in that of the waning moon. Days are also thought to be concrete and governed by their planets, and from this idea come all the superstitions about lucky and unlucky days. If a day had been from the beginning realised as a simple division of time no such superstitions could exist. Events, so far as they are conceived of, are also considered in a concrete sense. The reason why omens were so often drawn from birds114 is perhaps that birds fly from a distance and hence are able to see coming events on their way; and the hare and donkey were important animals of augury, perhaps because, on account of their long ears, they were credited with abnormally acute hearing, which would enable them to hear the sound of coming events before ordinary people. The proverb ‘Coming events cast their shadows before,’ appears to be a survival of this mode of belief, as it is obvious that that which has no substance cannot cast a shadow.

The whole category of superstitions about the evil eye arises from the belief that the glance of the eye is a concrete thing which strikes the person or object towards which it is directed like a dart. The theory that the injury is caused through the malice or envy of the person casting the evil eye seems to be derivative and explanatory. If a stranger’s glance falls on the food of a Rāmānuji Brāhman while it is being cooked, the food becomes polluted and must be buried in the ground. Here it is clear that the glance of the eye is equivalent to real contact of some part of the stranger’s body, which would pollute the food. In asking for leave in order [110]to nurse his brother who was seriously ill but could obtain no advantage from medical treatment, a Hindu clerk explained that the sick man had been pierced by the evil glance of some woman.

Similarly words were considered to have a concrete force, so that the mere repetition of words produced an effect analogous to their sense. The purely mechanical repetition of prayers was held to be a virtuous act, and this idea was carried to the most absurd length in the Buddhist’s praying-wheel, where merit was acquired by causing the wheel with prayers inscribed on its surface to revolve in a waterfall. The wearing of strips of paper, containing sacred texts, as amulets on the body is based on this belief, and some Muhammadans will wash off the ink from paper containing a verse of the Korān and drink the mixture under the impression that it will do them good. Here the belief in the concrete virtue and substance of the written word is very clear. The Hindus think that the continued repetition of the Gayatri or sacred prayer to the sun is a means of acquiring virtue, and the prayer is personified as a goddess. The enunciation of the sacred syllable Aum or Om is supposed to have the most powerful results. Homer’s phrase ‘winged words’ perhaps recalls the period when the words were considered as physical entities which actually travelled through the air from the speaker to the hearer and were called winged because they went so fast. A Korku clan has the name lobo which means a piece of cloth. But the word lobo also signifies ‘to leak.’ If a person says a sentence containing the word lobo in either signification before a member of the clan while he is eating, he will throw away the food before him as if it were contaminated and prepare a meal afresh. Here it is clear that the Korku pays no regard to the sense but solely to the word or sound. This belief in the concrete force of words has had the most important effects both in law and religion. The earliest codes of law were held to be commands of the god and claimed obedience on this ground. The binding force of the law rested in the words and not in the sense because the words were held to be those of the god and to partake of his divine nature. In ancient Rome the citizen had to take [111]care to know the words of the law and to state them exactly. If he used one wrong word the law gave him no assistance. “Gaius tells a story of a man whose neighbour had cut his vines; the facts were clear; he stated the law applying to his case, but he said vines, whereas the law said trees; he lost his suit.”115 The divine virtue attached to the sacred books of different religions rests on the same belief. Frequently the books themselves are worshipped, and it was held that they could not be translated because the sanctity resided in the actual words and would be lost if other words were used. The efficacy of spells and invocations seems to depend mainly on this belief in the concrete power of words. If one knows an efficacious form of words connoting a state of physical facts and repeats it with the proper accessory conditions, then that state of facts is actually caused to exist; and if one knows a man’s name and calls on him with a form of words efficacious to compel attendance, he has to come and his spirit can similarly be summoned from the dead. When a Malay wishes to kill an enemy he makes an image of the man, transfixes or otherwise injures it, and buries it on the path over which the enemy will tread. As he buries it with the impression that he will thereby cause the enemy to die and likewise be buried, he says:

and thinks that the repetition of these words produces the state of facts which they denote so that the guilt of the murder is removed from his own shoulders to those of the archangel Gabriel. Similarly when he has killed a deer and wishes to be free from the guilt of his action, or as he calls it to cast out the mischief from the deer, he says:

and so on, freeing himself in the same manner from responsibility for the death of the deer.116 Names also are regarded as concrete. Primitive man could not regard a [112]name as an abstract appellation, but thought of it as part of the person or thing to which it was applied and as containing part of his life, like his hair, spittle and the rest of his body. He would have used names for a long period before he had any word for a name, and his first idea of the name as a part of the substantive body to which it is applied has survived a more correct appreciation. Thus if one knew a person’s name one could injure him by working evil on it and the part of his life contained in it, just as one could injure him through the clippings of his hair, his spittle, clothes or the earth pressed by his foot. This is the reason for the common custom of having two names, one of which, the true name, is kept secret and only used on ceremonial occasions when it is essential, as at a wedding, while the other is employed for everyday life. The latter, not being the man’s true name, does not contain part of his life, and hence there is no harm in letting an enemy know it. Similarly the Hindus think that a child’s name should not be repeated at night, lest an owl might hear it, when this bird could injure the child through its name, just as if it got hold of a piece of cloth worn or soiled by the child. The practice of euphemism rests on this belief, as it was thought that if a person’s name was said and a part of him was thus caused to be present the rest would probably follow. Hence the rule of avoiding the use of the names of persons or things of which one does not desire the presence. Thus Sir E.B. Tylor says: “The Dayak will not speak of the smallpox by name, but will call it ‘The Chief,’ or ‘Jungle leaves,’ or say, ‘Has He left you?’ The euphemism of calling the Furies the Eumenides, or ‘Gracious Ones,’ is the stock illustration of this feeling, and the euphemisms for fairies and for the devil are too familiar to quote.”117 Similarly the name of a god was considered as part of him and hence partaking of his divine nature. It was thus so potent that it could not be mentioned on ordinary occasions or by common persons. Allah is only an epithet for the name of God among the Muhammadans and his True or Great Name is secret. Those who know it have power over all created things. Clearly then the [113]divine power is held to reside in the name itself. The concealment of the name of the tutelary deity of Rome, for divulging which Valerius Soranus is said to have paid the penalty of death, is a case in point.118 Sir E.B. Tylor gives many other interesting examples of the above ideas and points out the connection clearly existing in the savage mind between the name and the object to which it is applied. The Muhammadans think that Solomon’s name is very efficacious for casting out devils and evil spirits. The practice of naming children after gods or by the epithets or titles applied to the divine being, or after the names of saints, appears to be due to the belief that such names, by reason of their association with the god or saint, acquire a part of his divine life and virtue, which when given to children the names will in turn convey to them.119 On the other hand, when a Hindu mother is afraid lest her child may die, she sometimes gives it an opprobrious name as dirt, rubbish, sweepings, or sold for one or two cowries, so that the evil spirits who take the lives of children may be deceived by the name and think that such a valueless child is not worth having. The voice was also held to be concrete. The position of the Roman tribune was peculiar, as he was not a magistrate chosen by divine authority and hence could not summon people to his court; but the tribune had been dedicated to the city gods, and his person was sacrosanct. He could therefore lay hands on a man, and once the tribune touched him, the man was held to be in the magistrate’s power, and bound to obey him. This rule extended even to those who were within hearing of his voice; any one, even a patrician or consul, who heard the tribune’s voice was compelled to obey him. In this case it is clear that the voice and spoken words were held to be concrete, and to share in the sanctity attaching to the body.120 When primitive man could not think of a name as an abstraction but had to think of it as an actual part of the body and life of the person or visible object to which it belonged, it will be realised how impossible it was for him during a long period to conceive of any abstract [114]idea, which was only a word without visible or corporal reality.

Thus he could not at first have had any conception of a soul or spirit, which is an unseen thing. Savages generally may have evolved the conception of a soul or spirit as an explanation of dreams, according to the view taken by Mr. E. Clodd in Myths and Dreams,121 Mr. Clodd shows that dreams were necessarily and invariably considered as real events, and it could not have been otherwise, as primitive man would have been unable to conceive the abstract idea of a vision or fantasy. And since during dreams the body remained immobile and quiescent, it was thought that the spirit inside the body left it and travelled independently. Hence the reluctance often evinced to waking a sleeper suddenly from fear lest the absent spirit might not have time to return to the body before its awakening and hence the man might die. Savages, not having the conception of likeness or similarity,122 would confuse death and sleep, because the appearance of the body is similar in death and in sleep. Legends of the type of Rip Van Winkle and the Sleeping Beauty, and of heroes like King Arthur and Frederick Barbarossa lying asleep through the centuries in some remote cave or other hiding-place, from which they will one day issue forth to regenerate the world, perpetuate the primitive identification of death and sleep. And the belief long prevailed that after death the soul or spirit remained with the body in the place where it lay, leaving the body and returning to it as the spirit was held to do in sleep. The spirit was also thought to be able to quit the body and enter any other body, both during life and after death; most of the beliefs in spirit-possession and many of those about the power of witches arise from this view. The soul or spirit was commonly conceived of in concrete form; the Egyptians, Greeks and Hindus thought of it as a little mannikin inside the body. After death the Hindus often break the skull in order to allow the soul to escape. Often an insect or a stone is thought to harbour the spirit. As shown by Sir E. B. Tylor in Primitive Culture,123 the [115]breath, the shadow and the pupil of the eye were sometimes held to be or to represent the soul or spirit. Disembodied spirits are imprisoned in a tree or hole by driving nails into the tree or ground to confine them and prevent their exit. When a man died accidentally or a woman in childbirth and fear was felt that their spirits might annoy or injure the living, a stake might be driven through the body or a cairn of stones piled over it in order to keep the ghost down and prevent it from rising and walking. The genii of the Arabian Nights were imprisoned in sealed bottles, and when the bottle was opened they appeared in a cloud of vapour.

There seems every reason to suppose, as the same author suggests, that man first thought he had a spirit himself and as a consequence held that animals, plants and inanimate objects also contained spirits. Because the belief that the human body had a spirit can easily be accounted for, but there seems to be no valid reason why man should have thought that all other visible objects also contained spirits, except that at the period when he conceived of the existence of a soul or spirit he still held them to be possessed of life and self-conscious volition like himself. But certain beliefs, such as the universal existence of life, and of its distribution all over the body and transmission by contact and eating, the common life of the species, and possibly totemism itself, appear to have been pre-animistic or prior to any conception of or belief in a soul or spirit either in man himself or in nature.

Primitive man thought that the life and all qualities, mental and physical, were equally distributed over the body as part of the substance of the flesh. He thus came to think that they could be transferred from one body or substance to another in two ways: either by contact of the two bodies or substances, or by the eating or assimilation of one by the other. The transmission of qualities by contact could be indicated through simply saying the two names of the objects in contact together, and transmission by eating through saying the two names with a gesture of eating. Thus if one ate a piece of tiger’s flesh, one assimilated an equivalent amount of strength, ferocity, cruelty, yellowness, and any other qualities which might be attributed to the tiger. Warriors and youths are sometimes forbidden [116]to eat deer’s flesh because it will make them timid, but they are encouraged to eat the flesh of tigers, bears, and other ferocious animals, because it will make them brave. The Gonds, if they wish a child to be a good dancer, cause it to eat the flesh of a kind of hawk, which hangs gracefully poised over the water, with its wings continually flapping, on the look-out for its prey. They think that by eating the flesh the limbs of the child will become supple like the wings of the bird. If a child is slow in learning to speak, they give it to eat the leaves of the pīpal tree, which rustle continually in the wind and are hence supposed to have the quality of making a noise. All qualities, objective and instrumental, were conceived of in the same manner, because in the absence of verbs or abstract terms their proper relation to the subject and object could not be stated or understood. Thus if a woman’s labour in child-birth is prolonged she is given to drink water in which the charred wood of a tree struck by lightning has been dipped. Here it is clear that the quality of swiftness is held to have been conveyed by the lightning to the wood, by the wood to the water, and by the water to the woman, so as to give her a swift delivery. By a similar train of reasoning she is given to drink the water of a swiftly-flowing stream which thus has the quality of swiftness, or water poured through a gun-barrel in which the fouling of a bullet is left. Here the quality of swiftness appertaining to the bullet is conveyed by the soiling to the barrel and thence to the water and to the woman who drinks the water. In the above cases all the transfers except that to the woman are by contact. The belief in the transfer of qualities by contact may have arisen from the sensations of the body and skin, to which heat, cold and moisture are communicated by contact. It was applied to every kind of quality. A familiar instance is the worship of the marks on rocks or stone which are held to be the footprints left by a god. Here a part of the god’s divine virtue and power has been communicated through the sole of his foot to the rock dented by the latter. Touching for the king’s evil was another familiar case, when it was thought that a fraction of the king’s divine life and virtue was communicated by contact to the person touched and cured him of his ailment. [117]The wearing of amulets where these consist of parts of the bodies of animals is based on the same belief. When a man wears on his person the claws of a tiger in an amulet, he thinks that the claws being the tiger’s principal weapon of offence contain a concentrated part of his strength, and that the wearer of the claws will acquire some of this by contact. The Gonds carry the shoulder-bone of a tiger, or eat the powdered bone-dust, in order to acquire strength. The same train of reasoning applies to the wearing of the hair of a bear, a common amulet in India, the hair being often considered as the special seat of strength.124 The whole practice of wearing ornaments of the precious metals and precious stones appears to have been originally due to the same motive, as shown in the article on Sunār.

If the Gonds want a child to become fat, they put it in a pigsty or a place where asses have rolled, so that it may acquire by contact the quality of fatness belonging to the pigs or asses. If they wish to breed quarrels in an enemy’s house, they put the seeds of the amaltās or the quills of the porcupine in the thatch of the roof. The seeds in the dried pods of this tree rattle in the wind, while the fretful porcupine raises its quills when angry. Hence the seeds will impart the quality of noise to the house, so that its inmates will be noisy, while the quills of the porcupine will similarly breed strife between them. The effects produced by weapons and instruments are thought of in the same manner. We say that an arrow is shot from a bow with such force as to penetrate the body and cause a wound. The savage could not think or speak in this way, because he had no verbs and could not think of nouns in the objective case. He thought of the arrow as an animate thing having a cutting or piercing quality. When placed in a suitable position to exercise its powers, it flew, of its own volition, through the air to the target, and communicated to it by contact some of the above quality. The idea is more easily realised in the case of balls, pieces of bone or other missiles thrown by magicians. Here the person whom it is intended to injure may be miles away, so that the object could not possibly strike him merely through the force imparted to [118]it by the thrower. But when the magician has said charms over the missile, communicating to it the power and desire to do his will, he throws it in the proper direction and savages believe that it will go of its own accord to the person against whom it is aimed and penetrate his body. To pretend to suck pieces of bone out of the body, which are supposed to have been propelled into the victim by an enemy, is one of the commonest magical methods of curing an illness. The following instances of this idea are taken from the admirable collection in The Golden Bough125: “(In Suffolk) if a man cuts himself with a bill-hook or a scythe he always takes care to keep the weapon bright, and oils it to prevent the wound from festering. If he runs a thorn or, as he calls it, a bush into his hand, he oils or greases the extracted thorn. A man came to a doctor with an inflamed hand, having run a thorn into it while he was hedging. On being told that the hand was festering, he remarked: ‘That didn’t ought to, for I greased the bush well after I pulled it out’ If a horse wounds its foot by treading on a nail, a Suffolk groom will invariably preserve the nail, clean it and grease it every day to prevent the wound from festering.” Here the heat and festering of the wounds are held to be qualities of the axe, thorn or nail, which have been communicated to the person or animal wounded by contact. If these qualities of the instrument are reduced by cleaning and oiling it, then that portion of them communicated to the wound, which was originally held to be a severed part of the life and qualities of the instrument, will similarly be made cool and easy. It is not probable that the people of Suffolk really believe this at present, but they retain the method of treatment arising from the belief without being able to explain it. Similarly the Hindus must have thought that the results produced by the tools of artisans working on materials, and by the plough on the earth, were communicated by these instruments volitionally through contact; and this is why they worship once or twice a year the implements of their profession as the givers of the means of subsistence. All the stories of magic swords, axes, impenetrable shields, sandals, lamps, carpets and so on originally arose from the same belief. [119]

But primitive man not only considered the body as a homogeneous mass with the life and qualities distributed equally over it. He further, it may be suggested, did not distinguish between the individual and the species. The reason for this was that he could not count, and had no idea of numbers. The faculty of counting appears to have been acquired very late. Messrs. Spencer and Gillan remark of the aborigines of Central Australia:126 “While in matters such as tracking, which are concerned with their everyday life, and upon efficiency in which they actually depend for their livelihood, the natives show conspicuous ability, there are other directions in which they are as conspicuously deficient. This is perhaps shown most clearly in the matter of counting. At Alice Springs they occasionally count, sometimes using their fingers in doing so, up to five, but frequently anything beyond four is indicated by the word oknira, meaning ‘much’ or ‘great.’ One is nintha, two thrama or thera, three mapitcha, four therankathera, five therankathera-nintha.” The form of these words is interesting, because it is clear that the word for four is two and two, or twice two, and the word for five is two and two and one. These words indicate the prolonged and painful efforts which must have been necessary to count as far as five, and this though in other respects the Australian natives show substantial mental development, having a most complicated system of exogamy, and sometimes two personal names for each individual. Again, the Andamanese islanders, despite the extraordinary complexity of their agglutinative language, have no names for the numerals beyond two.127 It is said that the Majhwār tribe can only count up to three, while among the Bhatras the qualification for being a village astrologer, who foretells the character of the rainfall and gives auspicious days for sowing and harvest, is the ability to count a certain number of posts. The astrologer’s title is Meda Gantia, or Counter of Posts. The above facts demonstrate that counting is a faculty acquired with difficulty after considerable mental progress, and primitive man apparently did not feel the necessity for [120]it.128 But if he could not count, it seems a proper deduction that his eye would not distinguish a number of animals of the same species together, because the ability to do this, and to appraise distinct individuals of like appearance appears to depend ultimately on the faculty of counting. Major Hendley, a doctor and therefore a skilled observer, states that the Bhīls were unable to distinguish colours or to count numbers, apparently on account of their want of words to express themselves.129 Now it seems clearly more easy for the eye to discriminate between opposing colours than to distinguish a number of individuals of the same species together. There are a few things which we still cannot count, such as the blades of grass, the ears of corn, drops of rain, snowflakes, and hailstones. All of these things are still spoken of in the singular, though this is well known to be scientifically incorrect. We say an expanse of grass, a field of corn, and so on, as if the grass and corn were all one plant instead of an innumerable quantity of plants. Apparently when primitive man saw a number of animals or trees of the same species together, the effect on him must have been exactly the same as that of a field of grass or corn on us. He could be conscious only of an indefinite sense of magnitude. But he did not know, as we do in the cases cited, that the objects he saw were really a collection of distinct individuals. He would naturally consider them as all one, just as children would think a field of grass or corn to be one great plant until they were told otherwise. But there was no one to tell him, nor any means by which he could find out his mistake. He had no plural number, and no definite or indefinite articles. Whether he saw one or a hundred tigers together, he could only describe them by the one word tiger. It was a long time before he could even say ‘much tiger,’ as the Australian natives still have to do if they see more animals than five together, and the Andamanese if they see more than two. The hypothesis therefore seems reasonable that at first man considered each species of animals or plants which he distinguished to have a separate single life, [121]of which all the individuals were pieces or members. The separation of different parts of one living body presented no difficulties to his mind, since, as already seen, he believed the life to continue in severed fractions of the human body.

A connection between individuals, apparently based on the idea that they have a common life, has been noticed in other cases. Thus at the commencement of the patriarchal state of society, when the child is believed to derive its life from its father, any carelessness in the father’s conduct may injuriously affect the child. Sir E.B. Tylor notes this among the tribes of South America. After the birth of a child among the Indians of South America the father would eat no regular cooked food, not suitable for children, as he feared that if he did this his child would die.130 “Among the Arawaks of Surinam for some time after the birth of a child the father must fell no tree, fire no gun, hunt no large game; he may stay near home, shoot little birds with a bow and arrow, and angle for little fish; but his time hanging heavy on his hands the only comfortable thing he can do is to lounge in his hammock.”131 On another occasion a savage who had lately become a father, refused snuff, of which he was very fond, because his sneezing would endanger the life of his newly-born child. They believed that any intemperance or carelessness of the father, such as drinking, eating large quantities of meat, swimming in cold weather, riding till he was tired and sweated, would endanger the child’s life, and if the child died, the father was bitterly reproached with having caused its death by some such indiscretion.132 Here the idea clearly seems to be that the father’s and child’s life are one, the latter being derived from and part of the former. The custom of the Couvade may therefore perhaps be assigned to the early patriarchal stage. The first belief was that the child derived its life from its mother, and apparently that the weakness and debility of the mother after childbirth were due to the fact that she had given up a part of her life to the child. When the system of female descent changed to male descent, the woman was taken from another clan into her husband’s; the child, being born in its father’s clan, [122]obviously could not draw its life from its mother, who was originally of a different clan. The inference was that it drew its life from its father; consequently the father, having parted with a part of his life to his child, had to imitate the conduct of the mother after childbirth, abstain from any violent exertion, and sometimes feign weakness and lie up in the house, so as not to place any undue strain on the severed fraction of his life in his child, which would be simultaneously affected with his own, but was much more fragile.

Again, primitive man had no conception of likeness or similarity, nor did he realise an imitation as distinct from the thing imitated. Likeness or similarity and imitation are abstract ideas, for which he had no words, and consequently did not conceive of them. And clearly if one had absolutely no term signifying likeness or similarity, and if one wished to indicate say, that something resembled a goat, all one could do would be to point at the goat and the object resembling it and say ‘goat,’ ‘goat.’ Since the name was held to be part of the thing named, such a method would strengthen the idea that resemblance was equivalent to identity. This point of view can also be observed in children, who have no difficulty in thinking that any imitation or toy model is just as good as the object or animal imitated, and playing with it as such. Even to call a thing by the name of any object is sufficient with children to establish its identity with that object for the purposes of a game or mimicry, and a large part of children’s games are based on such pretensions. They also have not yet clearly grasped the difference between likeness and identity, and between an imitation of an object and the object itself. A large part of the category of substituted ceremonies and sacrifices are based on this confusion between similarity and identity. Thus when the Hindus put four pieces of stick into a pumpkin and call it a goat, they do not mean to cheat the god to whom it is offered, but fancy that when they have made a likeness of a goat and called it a goat, it is a goat, at any rate for the purpose of sacrifice. And when the Jains, desiring to eat after sunset against the rule of their religion, place a lamp under a sieve and call it the [123]sun, and eat by it, they are acting on the same principle and think they have avoided committing a sin. A Baiga should go to his wedding on an elephant, but as he cannot obtain a real elephant, two wooden cots are lashed together and covered with blankets, with a black cloth trunk in front, and this arrangement passes muster for an elephant. A small gold image of a cat is offered to a Brāhman in expiation for killing a cat, silver eyes are offered to the goddess to save the eyes of a person suffering from smallpox, a wisp of straw is burnt on a man’s grave as a substitute for cremating the body, a girl is married to an image of a man made of kusha grass, and so on. In rites where blood is required vermilion is used as a substitute for blood; on the other hand castes which abstain from flesh sometimes also decline to eat red vegetables and fruits, because the red colour is held to make them resemble and be equivalent to blood. These beliefs survive in religious ceremonial long after the hard logic of facts has dispelled them from ordinary life.133 Thus when an image of a god was made it was at once the god and contained part of his life. Primitive man had no idea of an imitation or an image nor of a lifeless object, and therefore could not conceive of the representation being anything else than the god. Only in later times was some ceremony of conveying life to the image considered requisite. The prohibition of sculpture among the Jews and of painting among the Muhammadans was based on this view,134 because sculptures and paintings were not considered as images or representations, but as living beings or gods, and consequently false gods. The world-wide custom of making an image of a man with intent to injure him arises from the same belief. Since primitive man could conceive neither of an imitation nor of an inanimate object, the image of a man was to his view the man; there was nothing else which it could be. And thus it contained part of the man’s life, just as every idol of a god was the god himself and contained part of the god’s life. Since the man’s life was common to himself and the image, by injuring or destroying the image it was held [124]that the man’s life would similarly be injured or destroyed, on the analogy already explained of injury to life being frequently observed to follow a hurt or wound of any part of the body. Afterwards the connection between the man and the image was strengthened by working into the material of the latter some fraction of his body, such as severed hair or the earth pressed by his foot. But this was not necessary to the original belief. The objection often raised by savages to having their photographs taken or pictures painted may be explained in the same manner. Here the photograph or picture cannot be realised as a simple imitation; it is held to be the man himself, and must therefore contain part of his life. Hence any one in whose possession it is can do him harm by injuring or destroying the photograph or picture, according to the method of reasoning already explained. The superstitions against looking in a mirror, especially after dark, or seeing one’s reflection in water, are analogous cases. Here the reflection in the mirror or water is held to be the person himself, because savages do not understand the nature of the reflected image. It is the person himself, but has no corporeal substance; therefore the reflection must be his ghost or spirit. But if the spirit appears once it is an omen that it will appear again; and in order that it may do so the man will have to die so that the spirit may be set free from the body in order to appear. The special reason for not looking into a mirror at night would thus be because the night is the usual time for the appearance of spirits. The fable of Narcissus, who fell in love with his own image reflected in the water and was drowned, probably arose from the superstition against seeing one’s image reflected in water. And similarly the belief was that a man’s clothes and other possessions contained part of his life by contact; this is the explanation of the custom of representing a person by some implement or article of clothing, such as performing the marriage ceremony with the bridegroom’s sword instead of himself, and sending the bride’s shoes home with the bridegroom to represent her. A barren woman will try to obtain a piece of a pregnant woman’s breast-cloth and will burn it and eat the ashes, thinking thereby to transfer the pregnant [125]woman’s quality of fertility to herself. When a Hindu widow is remarried her clothes and ornaments are sometimes buried on the boundary of her second husband’s village and she puts on new clothes, because it is thought that her first husband’s spirit will remain in the old clothes and give trouble.

A brief digression may be made here in order to suggest an explanation of another important class of primitive ideas. These arise from the belief that when something has happened, that same event, or some other resembling it, will again occur, or, more briefly, the belief in the recurrence of events. This view is the origin of a large class of omens, and appears to have been originally evolved simply from the recurring phenomena of day and night and of the months and climatic seasons. For suppose that one was in the position of primitive man, knowing absolutely nothing of the nature and constitution of the earth and the heavenly bodies, or of the most elementary facts of astronomy; then, if the question were asked why one expected the sun to rise to-morrow, the only possible answer, and the answer which one would give, would be because it had risen to-day and every day as long as one could remember. The reason so stated might have no scientific value, but would at any rate establish a strong general probability. But primitive man could not have given it in this form, because he had no memory and could not count. Even now comparatively advanced tribes like the Gonds have a hopelessly inaccurate memory for ordinary incidents; and, as suggested subsequently, the faculty of memory was probably acquired very slowly with the development of language. And since he could not count, the continuous recurrence of natural phenomena had no cumulative force with him, so that he might distinguish them from other events. His argument was thus simply “the sun will rise again because it rose before; the moon will wax and wane again because she waxed and waned before”; grass and leaves and fruit would grow again because they did so before; the animals which gave him food would come again as before; and so on. But these were the only events which his brain retained at all, and [126]that only because his existence depended upon them and they continually recurred. The ordinary incidents of life which presented some variation passed without record in his mind, as they still do very largely in those of primitive savages. And since he made no distinction between the different classes of events, holding them all to be the acts of volitional beings, he applied this law of the recurrence of events to every incident of life, and thought that whenever anything happened, reason existed for supposing that the same thing or something like it would happen again. It was sufficient that the second event should be like the first, since, as already seen, he did not distinguish between similarity and identity. Thus, to give instances, the Hindus think that if a man lies full length inside a bed, he is lying as if on a bier and will consequently soon be dead on a real bier; hence beds should be made so that one’s feet project uncomfortably over the end. By a similar reasoning he must not lie with his feet to the south because corpses are laid in this direction. A Hindu married woman always wears glass bangles as a sign of her state, and a widow may not wear them. A married woman must therefore never let her arms be without bangles or it is an omen that she will become a widow. She must not wear wholly white clothes, because a widow wears these. If a man places one of his shoes over the other in the house, it is an omen that he will go on a journey when the shoes will be in a similar position as he walks along. A Kolta woman who desires to ascertain whether she will have a son, puts a fish into a pot full of water and spreads her cloth by it. If the fish jumps into her lap, it is thought that her lap will shortly hold another living being, that is a son. At a wedding, in many Hindu castes, the bride and bridegroom perform the business of their caste or an imitation of it. Among the Kuramwār shepherds the bride and bridegroom are seated with the shuttle which is used for weaving blankets between them. A miniature swing is put up and a doll is placed in it in imitation of a child and swung to and fro. The bride then takes the doll out and gives it to the bridegroom, saying:—“Here, take care of it, I am now going to cook food”; while, after a time, the boy returns the doll to the [127]girl saying, “I must now weave the blanket and go to tend the flock.” Thus, having performed their life’s business at their wedding, it is thought that they will continue to do so happily as long as they live. Many castes, before sowing the real crop, make a pretence of sowing seed before the shrine of the god, and hope thus to ensure that the subsequent sowing will be auspicious. The common stories of the appearance of a ghost, or other variety of apparition, before the deaths of members of a particular family, are based partly on the belief in the recurrence of associated events. The well-known superstition about sitting down thirteen to dinner, on the ground that one of the party may die shortly afterwards, is an instance of the same belief, being of course based on the Last Supper. But the number thirteen is generally unlucky, being held to be so by the Hindus, Muhammadans and Persians, as well as Europeans, and the superstition perhaps arose from its being the number of the intercalary month in the soli-lunar calendar, which is present one year and absent the next year. Thirteen is one more than twelve, the auspicious number of the months of the year. Similarly seven was perhaps lucky or sacred as being the number of the planets which gave their names to the days of the week, and three because it represented the sun, moon and earth. When a gambler stakes his money on a number such as the date of his birth or marriage, he acts on the supposition that a number which has been propitious to him once will be so again, and this appears to be a survival of the belief in the recurrence of events.

But primitive man was not actuated by any abstract love of knowledge, and when he had observed what appeared to him to be a law of nature, he proceeded to turn it to advantage in his efforts for the preservation of his life. Since events had the characteristic of recurrence, all he had to do in order to produce the recurrence of any particular event which he desired, was to cause it to happen in the first instance; and since he did not distinguish between imitation and reality, he thought that if he simply enacted the event he would thus ensure its being brought to pass. And so he assiduously set himself to influence the course of nature to [128]his own advantage. When the Australian aborigines are performing ceremonies for the increase of witchetty grubs, a long narrow structure of boughs is made which represents the chrysalis of the grub. The men of the witchetty grub totem enter the structure and sing songs about the production and growth of the witchetty grub. Then one after another they shuffle out of the chrysalis, and glide slowly along for a distance of some yards, imitating the emergence and movements of the witchetty grubs. By thus enacting the production of the grubs they think to cause and multiply the real production.135 When the men of the emu totem wish to multiply the number of emus, they allow blood from their arms, that is emu blood, to fall on the ground until a certain space is covered. Then on this space a picture is drawn representing the emu; two large patches of yellow indicate lumps of its fat, of which the natives are very fond, but the greater part shows, by means of circles and circular patches, the eggs in various stages of development, some before and some after laying. Then the men of the totem, placing on their heads a stick with a tuft of feathers to represent the long neck and small head of the bird, stand gazing about aimlessly after the manner of the emu. Here the picture itself is held to be a living emu, perhaps the source or centre from which all emus will originate, and the men, pretending to be emus, will cause numbers of actual emus to be produced.136 Before sowing the crops, a common practice is to sow small quantities of grain in baskets or pots in rich soil, so that it will sprout and grow up quickly, the idea being to ensure that the real crop will have a similarly successful growth. These baskets are the well-known Gardens of Adonis fully described in The Golden Bough. They are grown for nine days, and on the tenth day are taken in procession by the women and deposited in a river. The women may be seen carrying the baskets of wheat to the river after the nine days’ fasts of Chait and Kunwār (March and September) in many towns of the Central Provinces, as the Athenian women carried the Gardens of Adonis to the sea on the day that the expedition under Nicias set sail for Syracuse.137 The fire [129]kindled at the Holi festival in spring is meant, as explained by Sir J.G. Frazer, to increase the power of the sun for the growth of vegetation. By the production of fire the quantity and strength of the heavenly fire is increased. He remarks:138—“The custom of throwing blazing discs, shaped like suns, into the air, is probably also a piece of imitative magic. In these, as in so many cases, the magic force is supposed to take effect through mimicry or sympathy; by imitating the desired result you actually produce it; by counterfeiting the sun’s progress through the heavens you really help the luminary to pursue his celestial journey with punctuality and despatch. The name ‘fire of heaven,’ by which the midsummer fire is sometimes popularly known, clearly indicates a consciousness of the connection between the earthly and the heavenly flame.” The obscene songs of the Holi appear to be the relic of a former period of promiscuous sexual debauchery, which, through the multiplied act of reproduction, was intended to ensure that nature should also reproduce on a generous scale. The red powder thrown over everybody at the Holi is said to represent the seed of life. The gifts of Easter eggs seem to be the vestige of a rite having the same object. At a wedding in the Lodhi caste the bride is seated before the family god while an old woman brings a stone rolling-pin wrapped up in a piece of cloth, which is supposed to be a baby, and the old woman imitates a baby crying. She puts the roller in the bride’s lap, saying, “Take this and give it milk.” The bride is abashed and throws it aside. The old woman picks it up and shows it to the assembled women, saying, “The bride has just had a baby,” amid loud laughter. Then she gives the stone to the bridegroom, who also throws it aside. This ceremony is meant to induce fertility, and it is supposed that by making believe that the bride has had a baby she will quickly have one. Similar rites are performed in several other castes, and when a girl becomes adult her lap is filled with fruits with the idea that this will cause it subsequently to be filled with the fruit of her womb. The whole custom of giving dolls to girls to play with, perhaps originated in the belief that by doing so they would afterwards come to play with children. [130]

The dances of the Kol tribe consist partly of symbolical enactments of events which they desired to be successfully accomplished. Some variations of the dance, Colonel Dalton states, represent the different seasons and the necessary acts of cultivation that each brings with it. In one the dancers, bending down, make a motion with their hands, as though they were sowing the grain, keeping step with their feet all the time. Then comes the reaping of the crop and the binding of the sheaves, all done in perfect time and rhythm, and making, with the continuous droning of the voices, a quaint and picturesque performance.139 The Karma dance of the Gonds and Oraons is also connected with the crops, and probably was once an enactment of the work of cultivation.140 The Bhīls danced at their festivals and before battles. The men danced in a ring, holding sticks and striking them against one another. Before a battle they had a war-dance in which the performers were armed and imitated a combat. To be carried on the shoulders of one of the combatants was a great honour, perhaps because it symbolised being on horseback. The object was to obtain success in battle by going through an imitation of a successful battle beforehand. This was also the common custom of the Red Indians, whose war-dances are well known; they brandished their weapons and killed their foe in mimicry in order that they might soon do so in reality. The Sela dance of the Gonds and Baigas, in which they perform the figure of the grand chain of the lancers, only that they strike their sticks together instead of clasping hands as they pass, was probably once an imitation of a combat. It is still sometimes danced before their communal hunting and fishing parties. In these mimetic rehearsals of events with the object of causing them to occur we may perhaps discern the origin of the arts both of acting and dancing. Another, and perhaps later form, was the reproduction of important events, or those which had influenced history. For to the primitive mind, as already seen, the results were not conceived of as instrumentally caused by the event, but as part of the event itself and of [131]its life and personality. Hence by the re-enactment of the event the beneficial results would be again obtained or at least preserved in undiminished potency and vigour. This was perhaps the root idea of the drama and the representation of sacred or heroic episodes on the stage.

Thus, resuming from paragraph 61, primitive man had no difficulty in conceiving of a life as shared between two or more persons or objects, and it does not seem impossible that he should have at first conceived it to extend through a whole species.141 A good instance of the common life is afforded by the gods of the Hindu and other pantheons. Each god was conceived of as performing some divine function, guiding the chariot of the sun, manipulating the thunder and so on; but at the same time thousands of temples existed throughout the country, and in each of these the god was alive and present in his image or idol, able to act independently, receive and consume sacrifices and offerings, protect suppliants and punish transgressors. No doubt at all can be entertained that each idol was in itself held to be a living god. In India food is offered to the idol, it goes through its ablutions, is fanned, and so on, exactly like a human king. The ideas of sanctuary and sacrilege appear to depend primarily on the belief in the actual presence of the god in his shrine. And in India no sanctity at all attaches to a temple from which the idol has been removed. Thus we see the life of the god distributed over a multitude of personalities. Again, the same god, as Vishnu or the sun, is held to have had a number of incarnations, as the boar, the tortoise, a man-lion, a dwarf, Rāma and Krishna, and these are venerated simultaneously as distinct deities. The whole Brāhman caste considered itself divine or as partaking in the life of the god, the original reason for this perhaps being that the Brāhmans obtained the exclusive right to perform sacrifices, and hence the life of the sacrificial animal or food passed to them, as in other societies it passed to the king who performed the sacrifice. A Brāhman further holds that the five gods, Indra, Brahma, Siva, Vishnu and Ganesh, are present in different parts of his body,142 and here again the [132]life of the god is seen to be divided into innumerable fragments. The priests of the Vallabhachārya sect, the Gokulastha Gosains, were all held to be possessed by the god Krishna, so that it was esteemed a high privilege to perform the most menial offices for them, because to touch them was equivalent to touching the god, and perhaps assimilating by contact a fragment of his divine life and nature.143 The belief in a common life would also explain the veneration of domestic animals and the prohibition against killing them, because to kill one would injure the whole life of the species, from which the tribe drew its subsistence. Similarly in a number of cases the first idea of seasonal fasts is that the people abstain from the grain or fruit which is growing or sown in the ground. Thus in India during the rains the vegetables growing at this period are not eaten, and are again partaken of for the first time after the sacrificial offering of the new crop. This rule could not possibly be observed in the case of grain, but instead certain single fast-days are prescribed, and on these days no cultivated grain or fruit, but only those growing wild, should be eaten. These rules seem to indicate that the original motive of the fast was to avoid injuring the common life of the grain or fruit, which injury would be caused by a consumption of any part of it, at a time when the whole of the common life and vigour was required for its reproduction and multiplication. This idea may have operated to enable the savage to restrain himself from digging up and eating the grain sown in the ground, or slaughtering his domestic animals for food, and a taboo on the consumption of grain and fruits during their period of ripening may have first begun in their wild state. The Intichiuma ceremonies of the Australian natives are carried out with the object of increasing the supply of the totem for food purposes. In the Ilpirla or Manna totem the members of the clan go to a large boulder surrounded by stones, which are held to represent masses of Ilpirla or the manna of the mulga tree. A Churinga stone is dug up, which is supposed to represent another mass of manna, and this is rubbed over the boulder, and the smaller stones are also rubbed over it. While the [133]leader does this, the others sing a song which is an invitation to the dust produced by the rubbing of the stones to go out and produce a plentiful supply of Ilpirla on the mulga trees.144 Then the dust is swept off the surface of the stones with twigs of the mulga tree. Here apparently the large boulder and other stones are held to be the centre or focus of the common life of the manna, and from them the seed issues forth which will produce a crop of manna on all the mulga trees. The deduction seems clear that the trees are not conceived of individually, but are held to have a common life. In the case of the hakea flower totem they go to a stone lying beneath an old tree, and one of the members lets his blood flow on to the stone until it is covered, while the others sing a song inciting the hakea tree to flower much and to the blossoms to be full of honey.145 The blood is said to represent a drink prepared from the hakea flowers, but probably it was originally meant to quicken the stone with the blood of a member of the totem, that is its own blood or life, in order that it might produce abundance of flowers. Here again the stone seems to be the centre of the common life of the hakea flower. The songs are sung with the idea that the repetition of words connoting a state of facts will have the effect of causing that state of facts to exist, in accordance with the belief already explained in the concrete virtue of words.

Sir E. B. Tylor states: “In Polynesia, if a village god were accustomed to appear as an owl, and one of his votaries found a dead owl by the roadside, he would mourn over the sacred bird and bury it with much ceremony, but the god himself would not be thought to be dead, for he remains incarnate in all existing owls. According to Father Geronimo Boscana, the Acagchemen tribe of Upper California furnish a curious parallel to this notion. They worshipped the panes bird, which seems to have been an eagle or vulture, and each year, in the temple of each village, one of them was solemnly killed without shedding blood, and the body buried. Yet the natives maintained and believed that it was the same individual bird they sacrificed each year, and more than this, that the same bird was slain by each of the villages.”146 An [134]account of the North American Indians quoted by the same author states that they believe all the animals of each species to have an elder brother, who is as it were the principle and origin of all the individuals, and this elder brother is marvellously great and powerful. According to another view each species has its archetype in the land of souls; there exists, for example, a manitu or archetype of all oxen, which animates all oxen.147

Generally in the relations between the totem-clan and its totem-animal, and in all the fables about animals, one animal is taken as representing the species, and it is tacitly assumed that all the animals of the species have the same knowledge and qualities and would behave in the same manner as the typical one. Thus when the Majhwār says that the tiger would run away if he met a member of the tiger-clan who was free from sin, but would devour any member who had been put out of caste for an offence, he assumes that every tiger would know a member of the clan on meeting him, and also whether that member was in or out of caste. He therefore apparently supposes a common knowledge and intelligence to exist in all tigers as regards the clan, as if they were parts of one mind or intelligence. And since the tigers know instinctively when a member of the clan is out of caste, the mind and intelligence of the tigers must be the same as that of the clan. The Kols of the tiger clan think that if they were to sit up for a tiger over a kill the tiger would not come and would be deprived of his food, and that they themselves would fall ill. Here the evil effects of the want of food on one tiger are apparently held to extend to all tigers and also to all members of the tiger clan.

The totem-clan held itself to partake of the life of its totem, and on the above hypothesis one common life would flow through all the animals and plants of the totem and all the members of the clan. An Australian calls his totem his Wingong (friend) or Tumang (flesh), and nowadays expresses his sorrow when he has to eat it.148 If a man wishes to injure any man of a certain totem, he kills any [135]animal of that man’s totem.149 This clearly shows that one common life is held to bind together all the animals of the totem-species and all the members of the totem-clan, and the belief seems to be inexplicable on any other hypothesis. The same is the case with the sex-totems of the Kurnai tribe. In addition to the clan-totems all the boys have the Superb Warbler bird as a sex-totem, and call it their elder brother; and all the girls the Emu-wren, and call it their elder sister. If the boys wish to annoy the girls, or vice versa, each kills or injures the other’s totem-bird, and such an act is always followed by a free fight between the boys and girls.150 Sex-totems are a peculiar development which need not be discussed here, but again it would appear that a common life runs through the birds of the totem and the members of the sex. Professor Robertson Smith describes the clan or kin as follows: “A kin was a group of persons whose lives were so bound up together, in what must be called a physical unity, that they could be treated as parts of one common life. The members of one kindred looked on themselves as one living whole, one single animated mass of blood, flesh and bones, of which no member could be touched without all the members suffering. This point of view is expressed in the Semitic tongue in many familiar forms of speech. In case of homicide Arabian tribesmen do not say, ‘The blood of M. or N. has been spilt’ (naming the man): they say, ‘Our blood has been spilt.’ In Hebrew the phrase by which one claims kinship is, ‘I am your bone and your flesh.’ Both in Hebrew and in Arabic flesh is synonymous with ‘clan’ or kindred group.”151 The custom of the blood-feud appears to have arisen from the belief in a common life of the clan. “The blood-feud is an institution not peculiar to tribes reckoning descent through females; and it is still in force. By virtue of its requirements every member of a kin, one of whom had suffered at the hands of a member of another kin, was bound to avenge the wrong upon the latter kin. Such is the solidarity between members of a kin that vengeance might be taken upon any member [136]of the offending kin, though he might be personally quite innocent. In the growth of civilisation vengeance has gradually come to be concentrated upon the offender only.”152 Thus the blood-feud appears to have originated from the idea of primary retributive justice between clan and clan. When a member of a clan had been killed, one of the offending clan must be killed in return. Who he might be, and whether the original homicide was justifiable or not, were questions not regarded by primitive man; motives were abstract ideas with which he had no concern; he only knew that a piece of the common life had been lopped off, and the instinct of self-preservation of the clan demanded that a piece of the life of the offending clan should be cut off in return. And the tie which united the kin was eating and drinking together. “According to antique ideas those who eat and drink together are by this very act tied to one another by a bond of friendship and mutual obligation.”153 This was the bond which first united the members of the totem-clan both among themselves and with their totem. And the relationship with the totem could only have arisen from the fact that they ate it. The belief in a common life could not possibly arise in the totem-clan towards any animal or plant which they did not eat or otherwise use. These they would simply disregard. Nor would savages, destitute at first of any moral ideas, and frequently on the brink of starvation, abstain from eating any edible animal from sentimental considerations; and, as already seen, the first totems were generally edible. They could not either have in the first place eaten the totem ceremonially, as there would be no reason for such a custom. But the ceremonial eating of the domestic animal, which was the tie subsequently uniting the members of the tribe,154 cannot be satisfactorily explained except on the hypothesis that it was evolved from the customary eating of the totem-animal. Primitive savages would only feel affection towards the animals which they ate, just as the affection of animals is gained by feeding them. The objection might be made that savages could not feel [137]affection and kinship for an animal which they killed and ate, but no doubt exists that they do.

“In British Columbia, when the fishing season commenced and the fish began coming up the rivers, the Indians used to meet them and speak to them. They paid court to them and would address them thus: ‘You fish, you fish; you are all chiefs, you are; you are all chiefs.’ Among the Northas when a bear is killed, it is dressed in a bonnet, covered with fine down, and solemnly invited to the chiefs presence.”155 And there are many other instances.156 Savages had no clear realisation of death, and they did not think that the life of the animal was extinguished but that it passed to them with the flesh. Moreover they only ate part of the life. In many cases also the totem-animal only appeared at a certain season of the year, in consequence of the habit of hibernation or migration in search of food, while trees only bore fruit in their season. The savage, regarding all animals and plants as possessed of self-conscious life and volition, would think that they came of their own accord to give him subsistence or life. Afterwards, when they had obtained the idea of a soul or spirit, and of the survival of the soul after death, and when, on the introduction of personal names, the personality of individuals could be realised and remembered after death, they frequently thought that the spirits of ancestors went back to the totem-animal, whence they derived their life. The idea of descent from the totem would thus naturally arise. As the means of subsistence increased, and especially in those communities which had domesticated animals or cultivated plants, the conception of the totem as the chief source of life would gradually die away and be replaced by the belief in descent from it; and when they also thought that the spirits of ancestors were in the totem, they would naturally abstain from eating it. Perhaps also the Australians consider that the members of the totem-clan should abstain from eating the totem for fear of injuring the common life, as more advanced communities abstained from eating the flesh of domestic [138]animals. This may be the ground for the rule that they should only eat sparingly of the totem. To the later period may be ascribed the adoption of carnivorous animals as totems; when these animals came to be feared and also venerated for their qualities of strength, ferocity and courage, warriors would naturally wish to claim kinship with and descent from them.

When the members of the totem-clan who lived together recognised that they owed something to each other, and that the gratification of the instincts and passions of the individual must to a certain degree be restrained if they endangered the lives and security of other members of the clan, they had taken the first step on the long path of moral and social progress. The tie by which they supposed themselves to be united was quite different from those which have constituted a bond of union between the communities who have subsequently lived together in the tribe, the city-state and the country. These have been a common religion, common language, race, or loyalty to a common sovereign; but the real bond has throughout been the common good or the public interest. And the desire for this end on the part of the majority of the members of the community, or the majority of those who were able to express their opinions, though its action was until recently not overt nor direct, and was not recognised, has led to the gradual evolution of the whole fabric of law and moral feeling, in order to govern and control the behaviour and conduct of the individual in his relations with his family, neighbours and fellow-citizens for the public advantage. The members of the totem-clan would have been quite unable to understand either the motives by which they were themselves actuated or the abstract ideas which have united more advanced communities; but they devised an even stronger bond than these, in supposing that they were parts or fractions of one common body or life. This was the more necessary as their natural impulses were uncontrolled by moral feeling. They conceived the bond of union in the concrete form of eating together. As language improved and passing events were recorded in speech and in the mind, the faculty of memory was perhaps concurrently developed. Then man began to realise the [139]insecurity of his life, the dangers and misfortunes to which he was subject, the periodical failure or irregularity of the supply of food, and the imminent risks of death. Memory of the past made him apprehensive for the future, and holding that every event was the result of an act of volition, he began to assume an attitude either of veneration, gratitude, or fear towards the strongest of the beings by whom he thought his destinies were controlled—the sun, moon, sky, wind and rain, the ocean and great rivers, high mountains and trees, and the most important animals of his environment, whether they destroyed or assisted to preserve his life. The ideas of propitiation, atonement and purification were then imparted to the sacrifice, and it became an offering to a god.157 But the primary idea of eating or drinking together as a bond of union was preserved, and can be recognised in religious and social custom to an advanced period of civilisation.

Again, Dr. Westermarck shows that the practice of exogamy or the avoidance of intermarriage did not at first arise between persons recognised as blood relations, but between those who lived together. “Facts show that the extent to which relatives are not allowed to intermarry is nearly connected with their close living together. Generally speaking the prohibited degrees are extended much further among savage and barbarous peoples than in civilised societies. As a rule the former, if they have not remained in the most primitive social condition of man, live not in separate families but in large households or communities, all the members of which dwell in very close contact with each other.”158 And later, after adducing the evil results of self-fertilisation in plants and close interbreeding in animals, Dr. Westermarck continues: “Taking all these facts into consideration, I cannot but believe that consanguineous marriages, in some way or other, are more or less detrimental to the species. And here I think we may find a quite sufficient explanation of the horror of incest; not because man at an early stage recognised the injurious influence of close intermarriage, but because the law of natural selection must [140]inevitably have operated. Among the ancestors of man, as among other animals, there was no doubt a time when blood relationship was no bar to sexual intercourse. But variations here, as elsewhere, would naturally present themselves; and those of our ancestors who avoided in-and-in breeding would survive, while the others would gradually decay and ultimately perish. Thus an instinct would be developed, which would be powerful enough as a rule to prevent injurious unions. Of course it would display itself simply as an aversion on the part of individuals to union with others with whom they lived; but these as a matter of fact would be blood relations, so that the result would be the survival of the fittest.”

The instinct of exogamy first developed in the totem-clan when it was migratory and lived by hunting, at least among the Australians and probably the American Indians.

The first condition of the clan was one of sexual promiscuity, and in Totemism and Exogamy Sir J.G. Frazer has adduced many instances of periodical promiscuous debauchery which probably recall this state of things.159 The evil results which would accrue from in-breeding in the condition of promiscuity may have been modified by such incidents as the expulsion of the young males through the spasmodic jealousy of the older ones, the voluntary segregation of the old males, fights and quarrels leading to the rearrangement of groups, and the frequent partial destruction of a group, when the survivors might attach themselves to a new group. Primitive peoples attached the utmost importance to the rule of exogamy, and the punishments for the breach of it were generally more severe than those for the violation of the laws of affinity in civilised countries. The Australians say that the good spirit or the wise men prescribed to them the rule that the members of each totem-clan should not marry with each other.160 Similarly the Gonds say that their divine hero, Lingo, introduced the rule of exogamy and the division into clans before he went to the gods.

At first, however, the exogamous clan was not constituted by descent through males, but through females. [141]The hypothesis that female everywhere preceded male descent is strongly supported by natural probability. In the first instance, the parentage of children was no more observed and remembered than that of animals. When first observed, it was necessarily through the mother, the identity of the father being wholly uncertain. The mother would also be the first parent to remember her children, her affection for them being based on one of the strongest natural instincts, whereas the father neither knew nor cared for his children until long afterwards. Sir J.G. Frazer has further shown that even now some of the Australian aborigines are ignorant of the physical fact of paternity and its relation to sexual intercourse. That such ignorance could have survived so long is the strongest evidence in favour of the universal priority of female to male descent. It is doubtful, however, whether even the mother could remember her children after they had become adult, prior to the introduction of personal names. Mr. M’Lennan states: “The tie between mother and child, which exists as a matter of necessity during infancy, is not infrequently found to be lost sight of among savages on the age of independence being reached.”161 Personal names were probably long subsequent to clan-names, and when they were first introduced the name usually had some reference to the clan. The Red Indians and other races have totem-names which are frequently some variant of the name of the totem.162 When personal names came to be generally introduced, the genesis of the individual family might soon follow, but the family could scarcely have come into existence in the absence of personal names. As a rule, in the exogamous clan with female descent no regard was paid to the chastity of women, and they could select their partners as they pleased. Mr. Hartland has shown in Primitive Paternity that in a large number of primitive communities the chastity of women was neither enforced nor desired by the men, this state of things being probably a [142]relic of the period of female descent. Thus exogamy first arose through the women of the clan resorting to men outside it. When we consider the extreme rigour of life and the frequent danger of starvation to which the small clans in the hunting stage must have been exposed, it does not seem impossible that the evil effects of marriage within the clan may have been noticed. At that time probably only a minority even of healthy children survived, and the slight congenital weakness produced by in-breeding might apparently be fatal to a child’s chance of life. Possibly some dim perception may have been obtained of the different fates of the children of women who restricted their sexual relations to men within the clan and those who resorted to strangers, even though the nature of paternity may not have been understood. The strength of the feeling and custom of exogamy seems to demand some such recognition for its satisfactory explanation, though, on the other hand, the lateness of the recognition of the father’s share in the production of children militates against this view. The suggestion may be made also that the belief that the new life of a child must be produced by a spirit entering the woman, or other extraneous source, does not necessarily involve an ignorance of the physical fact of paternity; the view that the spirits of ancestors are reborn in children is still firmly held by tribes who have long been wholly familiar with the results of the commerce of the sexes. The practice of exogamy was no doubt, as shown by Dr. Westermarck, favoured and supported by the influence of novelty in sexual attraction, since according to common observation and experience sexual love or desire is more easily excited between strangers or slight acquaintances than between those who have long lived together in the same household or in familiar intercourse. In the latter case the attraction is dulled by custom and familiarity.

The exogamous clan, with female descent, was, however, an unstable social institution, in that it had no regular provision for marriage nor for the incorporation of married couples. The men who associated with the women of the clan were not necessarily, nor as a rule, admitted to it, but [143]remained in their own clans. How this association took place is not altogether clear. At a comparatively late period in Arabia, according to Professor Robertson Smith,163 the woman would have a tent, and could entertain outside men for a shorter or longer period according to her inclination. The practice of serving for a wife also perhaps dates from the period of female descent. The arrangement would have been that a man went and lived with a woman’s family and gave his services in return for her conjugal society. Whether the residence with the wife’s family was permanent or not is perhaps uncertain. When Jacob served for Leah and Rachel, society seems to have been in the early patriarchal stage, as Laban was their father and he was Laban’s sister’s son. But it seems doubtful whether his right was then recognised to take his wives away with him, for even after he had served fourteen years Laban pursued him, and would have taken them back if he had not been warned against doing so in a vision. The episode of Rachel’s theft of the images also seems to indicate that she intended to take her own household gods with her and not to adopt those of her husband’s house. And Laban’s chief anxiety was for the recovery of the images. A relic of the husband’s residence with his wife’s family during the period of female descent may perhaps be found in the Banjāra caste, who oblige a man to go and live with his wife’s father for a month without seeing her face. Under the patriarchal system this rule of the Banjāras is meaningless, though the general practice of serving for a wife survives as a method of purchase.

Among the Australian aborigines apparently the clans, or sections of them, wander about in search of food and game, and meet each other for more or less promiscuous intercourse. This may perhaps be supposed to have been the general primitive condition of society after the introduction of exogamy combined with female descent. And its memory is possibly preserved in the tradition of the Golden Age, golden only in the sense that man was not troubled either by memory or anticipation, and lived only for the day. The entire insecurity of life and its frequent end by [144]starvation or a violent death did not therefore trouble him any more than is the case with animals. He took no thought for the morrow, nor did the ills of yesterday oppress his mind. As when one of a herd of deer is shot by a hunter and the others stand by it pityingly as it lies dying on the ground, uncertain of its mishap, though they would help it if they could; yet when they perceive the hunter they make quickly off and in a few minutes are again grazing happily a mile or two away: little or no more than this can primitive man be supposed to have been affected by the deaths of his fellows. But possibly, since he was carnivorous, the sick and old may have been killed for food, as is still the practice among some tribes of savages. In the natural course, however, more or less permanent unions, though perhaps not regular marriages, must have developed in the female exogamous clan, which would thus usually have men of other clans living with it. And since identification of individuals would be extremely difficult before the introduction of personal names, there would be danger that when two clans met, men and women belonging to the same totem-clan would have sexual intercourse. This offence, owing to the strength of the feeling for exogamy, was frequently held to entail terrible evils for the community, and was consequently sometimes punished with death as treason. Moreover, if we suppose a number of small clans, A, B, C, D and E, to meet each other again and again, and the men and women to unite promiscuously, it is clear that the result would be a mixture of relationships of a very incestuous character. The incest of brothers and sisters by the same father would be possible and of almost all other relations, though that of brothers and sisters by the same mother would not be caused. This may have been the reason for the introduction of the class system among the Australians and Red Indians, by which all the clans of a certain area were divided into two classes, and the men of any clan of one class could only marry or have intercourse with the women of a clan of the other class. By such a division the evil results of the mixture of totems in exogamous clans with female descent would be avoided. The class system was sometimes further strengthened by [145]the rule, in Australia, that different classes should, when they met, encamp on opposite sides of a creek or other natural division164; whilst among the Red Indians, the classes camp on opposite sides of the road, or live on different sides of the same house or street.165 In Australia, and very occasionally elsewhere, the class system has been developed into four and eight sub-classes. A man of one sub-class can only marry a woman of one other, and their children belong to one of those different from either the father’s or mother’s. This highly elaborate and artificial system was no doubt, as stated by Sir J. G. Frazer, devised for the purpose of preventing the intermarriage of parents and children belonging to different clans where there are four sub-classes, and of first cousins where there are eight sub-classes.166 The class system, however, would not appear to have been the earliest form of exogamy among the Australian tribes. Its very complicated character, and the fact that the two principal classes sometimes do not even have names, seem to preclude the idea of its having been the first form of exogamy, which is a strong natural feeling, so much so that it may almost be described as an instinct, though of course not a primitive animal instinct. And just as the totem clan, which establishes a sentiment of kinship between people who are not related by blood, was prior to the individual family, so exogamy, which forbids the marriage of people who are not related by blood, must apparently have been prior to the feeling simply against connections of persons related by blood or what we call incest. If the two-class system was introduced in Australia to prohibit the marriage of brothers and sisters at a time when they could not recognise each other in adult life, then on the introduction of personal names which would enable brothers and sisters to recognise and remember each other, the two-class system should have been succeeded by a modern table of prohibited degrees, and not by clan exogamy at all. It is suggested that the two-class system was a common and natural form of evolution of a society divided into exogamous totem clans with female descent, when a man was not taken into [146]the clan of the woman with whom he lived. The further subdivision into four and eight sub-classes is almost peculiar to the Australian tribes; its development may perhaps be attributed to the fact that these tribes have retained the system of female descent and the migratory hunting method of life for an abnormally long period, and have evolved this special institution to prevent the unions of near relatives which are likely to occur under such conditions. The remains of a two-class system appear to be traceable among the Gonds of the Central Provinces. In one part of Bastar all the Gond clans are divided into two classes without names, and a man cannot marry a woman belonging to any clan of his own class, but must take one from a clan of the other class. Elsewhere the Gonds are divided into two groups of six-god and seven-god worshippers among whom the same rule obtains. Formerly the Gonds appear in some places to have had seven groups, worshipping different numbers of gods from one to seven, and each of these groups was exogamous. But after the complete substitution of male for female kinship in the clan, and the settlement of clans in different villages, the classes cease to fulfil any useful purpose. They are now disappearing, and it is very difficult to obtain any reliable information about their rules. The system of counting kinship through the mother, or female descent, has long been extinct in the Central Provinces and over most of India. Some survival of it, or at least the custom of polyandry, is found among the Nairs of southern India and in Thibet. Elsewhere scarcely a trace remains, and this was also the condition of things with the classical races of antiquity; so much so, indeed, that even great thinkers like Sir Henry Maine and M. Fustel de Coulanges, with the examples only of India, Greece and Rome before them, did not recognise the system of female descent, and thought that the exogamous clan with male descent was an extension of the patriarchal family, this latter having been the original unit of society. The wide distribution of exogamy and the probable priority of the system of female to that of male descent were first brought prominently to notice by Mr. M’Lennan. Still a distinct trace of the prior form survives here in the special relationship sometimes found to [147]exist between a man and his sister’s children. This is a survival of the period when a woman’s children, under the rule of female descent, belonged to her own family and her husband or partner in sexual relations had no proprietary right or authority over them, the place and authority of a father belonging in such a condition of society to the mother’s brother or brothers. Among the Halbas a marriage is commonly arranged when practicable between a brother’s daughter and a sister’s son. And a man always shows a special regard and respect for his sister’s son, touching the latter’s feet as to a superior, while whenever he desires to make a gift as an offering of thanks and atonement, or as a meritorious action, the sister’s son is the recipient. At his death he usually leaves a substantial legacy, such as one or two buffaloes, to his sister’s son, the remainder of the property going to his own family. Similarly among the Kamārs the marriage of a man’s children with his sister’s children is considered the most suitable union. If a man’s sister is poor, he will arrange for the weddings of her children. He will never beat his sister’s children however much they may deserve it, and he will not permit his sister’s son or daughter to eat from the dish from which he eats. The last rule, it is said, also applies to the maternal aunt. The Kunbis, and other Marātha castes, have a saying: ‘At the sister’s house the brother’s daughter is a daughter-in-law.’ The Gonds call the wedding of a brother’s daughter to a sister’s son Dūdh lautāna, or ‘bringing back the milk.’ The reason why a brother was formerly anxious to marry his daughter to his sister’s son was that the latter would be his heir under the matriarchal system; but now that inheritance is through males, and girls are at a premium for marriage, a brother is usually more anxious to get his sister’s daughter for his son, and on the analogy of the opposite union it is sometimes supposed, as among the Gonds, that he also has a right to her. Many other instances of the special relation between a brother and his sister’s children are given by Sir J.G. Frazer in Totemism and Exogamy. In some localities also the Korkus build their villages in two long lines of houses on each side of the road, and it may be the case that this is a relic of the period when two or more clans with female [148]descent lived in the same village, and those belonging to each class who could not marry or have sexual relations among themselves occupied one side of the road.

The transfer of the reckoning of kinship and descent from the mother’s to the father’s side may perhaps be associated with the full recognition of the physical fact of paternity. Though they may not have been contemporaneous in all or even the majority of societies, it would seem that the former was in most cases the logical outcome of the latter, regard being had also to the man’s natural function as protector of the family and provider of its sustenance. But this transition from female to male kinship was a social revolution of the first importance. Under the system of female descent there had been generally no transfer of clanship; both the woman and her partner or husband retained their own clans, and the children belonged to their mother’s clan. In the totemic stage of society the totem-clan was the vital organism, and the individual scarcely realised his own separate existence, but regarded himself as a member of his totem-clan, being a piece or fraction of a common life which extended through all the members of the clan and all the totem animals of the species. They may have thought also that each species of animals and plants had a different kind of life, and consequently also each clan whose life was derived from, and linked to, that of its totem-species. For the name, and life, and qualities, and flesh and blood were not separate conceptions, but only one conception; and since the name and qualities were part of the life, the life of one species could not be the same as that of another, and every species which had a separate name must have been thought to have a different kind of life. Nor would man have been regarded as a distinct species in the early totem-stage, and there would be no word for man; but each totem-clan would regard itself as having the same life as its totem-species. With the introduction of the system of male kinship came also the practice of transferring a woman from her own clan to that of her husband. It may be suggested that this was the origin of the social institution of marriage. Primitive society had no provision for such a procedure, which was opposed to its one fundamental idea of its own constitution, and involved [149]a change of the life and personality of the woman transferred.

The view seems to have been long held that this transfer could only be effected by violence or capture, the manner in which presumably it was first practised. Marriage by capture is very widely prevalent among savage races, as shown by Mr. M’Lennan in Primitive Marriage, and by Dr. Westermarck in The History of Human Marriage. Where the custom has given place to more peaceable methods of procuring a wife, survivals commonly occur. In Bastar the regular capture of the girl is still sometimes carried out, though the business is usually arranged by the couple beforehand, and the same is the case among the Kolāms of Wardha. A regular part of the marriage procedure among the Gonds and other tribes is that the bride should weep formally for some hours, or a day before the wedding, and she is sometimes taught to cry in the proper note. At the wedding the bride hides somewhere and has to be found or carried off by the bridegroom or his brother. This ritualistic display of grief and coyness appears to be of considerable interest. It cannot be explained by the girl’s reluctance to marriage as involving the loss of her virginity, inasmuch as she is still frequently not a virgin at her wedding, and to judge from the analogy of other tribes, could seldom or never have been one a few generations back. Nor is affection for her family or grief at the approaching separation from them a satisfactory motive. This would not account for the hiding at all, and not properly for the weeping, since she will after all only live a few miles away and will often return home; and sometimes she does not only weep at her own house but at all the houses of the village. The suggestion may be made that the procedure really indicates the girl’s reluctance to be severed from her own clan and transferred to another; and that the sentiment is a survival of the resistance to marriage by capture which was at first imposed on the women by the men from loyalty to the clan totem and its common life, and had nothing to do with the conjugal relationship of marriage. But out of this feeling the sexual modesty of women, which had been non-existent in the matriarchal [150]condition of society, was perhaps gradually developed. The Chamārs of Bilāspur have sham fights on the approach of the wedding party, and in most Hindu castes the bridegroom on his arrival performs some militant action, such as striking the marriage-shed or breaking one of its festoons. After the marriage the bride is nearly always sent home with the bridegroom’s party for a few days, even though she may be a child and the consummation of the marriage impossible. This may be in memory of her having formerly been carried off, and some analogous significance may attach to our honeymoon. When the custom of capture had died down it was succeeded by the milder form of elopement, or the bride was sold or exchanged against a girl from the bridegroom’s family or clan, but there is usually a relic of a formal transfer, such as the Hindu Kanyadān or gift of the virgin, the Roman Traditio in manum or her transfer from her father’s to her husband’s power, and the giving away of the bride.

These customs seem to mark the transfer of the woman from her father’s to her husband’s clan, which was in the first instance effected forcibly and afterwards by the free gift of her father or guardian, and the change of surname would be a relic of the change of clan. Among the Hindus a girl is never called by her proper name in her husband’s house, but always by some other name or nickname. This custom seems to be a relic of the period when the name denoted the clan, though it no longer has any reference either to the girl’s clan or family. Another rite portraying the transfer in India is the marking of the bride’s forehead with vermilion, which is no doubt a substitute for blood. The ceremony would be a relic of participation in the clan sacrifice when the bride would in the first place drink the blood of the totem animal or tribal god with the bridegroom in sign of her admission to his clan and afterwards be marked with the blood as a substitute. This smear of vermilion a married woman always continues to wear as a sign of her state, unless she wears pink powder or a spangle as a substitute.167 Where this pink powder (kunku) or spangles are used they must always be given by the bridegroom [151]to the bride as part of the Sohāg or trousseau. At a Bhaina wedding the bride’s father makes an image in clay of the bird or animal of the groom’s sept and places it beside the marriage-post. The bridegroom worships the image, lighting a sacrificial fire before it, or offers to it the vermilion which he afterwards smears upon the forehead of the bride. The Khadāls at their marriages worship their totem animal or tree, and offer to it flowers, sandalwood, vermilion, uncooked rice, and the new clothes and ornaments intended for the bride, which she may not wear until this ceremony has been performed. Again, the sacrament of the Meher or marriage cakes is sometimes connected with the clan totem in India. These cakes are cooked and eaten sacramentally by all the members of the family and their relatives, the bride and bridegroom commencing first. Among the Kols the relatives to whom these cakes are distributed cannot intermarry, and this indicates that the eating of them was formerly a sacrament of the exogamous clan. The association of the totem with the marriage cakes is sometimes clearly shown. Thus in the Dahāit caste members of the clans named after certain trees, go to the tree at the time of their weddings and invite it to be present at the ceremony. They offer the marriage cakes to the tree. Those of the Nāgotia or cobra clan deposit the cakes at a snake’s hole. Members of the Singh (lion) and Bāgh (tiger) clans draw images of these animals on the wall at the time of their weddings and offer the cakes to them. The Basors of the Kulatia or somersault clan do somersaults at the time of eating the cakes; those of the Karai Nor clan, who venerate a well, eat the cakes at a well and not at home. Basors of the Lurhia clan, who venerate a grinding-stone, worship this implement at the time of eating the marriage cakes. M. Fustel de Coulanges states that the Roman Confarreatio, or eating of a cake together by the bride and bridegroom in the presence of the family gods of the latter, constituted their holy union or marriage. By this act the wife was transferred to the gods and religion of her husband.168 Here the gods referred to are clearly held to be the family gods, and in the historical period it [152]seems doubtful whether the Roman gens was still exogamous. But if the patriarchal family developed within the exogamous clan tracing descent through males, and finally supplanted the clan as the most important social unit, then it would follow that the family gods were only a substitute for the clan gods, and the bride came to be transferred to her husband’s family instead of to his clan. The marriage ceremony in Greece consisted of a common meal of a precisely similar character,169 and the English wedding cake seems to be a survival of such a rite. At their weddings the Bhīls make cakes of the large millet juāri, calling it Juāri Māta or Mother Juāri. These cakes are eaten at the houses of the bride and bridegroom by the members of their respective clans, and the remains are buried inside the house as sacred food. Dr. Howitt states of the Kurnai tribe: “By and by, when the bruises and perhaps wounds received in these fights (between the young men and women) had healed, a young man and a young woman might meet, and he, looking at her, would say, for instance, ‘Djiitgun!170 What does the Djiitgun eat?’ The reply would be ‘She eats kangaroo, opossum,’ or some other game. This constituted a formal offer and acceptance, and would be followed by the elopement of the couple as described in the chapter on Marriage.”171 There is no statement that the question about eating refers to the totem, but this must apparently have been the original bearing of the question, which otherwise would be meaningless. Since this proposal of marriage followed on a fight between the boys and girls arising from the fact that one party had injured the other party’s sex-totem, the fight may perhaps really have been a preliminary to the proposal and have represented a symbolic substitute for or survival of marriage by capture. Among the Santāls, Colonel Dalton says, “the social meal that the boy and girl eat together is the most important part of the ceremony, as by the act the girl ceases to belong to her father’s tribe and becomes a member of the husband’s family.” Since the terms tribe and family [153]are obviously used loosely in the above statement, we may perhaps substitute clan in both cases. Many other instances of the rite of eating together at a wedding are given by Dr. Westermarck.172 If, therefore, it be supposed that the wedding ceremony consisted originally of the formal transfer of the bride to the bridegroom’s clan, and further that the original tie which united the totem-clan was the common eating of the totem animal, then the practice of the bride and bridegroom eating together as a symbol of marriage can be fully understood. When the totem animal had ceased to be the principal means of subsistence, bread, which to a people in the agricultural stage had become the staff or chief support of life, was substituted for it, as argued by Professor Robertson Smith in The Religion of the Semites. If the institution of marriage was thus originally based on the forcible transfer of a woman from her own to her husband’s clan, certain Indian customs become easily explicable in the light of this view. We can understand why a Brāhman or Rājpūt thought it essential to marry his daughter into a clan or family of higher status than his own; because the disgrace of having his daughter taken from him by what had been originally an act of force, was atoned for by the superior rank of the captor or abductor. And similarly the terms father-in-law and brother-in-law would be regarded as opprobrious because they originally implied not merely that the speaker had married the sister or daughter of the person addressed, but had married her forcibly, thereby placing him in a position of inferiority. A Rājpūt formerly felt it derogatory that any man should address him either as father-or brother-in-law. And the analogous custom of a man refusing to take food in the house of his son-in-law’s family and sometimes even refusing to drink water in their village would be explicable on precisely the same grounds. This view of marriage would also account for the wide prevalence of female infanticide. Because in the primitive condition of exogamy with male descent, girls could not be married in their own clan, as this would transgress the binding law of exogamy, and they could not be transferred from their own totem-clan and [154]married in another except by force and rape. Hence it was thought better to kill girl children than to suffer the ignominy of their being forcibly carried off. Both kinds of female infanticide as distinguished by Sir H. Risley173 would thus originally be due to the same belief. The Khond killed his daughter because she could not be married otherwise than by forcible abduction; not necessarily because he was unable to protect her, but because he could not conceive of her being transferred from one totem-clan to another by any other means; and he was bound to resist the transfer because by acquiescing in it, he would have been guilty of disloyalty to his own totem, whose common life was injured by the loss of the girl. The Rājpūt killed his daughter because it was a disgrace to him to get her married at all outside his clan, and she could not be married within it. Afterwards the disgrace was removed by marrying her into a higher clan than his own and by lavish expenditure on the wedding; and the practice of female infanticide was continued to avoid the ruinous outlay which this primitive view of marriage had originally entailed. The Hindu custom of the Swayamvāra or armed contest for the hand of a Rājpūt princess, and the curious recognition by the Hindu law-books of simple rape as a legitimate form of marriage would be explained on the same ground.

It has been seen that the exogamous clan with female descent contained no married couples, and therefore it was necessary either that outside men should live with it, or that the clans should continually meet each other, or that two or more should live in the same village. With the change to male descent and the transfer of women to their husbands’ clans, this unstable characteristic was removed. Henceforth the clan was self-contained, having its married couples, both members of it, whose children would also be born in and belong to it. Since the clan was originally a body of persons who wandered about and hunted together, its character would be maintained by living together, and there is reason to suppose that the Indian exogamous clan with male descent took its special character because its [155]members usually lived in one or more villages. This fact would account for the large number and multiplication of clans in India as compared with other places. As already seen one of the names of a clan is khera, which also means a village, and a large number of the clan names are derived from, or the same, as those of villages. Among the Khonds all the members of one clan live in the same locality about some central village. Thus the Tupa clan are collected about the village of Teplagārh in Patna State, the Loa clan round Sindhekala, the Borga clan round Bangomunda and so on. The Nunias of Mīrzāpur, Mr. Crooke remarks,174 have a system of local subdivisions called dīh, each subdivision being named after the village which is supposed to be its home. The word dīh itself means a site or village. Those who have the same dīh do not intermarry. In the villages first settled by the Oraons, Father Dehon states,175 the population is divided into three khunts or branches, the founders of the three branches being held to have been sons of the first settler. Members of each branch belong to the same clan or got. Each khunt or branch has a share of the village lands. The Mochis or cobblers have forty exogamous sections or gotras, mostly named after Rājpūt clans, and they also have an equal number of kheras or groups named after villages. The limits of the two groups seem to be identical; and members of each group have an ancestral village from which they are supposed to have come. Marriage is now regulated by the Rājpūt sept-names, but the probability is that the kheras were the original divisions, and the Rājpūt gotras have been more recently adopted in support of the claims already noticed. The Parjas have totemistic exogamous clans and marriage is prohibited in theory between members of the same clan. But as the number of clans is rather small, the rule is not adhered to, and members of the same clan are permitted to marry so long as they do not come from the same village. The Mīnas of Rājputāna are divided into twelve exogamous pāls or clans; the original meaning of the word pāl was a [156]defile or valley suitable for defence, where the members of the clan would live together as in a Scotch glen.

Thus among the cultivating castes apparently each exogamous clan consisted originally of the residents of one village, though they afterwards spread to a number of villages. The servile labouring castes may also have arranged their clans by villages as the primitive forest-tribes did. How the menial castes formed exogamous clans is not altogether clear, as the numbers in one village would be only small. But it may be supposed that as they gradually increased, clans came into existence either in one large village or a number of adjacent ones, and sometimes traced their descent from a single family or from an ancestor with a nickname. As a rule, the artisan castes do not appear to have formed villages of their own in India, as they did in Russia, though this may occasionally have happened. When among the cultivating castes the lands were divided, separate joint families would be constituted; the head only of each family would be its representative in the clan, as he would hold the share of the village land assigned to the family, which was their joint means of subsistence, and the family would live in one household. Thus perhaps the Hindu joint family came into existence as a subdivision of the exogamous clan with male descent, on which its constitution was modelled. In Chhattīsgarh families still live together in large enclosures with separate huts for the married couples. A human ancestor gradually took the place of the totem as the giver of life to the clan. The members thought themselves bound together by the tie of his blood which flowed through all their veins, and frequently, as in Athens, Rome and Scotland, every member of the clan bore his name. In this capacity, as the source of the clan’s life, the original ancestor was perhaps venerated, and on the development of the family system within the clan, the ancestors of the family were held in a similar regard, and the feeling extended to the living ancestor or father, who is treated with the greatest deference in the early patriarchal family. Even now Hindu boys, though they may be better educated and more intelligent than their father, will not as a rule address him at meals unless he speaks to them first, on account of their [157]traditional respect for him. The regard for the father may be strengthened by his position as the stay and support of the family, but could scarcely have arisen solely from this cause.

Dr. Westermarck’s view that the origin of exogamy lay in the feeling against the marriage of persons who lived together, receives support from the fact that a feeling of kinship still subsists between Hindus living in the same village, even though they may belong to different castes and clans. It is commonly found that all the households of a village believe themselves in a manner related. A man will address all the men of the generation above his own as uncle, though they may be of different castes, and the children of the generation below his own as niece and nephew. When a girl is married, all the old men of the village call her husband ‘son-in-law.’ This extends even to the impure castes who cannot be touched. Yet owing to the fact that they live together they are considered by fiction to be related. The Gowāri caste do not employ Brāhmans for their weddings, but the ceremony is performed by the bhānja or sister’s son either of the girl’s father or the boy’s father. If he is not available, any one whom either the girl’s father or the boy’s father addresses as bhānja or nephew in the village, even though he may be no relation and may belong to another caste, may perform the ceremony as a substitute. Among the Oraons and other tribes prenuptial intercourse between boys and girls of the same village is regularly allowed. It is not considered right, however, that these unions should end in marriage, for which partners should be sought from other villages.176 In the Marātha country the villagers have a communal feast on the occasion of the Dasahra festival, the Kunbis or cultivators eating first and the members of the menial and labouring castes afterwards.

The Brāhmans and Rājpūts, however, and one or two other military castes, as the Marāthas and Lodhis, do not have the small exogamous clans (which probably, as has been seen, represented the persons who lived together in a village), but large ones. Thus the Rājpūts were divided into thirty-six royal races, and theoretically all these should [158]have been exogamous, marrying with each other. Each great clan was afterwards, as a rule, split into a number of branches, and it is probable that these became exogamous; while in cases where a community of Rājpūts have settled on the land and become ordinary cultivators, they have developed into an endogamous subcaste containing small clans of the ordinary type. It seems likely that the Rājpūt clan originally consisted of those who followed the chief to battle and fought together, and hence considered themselves to be related. This was, as a matter of fact, the case. Colonel Tod states that the great Rāthor clan, who said that they could muster a hundred thousand swords, spoke of themselves as the sons of one father. The members of the Scotch clans considered themselves related in the same manner, and they were probably of similar character to the Rājpūt clans.177 I do not know, however, that there is any definite evidence as to the exogamy of the Scotch clans, which would have disappeared with their conversion to Christianity. The original Rājpūt clan may perhaps have lived round the chiefs castle or headquarters and been supported by the produce of his private fief or demesne. The regular Brāhman gotras are also few in number, possibly because they were limited by the paucity of eponymous saints of the first rank. The word gotra means a stall or cow-pen, and would thus originally signify those who lived together in one place like a herd of cattle. But the gotras are now exceedingly large, the same ones being found in most or all of the Brāhman subcastes, and it is believed that they do not regulate marriage as a rule. Sometimes ordinary surnames have taken the place of clan names, and persons with the same surname consider themselves related and do not marry. But usually Brāhmans prohibit marriage between Sapindas or persons related to each other within seven degrees from a common ancestor. The word Sapinda signifies those who partake together of the pindas or funeral cakes offered to the dead. The Sapindas are also a man’s heirs in the absence of closer relations; the group of the Sapindas is thus an exact replica within the gotra of the primitive totem clan which [159]was exogamous and constituted by the tie of living and eating together. Similarly marriage at Rome was prohibited to seven degrees of relationship through males within the gens,178 and this exogamous group of kinsmen appear to have been the body of agnatic kinsmen within the gens who are referred to by Sir H. Maine as a man’s ultimate heirs.179 At Athens, when a contest arose upon a question of inheritance, the proper legal evidence to establish kinship was the proof that the alleged ancestor and the alleged heir observed a common worship and shared in the same repast in honour of the dead.180 The distant heirs were thus a group within the Athenian γένος corresponding to the Sapindas and bound by the same tie of eating together. Professor Hearn states that there is no certain evidence that the Roman gens and Greek γένος were originally exogamous, but we find that of the Roman matrons whose names are known to us none married a husband with her own Gentile name; and further, that Plutarch, in writing of the Romans, says that in former days men did not marry women of their own blood or, as in the preceding sentence he calls them, kinswomen συγγενίδας, just as in his own day they did not marry their aunts or sisters; and he adds that it was long before they consented to wed with cousins.181 Professor Hearn’s opinion was that the Hindu gotra, the Roman gens and the Greek γένος were originally the same institution, the exogamous clan with male descent, and all the evidence available, as well as the close correspondence in other respects of early Hindu institutions with those of the Greek and Latin cities would tend to support this view.

In the admirable account of the early constitution of the city-states of Greece and Italy contained in the work of M. Fustel de Coulanges, La Cité Antique, a close resemblance may be traced with the main strata of Hindu society given earlier in this essay. The Roman state was composed of a number of gentes or clans, each gens tracing its descent [160]from a common ancestor, whose name it usually bore. The termination of the Gentile name in ius signified descendant, as Claudius, Fabius, and so on. Similarly the names of the Athenian γένη or clans ended in ides or ades, as Butades, Phytalides, which had the same signification.182 The Gentile or clan name was the nomen or principal name, just as the personal names of the members of the totem-clans were at first connected with the totems. The members of the gens lived together on a section of the city land and cultivated it under the control of the head of the gens. The original ager Romanus is held to have been 115 square miles or about 74,000 acres,183 and this was divided up among the clans. The heads of clans originally lived on their estates and went in to Rome for the periodical feasts and other duties. The principal family or eldest branch of the gens in the descent from a common ancestor ranked above the others, and its head held the position of a petty king in the territory of the gens. In Greece he was called ἄναξ or βασιλεύς.184 Originally the Roman Senate consisted solely of the heads of gentes, and the consuls, flamens and augurs were also chosen exclusively from them; they were known as patres; after the expulsion of the kings, fresh senators were added from the junior branches of the gentes, of which there were at this period 160, and these were known as patres conscripti185. The distinction between the eldest and junior branches of the gentes may have corresponded to the distinction between the Kshatriyas and Vaishyas, though as practically nothing is known of the constitution of the original Kshatriyas, this can only be hypothetical.

Within the gens, and living in the household or households of its members, there existed a body of slaves, and also another class of persons called clients.186 The client was a servant and dependant; he might be assigned a plot of land by his patron, but at first could not transmit it nor hold it against his patron. It is probable that originally he had no right of property of his own, but he gradually acquired it. First he obtained a right of occupancy in his land and of its [161]devolution to his son if he had one. Finally he was given the power of making a will. But he was still obliged to contribute to such expenses of the patron as ransom in war, fines imposed by the courts, or the dowry of a daughter.187 The client was considered as a member of the family and bore its name.188 But he was not a proper member of the family or gens, because his pedigree never ascended to a pater or the head of a gens.189 It was incumbent on the patron to protect the client, and guard his interests both in peace and war. The client participated in the household and Gentile sacrifices and worshipped the gods of the gens.190 At first the people of Rome consisted of three classes, the patricians, the clients and the plebeians. In course of time, as the rights and privileges of the plebeians increased after the appointment of tribunes, their position, from having originally been much inferior, became superior to that of the clients, and the latter preferred to throw off the tie uniting them to their patrons and become merged in the plebeians. In this manner the intermediate class of clients at length entirely disappeared.191 These clients must not be confused with the subsequent class of the same name, who are found during the later period of the republic and the empire, and were the voluntary supporters or hangers-on of rich men. It would appear that these early clients corresponded very closely to the household servants of the Indian cultivators, from whom the village menial castes were developed. The Roman client was sometimes a freed slave, but this would not have made him a member of the family, even in a subordinate position. Apparently the class of clients may have to a great extent originated in mixed descent, as the Indian household and village menials probably did. This view would account satisfactorily for the client’s position as a member of the family but not a proper one. From the fact that they were considered one of the three principal divisions of the people it is clear that the clients must at one time have been numerous and important.

Below the clients came the plebeians, whose position, as [162]M. Fustel de Coulanges himself points out, corresponded very closely to that of the Sūdras. The plebeians had no religion and no ancestors; they did not belong to a family or a gens.192 They were a despised and abject class, who lived like beasts outside the proper boundary of the city. The touch of the plebeian was impure.193 “When tribunes were created a special law was necessary to protect their life and liberty, and it was promulgated as follows: ‘It is forbidden to strike or kill a tribune, as if he was an ordinary plebeian.’ It would appear then that a patrician had the right to strike or kill an ordinary plebeian, or at least that he was amenable to no legal punishment for doing so.”194 Similarly in the ancient Greek cities the citizens were known as ἀγαθοί or good, and the plebeians as κακοί or bad. This latter class is described by the poet Theognis as having had aforetime neither tribunals nor laws; they were not allowed even to enter the town, but lived outside like wild beasts. They had no part in the religious feasts and could not intermarry with the proper citizens.195

This position corresponds exactly with that of the Sūdras and the existing impure castes, who have to live outside the village and cannot enter or even approach Hindu temples.

M. de Coulanges considers that the plebeians were to a large extent made up of conquered and subjected peoples. An asylum was also established at Rome for broken men and outlaws from other cities, with a view to increasing the population and strength of the state. Subsequently the class of clients became absorbed among the plebeians.

Thus the gradation of society in the city-states of Greece and Italy, the account given above being typical of them all, is seen to correspond fairly closely with that of the Hindus, as exemplified in the Hindu classics and the microcosm of Hindu society, the village community. It is desirable, therefore, to inquire what was the tie which united the members of the gens, the curia or phratry, and the city, and which distinguished the patricians from the plebeians. [163]On this point M. Fustel de Coulanges leaves us in no doubt at all. The bond of union among all these bodies was a common sacrifice or sacrificial meal, at which all the members had to be present. “The principal ceremony of the religion of the household was a meal, which was called a sacrifice. To eat a meal prepared on an altar was, according to all appearance, the first form of religious worship.”196 “The principal ceremony of the religion of the city was also a public feast; it had to be partaken of communally by all the citizens in honour of the tutelary deities. The custom of holding these public feasts was universal in Greece; and it was believed that the safety of the city depended on their accomplishment.”197 M. de Coulanges quotes from the Odyssey an account of one of these sacred feasts at which nine long tables were set out for the people of Pylos; five hundred citizens were seated and nine bulls were slaughtered for each table. When Orestes arrived at Athens after the murder of his mother, he found the people, assembled round their king, about to hold the sacred feast. Similar feasts were held and numerous victims were slaughtered in Xenophon’s time.198 At these meals the guests were crowned with garlands and the vessels were of a special form and material, such as copper or earthenware, no doubt dating from the antique past.199 As regards the importance and necessity of being present at the Gentile sacrificial feast, the same author states: “The Capitol was blockaded by the Gauls; but Fabius left it and passed through the hostile lines, clad in religious garb, and carrying in his hand the sacred objects; he was going to offer a sacrifice on the altar of his gens which was situated on the Quirinal. In the second Punic war another Fabius, he who was called the buckler of Rome, was holding Hannibal in check; it was assuredly of the greatest importance to the Republic that he should not leave his army; he left it, however, in the hands of the imprudent Minucius; it was because the anniversary day of the sacrifice of his gens had come and it was necessary that he should hasten to Rome to perform the sacred rite.” In Greece the members of the gens were [164]known by the fact that they performed communal sacrifices together from a remote period.200 As already seen, a communal sacrifice meant the eating together of the sacred food, whether the flesh of a victim or grain.

The Roman city sacrifice of the Suovetaurilia, as described by M. de Coulanges, is of the greatest interest. The magistrate whose duty it was to accomplish it, that is in the first place the king, after him the consul, and after him the censor, had first to take the auspices and ascertain that the gods were favourable. Then he summoned the people through a herald by a consecrated form of words. On the appointed day all the citizens assembled outside the walls; and while they stood silent the magistrate proceeded three times round the assembly, driving before him three victims—a pig, a ram and a bull. The combination of these three victims constituted with the Greeks as well as the Romans an expiatory sacrifice. Priests and attendants followed the procession: when the third round had been accomplished, the magistrate pronounced a prayer and slaughtered the victims. From this moment all sins were expiated, and neglect of religious duties effaced, and the city was at peace with its gods.

There were two essential features of this ceremony: the first, that no stranger should be present at it; and the second, that no citizen should be absent from it. In the latter case the whole city might not have been freed from impurity. The Suovetaurilia was therefore preceded by a census, which was conducted with the greatest care both at Rome and Athens. The citizen who was not enrolled and was not present at the sacrifice could no longer be a member of the city. He could be beaten and sold as a slave, this rule being relaxed only in the last two centuries of the Republic. Only male citizens were present at the sacrifice, but they gave a list of their families and belongings to the censor, and these were considered to be purified through the head of the family.201

This sacrifice was called a lustratio or purification, and in the historical period was considered to be expiatory. But it does not seem probable that this was its original [165]significance. For there would not in that case have been the paramount necessity for every citizen to be present. All females and children under power were purified through the list given to the censor, and there seems no reason why absent citizens could not have been purified in the same manner. But participation in this sacrifice was itself the very test and essence of citizenship. And it has been seen that a public meal was the principal religious rite of the city. The conclusion therefore seems reasonable that the Suovetaurilia was originally also a sacrificial meal of which each citizen partook, and that the eating of the deified domestic animals in common was the essence of the rite and the act which conferred the privilege of citizenship. The driving of the sacrificial animals round the citizens three times might well be a substitute for the previous communal meal, if for any reason, such as the large number of citizens, the practice of eating them had fallen into abeyance. The original ground for the taking of a census was to ensure that all the citizens were present at the communal sacrifice; and it was by the place which a man occupied on this day that his rank in the city was determined till the next sacrifice. If the censor counted him among the senators, he remained a senator; if among the equites, he remained a knight; if as a simple member of a tribe, he belonged henceforward to the tribe in which he was counted. If the censor refused to enumerate him, he was no longer a citizen.202 Such was the vital importance of the act of participation in the sacrifice.

The Roman sacrifice of the Suovetaurilia was in no way peculiar, similar rites being found in other Greek and Latin cities. Some instances are recorded in the article on Kasai, and in Themis203 Miss Jane Harrison gives an account of a sacrifice at Magnesia in which a bull, ram and he- and she-goats were sacrificed to the gods and partaken of communally by the citizens. As already seen, the act of participation in the sacrifice conferred the status of citizenship. The domestic animals were not as a rule eaten, but their milk was drunk, and they were used for transport, and [166]clothes were perhaps sometimes made from their hair and skins. Hence they were the principal source of life of the tribe, as the totem had been of the clan, and were venerated and deified. One common life was held to run through all the members of the tribe and all the domestic animals of the species which was its principal means of support. In the totem or hunting stage the clan had necessarily been small, because a large collection of persons could not subsist together by hunting and the consumption of roots and fruits. When an additional means of support was afforded by the domestication of an important animal, a much larger number of persons could live together, and apparently several clans became amalgamated into a tribe. The sanctity of the domestic animals was much greater than that of the totem because they lived with man and partook of his food, which was the strongest tie of kinship; and since he still endowed them with self-consciousness and volition, he thought they had come voluntarily to aid him in sustaining life. Both on this account and for fear of injuring the common life they were not usually killed. But it was necessary to primitive man that the tie should take a concrete form and that he should actually assimilate the life of the sacred animal by eating its flesh, and this was accordingly done at a ceremonial sacrifice, which was held annually, and often in the spring, the season of the renewal and increase of life. Since this renewal of the communal life was the concrete tie which bound the tribe together, any one who was absent from it could no longer be a member of the tribe. The whole of this rite and the intense importance attached to it are inexplicable except on the supposition that the tie which had originally constituted the totem-clan was the eating of the totem-animal, and that this tie was perpetuated in the tribe by the communal eating of the domestic animal. The communal sacrifice of the domestic animal was, as already seen, typical of society in the tribal or pastoral stage. But one very important case, in addition to those given above and in the article on Kasai, remains for notice. The Id-ul-Zoha or Bakr-Id festival of the Muhammadans is such a rite. In pre-Islamic times this sacrifice was held at Mecca and all the Arab tribes went to Mecca to celebrate it. The month in which the [167]sacrifice was held was one of those of truce, when the feuds between the different clans were in abeyance so that they could meet at Mecca. Muhammad continued the sacrifice of the Id-ul-Zoha and it is this sacrifice which a good Muhammadan takes the pilgrimage to Mecca to perform. He must be at Mecca on the tenth day of the month of Z’ul Hijjah and perform the sacrifice there, and unless he does this there is no special merit in making the journey to Mecca. It is incumbent on every Muhammadan who can afford it to make the pilgrimage to Mecca or the Hajj once in his life and perform the sacrifice there; and though as a matter of fact only a very small minority of Muhammadans now carry out the rule, the pilgrimage and sacrifice may yet be looked upon as the central and principal rite of the Muhammadan religion. All Muhammadans who cannot go to Mecca nevertheless celebrate the sacrifice at home at the Indian festival of the Id-ul-Zoha and the Turkish and Egyptian Idu-Bairām. At the Id-ul-Zoha any one of four domestic animals, the camel, the cow, the sheep or the goat, may be sacrificed; and this rule makes it a connecting link between the two great Semitic sacrifices described in the article on Kasai, the camel sacrifice of the Arabs in pre-Islamic times and the Passover of the Jews. At the present time one-third of the flesh of the sacrificial animal should be given to the poor, one-third to relations, and the remainder to the sacrificer’s own family.204 Though it has now become a household sacrifice, the communal character thus still partly survives.

Both in Athens and Rome there was a division known as phratry or curia. This apparently consisted of a collection of gentes, γένη, or clans, and would correspond roughly to a Hindu subcaste. The evidence does not show, however, that it was endogamous. The bond which united the phratry or curia was precisely the same as that of the gens or clan and the city. It consisted also in a common meal, which was prepared on the altar, and was eaten with the recitation of prayers, a part being offered to the god, who was held to be present. At Athens on feast-days the members [168]of the phratry assembled round their altar. A victim was sacrificed and its flesh cooked on the altar, and divided among the members of the phratry, great care being taken that no stranger should be present. A young Athenian was presented to the phratry by his father, who swore that the boy was his son. A victim was sacrificed and cooked on the altar in the presence of all the members of the phratry; if they were doubtful of the boy’s legitimacy, and hence wished to refuse him admittance, as they had the right to do, they refused to remove the flesh from the altar. If they did not do this, but divided and partook of the flesh with the candidate, he was finally and irrevocably admitted to the phratry. The explanation of this custom, M. de Coulanges states, is that food prepared on an altar and eaten by a number of persons together, was believed to establish between them a sacred tie which endured through life.205 Even a slave was to a certain degree admitted into the family by the same tie of common eating of food. At Athens he was made to approach the hearth; he was purified by pouring water on his head, and ate some cakes and fruit with the members of the family. This ceremony was analogous to those of marriage and adoption. It signified that the new arrival, hitherto a stranger, was henceforth a member of the family and participated in the family worship.206

The analogy of Greece and Rome would suggest the probability that the tie uniting the members of the Indian caste or subcaste is also participation in a common sacrificial meal, and there is a considerable amount of evidence to support this view. The Confarreatio or eating together of the bride and bridegroom finds a close parallel in the family sacrament of the Meher or marriage cakes, which has already been described. This would appear formerly to have been a clan rite, and to have marked the admission of the bride to the bridegroom’s clan. It is obligatory on relations of the families to attend a wedding and they proceed from great distances to do so, and clerks and other officials are much aggrieved if the exigencies of Government business prevent them from obtaining leave. The obligation seems to be of the same character as that which caused Fabius to leave the [169]army in order to attend his Gentile sacrifice at Rome. If he did not attend the Gentile sacrifice he was not a member of the gens, and if a Hindu did not attend the feast of his clan in past times perhaps he did not remain a member of the clan. Among the Marātha Brāhmans the girl-bride eats with her husband’s relations on this day only to mark her admission into their clan, and among the Bengali Brāhmans, when the wedding guests are collected, the bride comes and puts a little sugar on each of their leaf-plates, which they eat in token of their recognition of her in her new status of married woman. The members of the caste or subcaste also assemble and eat together on three occasions: at a marriage, which will have the effect of bringing new life into the community; at a death, when a life is lost; and at the initiation of a new member or the readmission of an offender temporarily put out of caste. It is a general rule of the caste feasts that all members of the subcaste in the locality must be invited, and if any considerable number of them do not attend, the host’s position in the community is impugned. For this reason he has to incur lavish expenditure on the feast, so as to avoid criticism or dissatisfaction among his guests. These consider themselves at liberty to comment freely on the character and quality of the provisions offered to them. In most castes the feast cannot begin until all the guests have assembled; the Maheshri Banias and one or two other castes are distinguished by the fact that they allow the guests at the pangat or caste feast to begin eating as they arrive. Those who bear the host a grudge purposely stay away, and he has to run to their houses and beg them to come, so that his feast can begin. When the feast has begun it was formerly considered a great calamity if any accident should necessitate the rising of the guests before its conclusion. Even if a dog or other impure animal should enter the assembly they would not rise. The explanation of this rule was that it would be disrespectful to Um Deo, the food-god, to interrupt the feast. At the feast each man sits with his bare crossed knees actually touching those of the men on each side of him, to show that they are one brotherhood and one body. If a man sat even a few inches apart from his fellows, people would say he was out of caste; [170]and in recent times, since those out of caste have been allowed to attend the feasts, they sit a little apart in this manner. The Gowāris fine a man who uses abusive language to a fellow-casteman at a caste feast, and also one who gets up and leaves the feast without the permission of the caste headman. The Hatkars have as the names of two exogamous groups Wakmār, or one who left the Pangat or caste feast while his fellows were eating; and Polya, or one who did not take off his turban at the feast. It has been seen also207 that in one or two castes the exogamous sections are named after the offices which their members hold or the duties they perform at the caste feast. Among the Halbas the illegitimate subcaste Surāit is also known as Chhoti Pangat or the inferior feast, with the implication that its members cannot be admitted to the proper feast of the caste, but have an inferior one of their own.

When an outsider is admitted to the caste the rite is usually connected with food. A man who is to be admitted to the Dahāit caste must clean his house, break his earthen cooking-vessels and buy new ones, and give a feast to the caste-fellows in his house. He sits and takes food with them, and when the meal is over he takes a grain of rice from the leaf-plate of each guest and eats it, and drinks a drop of water from his leaf-cup. After this he cannot be readmitted to his own caste. A new Mehtar or sweeper gives water to and takes bread from each casteman. In Mandla a new convert to the Panka caste vacates his house and the caste panchāyat or committee go and live in it, in order to purify it. He gives them a feast inside the house, while he himself stays outside. Finally he is permitted to eat with the panchāyat in his own house in order to mark his admission into the caste. A candidate for admission in the Mahli caste has to eat a little of the leavings of the food of each of the castemen at a feast. The community of robbers known as Badhak or Baoria formerly dwelt in the Oudh forests. They were accustomed to take omens from the cry of the jackal, and they may probably have venerated it as representing the spirit of the forest and as a fellow-hunter. They were called jackal-eaters, and it was [171]said that when an outsider was admitted to one of their bands he was given jackal’s flesh to eat.

Again, the rite of initiation or investiture with the sacred thread appears to be the occasion of the admission of a boy to the caste community. Before this he is not really a member of the caste and may eat any kind of food. The initiation is called by the Brāhmans the second birth, and appears to be the birth of the soul or spirit. After it the boy will eat the sacrificial food at the caste feasts and be united with the members of the caste and their god. The bodies of children who have not been initiated are buried and not burnt. The reason seems to be that their spirits will not go to the god nor be united with the ancestors, but will be born again. Formerly such children were often buried in the house or courtyard so that their spirits might be born again in the same family. The lower castes sometimes consider the rite of ear-piercing as the initiation and sometimes marriage. Among the Panwār Rājpūts a child is initiated when about two years old by being given cooked rice and milk to eat. The initiation cannot for some reason be performed by the natural father, but must be done by a guru or spiritual father, who should thereafter be regarded with a reverence equal to or even exceeding that paid to the natural father.

When a man is readmitted to caste after exclusion for some offence, the principal feature of the rite is a feast at which he is again permitted to eat with his fellows. There are commonly two feasts, one known as the Maili Roti or impure meal, and the other as Chokhi or pure, both being at the cost of the offender. The former is eaten by the side of a stream or elsewhere on neutral ground, and by it the offender is considered to be partly purified; the latter is in his own house, and by eating there the castemen demonstrate that no impurity attaches to him, and he is again a full member. Some castes, as the Dhobas, have three feasts: the first is eaten at the bank of a stream, and at this the offender’s hair is shaved and thrown into the stream; the second is in his yard; and the third in his house. The offender is not allowed to partake of the first two meals himself, but he joins in the third, and before it [172]begins the head of the panchāyat gives him water to drink in which gold has been dipped as a purificatory rite. Among the Gonds the flesh of goats is provided at the first meal, but at the second only grain cooked with water, which they now, in imitation of the Hindus, consider as the sacred sacrificial food. Frequently the view obtains that the head of the caste panchāyat takes the offender’s sins upon himself by commencing to eat, and in return for this a present of some rupees is deposited beneath his plate. Similarly among some castes, as the Bahnas, exclusion from caste is known as the stopping of food and water. The Gowāris readmit offenders by the joint drinking of opium and water. One member is especially charged with the preparation of this, and if there should not be enough for all the castemen to partake of it, he is severely punished. Opium was also considered sacred by the Rājpūts, and the chief and his kinsmen were accustomed to drink it together as a pledge of amity.208

Grain cooked with water is considered as sacred food by the Hindus. It should be eaten only on a space within the house called chauka purified with cowdung, and sometimes marked out with white quartz-powder or flour. Before taking his meal a member of the higher castes should bathe and worship the household gods. At the meal he should wear no sewn clothes, but only a waist-cloth made of silk or wool, and not of cotton. The lower castes will take food cooked with water outside the house in the fields, and are looked down upon for doing this, so that those who aspire to raise their social position abandon the practice, or at least pretend to do so. Sir J.G. Frazer quotes a passage showing that the ancient Brāhmans considered the sacrificial rice-cakes cooked with water to be transformed into human bodies.209 The Urdu word bali means a sacrifice or offering, and is applied to the portion of the daily meal which is offered to the gods and to the hearth-fire. Thus all grain cooked with water is apparently looked upon as sacred or sacramental food, and it is for this reason that it can only be eaten after the purificatory rites already described. The grain is [173]venerated as the chief means of subsistence, and the communal eating of it seems to be analogous to the sacrificial eating of the domestic animals, such as the camel, horse, ox and sheep, which is described above and in the article on Kasai. Just as in the hunting stage the eating of the totem-animal, which furnished the chief means of subsistence, was the tie which united the totem-clan: and in the pastoral stage the domestic animal which afforded to the tribe its principal support, not usually as an article of food, but through its milk and its use as a means of transport, was yet eaten sacrificially owing to the persistence of the belief that the essential bond which united the tribe was the communal eating of the flesh of the animal from which the tribe obtained its subsistence: so when the community reaches the agricultural stage the old communal feast is retained as the bond of union, but it now consists of grain, which is the principal support of life.

The totem-animal was regarded as a kinsman, and the domestic animal often as a god.210 But in both these cases the life of the kinsman and god was sacrificed in order that the community might be bound together by eating the body and assimilating the life. Consequently, when grain came to be the sacrificial food, it was often held that an animal or human being must be sacrificed in the character of the corn-god or spirit, whether his own flesh was eaten or the sacred grain was imagined to be his flesh. Numerous instances of the sacrifice of the corn-spirit have been adduced by Sir J.G. Frazer in The Golden Bough, and it was he who brought this custom prominently to notice. One of the most important cases in India was the Meriah-sacrifice of the Khonds, which is described in the article on that tribe.

Two features of the Khond sacrifice of a human victim as a corn-spirit appear to indicate its derivation from the sacrifice of the domestic animal and the eating of the totem-animal, the ties uniting the clan and tribe: first, that the flesh was cut from the living victim, and, second, that the sacrifice was communal. When the Meriah-victim was bound the Khonds hacked at him with their knives while life remained, leaving only the head and bowels untouched, so [174]that each man might secure a strip of flesh. This rite appears to recall the earliest period when the members of the primitive group or clan tore their prey to pieces and ate and drank the raw flesh and blood. The reason for its survival was apparently that it was the actual life of the divine victim, existing in concrete form in the flesh and blood which they desired to obtain, and they thought that this end was more certainly achieved by cutting the flesh off him while he was still alive. In the sacrifice of the camel in Arabia the same procedure was followed; the camel was bound on an altar and the tribesmen cut the flesh from the body with their knives and swallowed it raw and bleeding.211 M. Salomon Reinach shows how the memory of similar sacrifices in Greece has been preserved in legend:212 “Actaeon was really a great stag sacrificed by women devotees, who called themselves the great hind and the little hinds; he became the rash hunter who surprised Artemis at her bath and was transformed into a stag and devoured by his own dogs. The dogs are a euphemism; in the early legend they were the human devotees of the sacred stag who tore him to pieces and devoured him with their bare teeth. These feasts of raw flesh survived in the secret religious cults of Greece long after uncooked food had ceased to be consumed in ordinary life. Orpheus (ophreus, the haughty), who appears in art with the skin of a fox on his head, was originally a sacred fox devoured by the women of the fox totem-clan; these women call themselves Bassarides in the legend, and bassareus is one of the old names of the fox. Hippolytus in the fable is the son of Theseus who repels the advances of Phaedra, his stepmother, and was killed by his runaway horses because Theseus, deceived by Phaedra, invoked the anger of a god upon him. But Hippolytus in Greek means ‘one torn to pieces by horses.’ Hippolytus is himself a horse whom the worshippers of the horse, calling themselves horses and disguised as such, tore to pieces and devoured.” All such sacrifices in which the flesh was taken from the living victim may thus perhaps be derived from the common origin of totemism. The second point about the Khond sacrifice is that it was communal; every householder [175]desired a piece of the flesh, and for those who could not be present at the sacrifice relays of messengers were posted to carry it to them while it was still fresh and might be supposed to retain the life. They did not eat the strips of flesh, but each householder buried his piece in his field, which they believed would thereby be fertilised and caused to produce the grain which they would eat. The death of the victim was considered essential to the life of the tribe, which would be renewed and strengthened by it as in the case of the sacrifice of the domestic animal. Lord Avebury gives in The Origin of Civilisation213 an almost exact parallel to the Khond sacrifice in which the flesh of the victim actually was eaten. This occurred among the Marimos, a tribe of South Africa much resembling the Bechuanas. The ceremony was called ‘the boiling of the corn.’ A young man, stout but of small stature, was usually selected and secured by violence or by intoxicating him with yaala. “They then lead him into the fields, and sacrifice him in the fields, according to their own expression, for seed. His blood, after having been coagulated by the rays of the sun, is burned along with the frontal bone, the flesh attached to it and the brain. The ashes are then scattered over the fields to fertilise them and the remainder of the body is eaten.” In other cases quoted by the same author an image only was made of flour and eaten instead of a human being:214 “In Mexico at a certain period of the year the priest of Quetzalcoatl made an image of the Deity, of meal mixed with infants’ blood, and then, after many impressive ceremonies, killed the image by shooting it with an arrow, and tore out the heart, which was eaten by the king, while the rest of the body was distributed among the people, every one of whom was anxious to procure a piece to eat, however small.” Here the communal sacrificial meal, the remaining link necessary to connect the sacrifice of the corn-spirit with that of the domestic animal and clan totem, is present. Among cases of animals sacrificed as the corn-spirit in India that of the buffalo at the Dasahra festival is the most important. The rite extends over most of India, and a full and interesting account of it has recently been published [176]by Mr. W. Crooke.215 The buffalo is probably considered as the corn-spirit because it was the animal which mainly damaged the crops in past times. Where the sacrifice still survives the proprietor of the village usually makes the first cut in the buffalo and it is then killed and eaten by the inferior castes, as Hindus cannot now touch the flesh. In the Deccan after the buffalo is killed the Mahārs rush on the carcase and each one secures a piece of the flesh. This done they go in procession round the walls, calling on the spirits and demons, and asking them to accept the pieces of meat as offerings, which are then thrown to them backwards over the wall.216 The buffalo is now looked upon in the light of a scape-goat, but the procedure described above cannot be satisfactorily explained on the scape-goat theory, and would appear clearly to have been substituted for the former eating of the flesh. In the Marātha Districts the lower castes have a periodical sacrifice of a pig to the sun; they eat the flesh of the pig together, and even the Panwār Rājpūts of the Waringanga Valley join in the sacrifice and will allow the impure caste of Mahars to enter their houses and eat of this sacrifice with them, though at other times the entry of a Mahār would defile a Panwār’s house.217 The pig is sacrificed either as the animal which now mainly injures the crops or because it was the principal sacrificial animal of the non-Aryan tribes, or from a combination of both reasons. Probably it may be regarded as the corn-spirit because pigs are sacrificed to Bhanisasur or the buffalo demon for the protection of the crops.

When the community reached the national or agricultural stage some central executive authority became necessary for its preservation. This authority usually fell into the hands of the priest who performed the sacrifice, and he became a king. Since the priest killed the sacrificial animal in which the common life of the community was held to be centred, it was thought that the life passed to him and centred in his person. For the idea of the extinction of life was not properly understood, and the life of a human being or animal [177]might pass by contact, according to primitive ideas, to the person or even the weapon which killed it, just as it could pass by assimilation to those who ate the flesh. In most of the city-states of Greece and Italy the primary function of the kings was the performance of the communal or national sacrifices. Through this act they obtained political power as representing the common life of the people, and its performance was sometimes left to them after their political power had been taken away.218 After the expulsion of the kings from Rome the duty of performing the city sacrifices devolved on the consuls. In India also the kings performed sacrifices. When a king desired to be paramount over his neighbours he sent a horse to march through their territories. If it passed through them without being captured they became subordinate to the king who owned the horse. Finally the horse was sacrificed at the Ashva-medha, the king paramount making the sacrifice, while the other kings performed subordinate parts at it.219 Similarly the Rāja of Nāgpur killed the sacrificial buffalo at the Dasahra festival. But the common life of the people was sometimes conveyed from the domestic animal to the king by other methods than the performance of a sacrifice. The king of Unyoro in Africa might never eat vegetable food but must subsist on milk and beef. Mutton he might not touch, though he could drink beer after partaking of meat. A sacred herd was kept for the king’s use, and nine cows, neither more nor less, were daily brought to the royal enclosure to be milked for his majesty. The boy who brought the cows from the pasture to the royal enclosure must be a member of a particular clan and under the age of puberty, and was subject to other restrictions. The milk for the king was drawn into a sacred pot which neither the milkman nor anybody else might touch. The king drank the milk, sitting on a sacred stool, three times a day, and any which was left over must be drunk by the boy who brought the cows from pasture. Numerous other rules and restrictions are detailed by Sir J.G. Frazer, and it may be suggested that their object was to ensure that the life of the domestic animal and with [178]it the life of the people should be conveyed pure and undefiled to the king through the milk. The kings of Unyoro had to take their own lives while their bodily vigour was still unimpaired. When the period for his death arrived the king asked his wife for a cup of poison and drank it. “The public announcement of the death was made by the chief milkman. Taking a pot of the sacred milk in his hands he mounted the house-top and cried, ‘Who will drink the milk?’ With these words he dashed the pot on the roof; it rolled off and falling to the ground was broken in pieces. That was the signal for war to the death between the princes who aspired to the throne. They fought till only one was left alive. He was the king.”220 After completing the above account, of which only the principal points have been stated, Sir J.G. Frazer remarks: “The rule which obliged the kings of Unyoro to kill themselves or be killed before their strength of mind and body began to fail through disease or age is only a particular example of a custom which appears to have prevailed widely among barbarous tribes in Africa and to some extent elsewhere. Apparently this curious practice rests on a belief that the welfare of the people is sympathetically bound up with the welfare of their king, and that to suffer him to fall into bodily or mental decay would be to involve the whole kingdom in ruin.”221 Other instances connecting the life of the king with the ox or other domestic animal are given in Totemism and Exogamy and The Golden Bough222 Among the Hereros the body of a dead chief was wrapped up in the hide of an ox before being buried.223 In the Vedic horse-sacrifice in India the horse was stifled in robes. The chief queen approached him; a cloak having been thrown over them both, she performed a repulsively obscene act symbolising the transmission to her of his fructifying powers.224 In other cases the king was identified with the corn-spirit, and in this manner he also, it may be suggested, represented the common life of the people.

The belief that the king was the incarnation of the [179]common life of the people led to the most absurd restrictions on his liberty and conduct, a few instances of which from the large collection in The Golden Bough have been quoted in the article on Nai. Thus in an old account of the daily life of the Mikado it is stated: “In ancient times he was obliged to sit on the throne for some hours every morning, with the imperial crown on his head, but to sit altogether like a statue, without stirring either hands or feet, head or eyes, nor indeed any part of his body, because, by this means, it was thought that he could preserve peace and tranquillity in his empire; for if, unfortunately, he turned himself on one side or the other, or if he looked a good while towards any part of his dominions, it was apprehended that war, famine, fire or some great misfortune was near at hand to desolate the country.”225 Here it would appear that by sitting absolutely immobile the king conferred the quality of tranquillity on the common life of his people incarnate in his person; but by looking too long in any one direction he would cause a severe disturbance of the common life in the part to which he looked. And when the Israelites were fighting with the Amalekites, so long as Moses held up his hands the Israelites prevailed; but when his hands hung down they gave way before the enemy. Here apparently the common life was held to be centred in Moses, and when he held his arms up it was vigorous, but declined as he let them down. Similarly it was often thought that the king should be killed as soon as his bodily strength showed signs of waning, so that the common life might be renewed and saved from a similar decay. Even the appearance of grey hair or the loss of a tooth were sometimes considered sufficient reasons for putting the king to death in Africa.226 Another view was that any one who killed the king was entitled to succeed him, because the life of the king, and with it the common life of the people, passed to the slayer, just as it had previously passed from the domestic animal to the priest-king who sacrificed it. One or two instances of succession by killing the king are given [180]in the article on Bhīl. Sometimes the view was that the king should be sacrificed annually, or at other intervals, like the corn-spirit or domestic animal, for the renewal of the common life. And this practice, as shown by Sir J.G. Frazer, tended to result in the substitution of a victim, usually a criminal or slave, who was identified with the king by being given royal honours for a short time before his death. Sometimes the king’s son or daughter was offered as a substitute for him, and such a sacrifice was occasionally made in time of peril, apparently as a means of strengthening or preserving the common life. When Chitor, the home of the Sesodia clan of Rājpūts, was besieged by the Muhammadans, the tradition is that the goddess of their house appeared and demanded the sacrifice of twelve chiefs as a condition of its preservation. Eleven of the chiefs sons were in turn crowned as king, and each ruled for three days, while on the fourth he sallied out and fell in battle. Lastly, the Rāna offered himself in order that his favourite son, Ajeysi, might be spared and might perpetuate the clan. In reality the chief and his sons seem to have devoted themselves in the hope that the sacrifice of the king might bring strength and victory to the clan. The sacrifice of Iphigenia and possibly of Jephthah’s daughter appear to be parallel instances. The story of Alcestis may be an instance of the substitution of the king’s wife. The position of the king in early society and the peculiar practices and beliefs attaching to it were brought to notice and fully illustrated by Sir J.G. Frazer. The argument as to the clan and the veneration of the domestic animal follows that outlined by the late Professor Robertson Smith in The Religion of the Semites.

Some other instances of the communal eating of grain or other food as a sacramental rite and bond of union have been given in the articles. Thus at a Kabīrpanthi Chauka or religious service the priest breaks a cocoanut on a stone, and the flesh is cut up and distributed to the worshippers with betel-leaf and sugar. Each receives it on his knees, taking the greatest care that none falls on the ground. The cocoanut is commonly regarded by the Hindus as a substituted offering for a human head. The betel-leaves [181]which are distributed have been specially consecrated by the head priest of the sect, and are held to represent the body of Kabīr.227

Similarly, Guru Govind Singh instituted a prasād or communion among the Sikhs, in which cakes of flour, butter and sugar are made and consecrated with certain ceremonies while the communicants sit round in prayer, and are then distributed equally to all the faithful present, to whatever caste they may belong. At a Guru-Māta or great council of the Sikhs, which was held at any great crisis in the affairs of the state, these cakes were laid before the Sikh scriptures and then eaten by all present, who swore on the scriptures to forget their internal dissensions and be united. Among the Rājpūts the test of legitimacy of a member of the chief’s family was held to depend on whether he had eaten of the chief’s food. The rice cooked at the temple of Jagannāth in Orissa may be eaten there by all castes together, and, when partaken of by two men together, is held to establish a bond of indissoluble friendship between them.

Members of several low castes of mixed origin will only take food with their relatives, and not with other families of the caste with whom they intermarry.228 The Chaukhutia Bhunjias will not eat food cooked by other members of the same community, and will not take it from their own daughters after the latter are married. At a feast among the Dewars uncooked food is distributed to the guests, who cook it for themselves; parents will not accept cooked food either from married sons or daughters, and each family with its children forms a separate commensal group. Thus the taking of food together is a more important and sacred tie than intermarriage. In most Hindu castes a man is not put out of caste for committing adultery with a woman of low caste, but for taking cooked food from her hands; though it is assumed that if he lives with her openly he must necessarily have accepted cooked food from her. Opium and alcoholic liquor or wine, being venerated on account of their intoxicating qualities, were sometimes regarded [182]as substitutes for the sacrificial food and partaken of sacramentally.229

An important class of communal meals remaining for discussion consists in the funeral feasts. The funeral feast seems a peculiar and unseasonable observance, but several circumstances point to the conclusion that it was originally held in the dead man’s own interest. He or his spirit was indeed held to participate in the feast, and it seems to have been further thought that unless he did so and ate the sacred food, his soul would not proceed to the heaven or god, but would wander about as an unquiet spirit or meet with some other fate. Many of the lower Hindu castes, such as the Kohlis and Bishnois, take food after a funeral, seated by the side of the grave. This custom is now considered somewhat derogatory, perhaps in consequence of a truer realisation of the fact of death. At a Baiga funeral the mourners take one white and one black fowl to a stream and kill and eat them there, setting aside a portion for the dead man. The Gonds also take their food and drink liquor at the grave. The Lohārs think that the spirit of the dead man returns to join in the funeral feast. Among the Telugu Koshtis the funeral party go to the grave on the fifth day, and after the priest has worshipped the image of Vishnu on the grave, the whole party take their food there. After a Panka funeral the mourners bathe and then break a cocoanut over the grave and distribute it among themselves. On the tenth day they go again and break a cocoanut, and each man buries a little piece of it in the earth over the grave. Among the Tameras, at the feast with which mourning is concluded, a leaf-plate containing a portion for the deceased is placed outside the house with a pot of water and a burning lamp to guide his spirit to the food. On the third day after death the Kolhātis sometimes bring back the skull of a corpse and, placing it on the bed, offer to it powder, dates and betel-leaves, and after a feast lasting for three days it is again buried. It is said that the members of the Lingāyat sect formerly set up the corpse in their midst at the funeral feast and sat round it, taking their food, but the custom is not known to exist at present. [183]Among the Bangalas, an African negro tribe, at a great funeral feast lasting for three days in honour of the chief’s son, the corpse was present at the festivities tied in a chair.230

Thus there seems reason to suppose that the caste-tie of the Hindus is the same as that which united the members of the city-states of Greece and Italy, that is the eating of a sacramental food together. Among the Vedic Aryans that country only was considered pure and fit for sacrifice in which the Aryan gods had taken up their residence.231 Hindustān was made a pure country in which Aryans could offer sacrifices by the fact that Agni, the sacrificial god of fire, spread himself over it. But the gods have changed. The old Vedic deities Indra, the rain-god, Varuna, the heaven-god, the Marūts or winds, and Soma, the divine liquor, have fallen into neglect. These were the principal forces which controlled the existence of a nomad pastoral people, dependent on rain to make the grass grow for their herds, and guiding their course by the sun and stars. The Soma or liquor apparently had a warming, exhilarating effect in the cold climate of the Central Asian steppes, and was therefore venerated. Since in the hot plains of India abstinence from alcoholic liquor has become a principal religious tenet of high-caste Hindus, Soma is naturally no more heard of. Agni, the fire-god, was also one of the greatest deities to the nomads of the cold uplands, as the preserver of life against cold. But in India, except as represented by the hearth, for cooking, little regard is paid to him, since fires are not required for warmth. New gods have arisen in Hinduism. The sun was an important Vedic deity, both as Mitra and under other names. Vishnu as the sun, or the spirit of whom the sun is the visible embodiment, has become the most important deity in his capacity of the universal giver and preserver of life. He is also widely venerated in his anthropomorphic forms of Rāma, the hero-prince of Ajodhia and leader of the Aryan expedition to Ceylon, and Krishna, the divine cowherd, perhaps some fabled hero sprung from the indigenous tribes. Siva is the mountain-god of the Himalayas and a moon-deity, [184]and in his character of god of destruction the lightning and cobra are associated with him. But he is really worshipped in his beneficent form of the phallic emblem as the agent of life, and the bull, the fertiliser of the soil and provider of food. Devi, the earth, is the great mother goddess. Sprung from her are Hanumān, the monkey-god, and Ganpati, the elephant-god, and in one of her forms, as the terrible goddess Kāli, she is perhaps the deified tiger.232 Lachmi, the goddess of wealth, and held to have been evolved from the cow, is the consort of Vishnu. It was thus not the god to whom the sacrifice was offered, but the sacrifice itself that was the essential thing, and participation in the common eating of the sacrifice constituted the bond of union. In early times a sacrifice was the occasion for every important gathering or festivity, as is shown both in Indian history and legend. And the caste feasts above described seem to be the continuation and modern form of the ancient sacrifice.

The Roman population, as already seen, consisted of a set of clans or gentes. The clans were collected in tribal groups such as the curia, but it does not appear that these latter were endogamous. The rite which constituted a Roman citizen was participation in the Suovetaurilia, the communal sacrifice of the domestic animals, the pig, the ram, and the bull. Since all the Roman citizens at first lived in a comparatively small area, they were all able to be present at the sacrifice. The other states of Greece and Italy had an analogous constitution, as stated by M. Fustel de Coulanges. It may be supposed that the Aryans were similarly divided into clans and tribes. The word visha, the substantive root of Vaishya, originally meant a clan.233 But as pointed out by M. Senart, they did not form city-states in India, but settled in villages over a large area of country. Their method of government was by small states under kings, and probably they had a kind of national constitution, of which the king was the centre and embodiment. But these states gradually lost their individuality, [185]and were merged in large empires, where the king could no longer be the centre of the state or of the common life of his people, nor perform a sacrifice at which they could all be present, as the Roman kings did. This religious idea of nationality, based on participation in a common sacrifice, was the only one which existed in early times. Thus apparently the Aryans retained their tribal constitution instead of expanding it into a national one, and the members of clans within a certain local area gathered for a communal sacrifice. But there was a great class, that of the Sūdras or indigenous inhabitants, who could not join in the sacrifices at all. And between the Sūdras and the Vaishyas or main body of the Aryans there gradually grew up another mixed class, which also could not properly participate in them. The priests and rulers, Brāhmans and Kshatriyas, tended to form exclusive bodies, and in this manner a classification by occupation gradually grew up, the distinction being marked by participation in separate sacrificial feasts. The cause which ultimately broke down the religious distinctions of the Roman and Greek states was the development of a feeling of nationality. In the common struggle for the preservation of the city the prejudices of the patricians weakened, and after a long internal conflict, the plebeians were admitted to full rights of citizenship. The plebeians were employed as infantry in the Roman armies, while the patricians rode, and the increased importance of infantry in war was one great cause of the improvement in the position of the plebeians.234 In India, in the absence of any national feeling, and with the growth of a large and powerful priestly order, religious barriers and prejudices became accentuated rather than weakened. The class distinctions grew more rigid, and gradually, as the original racial line of cleavage was fused by intermarriage and the production of groups of varying status, these came to arrange themselves on a basis of occupation. This is the inevitable and necessary rule in all societies whose activities and mode of life are at all complicated. Racial distinctions cannot be preserved unless in the most exceptional cases, where they are accentuated by [186]the difference of colour, and such a moral and social gulf as that which exists between the whites and negroes in North America. In primitive society there is no such mental cleavage to render the idea of fusion abhorrent to the superior race; the bar is religious, and while it places the inferior race in a despised and abject position, there is no prohibition of illicit unions nor any such moral feeling or principle as would tend to restrict them. The ideas of the responsibilities and duties of parentage in connection with heredity, or the science of eugenics, are entirely modern, and have no place at all in ancient society. As racial and religious distinctions fade away, and social progress takes place, a fresh set of divisions by wealth and occupation grows up. But though this happened also in the Greek and Italian cities, the old religious divisions were not transferred to the new occupational groups, but fell slowly into abeyance, and the latter assumed the simply social character which they have in modern communities. The main reason for the obliteration of religious barriers, as already stated, was the growth of the idea of nationality and the public interest. But in India the feeling of nationality never arose. The Hindu states and empires had no national basis, since at the period in question the only way in which the idea of nationality could be conceived, was by participation of the citizens in a common sacrifice, and this participation is only possible to persons living in a small local area. Hence Hindu society developed on its own lines independently of the form of government to which it was subject, and in the new grouping by occupation the old communal sacrifices were preserved and adapted to the fresh divisions. The result was the growth of the system of occupational castes which still exists. But since the basis of society was the participation of each social group in a communal meal, the group could not be extended to take in persons of the same occupation over a large area, and as a result the widely ramified system of subcastes came into existence. The subcaste or commensal group was the direct evolutionary product of the pre-existing tribe. Its size was limited by the fact that its members had to meet at the periodical sacrificial feasts, by which their unity and the tie which bound them [187]together was cemented and renewed. As already seen, when members of a subcaste migrated to a fresh local area, and were cut off from communication with those remaining behind, they tended as a rule to form a fresh endogamous and commensal group. Since the tie between the members of the subcaste was participation in a sacrificial meal of grain cooked with water, and as this food was held to be sacred, the members of the subcaste came to refuse to eat it except with those who could join in the communal feast; and as the idea gradually gained acceptance, that a legitimate child must be the offspring of a father and mother both belonging to the commensal group, the practice of endogamy within the subcaste became a rule.

Since all the citizens of the Roman State participated in a common sacrifice, they might be considered as a single caste, or even a subcaste or commensal group. The Hindu castes have a common ceremony which presents some analogy to that of the Roman state. They worship or pay homage once or twice a year to the implements of their profession. The occasions for this rite are usually the Dasahra festival in September and the fast after the Holi festival in March. Both these are festivals of the goddess Devi or Mother Earth, when a fast is observed in her honour, first before sowing the spring crops and secondly before reaping them. On each occasion the fast lasts for nine days and the Jawaras or pots of wheat corresponding to the Gardens of Adonis are sown. The fasts and festivals thus belong primarily to the agricultural castes, and they worship the earth-mother, who provides them with subsistence. But the professional and artisan castes also take the occasion to venerate the implements of their profession. Thus among the Kasārs or brass-workers, at the festival of Mando Amāwas or the new moon of Chait (March), every Kasār must return to the community of which he is a member and celebrate the feast with them. And in default of this he will be expelled from the caste until the next Amāwas of Chait comes round. They close their shops and worship the implements of their profession on this day. The rule is thus the same as that of the Roman Suovetaurilia. He who does not join in the sacrificial feast ceases to be a member of the community. [188]And the object of veneration is the same; the Romans venerated and sacrificed the domestic animals which in the pastoral stage had been their means of subsistence. The Kasārs and other occupational castes worship the implements of their profession which are also their means of livelihood, or that which gives them life. Formerly all these implements were held to be animate, and to produce their effect by their own power and volition. The Nats or acrobats of Bombay say that their favourite and only living gods are those which give them their bread: the drum, the rope and the balancing-pole. The Murha or earth-digger invokes the implements of his trade as follows: “O, my lord the basket, my lord the pickaxe shaped like a snake, and my lady the hod! Come and eat up those who do not pay me for my work!” Similarly the Dhīmar venerates his fishing-net, and will not wear shoes of sewn leather, because he thinks that the sacred thread which makes his net is debased if used for shoes. The Chamār worships his currier’s knife; the Ghasia or groom his horse and the peg to which the horse is secured in the stable; the Rājpūt his horse and sword and shield; the writer his inkpot, and so on. The Pola festival of the Kunbis has a feature resembling the Suovetaurilia. On this occasion all the plough-bullocks of the cultivators are mustered and go in procession to a toran or arch constructed of branches and foliage. The bullock of the village proprietor leads the way, and has flaming torches tied to his horns. The bullocks of the other cultivators follow according to the status of each cultivator in the village, which depends upon hereditary right and antiquity of tenure, and not on mere wealth. A Kunbi feels bitterly insulted if his bullocks are not awarded the proper place in the procession. A string across the arch is broken by the leading bullock, and the cattle are then all driven helter-skelter through the arch and back to the village. The rite would appear to be a relic of the communal sacrifice of a bullock, the torches tied to the proprietor’s bullock signifying that he was formerly killed and roasted. It is now said that this bullock is full of magic, and that he will die within three years. The rite may be compared to the needfire as practised in Russia when all the horses of the village were driven between [189]two fires, or through fire, and their bridles thrown into the fire and burnt. The burning of the bridles would appear to be a substitute for the previous sacrifice of the horse.235 The Pola ceremony of the Kunbis resembles the Roman Suovetaurilia inasmuch as all the cultivators participate in it according to their status, just as the rank of Roman citizens was determined by their position at the ceremony. Formerly, if a bull was sacrificed and eaten sacramentally it would have been practically an exact parallel to the Roman rite.

The tribunal for the punishment of caste offences is known as the panchāyat, because it usually consists of five persons (pānch, five). As a rule a separate panchāyat exists for every subcaste over an area not too large for all the members of it to meet. In theory, however, the panchāyat is only the mouthpiece of the assembly, which should consist of all the members of the subcaste. Some castes fine a member who absents himself from the meeting. The panchāyat may perhaps be supposed to represent the hand acting on behalf of the subcaste, which is considered the body. The panchāyat, however, was not the original judge. It was at first the god before whom the parties pleaded their cause, and the god who gave judgment by the method of trial by ordeal. This was probably the general character of primitive justice, and in some of the lower castes the ordeal is still resorted to for decisions. The tribe or subcaste attended as jurors or assessors, and carried out the proceedings, perhaps after having united themselves to the god for the purpose by a sacrificial meal. The panchāyat, having succeeded the god as the judge, is held to give its decisions by divine inspiration, according to the sayings: ‘God is on high and the panch on earth,’ and ‘The voice of the panchāyat is the voice of God.’236 The headship of the panchāyat and the subcaste commonly descends in one family, or did so till recently, and the utmost deference is shown to the person holding it, even though he may be only a boy for the above reason. The offences involving [190]temporary or permanent excommunication from caste are of a somewhat peculiar kind. In the case of both a man and woman, to take food from a person of a caste from whom it is forbidden to do so, and especially from one of an impure caste, is a very serious offence, as is also that of being beaten by a member of an impure caste, especially with a shoe. It is also a serious offence to be sent to jail, because a man has to eat the impure jail food. To be handcuffed is a minor offence, perhaps by analogy with the major one of being sent to jail, or else on account of the indignity involved by the touch of the police. As regards sexual offences, there is no direct punishment for a man as a rule, but if he lives with a low-caste woman he is temporarily expelled because it is assumed that he has taken food from her hands. Sometimes a man and woman of the caste committing adultery together are both punished. A married woman who commits adultery should in the higher and middle castes, in theory at least, be permanently expelled, but if her husband does not put her away she is sometimes readmitted with a severe punishment. A girl going wrong with an outsider is as a rule expelled unless the matter can be hushed up, but if she becomes pregnant by a man of the caste, she can often be readmitted with a penalty and married to him or to some other man. There are also some religious crimes, such as killing a cow or a cat or other sacred domestic animal; and in the case of a woman it is a very serious offence to get the lobe of her ear torn apart at the large perforation usually made for earrings;237 while for either a man or a woman to get vermin in a wound is an offence of the first magnitude, entailing several months’ exclusion and large expenditure on readmission. Offences against ordinary morality are scarcely found in the category of those entailing punishment. Murder must sometimes be expiated by a pilgrimage to the Ganges, but other criminal offences against the person and property are not taken cognisance of by the caste committee unless the offender is sent to jail. Both in its negative and positive aspects the category of offences affords interesting deductions on the basis of the explanation of the caste system already given. The reason why there is [191]scarcely any punishment for offences against ordinary morality is that the caste organisation has never developed any responsibility for the maintenance of social order and the protection of life and property. It has never exercised the function of government, because in the historical Hindu period India was divided into large military states, while since then it has been subject to foreign domination. The social organisation has thus maintained its pristine form, neither influenced by the government nor affording to it any co-operation or support. And the aims of the caste tribunal have been restricted to preserving its own corporate existence free from injury or pollution, which might arise mainly from two sources. If a member’s body was rendered impure either by eating impure food or by contact with a person of impure caste it became an unfit receptacle for the sacred food eaten at the caste feast, which bound its members together in one body. This appears to be the object of the rules about food. And since the blood of the clan and of the caste is communicated by descent through the father under the patriarchal system, adultery on the part of a married woman would bring a stranger into the group and undermine its corporate existence and unity. Hence the severity of the punishment for the adultery of a married woman, which is a special feature of the patriarchal system. It has already been seen that under the rule of female descent, as shown by Mr. Hartland in Primitive Paternity, the chastity of women was as a rule scarcely regarded at all or even conceived of. After the change to the patriarchal system a similar laxity seems to have prevailed for some period, and it was thought that any child born to a man in his house or on his bed was his own, even though he might not be the father. This idea obtained among the Arabs, as pointed out by Professor Robertson Smith in Kinship and Marriage in Early Arabia, and is also found in the Hindu classics, and to some extent even in modern practice. It was perhaps based on the virtue assigned to concrete facts; just as the Hindus think that a girl is properly married by going through the ceremony with an arrow or a flower, and that the fact of two children being suckled by the same woman, though she is not their mother, establishes a tie akin to consanguinity [192]between them, so they might have thought that the fact of a boy being born in a man’s house constituted him the man’s son. Subsequently, however, the view came to be held that the clan blood was communicated directly through the father, to whom the life of the child was solely assigned in the early patriarchal period. And the chastity of married women then became of vital importance to the community, because the lack of it would cause strangers to be born into the clan, which now based its tie of kinship on descent from a common male ancestor. Thus the adultery of women became a crime which would undermine the foundations of society and the state, and as such was sometimes punished with death among communities in the early patriarchal stage. It is this view, and not simply moral principle, which has led to the severe caste penalties for the offence. Some of the primitive tribes care nothing about the chastity of unmarried girls, but punish unfaithful wives rigorously. Among the Māria Gonds a man will murder his wife for infidelity, but girls are commonly unchaste. Another rule sometimes found is that an unmarried girl becoming with child by an outsider is put out of caste for the time. When her child, which does not belong to the caste, has been born, she must make it over to some outside family, and she herself can then be readmitted to the community. Out of the view of adultery as a religious and social offence, a moral regard for chastity is however developing among the Hindus as it has in other societies.

It has been seen that the Sūdras as well as the plebeians were regarded as impure, and the reason was perhaps that they were considered to belong to a hostile god. By their participation in the sacrifice and partaking of the sacrificial food, the Indian Aryans and other races considered that they were not only in fellowship with, but actually a part of the god. And similarly their enemies were part of the substance of a hostile god, whose very existence and contact were abhorrent to their own. Hence their enemies should as far as possible be completely exterminated, but when this was impossible they must dwell apart and not pollute by contact of their persons, or in any other way, the sacred soil on which the gods dwelt, nor the persons of those who [193]became part of the substance of the god by participation in the sacrificial meal. For this reason the plebeians had to live outside the Roman city, which was all sacred ground, and the Sūdras and modern impure castes have to live outside the village, which is similarly sacred as the abode of the earth-goddess in her form of the goddess of the land of that village. For the same reason their contact had to be avoided by those who belonged to the village and were united to the goddess by partaking of the crops which she brought forth on her land. As already seen, the belief existed that the life and qualities could be communicated by contact, and in this case the worshippers would assimilate by contact the life of a god hostile to their own. In the same manner, as shown by M. Salomon Reinach in Cults, Myths and Religions, all the weapons, clothes and material possessions of the enemy were considered as impure, perhaps because they also contained part of the life of a hostile god. As already seen,238 a man’s clothing and weapons were considered to contain part of his life by contact, and since the man was united to the god by partaking of the sacrificial feast, all the possessions of the enemy might be held to participate in the life of the hostile god, and hence they could not be preserved, nor taken by the victors into their own houses or dwellings. This was the offence which Achan committed when he hid in his tent part of the spoils of Jericho; and in consequence Jehovah ceased to be with the children of Israel when they went up against Ai, that is ceased to be in them, and they could not stand before the enemy. Achan and his family were stoned and his property destroyed by fire and the impurity was removed. For the same reason the ancient Gauls and Germans destroyed all the spoils of war or burned them, or buried them in lakes where they are still found. At a later stage the Romans, instead of destroying the spoils of war, dedicated them to their own gods, perhaps as a visible sign of the conquest and subjection of the enemy’s gods; and they were hung in temples or on oak-trees, where they could not be touched except in the very direst need, as when Rome was left without arms after Cannae. Subsequently the spoils were [194]permitted to decorate the houses of the victorious generals, where they remained sacred and inviolable heirlooms.239

In The Religions of India M. Barth defined a Hindu as a man who has a caste: ‘The man who is a member of a caste is a Hindu; he who is not, is not a Hindu.’ His definition remains perhaps the best. There is practically no dogma which is essential to Hinduism, nor is the veneration of any deity or sacred object either necessary or heretical. As has often been pointed out, there is no assembly more catholic or less exclusive than the Hindu pantheon. Another writer has said that the three essentials of a Hindu are to be a member of a caste, to venerate Brāhmans, and to hold the cow sacred. Of the latter two, the veneration of Brāhmans cannot be considered indispensable; for there are several sects, as the Lingāyats, the Bishnois, the Mānbhaos, the Kabīrpanthis and others, who expressly disclaim any veneration for Brāhmans, and, in theory at least, make no use of their services; and yet the members of these sects are by common consent acknowledged as Hindus. The sanctity of the bull and cow is a more nearly universal dogma, and extends practically to all Hindus, except the impure castes. These latter should not correctly be classed as Hindus; the very origin of their status is, as has been seen, the belief that they are the worshippers of gods hostile to Hinduism. But still they must now practically be accounted as Hindus. They worship the Hindu gods, standing at a distance when they are not allowed to enter the temples, perform their ceremonies by Hindu rites, and employ Brāhmans for fixing auspicious days, writing the marriage invitation and other business, which the Brāhman is willing to do for a consideration, so long as he does not have to enter their houses. Some of the impure castes eat beef, while others have abandoned it in order to improve their social position. At the other end of the scale are many well-educated Hindu gentlemen who have no objection to eat beef and may often have done so in England, though in India they may abstain out of deference to the prejudices of their relatives, especially the women. And Hindus of all [195]castes are beginning to sell worn-out cattle to the butchers for slaughter without scruple—an offence which fifty years ago would have entailed permanent expulsion from caste. The reverence for the cow is thus not an absolutely essential dogma of Hinduism, though it is the nearest approach to one. As a definition or test of Hinduism it is, however, obviously inadequate. Caste, on the other hand, regulates the whole of a Hindu’s life, his social position and, usually, his occupation. It is the only tribunal which punishes religious and social offences, and when a man is out of caste he has, for so long as this condition continues, no place in Hinduism. Theoretically he cannot eat with any other Hindu nor marry his child to any Hindu. If he dies out of caste the caste-men will not bury or burn his body, which is regarded as impure. The binding tie of caste is, according to the argument given above, the communal meal or feast of grain cooked with water, and this, it would therefore seem, may correctly be termed the chief religious function of Hinduism. Caste also obtains among the Jains and Sikhs, but Sikhism is really little more than a Hindu sect, while the Jains, who are nearly all Banias, scarcely differ from Vaishnava Hindu Banias, and have accepted caste, though it is not in accordance with the real tenets of their religion. The lower industrial classes of Muhammadans have also formed castes in imitation of the Hindus. Many of these are however the descendants of converted Hindus, and nearly all of them have a number of Hindu practices.

There have not been wanting reformers in Hinduism, and the ultimate object of their preaching seems to have been the abolition of the caste system. The totem-clans, perhaps, supposed that each species of animals and plants which they distinguished had a different kind of life, the qualities of each species being considered as part of its life. This belief may have been the original basis of the idea of difference of blood arising from nobility of lineage or descent, and it may also have been that from which the theory of caste distinctions was derived. Though the sacrificial food of each caste is the same, yet its members may have held themselves to be partaking of a different sacrificial feast and absorbing a different life; just as the sacrificial feasts and the gods of [196]the different Greek and Latin city-states were held to be distinct and hostile, and a citizen of one state could not join in the sacrificial feast of another, though the gods and sacrificial animals might be as a matter of fact the same. And the earth-goddess of each village was a separate form or part of the goddess, so that her land should only be tilled by the descendants of the cultivators who were in communion with her. The severe caste penalties attached to getting vermin in a wound, involving a long period of complete ostracism and the most elaborate ceremonies of purification, may perhaps be explained by the idea that the man so afflicted has in his body an alien and hostile life which is incompatible with his forming part of the common life of the caste or subcaste. The leading feature of the doctrines of the Hindu reformers has been that there is only one kind of life, which extends through the whole of creation and is all equally precious. Everything that lives has a spark of the divine life and hence should not be destroyed. The belief did not extend to vegetable life, perhaps because the true nature of the latter was by then partly realised, while if the consumption of vegetable life had been prohibited the sect could not have existed. The above doctrine will be recognised as a comparatively simple and natural expansion of the beliefs that animals have self-conscious volitional life and that each species of animals consists of one common life distributed through its members. If the true nature of individual animals and plants had been recognised from the beginning, it is difficult to see how the idea of one universal life running through them all could have been conceived and have obtained so large a degree of acceptance. As the effect of such a doctrine was that all men were of the same blood and life, its necessary consequence was the negation of caste distinctions. The transmigration of souls followed as a moral rule apportioning reward and punishment for the actions of men. The soul passed through a cycle of lives, and the location or body of its next life, whether an animal of varying importance or meanness, or a human being in different classes of society, was determined by its good or evil actions in previous lives. Finally, those souls which had been purified of all the gross qualities appertaining to the [197]body were released from the cycle of existence and reabsorbed into the divine centre or focus of life. In the case of the Buddhists and Jains the divine centre of life seems to have been conceived of impersonally. The leading authorities on Buddhism state that its founder’s doctrine was pure atheism, but one may suggest that the view seems somewhat improbable in the case of a religion promulgated at so early a period. And on such a hypothesis it is difficult to understand either the stress laid on the escape from life as the highest aim or the sanctity held to attach to all kinds of animal life. But these doctrines follow naturally on the belief in a divine centre or focus of life from which all life emanates for a time, to be ultimately reabsorbed. The Vaishnava reformers, who arose subsequently, took the sun or the spirit of the sun as the divine source of all life. They also preached the sanctity of animal life, the transmigration of souls, and the final absorption of the purified soul into the divine centre of life. The abolition of caste was generally a leading feature of their doctrine and may have been its principal social aim. The survival of the individual soul was not a tenet of the earlier reformers, though the later ones adopted it, perhaps in response to the growing perception of individuality. But even now it is doubtful how far the separate existence of the individual soul after it has finally left the world is a religious dogma of the Hindus. The basis of Hindu asceticism is the necessity of completely freeing the soul or spirit from all the appetites and passions of the body before it can be reabsorbed into the god. Those who have so mortified the body that the life merely subsists in it, almost unwillingly as it were, and absolutely unaffected by human desires or affections or worldly events, have rendered their individual spark of life capable of being at once absorbed into the divine life and equal in merit to it, while still on earth. Thus Hindu ascetics in the last or perfect stage say, ‘I am God,’ or ‘I am Siva,’ and are revered by their disciples and the people as divine. Both the Buddhists and Jains lay the same stress on the value of asceticism as enabling the soul to attain perfection through complete detachment from the appetites and passions of the body and the cares of the world; and the deduction therefore seems warranted that the end of the [198]perfect soul would be a similar reabsorption in the divine soul.

The caste system has maintained its vigour unimpaired either by the political vicissitudes and foreign invasions of India or by Muhammadan persecution. Except where it has been affected by European education and inventions, Hindu society preserved until recently a remarkably close resemblance to that of ancient Greece and Rome in the classical period. But several signs point to the conclusion that the decay of caste as the governing factor of Indian society is in sight. The freedom in selection of occupation which now obtains appears to strike at the root of the caste system, because the relative social status and gradation of castes is based on their traditional occupations. When in a large number of the principal castes the majority of the members have abandoned their traditional occupation and taken freely to others, the relative status of castes becomes a fiction, which, though it has hitherto subsisted, cannot apparently be indefinitely maintained. The great extension of education undertaken by Government and warmly advocated by the best Indian opinion exercises an analogous influence. Education is free to all, and, similarly, in the careers which it opens to the most successful boys there is no account of caste. Thus members of quite low castes obtain a good social position and, as regards them personally, the prejudices and contempt for their caste necessarily fall into abeyance. The process must, probably, in time extend to general social toleration. The educated classes are also coming to regard the restrictions on food and drink, and on eating and drinking with others, as an irksome and unnecessary bar to social intercourse, and are gradually abandoning them. This tendency is greatly strengthened by the example and social contact of Europeans. Finally, the facilities for travelling and the democratic nature of modern travel have a very powerful effect. The great majority of Hindus of all castes are obliged by their comparative poverty to avail themselves of the cheap third-class fares, and have to rub shoulders together in packed railway carriages. Soon they begin to realise that this does them no harm, and get accustomed to it, with the result [199]that the prejudices about bodily contact tend to disappear. The opinion has been given that the decline of social exclusiveness in England was largely due to the introduction of railway travelling. Taking account of all these influences, and assuming their continuance, the inference may safely be drawn that the life of the Indian caste system is limited, though no attempt can be made to estimate the degree of its vitality, nor to predict the form and constitution of the society which will arise on its decay. [200]

Arya Samāj Religion.—This important reforming sect of Hinduism numbered nearly 250,000 persons in India in 1911, as against 92,000 in 1901. Its adherents belong principally to the Punjab and the United Provinces. In the Central Provinces 974 members were returned. The sect was founded by Pandit Dayānand Sāraswati, a Gujarāti Brāhman, born in 1824. According to his own narrative he had been carefully instructed in the Vedas, which means that he had been made to commit a great portion of them to memory, and had been initiated at an early age into the Saiva sect to which his family belonged; but while still a mere boy his mind had revolted against the practices of idolatry. He could not bring himself to acknowledge that the image of Siva seated on his bull, the helpless idol, which, as he himself observed in the watches of the night, allowed the mice to run over it with impunity, ought to be worshipped as the omnipotent deity.1 He also conceived an intense aversion to marriage, and fled from home in order to avoid the match which had been arranged for him. He was attracted by the practice of Yoga, or ascetic philosophy, and [202]studied it with great ardour, claiming to have been initiated into the highest secrets of Yoga Vidya. He tells in one of his books of his many and extensive travels, his profound researches in Sanskritic lore, his constant meditations and his ceaseless inquirings. He tells how, by dissecting in his own rough way a corpse which he found floating on a river, he finally discerned the egregious errors of the Hindu medical treatises, and, tearing up his books in disgust, flung them into the river with the mutilated corpse. By degrees he found reason to reject the authority of all the sacred books of the Hindus subsequent to the Vedas. Once convinced of this, he braced himself to a wonderful course of missionary effort, in which he formulated his new system and attacked the existing orthodox Hinduism.2 He maintained that the Vedas gave no countenance to idolatry, but inculcated monotheism, and that their contents could be reconciled with all the results of modern science, which indeed he held to be indicated in them. The Arya Samāj was founded in Lahore in 1877, and during the remainder of his life Dayānand travelled over northern India continually preaching and disputing with the advocates of other religions, and founding branches of his sect. In 1883 he died at Ajmer, according to the story of his followers, from the effects of poison administered to him at the instigation of a prostitute against whose profession he had been lecturing.3

Dayānand’s attempt to found a sect which, while not going entirely outside Hinduism, should prove acceptable to educated Hindus desiring a purer faith, appears to have been distinctly successful. The leaders of the Brahmo Samāj were men of higher intelligence and ability than he, and after scrupulously fair and impartial inquiry were led to deny the infallibility of the Vedas, while they also declined to recognise caste. But by so doing they rendered it impossible for a man to become a Brahmo and remain a Hindu, and their movement has made little headway. By retaining the tenet of the divine authority of the Vedas, Dayānand made it possible for educated Hindus to join his sect without absolutely cutting themselves adrift from their old faith. [203]But Dayānand’s contention that the Vedas should be figuratively interpreted, and are so found to foreshadow the discoveries of modern science, will naturally not bear examination. The following instances of the method are given by Professor Oman: “At one of the anniversary meetings of the society a member gravely stated that the Vedas mentioned pure fire, and as pure fire was nothing but electricity, it was evident that the Indians of the Vedic period were acquainted with electricity. A leading member of the sect, who had studied science in the Government college, discovered in two Vedic texts, made up of only eighteen words in all, that oxygen and hydrogen with their characteristic properties were known to the writers of the Rig Veda, who were also acquainted with the composition of water, the constitution of the atmosphere, and had anticipated the modern kinetic theory of gases.”4 Mr. Burn gives the following parallel versions of a verse of the Rig Veda by Professor Max Müller and the late Pandit Guru Datt, M.A., of the Arya Samaj:

Professor Max Müller.—“May Mitra, Varūna, Aryaman, Ayu, Indra, the Lord of the Ribhus, and the Marūts not rebuke us because we shall proclaim at the sacrifice the virtues of the swift horse sprung from the Gods.”

Pandit Guru Datt.—“We shall describe the power-generating virtues of the energetic horses endowed with brilliant properties (or the virtues of the vigorous force of heat) which learned or scientific men can evoke to work for purposes of appliances. Let not philanthropists, noble men, judges, learned men, rulers, wise men and practical mechanics ever disregard these properties.” In fact, the learned Pandit has interpreted horse as horse-power.

Nevertheless the Arya Samāj does furnish a haven for educated Hindus who can no longer credit Hindu mythology, but do not wish entirely to break away from their religion; a step which, involving also the abandonment of caste, would in their case mean the cessation to a considerable extent of social and family intercourse. The present tenets and position of the Arya Samāj as given to Professor Oman by Lāla Lājpat Rai5 indicate that, while tending towards the [204]complete removal of the over-swollen body of Hindu ritual and the obstacles to social progress involved in the narrow restrictions of the caste system, the sect at present permits a compromise and does not require of its proselytes a full abjuration. In theory members of any religion may be admitted to the Samāj, and a few Muhammadans have been initiated, but unless they renounce Islam do not usually participate in social intercourse. Sikhs are freely admitted, and converts from any religion who accept the purified Hinduism of the Samāj are welcome. Such converts go through a simple ceremony of purification, for which a Brāhman is usually engaged, though not required by rule. Those who, as Hindus, wore the sacred thread are again invested with it, and it has also been conferred on converts, but this has excited opposition. A few marriages between members of different subcastes have been carried out, and in the case of orphan girls adopted into the Samāj caste, rules have been set aside and they have been married to members of other castes. Lavish expenditure on weddings is discouraged. Vishnu and Siva are accepted as alternative names of the one God; but their reputed consorts Kāli, Durga, Devi, and so on, are not regarded as deities. Brāhmans are usually employed for ceremonies, but these may also, especially birth and funeral ceremonies, be performed by non-Brāhmans. In the Punjab members of the Samāj of different castes will take food together, but rarely in the United Provinces. Dissension has arisen on the question of the consumption of flesh, and the Samāj is split into two parties, vegetarians and meat-eaters. In the United Provinces, Mr. Burn states, the vegetarian party would not object to employ men of low caste as cooks, excepting such impure castes as Chamārs, Doms and sweepers, so long as they were also vegetarians. The Aryas still hold the doctrine of the transmigration of souls and venerate the cow, but they do not regard the cow as divine. In this respect their position has been somewhat modified from that of Dayānand, who was a vigorous supporter of the Gaoraksha or cow-protection movement.

Again Dayānand enunciated a very peculiar doctrine on Niyoga or the custom of childless women, either married or [205]widows, resorting to men other than their husbands for obtaining an heir. This is permitted under certain circumstances by the Hindu lawbooks. Dayānand laid down that a Hindu widow might resort in succession to five men until she had borne each of them two children, and a married woman might do the same with the consent of her husband, or without his consent if he had been absent from home for a certain number of years, varying according to the purpose for which he was absent.6 Dayānand held that this rule would have beneficial results. Those who could restrain their impulses would still be considered as following the best way; but for the majority who could not do so, the authorised method and degree of intimacy laid down by him would prevent such evils as prostitution, connubial unfaithfulness, and the secret liaisons of widows, resulting in practices like abortion. The prevalence of such a custom would, however, certainly do more to injure social and family life than all the evils which it was designed to prevent, and it is not surprising to find that the Samāj does not now consider Niyoga an essential doctrine; instead of this they are trying in face of much opposition to introduce the natural and proper custom of the remarriage of widows. The principal rite of the Samāj is the old Hom sacrifice of burning clarified butter, grain, and various fragrant gums and spices on the sacred fire, with the repetition of Sanskrit texts. They now explain this by saying that it is a sanitary measure, designed to purify the air.

The Samāj does not believe in any literal heaven and hell, but considers these as figurative expressions of the state of the soul, whether in this life or the life to come. The Aryas therefore do not perform the shrādhh ceremony nor offer oblations to the dead, and in abolishing these they reduce enormously the power and influence of the priesthood.

The above account indicates that the Arya Samāj is tending to become a vaguely theistic sect. Its religious observances will probably fall more and more into the background, and its members will aspire to observe in their conduct the code of social morality obtaining in Europe, and to regulate their habit of life by similar considerations [206]of comfort and convenience. Already the principal aims of the Samāj tend mainly to the social improvement of its members and their fellow-Indians. It sets its face against child-marriage, and encourages the remarriage of widows. It busies itself with female education, with orphanages and schools, dispensaries and public libraries, and philanthropic institutions of all sorts.7 Its avowed aim is to unite and regenerate the peoples of Aryavārrta or India.

As one of its own poets has said:8

Its principal educational institutions are the Dayānand Anglo-Vedic College at Lahore and the Anglo-Vedic School at Meerut, a large orphanage at Bareilly, smaller ones at Allahabad and Cawnpore, and a number of primary schools. It employs a body of travelling teachers or Upadeshaks to make converts, and in the famine of 1900 took charge of as many famine orphans as the Local Governments would entrust to it, in order to prevent them from being handed over to Christian missionaries. All members of the Samāj are expected to contribute one per cent of their incomes to the society, and a large number of them do this. The Arya Samāj has been accused of cherishing political aims and of anti-British propaganda, but the writers quoted in this article unite in acquitting it of such a charge as an institution, though some of its members have been more or less identified with the Extremist party. From the beginning, however, and apparently up to the present time, its religious teaching has been directed to social and not to political reform, and so long as it adheres to this course its work must be considered to be useful and praiseworthy. Nevertheless some danger may perhaps exist lest the boys educated in its institutions may with youthful intemperance read into the instruction of their teachers more than it is meant to convey, and divert exhortations for social improvement and progress to political ends.

The census of 1911 showed the Arya Samāj to be in a flourishing and progressive condition. There seems good [207]reason to suppose that its success may continue, as it meets a distinct religious and social requirement of educated Hindus. Narsinghpur is the principal centre of the sect in the Central Provinces, and here an orphanage is maintained with about thirty inmates; the local members have an āta fund, to which they daily contribute a handful of flour, and this accumulates and is periodically made over to the orphanage. There is also a Vedic school at Narsinghpur, and a Sanskrit school has been started at Drūg.9 [208]