Kathasaritsagara (the Ocean of Story)

by Somadeva | 1924 | 1,023,469 words | ISBN-13: 9789350501351

This is the English translation of the Kathasaritsagara written by Somadeva around 1070. The principle story line revolves around prince Naravāhanadatta and his quest to become the emperor of the Vidhyādharas (‘celestial beings’). The work is one of the adoptations of the now lost Bṛhatkathā, a great Indian epic tale said to have been composed by ...

Appendix 1.4 - Sacred Prostitution

The story of Rūpiṇikā (p. 138 et seq.) is laid in “a city named Mathurā, the birthplace of Kṛṣṇa.” The lady herself is described as a courtesan who at the time of worship went into the temple to perform her duty.

From this passage it is quite clear that Rūpiṇikā combined the professions of prostitution and temple servant, which latter consisted chiefly in dancing, fanning the idol and keeping the temple clean. She was, in fact, a dēva-dāsī, or “handmaid of the god.” As we shall see in the course of this appendix, the name applied to these so-called “sacred women” varied at different times and in different parts of India.

Mathurā is the modern Muttra, situated on the right bank of the Jumna, thirty miles above Agra. From at least 300 B.C. (when Megasthenes wrote) it had been sacred to Kṛṣṇa, and we hear from reliable Chinese travellers that in A.D. 400 and 650 it was an important centre of Buddhism and at a later date again became specially associated with the worship of Kṛṣṇa, owing to the fact that Mathurā was the scene of the adventures and miracles of his childhood as described in the Viṣṇu Purāṇa. Thus Mathurā has always been one of the most sacred spots in Hindu mythology.[1]

It has suffered from the Mohammedan invaders more than any city of Northern India, or nearly so, for it was first of all sacked in 1017-1018 by the Sultan Mahmud of Ghazni, and again in 1500 by Sikander Lodi, in 1636 by Ṣāh Jahān, in 1669-1670 by Aurangzēb, by whose commands the magnificent temple of Këśavadēva was levelled to the ground, and by Atimad Ṣāh in 1756. By this time every temple, image and shrine had been destroyed and a large part of the population had embraced Mohammedanism. The history of Mathurā is typical of what has occurred in many cities of Northern India, and I consider it is an important factor in the explanation of why sacred prostitution is much more developed in Southern India.

At the date when Somadeva wrote the city must have recovered from its first sacking and the religious life have been assuming its normal course. It was after our author’s day that the systematic and thorough destruction began, and in consequence we hear less about Hindu temples of Northern India.

In view of the anthropological importance of the connection of religion and prostitution, and of the interesting ritual, customs and ceremonies which it embodies, I shall endeavour to lay before my readers what data I have been able to collect, with a few suggestions as to the possible explanation of the curious institution of the dēva-dāsīs.


Ancient India

Owing to the lack of early historical evidence it is impossible to say to what extent sacred prostitution existed in ancient India.

Even in modern times it is often hard to differentiate between secular and sacred prostitution, while, through the clouds of myth and mystery which cover the dawn of Indian history, any distinction must be looked upon as little more than conjecture. In common with so many other parts of the world secular prostitution in India dates from the earliest times and is mentioned in the Ṛg-Veda, where terms meaning “harlot,” “son of a maiden,” “son of an unmarried girl,” etc., occur. In the Vājasaneyi Saṃhitā it seems to be recognised as a profession,[2] while in the law-books the prostitute is regarded with disfavour. (Manu, ix, 259 ; iv, 209, 211, 219, 220 ; v, 90.) In the Buddhist age Brāhmans were forbidden to be present at displays of dancing or music, owing to their inseparable connection with prostitution; yet on the other hand we see in the Jātakas (tales of the previous births of the Buddha) that prostitutes were not only tolerated, but held in a certain amount of respect.[3]

We also hear of the great wealth of some of the women and the valuable gifts made to the temples, which reminds us of similar donations among the ἑταῖραι of ancient Greece. In his article on “Indian Prostitution” in Hastings’ Encyclopædia of Religion and Ethics (vol. x, p. 407) W. Crooke quotes Somadeva as saying that prostitutes are occasionally of noble character and in some cases acquire enormous wealth. He also gives other references apart from those already quoted.

As literary historical evidence on the subject under discussion is so scarce, the discovery in 1905 of a work on Hindu polity was of the utmost importance. It is known as the Arthaśāstra, and gives full details of the social, administrative, fiscal and land systems of the Maurya age. The author is Kauṭilya (Chāṇakya, or Viṣṇugupta), who wrote about 300 B.C.[4] Book II, chap. xxvii, deals with the duties of the superintendent of prostitutes {gaṇikās), who held a highly paid post at the Court of Candragupta. The women enjoyed a privileged position and held the royal umbrella, fan and golden pitcher. They were, however, subject to strict official control, and Kauṭilya gives a long list of penalties for any breach of the regulations—for instance, a gaṇikā who refused her favours to anyone whom the king might choose received a thousand lashes with a whip or else had to pay five thousand paṇas. A further clause states that all the rules prescribed for the gaṇikās are also to apply to dancers, actors, singers, musicians, pimps, etc. There is no mention of temples, but the fact that the dancer, musician and prostitute are all put on the same basis is important in attempting to trace the history of sacred prostitution.

The corruption of the Court at this period is partly shown by the fact that every gaṇikā had to pay to the government each month the amount of two days’ earnings. They were, moreover, sometimes used as secret service agents and acquired position and wealth.

We shall see later that a similar state of affairs existed at the great city of Vijayanagar in the sixteenth century.


The Christian Era (First Eleven Centuries)

In the first eleven centuries of the Christian era more attention seems to have been paid to what we may politely call the Science of Erotics, and many such works were written.[5] Very few, however, are now extant, and it is of interest to note that those which do exist usually mention numerous other similar writings from which they have largely drawn. In most cases they deal in all seriousness with some quite trivial point (such as the best way for a courtesan to rid herself of a lover whose wealth is nearly spent) by listing the various opinions of previous writers and then giving their own opinion as the most acceptable.

It was a method used in 300 b.c. by Kauṭilya, and again by Vātsyāyana, who was the earliest and most important erotic writer of the Christian era. His work, the Kāma Sūtra, dates from about a.d. 250, and has been translated into most European languages, including English.[6] Although Vātsyāyana devotes a whole book (six chapters) to courtesans, there is no direct reference to sacred prostitution. He mentions, however, dancing, singing and the playing of musical instruments as among the chief requirements not only for a prostitute, but also for any married woman wishing to keep her husband’s affections. He divides prostitutes into nine classes,[7] the most honourable of which is the gaṇikā, which, as we have already seen, was the name used by Kauṭilya. “Such a woman,” says Vātsyāyana, “will always be rewarded by kings and praised by gifted persons, and her connection will be sought by many people.”

The next work of importance was by Daṇḍin, who ranks among the greatest poets of India. He flourished in the sixth century. Two of his works give a vivid, though perhaps rather exaggerated, picture of the luxury and depravity of his day. The first is the Baśa Kumāra Charitct,[8] or Adventures of the Ten Princes, while the second (whose authorship is doubtful, though sometimes ascribed to Daṇḍin) is the Mṛcchakatika,[9] or Clay Cart, which treats of the courtship and marriage of a poor Brāhman and a wealthy and generous prostitute. Both works are important in our discussion as giving some idea of the social condition of middle and low class life of the sixth century.

A certain passage in the Daéa Kumāra Charita is of special interest as showing how all female accomplishments were to be found in the courtesan, whose education and conversational powers would certainly be more attractive than the uneducated and paltry household chatter of the wife.

The story goes that a famous dancer, who was, of course, also a prostitute, suddenly pretended to feel the desire to become a devotee. She accordingly went to the abode of an ascetic to carry out her purpose.

Soon, however, her mother follows to dissuade her from her intention, and addresses the holy man as follows: —

“Worthy sir, this daughter of mine would make it appear that I am to blame, but, indeed, I have done my duty, and have carefully prepared her for that profession for which by birth she was intended. From earliest childhood I have bestowed the greatest care upon her, doing everything in my power to promote her health and beauty. As soon as she was old enough I had her carefully instructed in the arts of dancing, acting, playing on musical instruments, singing, painting, preparing perfumes and flowers, in writing and conversation, and even to some extent in grammar, logic and philosophy. She was taught to play various games with skill and dexterity, how to dress well, and show herself off to the greatest advantage in public; I hired persons to go about praising her skill and her beauty, and to applaud her when she performed in public, and I did many other things to promote her success and to secure for her liberal remuneration; yet after all the time, trouble and money which I have spent upon her, just when I was beginning to reap the fruit of my labours, the ungrateful girl has fallen in love with a stranger, a young Brāhman, without property, and wishes to marry him and give up her profession, notwithstanding all my entreaties and representations of the poverty and distress to which all her family will be reduced, if she persists in her purpose; and because I oppose this marriage she declares that she will renounce the world and become a devotee.”[10]

It transpires in the course of the tale that the dancing-girl stays with the ascetic, who falls madly in love with her. She leads him to her home and finally to the palace of the king, where he learns to his great consternation that the whole thing was merely the result of a wager between two court beauties. The participation of the king in the joke and his rewarding the winner clearly shows the importance of the courtesan in this age.

Passing on to the eighth century we have Dāmodara-gupta’s Kuṭṭanīmatam, which resembles Vātsyāyana’s Kāma Sūtra. Besides a German translation, it has also been translated into French.[11]

This was followed in the tenth or eleventh centuries by Kalyāna Malla’s Anaṅga-Raṅga, which is a general guide to ars amoris indica. It is very well known in India and has been translated into numerous European languages.[12]

The only other work worthy of mention is Kṣemendra’s Samayamātṛkā. It can best be described as a guide or handbook for the courtesan, but its chief value lies in the fact that the author was a contemporary of Somadeva. His work has been translated into German[13] and French.[14]

The connection between Kṣemendra and Somadeva is strengthened by the fact that, besides being contemporary Kashmirian court poets, they both wrote a great collection of stories from a common source—the Bṛhat-Kathā. Soma-deva’s collection was the Kathā Sarit Sāgara, while that by Kṣemendra was the Bṛhat-Kathā-Mañjarī. The latter work was, however, only a third as long as the former and cannot compare in any way with the Ocean of Story as regards its style, metrical skill and masterly arrangement and handling of the stories. I shall have more to say about Kṣemendra in Vol. X of the present work.

It is practically impossible to say to what extent the above-mentioned works have bearing on sacred prostitution. I have merely endeavoured to acquaint the reader with such literature as exists dealing with the social life of women of these early times. It seems, however, quite safe to assert that from Buddhist times onwards the prostitute, especially the more learned classes, was held in a certain amount of esteem. She was an important factor in the palace and often acquired great wealth. Dancing and singing were among her accomplishments, but to what extent she was connected with temples we are not told. Soon after the twelfth century historical and literary evidence increases and it becomes possible to examine our data under definite geographical headings. Although Southern India yields by far the most material for our discussion, we will begin in the north, and work slowly southwards.


Northern India

In the introductory remarks to this appendix it has been shown to what extent Mathurā suffered from Mohammedan invasion. The whole of Northern India was similarly affected, and the bloody battles, enforced slavery, terrible tortures and complete destruction of Hindu temples and other public buildings during the Mohammedan Sultanate of Delhi (1175-1340) clearly show that the great upheavals so caused made any continual religious practices of the Hindus an impossibility. By 1340 the Sultanate of Delhi was breaking up and in the south Vijayanagar was already a powerful kingdom. I shall have more to say about Vijayanagar in the section on Southern India.

The destruction of the Hindu temples was continued with unabated zeal in the Mogul Empire. In the reign of Akbar (1556-1605) we are told by his most intimate friend, Abu-1 Fazl,[15] that the prostitutes of the realm (who had collected at the capital, and could scarcely be counted, so large was their number) had a separate quarter of the town assigned to them, which was called Shaiṭānpūrah, or Devilsville. A Dāroghah (superintendent) and a clerk were also appointed for it, who registered the names of such as went to prostitutes, or wanted to take some of them to their houses. People might indulge in such connections provided the toll-collectors heard of it. But, without permission, no one was allowed to take dancing-girls to his house.

The celebrated musician Tānsen, who was attached to Akbar’s Court, became a kind of patron saint of dancing-girls. It is believed that chewing the leaves of the tree above Tānsen’s grave at Gwālior imparts a wonderful melody to the voice, and consequently girls make pilgrimages there for that purpose.[16]

In the reigns of the next two Emperors, Jahāṅgīr (1605-1627) and Ṣāh Jahān (1628-1658), the luxury, ostentation, extravagance and depravity increased,[17] and it was not till the reign of Aurangzëb (1659-1707) that any attempt was made to check the ruthless waste which was slowly draining the resources of the country. Aurangzëb was a Mohammedan Puritan who lived and died an ascetic. During his long reign thousands of Hindu temples were demolished by his orders, and every effort was made to wipe out prostitution and everything pertaining thereto.

Khāfī Khān,[18] the historian, tells rather a pathetic story. It appears that Aurangzēb issued public proclamations prohibiting singing and dancing, and at the same time ordered all the dancing-girls to marry or be banished from the kingdom. They did not, however, submit to this treatment without a protest. One Friday as the Emperor was going to the mosque (another account says he was sitting at his audience window) he suddenly saw about a thousand women carrying over twenty highly ornamented biers. Their piercing cries and lamentations filled the air. The Emperor, surprised at such a display of grief, asked the cause of so great sorrow. He was told that Music, the mother of the dancing-girls, was now dead, and they were burying her. “Bury her deep,” cried the unmoved Emperor; “she must never rise again.”

After the death of Aurangzēb there followed an anarchical period which lasted till the advent of the British. During this time the standard of morality among the princes and public men sank lower and lower. Their lives were vicious and cruel in the extreme, and their gross sensuality naturally affected their courts and, through them, the populace. Prostitution had increased to huge dimensions, and appears to have been entirely secular. Thus we see how, partly owing to foreign conquest and partly to the general spread of immorality, the “religious” element in the temple dancers dropped out and they became ordinary prostitutes, who danced when occasion demanded. They would naturally be called upon if any dancing was wanted for a wedding feast or other private entertainment, for dancing and prostitution had been inseparable in India from the earliest times.

In modern accounts of the tribes and castes of Northern India (which are few enough) we find, therefore, practically no mention of temples or sacred prostitution.

Certain castes such as the ṭawāif and gandharb consist entirely of dancers, singers and prostitutes, but their sub-castes are so numerous that it is quite impossible to distinguish or describe them by any definite principle. Details of the ṭawāif and similar castes were given by Crooke[19] in 1896, and when writing on the same subject in 1918[20] he apparently had nothing further to add. The following details are taken from his former work.

The term ṭawāif is a general one, but is chiefly used for Mohammedan girls, while the Hindu branch is usually called pātar, pātur, pāturiyā (from the Sanskrit pātra, an actor). When they are nubile, the pātar girls marry a pīpal tree and then commence their career of prostitution. One of the numerous sub-castes is known as rājkanya, which appears to be the only one whose members actually dance in the Hindu temples. Prostitution is said to be rare among them. The pātars have Kṛṣṇa as their personal god and Śiva, in the form of Mahādēva, as their guardian deity. Among the ṭawāifs the rites are interesting. The girl is taught to dance and sing when about seven or eight years old. At the commencement of her training sweets are offered at a mosque and then distributed among Mohammedan faqlrs. At the first lesson the master receives a present of sweetmeats besides his pay. When the girl reaches puberty and her breasts begin to develop the rite of aṅgiya, or “the assumption of the bodice,” is performed. Certain of the brethren are feasted and the girl is ready for her first paramour. After the price is fixed she goes to him, which rite is known as sir dhaṅkāi, or “the covering of the head.” When she returns after the first visit, the brethren are again given sweetmeats, after which follows the rite of missi, or “blackening of the teeth.” She is dressed like a bride and paraded through the streets, afterwards attending a party with singing and dancing. The teeth cannot be stained until this feast is held, but Crooke says that at Lucknow the rule was relaxed. After the rite of missi the girl ceases to wear the nose-ring, and hence the ceremony is sometimes known as nathnī utārnā, or “the taking-off of the nose-ring.”

Somewhat similar ceremonies exist among the gandharbs, or gandharvs, who take their name from the heavenly musicians who attend the gods at Indra’s Court. In Northern India they are found only in Benares, Allahābād and Ghāzi-pur. They are Hindus of the Vaiṣṇava sect. Gaṇeśa is the patron of the dancing-girls since he is regarded by them as the author of music. They offer him wreaths of flowers and a sweetmeat made of sesamum and sugar every Wednesday. There are also certain gypsy tribes, such as the beḍiyās and naṭs, who are dancers, acrobats and prostitutes. They are divided into a large number of clans whose occupation is, nevertheless, the same. As they have no connection with temple worship, further details here would be superfluous. They have been fully described by B. R. Mitra[21] and W. Crooke.[22]


Central India

As the ancient kingdoms of India were confined either to the North or South, early travellers were naturally drawn to the most important cities, and tell us but little of Central India, especially as regards the religious practices and social conditions of the towns.

The earliest direct reference to the dancing-girls of Central India which I can find is made by the Chinese traveller Chau Ju-Kwa in his work, Chu-fan-cM, dealing with the Chinese and Arab trade of the twelfth and thirteenth centuries.[23] Speaking of Guzerat (p. 92) he mentions

“four thousand Buddhist temple buildings, in which live over twenty thousand dancing-girls who sing twice daily while offering food to the Buddha (i.e. the idols) and while offering flowers.”

He also speaks of similar customs in Cambodia (p. 53). They are here called a-nan, derived from the Sanskrit word ānanda, meaning “joy” or “happiness.”[24]

We hear little more on the subject till the seventeenth century, when the French traveller Jean Baptiste Tavernier[25] made his second journey to the East (1638-1643). In describing Golconda (five miles west of the modern city of Hyderabad) he says there are over 20,000 public women entered in the Daroglia’s [sic] register. They danced before the king every Friday. In the evenings they stood before the doors of their houses and as soon as they lighted a lamp or candle all the drinking-places were opened. No tax was levied on the women, for they were looked upon as the chief cause of the large consumption of tari, which was a Government monopoly. No mention is made of the women dancing in the temples, but from the evidence of other writers it seems very probable they did this in their spare time!

We shall return to Hyderabad (Nizam’s dominions) later when giving the most recent information, but we now pass on to the east coast and examine the evidence given by W. Ward, the Baptist missionary, who wrote at the beginning of the nineteenth century.[26]

He is speaking of the temple of Jagannātha (usually called Puri), in Orissa.

“It is a well-authenticated fact,” he says, “that at this place a number of females of infamous character are employed to dance and sing before the god. They live in separate houses, not in the temple. Persons going to see Jugünnat’hu [sic] are often guilty of criminal actions with these females.”

Then in a note he adds:

“The officiating Brāhmans there continually live in adulterous connection with them.”

Puri is to-day one of the most sacred spots in India. The name Juggernaut, the anglicised corruption of Jagannātha (Lord of the World), is that given to the form of Viṣṇu worshipped there. The legend of the sacred blue-stone image, details of the famous Car Festival and the truth about the suicides under its great wooden wheels have been told by Hunter.[27] The present temple is built in the shape of a pyramid, and is surmounted with the mystic wheel and flag of Viṣṇu. The annual rent-roll of the temple was put at no less than £68,000. Since Ward’s days little has been written on the dēva-dāsī of Central India. Anything of importance was reproduced by R. V. Russell in his work on the tribes and castes of the Central Provinces.[28]

He says:

“When a dancing-girl attains adolescence, her mother makes a bargain with some rich man to be her first consort. Oil and turmeric are rubbed on her body for five days as in the case of a bride. A feast is given to the caste and the girl is married to a dagger, walking seven times round the sacred post with it. Her human consort then marks her forehead with vermilion and covers her head with her head-cloth seven times. In the evening she goes to live with him for as long as he likes to maintain her, and afterwards takes up the practice of her profession. In this case it is necessary that the man should be an outsider and not a member of the kasbi caste, because the quasi-marriage is the formal commencement on the part of the woman of her hereditary trade.... In the fifth or seventh month of the first pregnancy of a kasbi woman 108[29] fried wafers of flour and sugar, known as gūjahu, are prepared, and are eaten by her as well as distributed to friends and relatives who are invited to the house. After this they, in return, prepare similar wafers and send them to the pregnant woman. Some little time before the birth the mother washes her head with gram flour, puts on new clothes, and jewels, and invites all her friends to the house, feasting them with rice boiled in milk, cakes and sweetmeats.”

The term kasbi, derived from the Arabic kasab= prostitution, denotes rather a profession than a caste. The term is only used for Hindus, as is also gāyan. The Mohammedan dancing-girls are known, as in Northern India, by the name of ṭawāif.

In Bengal this class of women become so-called religious mendicants, who join the Vaiṣṇavī or Bairāgī community. They wander about the country, and, under the cloak of religion, carry on a large trade in kidnapping. They are notoriously licentious, and infanticide is apparently common.[30]

The following description of the dress and dancing of the better class of kasbi women is given by Russell.[31]


“are conspicuous by their wealth of jewellery and their shoes of patent leather or other good material. Women of other castes do not commonly wear shoes in the streets. The kasbis are always well and completely clothed, and it has been noticed elsewhere that the Indian courtesan is more modestly dressed than most women. No doubt in this matter she knows her business. A well-to-do dancing-girl has a dress of coloured muslin or gauze trimmed with tinsel lace, with a short waist, long straight sleeves, and skirts which reach a little below the knee, a shawl falling from the head over the shoulders and wrapped round the body, and a pair of tight satin trousers, reaching to the ankles. The feet are bare, and strings of small bells are tied round them. They usually dance and sing to the accompaniment of the tabla, sārangi and majīra. The tabla or drum is made of two half-bowls—one brass or clay for the bass, and the other of wood for the treble. They are covered with goat-skin and played together. The sārangi is a fiddle. The majīra (cymbals) consist of two metallic cups slung together and used for beating time. Before a dancing-girl begins her performance she often invokes the aid of Sārasvati, the Goddess of Music. She then pulls her ear as a sign of remembrance of Tānsen, India’s greatest musician, and a confession to his spirit of the imperfection of her own sense of music. The movements of the feet are accompanied by a continual opening and closing of henna-dyed hands; and at intervals the girl kneels at the feet of one or other of the audience. On the festival of Basant Panchmi, or the commencement of spring, these girls worship their dancing-dress and musical instruments with offerings of rice, flowers and a cocoanut.”

Proceeding southwards we find that in Hyderabad (Nizam’s dominions) the usual term used for Telugu dancing-girls is bogam, although several others, including those with which we are already acquainted, are found. The bogams are divided into two classes, according as to whether they are Hindus or Mohammedans. If they are the former, the titles sāni or nāyaka are attached to their names; if the latter, they are called jān or nāyakan. Siraj Ul Hassan[32] describes them as having been originally attached to the temples of Śiva and Viṣṇu as “servants of the gods,” most of whom now earn their living by dancing, singing and prostitution. The initiation ceremonies of a bogam sāni include the marriage of the girl to an idol of Kṛṣṇa, and those of a bogam jān to a dagger. In the former case a marriage-booth of sixteen pillars is put up at the girl’s house, whither the idol is brought on an auspicious day.

“The girl is made to stand before the idol as if it were the bridegroom, a curtain is held between them and the officiating Brāhman, reciting the Maṅgalashtaka, or marriage stanzas, weds them in the orthodox fashion. The ceremonies that follow correspond in every particular to those of a Kapu or Munnur marriage. On the Nagveli day the girl is seated by the side of the idol and made to offer puja to Gaurl, the consort of Śiva. Betel leaves, areca nuts and kuṅkum (red powder) are distributed to the assembly of dancing-girls, who sing songs, and, after blessing the bride, retire to their houses.”

In the case of a bogam jān when a girl is married to a dagger the ceremony resembles that above described, with the addition that the rite of missi is also performed. It includes not only the blackening of the teeth, as among the ṭawāif of Northern India, but also the tying of a string of glass beads round the neck. Girls thus married are to a certain extent envied, for, as their husband is immortal, they can never become widows—a thing to be avoided at any cost! The bogams belong to both the Vaiṣṇava and Saiva sects. Their chief gods are Kṛṣṇa and Gaṇeśa, and in the light tenth of Aswin (October) they worship their dancing dresses, instruments, etc.[33] Their ranks are recruited to a certain extent from girls who have been vowed to temple service by their parents on their recovery from sickness, or on some other similar occasion when they wish to show gratitude to their gods. The training of the bogams is most thorough and complete.

“Commencing their studies at the early age of seven or eight, they are able to perform at twelve or thirteen years of age and continue dancing till they are thirty or forty years old. Dancing-girls attached to temples are required to dance daily before the idols, while the priests are officiating and offering puja to them: but the majority of these are trained to appear in public, when they are profusely ornamented with gold and jewels and sumptuously dressed in silk and muslin.”[34]

Their dress, mode of dancing and details of accompanying instruments are the same as already described by Russell. Most of their songs are lewd in character, usually relating to the amorous life of Kṛṣṇa.

Turning westwards to Bombay there is in the Ratnāgiri and Kānara districts and in the Sāvantvāḍi State a Śūdra caste in which the men are known as devlis or nāiks, and the women as bhāvins or nāikins. The majority trace their descent from the female servants of the Sāvantvāḍi or Mālvan chiefs who were regularly dedicated to the service of the local gods. Women from other Śūdra castes can become bhāvins by simply pouring oil on their heads from the god’s lamp in the temple. When a bhāvin girl attains puberty she has to under-go a form of marriage known as the śeṣa. The bridegroom is represented by a god from the temple. On an auspicious day Gaṇapati is worshipped and the ceremony of Punyāha - vāchana (holy-day blessing) is performed at the girl’s house, and also in a temple, by the Gurav or Rāul of the temple. The Gurav and other servants of the temple then go in procession to the girl’s house, taking with them a dagger and the mask of the god. The marriage ceremony is performed with the same details as an ordinary marriage, the mask taking the place of the bridegroom. The homa, or marriage sacrifice, is also performed. The ceremony ends with a feast to those assembled, but is frequently dispensed with owing to the expenditure involved. In such cases the young girl performs the worship of Gaṇapati, and dressing herself in her best attire goes to a temple to the beating of drums, accompanied by a party of bhāvins and temple servants, taking in her hands a cocoanut and a packet of sugar. She places the cocoanut and sugar before the image of the god and bows to him. The Gurav and other temple servants then invoke on her the blessings of the god, and the ceremony ends. Her temple duties are confined to sweeping the floor, sprinkling it with fresh cow-dung, and waving the fly-whisk before the god. She practises prostitution promiscuously, and only differs from the secular variety by her being a dēva-dāsī.

It is, however, interesting to note that the bhāvin is not allowed to dance or sing in public. The devlis also serve in the temples, their chief duties being the blowing of horns and trumpets morning and evening. The daughters of bhāvins usually follow their mothers’ calling; if not, they are married to the sons of other bhāvinsi.e. to the devlis.[35]

In the Karnāṭak, Kolhāpur and the States of the Southern Mahrāṭha country the dāsa caste dedicate their men to the temple, and their women only in a lesser degree. Contrary to the usual rule the women so dedicated are not allowed in the temple at all, their duties being only to sweep the temple yard. They live by prostitution.


Southern India

As has already been mentioned, it is in Southern India that the tenets of the Hindu faith have suffered less from the devastating hand of the invader. Consequently details of ritual have become deeply rooted in the minds of the people, so that in many cases we may expect to find earlier and more original forms of any particular custom or ceremony. Furthermore, the love of building innumerable temples and constantly increasing the Hindu pantheon always appears to have been greater in the South. It is here, therefore, that we get much fuller accounts of sacred prostitution, and nearly all the writings of missionaries and travellers have something to say of the dēva-dāsīs of Madras, Mysore or Travancore.

The earliest direct reference to the subject I can find appears in certain Tamil inscriptions dating back to the time of Rājarāja the Great, the most prominent of the Chōḷa monarchs. He came to the throne in a.d. 985 and, like all the Chōḷa kings, was a votary of Śiva. One inscription[36] shows that in a.d. 1004 the chief temple at Tañjore had four hundred tali-cheri-peṇḍugaḷ, or “women of the temple,” attached to it. They lived in the streets surrounding the temple and in return for their service received one or more shares, each of which consisted of the produce of one vēli[37] of land, calculated at 100 kalam of paddy. The whole Chōḷa country was full of temples with dēva-dāsīs in attendance, as is clear from this inscription, which gives a long list of the dancing-girls who had been transferred to the Tañjāvūr (Tañjore) temple. After each name details are added showing from what temple the girl originally came, and the number of shares she was now to receive. Finally the names and shares of the eunuchs, musicians, dancing-masters, singers, parasol-bearers, barbers and other men connected with the temple are given. It is interesting to note that although Rājarāja was a Saiva, the temple girls imported came from both Saiva and Vaiṣṇava temples.

The next mention of the dēva-dāsīs is made by the greatest of mediaeval travellers, Marco Polo. About 1290 he was on the Coromandel coast, and in describing the inhabitants of the “Province of Maabar” (i.e. Tañjore) he says[38]:

“They have certain abbeys in which are gods and goddesses to whom many young girls are consecrated; their fathers and mothers presenting them to that idol for which they entertain the greatest devotion. And when the [monks] of a convent desire to make a feast to their god, they send for all those consecrated damsels and make them sing and dance before the idol with great festivity. They also bring meats to feed their idol withal; that is to say, the damsels prepare dishes of meat and leave it there a good while, and then the damsels all go to their dancing and singing and festivity for about as long as a great Baron might require to eat his dinner. By that time they say the spirit of the idols has consumed the substance of the food, so they remove the viands to be eaten by themselves with great jollity. This is performed by these damsels several times every year until they are married.

“The reason assigned for summoning the damsels to these feasts is, as the monks say, that the god is vexed and angry with the goddess, and will hold no communication with her; and they say that if peace be not established between them things will go from bad to worse, and they never will bestow their grace and benediction. So they make those girls come in the way described, to dance and sing, all but naked, before the god and the goddess. And those people believe that the god often solaces himself with the society of the goddess.”

As Yule says in a note on this passage (p. 351), Polo does not seem to have quite understood the nature of the institution of the temple dancing-girls, for there was no question of marriage as they were already married—either to the god or to some substitute for a bridegroom such as a sword, dagger or drum. Another point to notice is that Polo describes the girls as “all but naked.” This is in strict contradiction to all accounts which came later; in fact travellers have drawn special attention to the fact that the attraction of the covered body was fully realised by the dancers.

At the beginning of the section on Northern India we saw that by 1340 the Sultanate of Delhi was breaking up and that in the South Vijayanagar was already a powerful kingdom. The story of the foundation of this great Hindu monarchy, formed to check the onrush of the Moslem hordes which were sweeping gradually southwards, makes a thrilling page of Indian history. The glories of the magnificent capital have been fully described by many travellers,[39] but a complete history of the kingdom has yet to be written. It was not until 1565 that Vijayanagar was destroyed by the Moslems, and even then the peninsula to the south of Tuṅgabhadrā remained unaffected as far as its dharma (religion and morality) were concerned. Of the various writers who have described the kingdom the two who give the best description of the social conditions are ‘Abdu-r Razzāq, the ambassador from Persia, and Domingos Paes, the Portuguese. ‘Abdu-r Razzāq explains how the prostitution of the dancing-girls was a great source of revenue to the kingdom; in fact the entire upkeep of the police (12,000 in number) was paid out of the proceeds of the women.

He gives a description of the wealth and splendour of the girls, and says:

“After the time of mid-day prayers, they place at the doors of these houses, which are beautifully decorated, chairs and settees on which the courtesans seat themselves. Every one is covered with pearls, precious stones and costly garments. They are all exceedingly young and beautiful. Each has one or two slave girls standing before her, wiio invite and allure indulgence and pleasure.”

We get, however, a better account from Paes. He is speaking of the idols in the temples, and after giving some description of Gaṇeśa says:

“They feed the idol every day, for they say that he eats; and when he eats, women dance before him who belong to that pagoda, and they give him food and all that is necessary, and all the girls born of these women belong to the temple. These women are of loose character, and live in the best streets that are in the city; it is the same in all their cities, their streets have the best rows of houses. They are very much esteemed, and are classed among those honoured ones who are the mistresses of the captains; any respectable man may go to their houses without any blame attaching thereto. These women are allowed even to enter the presence of the wives of the king, and they stay with them and eat betel with them, a thing which no other person may do, no matter what his rank may be.”

He also makes special mention of their great wealth:

“Who can fitly describe to you the great riches these women carry on their persons? —collars of gold with many diamonds and rubies and pearls, bracelets also on their arms and on their upper arms, girdles below, and of necessity anklets on the feet. The marvel should be otherwise, namely that women of such a profession should obtain such wealth; but there are women among them who have lands that have been given to them, and litters, and so many maid-servants that one cannot number all their things. There is a woman in this city who is said to have a hundred thousand pardaos, and I believe this from what I have seen of them.”

It seems obvious from the above accounts that in wealthy and powerful kingdoms, such as Vijayanagar was in the fifteenth and sixteenth centuries, secular and “religious” prostitution practically coincide.[40] If the diamonds were replaced by cheap and tawdry jewellery made in Birmingham, ‘Abdu-r Razzāq’s description might almost refer to one of the courtesan streets in the Esbekiya quarter of Cairo or to similar ones in Algiers. He is describing only the “prostitute” part of the girl’s business and makes no mention of her duties in the temple. They certainly must have been quite unimportant, and the powers of their “protectors” could in all probability regulate the amount of “service” in the temple. Paes, on the other hand, speaks of their temple duties, but also says that they live in the best streets.

We saw that in Maurya times, when Candragupta was at the zenith of his power in Pāṭaliputra (circa 300 B.C.), a similar state of affairs prevailed. Again in the early eighteenth century the reaction which occurred after the death of the Puritan Aurangzēb caused an enormous laxity of morals, and in consequence the “temple” part of the dēva-dāsīs entirely dropped out. In the seventeenth and eighteenth centuries travellers gave no detailed descriptions of the dēvi-dāsīs, and we get only scanty mentions in the various works of travel. The chief of these are Linschoten (1598), De Bry (1599), Gouvea (1606), Bernier (1660), Thévenot (1661), Fryer (1673), Wheeler (1701), a writer in Lettres Edificantes (1702), Orme (1770), Sonnerat (1782), and Moor (1794).[41]

At the end of the eighteenth and early nineteenth centuries the accounts become more detailed, the two most reliable of which are those of the Abbé J. A. Dubois and Francis Hamilton (formerly Buchanan). Dubois worked in the Madras Presidency in 1792 and went to Mysore in 1799 to reorganise the Christian community. The outcome of this work was his famous Hindu Manners, Customs and Ceremonies, which was translated into English in 1816 direct from the French MS. His remarks on the dancing-girls are interesting.

He says[42] that at first they were reserved exclusively for the Brāhmans, and proceeds:

“And these lewd women, who make a public traffic of their charms, are consecrated in a special manner to the worship of the divinities of India. Every temple of any importance has in its service a band of eight, twelve, or more. Their official duties consist in dancing and singing within the temple twice a day, morning and evening, and also at all public ceremonies. The first they execute with sufficient grace, although their attitudes are lascivious and their gestures indecorous. As regards their singing, it is almost always confined to obscene verses describing some licentious episode in the history of their gods. Their duties, however, are not confined to religious ceremonies. Ordinary politeness (and this is one of the characteristic features of Hindu morality) requires that when persons of any distinction make formal visits to each other they must be accompanied by a certain number of these courtesans. To dispense with them would show a want of respect towards the persons visited, whether the visit was one of duty or of politeness. [This custom is certainly not observed at the present day.—Beauchamp.]

“These women are also present at marriages and other solemn family meetings. All the time which they have to spare in the intervals of the various ceremonies is devoted to infinitely more shameful practices; and it is not an uncommon thing to see even sacred temples converted into mere brothels. They are brought up in this shameful licentiousness from infancy, and are recruited from various castes, some among them belonging to respectable families. It is not unusual for pregnant women, with the object of obtaining a safe delivery, to make a vow, with the consent of their husbands, to devote the child that they carry in their womb, if it should turn out a girl, to the temple service. They are far from thinking that this infamous vow offends in any way the laws of decency, or is contrary to the duties of motherhood. In fact no shame whatever is attached to parents whose daughters adopt this career.

“The courtesans are the only women in India who enjoy the privilege of learning to read, to dance, and to sing. A well-bred and respectable woman would for this reason blush to acquire any one of these accomplishments. [In these days female education is slowly extending to all classes, and the prejudice which formerly existed no longer applies to women learning to read and sing, though dancing is still restricted to the professional dancing-girls, and is not considered respectable. —Beauchamp.]

“The dēva-dāsīs receive a fixed salary for the religious duties which they perform; but as the amount is small they supplement it by selling their favours in as profitable a manner as possible.”

Like several other writers he mentions the special care taken by the dēva-dāsīs not to expose any part of their body, because they fully realise that the imagination is more easily captivated than the eye. Dubois says in the above extract that they dance “twice a day, morning and evening.” This agrees with the remarks of the Chinese traveller Chau Ju-Kwa of the thirteenth century, but differs from the description to be given by Shortt below.

Francis Hamilton,[43] writing nearly the same time as Dubois, gives a similar account of the dēva-dāsīs. He says, however, that if a girl is pretty she is almost certain to be taken from the temple by some “officer of revenue,” and seldom permitted to return except in his presence. When a dancing-girl grew too old to be attractive she was turned out of the temple without any means of support given her, and for this reason she always tried to get a good-looking daughter to succeed—and support her. Speaking of the temples at Tulava he says: “There prevails a very singular custom, which has given origin to a caste named moylar. Any woman... who is tired of her husband, or who (being a widow, and consequently incapable of marriage) is tired of a life of celibacy, goes to a temple, and eats some of the rice that is offered to the idol. She is then taken before the officers of Government, who assemble some people of her caste to inquire into the cause of her resolution; and, if she be of the Brāhman caste, to give her an option of either living in the temple or out of its precincts. If she choose the former, she gets a daily allowance of rice, and annually a piece of cloth. She must sweep the temple, fan the idol with a Tibet cow’s tail (bos grunniens), and confine her amours to the Brāhmans.... The Brāhmany women who do not choose to live in the temple, and the women of the three lower castes, cohabit with any man of pure descent that they please; but they must pay annually to the temple from one sixteenth to half a pagoda.”

No further information on the dēva-dāsīs appears to have been published till 1868, when Dr John Shortt read a most interesting paper before the Anthropological Society, entitled

“The Bayadère: or, Dancing Girls of Southern India.”[44]

His investigations confirm previous accounts, but owing to advantages gained in his medical capacity he was able to obtain details which the ordinary traveller finds so hard to acquire. He differs from Dubois in saying that the girls dance six times a day, but in turns.

They never marry, and begin a strenuous three-year course of singing and dancing at the early age of five.

“When these girls are attached to pagodas, they receive certain sums as wages, the amount of which is dependent on the worth, sanctity, and popularity of the particular temple which they have joined. The money salary they receive is nominal—seldom exceeding a few annas, and sometimes a rupee or two a month. The chief object in being paid this sum as a salary is to indicate that they are servants of the temple; in addition to this, one or more of them receive a meal a day, consisting merely of a mass of boiled rice rolled into a ball.”

He gives full details of their dress. It differs from that described by Thurston as worn by the girls in Central India.

Instead of tinsel-covered dress with skirts reaching below the knees and tight satin trousers, Shortt says:

“Their dancing dress comprises usually the short jacket or choolee, a pair of string drawers tied at the waist, termed jpyjamas —both these are generally of silk, and a white or coloured wrapper or saree: one end of the saree is wound around the waist, and two, three, or more feet, according to the length, is gathered and inserted into the portion encircling the waist, and permitting of a folding fringe or gathering of the cloth in front, and the other end, taken after the usual native fashion over the left shoulder, descends towards the waist, when the end, or moonthanee, is opened out and allowed to drop in front, one end of it being inserted in the waist on the side, and the other left free. This portion of the saree is usually highly ornamented with golden thread, tinsel, etc.—the free end descends to the middle or lower part of the thighs, the other free end of the saree hanging down towards the legs is now got hold of, passed between the legs and fastened to the tie around the waist at the back, and the whole encircled by a gold or silver waist belt. By this mode of dress a fold of the muslin saree forms a loop round each leg, and descends nearly to the ankles, whilst the gathering hangs in front between the legs free.”

They had their own special laws for adoption and inheritance, and were treated with respect and consideration. At one time their ranks were largely increased by kidnapping, but even in Shortt’s day this was quite a rare occurrence. This was often done by an aged dancer in order to procure a successor and a maintenance.

Once again we see the worst side of a depraved priesthood, for

“as soon as a girl attains maturity, her virginity, if not debauched by the pagoda brāhmins, is sold to outsiders in proportion to the wealth of the party seeking the honour, if such it may be termed, after which she leads a continuous course of prostitution—prostituting her person at random, to all but outcasts, for any trifling sum.”

Details of the musical instruments and dances are given, special attention being drawn to the surprising feats of strength and bodily powers of endurance the girls undergo.

“In what is called the sterria coothoo, athletic feats are performed, resting their hands on the ground and flinging their feet in the air with great rapidity, and thus twirling round and round successively performing various somersaults; lying full length on the ground with their hands and feet resting, contorting, twirling, and twisting their bodies in various ways, or whilst resting on the hands and legs, with their backs to the ground and their chest and abdomen turned upwards, drawing the hands and feet as close together as possible; whilst their bodies are thus arched, they, with their mouths, pick up rupees from the ground. In this arched position, beating time with their hands and feet, they work round and round in a circle. During their performance they join their attendants in the songs that are sung, and regulate the various movements of their bodies to the expressions given vent to in the song.”

In the remainder of his article Shortt confirms what we have already seen—the girls are far more educated than the married women, their songs are lewd, they get most of their wealth outside the temple, they are considered an acquisition in a town and form the chief magnet of Hindu society; a wife considers it honourable for her husband to patronise them, and, finally, they are more sinned against than sinning. This is obviously true, for what chance can a child of five have when everything is arranged for her —probably before her birth! Owing to the wise guidance of British rule female education and enlightenment have made great strides since 1868 and we are likely to hear less and less of the dēva-dāsīs. Secular prostitution always has existed and always will exist, for the simple reason that, where there is a certain and constant demand, so also is there an equally certain and constant supply.

We have now to consider a class of women who, although being sacred prostitutes, are hardly ever dancing-girls. Their existence is due to circumstance alone. Among women of the lower Śūdra castes of Southern India, when there is no son to perform the obsequies of the parents, it is customary to endow a daughter with masculine privileges by dedicating her to a deity. Such a woman is known by the name of basivi. As is often the custom among dēva-dāsīs, girls are frequently dedicated as basivis by promise before their birth, or owing to a vow during illness.

Detailed investigations on the basivis have been carried out by Mr Fawcett[45] in the western part of the Bellary district of Madras, and in the portions of Dharwar and Mysore which adjoin it. Although variations of the dedication ceremony occur in different localities, the following description by Mr Fawcett can be taken as generally representative.

After the girl has been conducted with music to the temple by her parents, she is dressed in new clothes, usually white, and two seers of rice, five dates, five cocoanuts, five[46] betel leaves, and the same number of betel nuts, also turmeric[47] and plantains and areca nuts, a gold tali, a silver bangle, and two silver toe-rings are borne in a tray or basket. On arrival at the temple reverence is made to the idol, and, if he is present, to the guru, or high priest, and he, as the officiating priest, receives a fee and the tray or basket of things, and the ceremony is begun. If the guru is present he orders the priest and disciples who may be present “to bring the god to the girl,” and they proceed with the ceremony. She is conducted to that part of the temple where such ceremonies are generally performed, usually in front of the idol, and is made to sit on a black cambly, or country-made blanket (never on a white one), facing east, right knee raised and right elbow resting on it, head bent and covered. In front of her is spread some rice, on which are placed the kernels of five cocoanuts, one at each corner and one in the centre, and similarly five betel nuts, five pieces of turmeric, five dried dates, and five duḍḍus and a ṭaṅkam in a bran vessel (a duḍḍu=l anna 8 pies, and a ṭaṅkam=5 annas 4 pies). Kankanam, a yellow thread, such as is used in Hindu marriages and once to be used in satis, to which a betel leaf is fastened, is tied on her right wrist by the senior basivi present. A marriage song is then sung by the basivi and married women (not widows), who throw yellow rice over the girl. They put the bangle on her right wrist, and tie the tali, on which is depicted the irāman of Viṣṇu, and which is fastened to a necklace of black beads, round her neck, and they make the girl put on the toe-rings. These marriage tokens, which are worn by Hindu women until their husbands’ death, are worn by the basivi until her own death. She is given, by way of insignia, a cane about three feet long, as a wand, carried in the right hand, and a gopālam, or begging basket, slung on the left arm. She is then branded with a heated brass instrument, with a cakra on the right shoulder, in front, similarly on the left shoulder with a shenk (chank) and over the right breast with a cakra. As well known, these are the emblems of Viṣṇu. The third mark, over the breast, is never done if there is any suspicion the girl is not a virgin. Sometimes girls are dedicated after maturity. It may be mentioned that, if he is present, the guru heats the instruments or holds them a moment ere they are used. After being branded, the girl’s forehead is marked with kunkam, a red powder commonly used in feminine adornment. A seer and a quarter of rice, two dried cocoanuts minus the shells, betel leaves, a few areca nuts, five pieces of turmeric and five dates are then tied in her cloth, in front, below the waist, and she is made to rise, taken thrice round the temple and into the god’s sanctuary, where she prostrates herself before the image.

Alms are distributed, certain sums, determined by the girl’s parents, are given to the officiating priest and to the guru, and the ceremony is concluded by the priest whispering a mantram in the girl’s ear. She is told to be good and think of god “Rāma Kṛṣṇa,” “Govind.” For the next five weeks she is required to beg in the village, carrying her insignia and shouting “Rām! Rām!” “Govind!” as she approaches each house. After this there is the hemm ceremony to mark the girl’s puberty, which corresponds with the garbhādhāna ceremony of the Hindus when the bride is of an age for the fulfilment of marriage. An auspicious day is chosen and fixed on if the parents of the girl are not needy; if they are, they wait until they can find the money or some man who, for the sake of securing the girl, will bear the expenses. The girl is given an oil-bath during the day, and in the evening the initiatory ceremony is repeated, with some additions. A sword with a lime stuck on its point is placed upright beside the novice, and it is held in her right hand. It represents the bridegroom, who in the corresponding ceremony of the Hindu marriage sits on the bride’s right. If the basivi happens to be a dancing-girl the object representing the bridegroom is a drum, and the girl’s insignia consists of a drum and bells. A tray, on which is a kalasyam and a lamp, is then produced and moved thrice in front of the girl from right to left. She rises and, carrying the sword in her right hand, places it in the god’s sanctuary. The ceremony is concluded between nine and ten p.m. The actual religious duties of a basivi are few. They are entirely confined to the temple of her dedication, and consist of fasting on Saturdays, attending the temple for worship, and accompanying processions with her insignia during festivals. Their superior position over married women is due to their bearing the god’s mark on their bodies, and by having no widowhood.

Among the Kakatias, a sect of weavers in Conjeeveram (and perhaps the custom obtains elsewhere), the eldest daughter is always dedicated to a deity, but she does not thereby attain any superior right to property. She is taken to a temple, with rice, cocoanuts, sugar, etc., a plantain leaf is placed on the ground, and on it some raw rice, and on that a brass vessel containing water; mango leaves and darbha grass are put into the vessel, a cocoanut and some flowers are placed on the top of it, and the water is purified by mantrams, and the leaves, grass and water are lightly thrown over the girl. A thread is then tied to her left wrist, and she swallows a pill of the five products of the cow for purification. She is then branded with a cakra on the right shoulder and with a shenk or chank on the left, and her forehead is marked with the god’s irāmam; the priest prays for her, and she distributes alms and presents. A tali, which has been lying at the god’s feet, is then placed on her neck by a senior dancing-girl (there are no basivis there), to whom she makes obeisance. She is given tridham to drink, a piece of cloth is tied on her head, she is decked with flowers and crowned with the god’s cap or mitre, she offers worship through the priest, and is taken home with music. At night she comes to the temple and dances before the idol with bells on her feet. She is not a vestal, and she may ply her music; but she is the god’s, and if not dedicated would soon be cut off from the living; so for her own benefit, and chiefly for the benefit of her family, she is dedicated. To avoid legal complications the public ceremony takes place after puberty.

In Mysore the castes among which the dedication of basivis is common are the Killēkyātas, Madiga, Dombar, Vadda, Beda, Kuruba and Golla. Details will be found in the pamphlets on these castes by H. V. Nañjundayya.[48] There is a certain amount of variation in ceremonies, but the general idea is the same in all cases. In his long article on the dēva-dāsīs Thurston[49] gives interesting samples of petitions presented to a European magistrate or superintendent of police by girls or mothers of girls who are about to become basivis.

One reads as follows:—

“I have got two daughters, aged fifteen and twelve respectively. As I have no male issues, I have got to necessarily celebrate [sic] the ceremony in the temple in connection with the tying of the goddess’s tali to my two daughters under the orders of the guru, in accordance with the customs of my caste. I therefore submit this petition for fear that the authorities may raise any objection (under the Age of Consent Act). I therefore request that the Honourable Court may be pleased to give permission to the tying of the tali to my daughters.”

The most recent account of the dēva-dāsīs is that by Thurston already mentioned. It is drawn mainly from articles in the census reports and gazetteers. Many of the customs have already been discussed in this appendix. There are, however, several important points in the Madras Census Reports for 1901, prepared by Mr Francis, which deserve including.

The profession is not now held in the consideration it once enjoyed.... It is one of the many inconsistencies of the Hindu religion that, though their profession is repeatedly and vehemently condemned by the śāstras, it has always received the countenance of the Church.... At the present day they form a regular caste, having its own laws of inheritance, its customs and rules of etiquette, and its own pañcāyats (councils) to see that all these are followed, and thus hold a position which is perhaps without a parallel in any other country. Dancing-girls, dedicated to the usual profession of the caste, are formally married in a temple to a sword or a god, the tali (marriage badge) being tied round their necks by some men of their caste. It was a standing puzzle to the census-enumerators whether such women should be entered as married in the column referring to civil condition.

Among the dāsīs, sons and daughters inherit equally, contrary to ordinary Hindu usage. Some of the sons remain in the caste, and live by playing music for the women to dance to, and accompaniments to their songs, or by teaching singing and dancing to the younger girls, and music to the boys. These are called naṭṭuvar. Others marry some girl of the caste who is too plain to be likely to be a success in the profession, and drift out of the community. Some of these affix to their names the terms piḷḷai and mudali, which are the usual titles of the two castes (veḷḷāḷa and kaikōla) from which most of the dāsīs are recruited, and try to live down the stigma attaching to their birth. Others join the mēlakkārar, or professional musicians. Cases have occurred in which wealthy sons of dancing-women have been allowed to marry girls of respectable parentage of other castes, but they are very rare. The daughters of the caste, who are brought up to follow the caste profession, are carefully taught dancing, singing, the art of dressing well, and the ars amoris, and their success in keeping up their clientèle is largely due to the contrast which they thus present to the ordinary Hindu housewife, whose ideas are bounded by the day’s dinner and the babies.

The dancing-girl castes and their allies, the mēlakkārar, are now practically the sole repository of Indian music, the system of which is probably one of the oldest in the world. Besides them and the Brāhmans few study the subject....

There are two divisions among the dāsīs, called valaṅgai (right-hand) and iḍaṅgai (left-hand). The chief distinction between them is that the former will have nothing to do with the kammālar (artisans) or any other of the left-hand castes, or play or sing in their houses. The latter division is not so particular, and its members are consequently sometimes known as the kcimmāla dāsīs. Neither division, however, is allowed to have any dealings with men of the lowest castes, and violation of this rule of etiquette is tried by a jpañcāyat of the caste, and visited with excommunication....

Among the kaikōlan musicians of Coimbatore at least one girl in every family should be set apart for the temple service, and she is instructed in music and dancing. At the tāli-tymg ceremony she is decorated with jewels and made to stand on a heap of paddy (unhusked rice). A folded cloth is held before her by two dāsīs, who also stand on heaps of paddy. The girl catches hold of the cloth, and her dancing-master, who is seated behind her, grasping her legs, moves them up and down in time with the music which is played. In the evening she is taken, astride a pony, to the temple, where a new cloth for the idol, the tāli, and other articles required for doing pūjā (worship) have been got ready. The girl is seated facing the idol, and the officiating Brāhman gives the sandal and flowers to her, and ties the tāli, which has been lying at the feet of the idol, round her neck. The tāli consists of a golden disc and black beads. She continues to learn music and dancing, and eventually goes through the form of a nuptial ceremony. The relations are invited on an auspicious day, and the maternal uncle, or his representative, ties a golden band on the girl’s forehead, and, carrying her, places her on a plank before the assembled guests. A Brāhman priest recites mantrams (prayers), and prepares the sacred fire (hōmam). For the actual nuptials a rich Brāhman, if possible, or, if not, a Brāhman of more lowly status, is invited. A Brāhman is called in, as he is next in importance to, and the representative of, the idol. As a dāsī can never become a widow, the beads in her tāli are considered to bring good luck to women who wear them. And some people send the tāli required for a marriage to a dāsī, who prepares the string for it, and attaches to it black beads from her own tali. A dāsī is also deputed to walk at the head of Hindu marriage-processions. Married women do not like to do this, as they are not proof against evil omens, which the procession may meet. And it is believed that dāsīs, to whom widowhood is unknown, possess the power of warding off the effects of inauspicious omens. It may be remarked, en passant, that dāsīs are not at the present day so much patronised at Hindu marriages as in olden times. Much is due in this direction to the progress of enlightened ideas, which have of late been strongly put forward by Hindu social reformers. When a kaikolan dāsī dies, her body is covered with a new cloth removed from the idol, and flowers are supplied from the temple to which she belonged. No pūjā is performed in the temple till the corpse is disposed of, as the idol, being her husband, has to observe pollution.

In Travancore the institution of the dēva-dāsīs affords an interesting comparison with that existing in other parts of India. The following account is taken from data collected by Mr N. S. Aiyer.

While the dāsīs of Kartikappalli, Ambalapuzha and Shertallay belonged originally to the Konkan coast, those of Shenkottah belonged to the Pāṇḍiyan country. But the South Travancore dāsīs are an indigenous class. The female members of the caste are, besides being known by the ordinary name of tēvaḍiyāḷ and dāsī, both meaning “servant of god,” called kuḍikkar, meaning “those belonging to the house” (i.e. given rent free by the Sirkar), and peṇḍukaḷ, or women, the former of these designations being more popular than the latter. Males are called tēvaḍiyan, though many prefer to be known as Nanchināṭ Veḷḷāḷas. Males, like these Veḷḷāḷas, take the title of Piḷḷai. In ancient days dēva-dāsīs who became experts in singing and dancing received the title of Rāyar (king), which appears to have been last conferred in A.D. 1847. The South Travancore dāsīs neither interdine nor intermarry with the dancing-girls of the Tamil-speaking districts. They adopt girls only from a particular division of the Nāyars, the Tamil Padam, and dance only in temples. Unlike their sisters outside Travancore, they do not accept private engagements in houses on the occasion of marriage. The males, in a few houses, marry the Tamil Padam and Padamaṅgalam Nāyars, while some Padamaṅgalam Nāyars and Nanchināṭ Veḷḷāḷas in their turn take their women as wives.

When a dancing-woman becomes too old or diseased, and thus unable to perform her usual temple duties, she applies to the temple authorities for permission to remove her ear-pendants (tōḍu). The ceremony takes place at the palace of the Mahārāja. At the appointed spot the officers concerned assemble, and the woman, seated on a wooden plank, proceeds to unhook the pendants, and places them, with a nazar (gift) of twelve paṇams (coins), on the plank. Directly after this she turns about, and walks away without casting a second glance at the ear-ornaments which have been laid down. She becomes immediately a taikkizhavi, or old mother, and is supposed to lead a life of retirement and resignation. By way of distinction, a dāsī in active service is referred to as āṭum-pātram. Though the ear-ornaments are at once returned to her from the palace, the woman is never again permitted to put them on, but only to wear the pampadam, or antiquated ear-ornament of Tamil Śūdra women. Her temple wages undergo a slight reduction, consequent on her proved incapacity.

In some temples, as at Kēralapuram, there are two divisions of dancing-girls, one known as the muzakkuḍi, to attend to the daily routine, the other as the chirappukuḍi, to serve on special occasions.

The special duties that may be required of the South Travancore dāsīs are:

  1. to attend the two Utsavas at Padmanābhaswāmi’s temple, and the Dusserah at the capital;
  2. to meet and escort members of the royal family at their respective village limits;
  3. to undertake the prescribed fasts for the apamārga ceremony in connection with the annual festival of the temple.

On these days strict continence is enjoined, and they are fed at the temple, and allowed only one meal a day.

The principal deities of the dancing-girls are those to whom the temples, in which they are employed, are dedicated. They observe the new and full moon days, and the last Friday of every month, as important. The Oṇam, Śivarātri, Tai-Pongal, Dīpāvali and Citrapūrṇami are the best recognised religious festivals. Minor deities, such as Bhadrakāli, Yakṣi and Gandharva are worshipped by the figure of a trident or sword being drawn on the wall of the house, to which food and sweetmeats are offered on Fridays. The priests on these occasions are ōccans. There are no recognised headmen in the caste. The services of Brāhmans are resorted to for the purpose of purification, of nampiyans and Śaiva Veḷḷāḷas for the performance of funeral rites, and of gurus on occasions of marriage and for the final ceremonies on the sixteenth day after death.

Girls belonging to this caste may either be dedicated to temple service or married to a male member of the caste. No woman can be dedicated to the temple after she has reached puberty. On the occasion of marriage a sum of from fifty to a hundred and fifty rupees is given to the bride’s house, not as a bride-price, but for defraying the marriage expenses. There is a preliminary ceremony of betrothal, and the marriage is celebrated at an auspicious hour. The guru recites a few hymns, and the ceremonies, which include the tying of the tāli, continue for four days. The couple commence joint life on the sixteenth day after the girl has reached puberty. It is easy enough to get a divorce, as this merely depends upon the will of one of the two parties, and the woman becomes free to receive clothes from another person in token of her having entered into a fresh matrimonial alliance.

All applications for the presentation of a girl to the temple are made to the temple authorities by the senior dancing-girl of the temple, the girl to be presented being in all cases from six to eight years of age. If she is closely related to the applicant no inquiries regarding her status and claim need be made. In all other cases formal investigations are instituted, and the records taken are submitted to the chief revenue officer of the division for orders. Some paddy (rice) and five paṇams are given to the family from the temple funds towards the expenses of the ceremony. The practice at the Sucindram temple is to convene, on an auspicious day, a yoga, or meeting, composed of the Valiya Śrl-kāriyakkar, the Yogattil Potti, the Vattappalli Muttatu, and others, at which the preliminaries are arranged. The girl bathes, and goes to the temple on the morning of the selected day with two new cloths, betel leaves and nuts. The temple priest places the cloths and the tāli at the feet of the image and sets apart one for the divine use. The tāli consists of a triangular bottu, bearing the image of Gaṇeśa, with a gold bead on either side. Taking the remaining cloth and the tāli, and sitting close to the girl, the priest, facing to the north, proceeds to officiate. The girl sits, facing the deity, in the inner sanctuary. The priest kindles the fire, and performs all the marriage ceremonies, following the custom of the Tirukkalyāṇam festival, when Śiva is represented as marrying Pārvatī. He then teaches the girl the Pañcākṣara hymn if the temple is Śaivite, and Aṣṭākṣara if it is Vaiṣṇavite, presents her with the cloth, and ties the tali round her neck. The naṭṭuvan, or dancing-master, instructs her for the first time in his art, and a quantity of raw rice is given to her by the temple authorities. The girl, thus married, is taken to her house, where the marriage festivities are celebrated for two or three days. As in Brāhmanical marriages, the rolling of a cocoanut to and fro is gone through, the temple priest or an elderly dāsī, dressed in male attire, acting the part of the bridegroom. The girl is taken in procession through the streets.

The birth of male children is not made an occasion for rejoicing, and, as the proverb goes, the lamp on these occasions is only dimly lighted. Inheritance is in the female line, and women are the absolute owners of all property earned. When a dancing-girl dies some paddy and five paṇams are given to the temple to which she was attached, to defray the funeral expenses. The temple priest gives a garland, and a quantity of ashes for decorating the corpse. After this a nampiyan, an ōccan, some Veḷḷäḷa headmen and a kuḍik-kāri, having no pollution, assemble at the house of the deceased. The nampiyan consecrates a pot of water with prayers, the ōccan plays on his musical instrument, and the Veḷḷāḷas and kuḍikkāri powder the turmeric to be smeared over the corpse. In the case of temple devotees, their dead bodies must be bathed with this substance by the priest, after which alone the funeral ceremonies may proceed. The kartā (chief mourner), who is the nearest male relative, has to get his whole head shaved. When a temple priest dies, though he is a Brāhman, the dancing-girl on whom he has performed the vicarious marriage rite has to go to his death-bed and prepare the turmeric powder to be dusted over his corpse. The anniversary of the death of the mother and maternal uncle are invariably observed.

The adoption of a dancing-girl is a lengthy ceremony. The application to the temple authorities takes the form of a request that the girl to be adopted may be made heir to both kuḍi and pati —that is, to the house and temple service of the person adopting. The sanction of the authorities having been obtained, all concerned meet at the house of the person who is adopting, a document is executed, and a ceremony, of the nature of the Jātakarma, performed. The girl then goes through the marriage-rite, and is handed over to the charge of the music-teacher to be regularly trained in her profession.

In concluding his article, Thurston gives a number of cases about the initiation, laws of inheritance, etc., which have been argued in the Madras High Court, besides a selection of current proverbs relating to dancing-girls. These will be found on pp. 145-153 of the above-mentioned article.

We have now become acquainted with all the important data on the subject under discussion so far as India is concerned.

In summarising we notice the following points:—

In Vedic times reliable evidence is insufficient to enable us to form any definite conclusion as to the possibility of distinct connection between religion and prostitution.

Although the law-books regarded the latter with disfavour, and in the Buddhist age Brāhmans were not even allowed to hear music or witness dances owing to their inseparable connection with prostitution, yet it appears that the letter of the law was not carried out in any great strictness. This is especially evident when in the collection of the birth-stories of Buddha (the Jātakas) we read of the high esteem in which such women were held, and of the important positions—sometimes even in the king’s palace—which they occupied.

It is quite a feasible suggestion that this State approval of prostitutes may have been, even at this early date, largely due to their taking part (however small) in the ritual at the neighbouring temples. Direct historical evidence of the privileges which these women enjoyed is afforded by Kauṭilya’s Arthaśāstra (circa 300 B.C.), where we learn that, although under strict regulations, the prostitutes often acquired great position and wealth.

In the early Christian era we still find no direct reference to the dēva-dāsī, but literary evidence distinctly refers to dancing as one of the chief accomplishments of the courtesan. After about the twelfth century our evidence becomes more definite and geographical.

In the time of Akbar rules were issued relating to the superintendence of the prostitute dancing-girls, and, as the oppression of the Mohammedans increased, so, in inverse ratio, did the “religious” element in the institution of the dēva-dāsīs become less and less. After the death of the Puritan Aurangzēb the general morality sank to a very low level, and prostitution, now entirely secular, reached huge dimensions.

In modern days the prostitute dancing-castes divide themselves into two branches, according as to whether they are Hindus or Mohammedans. Only one sub-caste, the rāj-kanyā, has any definite connection with the temples. Further evidence shows that there is no system of dev a-dāsīs as there is in the South, which state of things is due mainly to the Mohammedan conquest in earlier days.

As we proceed southwards direct references to the dēva-dāsīs become more common. In Central India we find the system fully developed at Jagannātha, in Orissa, where the sincerity of the worshippers was as undoubted as the viciousness of the priesthood. Thus there existed side by side religion and prostitution. As the latter was recognised and approved by both Church and State, its acceptance by the worshippers of Viṣṇu, who looked to the Brāhman priests for guidance, can be readily understood.

We now come across accounts of the so-called marriage ceremonies of the dēva-dāsīs which attach to them a certain amount of envy, owing to the fact that, as they are married to a god, or an emblem of a divine husband, they can never become widows. This fact and the stamping of the bodies of the women with the symbols of the gods are the chief reasons which cause the dēva-dāsī to be approved by the ordinary married women and resorted to by their husbands.

Although British rule has done much to suppress the element of vice in the institution of the dēva-dāsīs, it is much too deeply rooted to extirpate. We find the ritual still prevalent in parts of Central India and still more so in the South.

It is here that our accounts are much fuller and reliable, and even as early as a.d. 985 we find the system flourishing under the Chōja monarchs. Mediæval travellers confirm these accounts.

It seems clear, however, that when the wealth and splendour of a kingdom reached its height, as in the case of Vijayanagar in the fifteenth and sixteenth centuries, the “service” of the dēva-dāsī became almost entirely confined to the streets, while her temple duties were practically nonexistent.

Farther south the religious observances had been more closely maintained, and travellers of the seventeenth, eighteenth and nineteenth centuries found the temple-women taking a prominent part at all the chief temples. It is obvious to see from the more detailed accounts that here we have the fuller and more developed form of the system of sacred prostitution as compared with what we find farther north.

The privileges of dedicating a girl to the deity are fully realised by the lower Śūdra castes and, as we see by the strange system of basivis, they can actually perform the obsequies of the parents in the place of the son. As the duty to the dead is of such great importance to the Hindu, it can at once be realised that not only are the dedicated prostitutes regarded with favour, but in many cases are entrusted with the performing of the most sacred duties, thus enabling their parents to die in peace.

On the other hand, the status of the dèva-dāsī is not held in the high consideration it once was, and modern education in India has done much to open the eyes of a more enlightened generation.

Surveying the total evidence here collected, the reader naturally asks himself how it was that the sacred and profane became thus united; or, in other words, what was the real origin of “sacred” prostitution. Numerous theories exist as to the true explanation of this strange custom, but none is entirely satisfactory. It will, however, help us in our inquiry to list the chief:

  1. It is a substitute for human sacrifice, being an offering to the deity in order to appease him or to secure blessings for the country in question and its inhabitants.
  2. It is an expiation for individual marriage as a temporary recognition of pre-existing communal marriage.
  3. It springs from the custom of providing sexual hospitality for strangers; and if such hospitality be offered by the mortal wife of a deity, good would be bound to result.
  4. It is a rite to ensure the fruitfulness of the ground and the increase of man and beast on the principle of homoeopathic magic.
  5. It arises from the secular and precautionary practice of destroying a bride’s virginity by someone other than the bridegroom.
  6. It merely represents the licentious worship of a people, subservient to a degraded and vicious priesthood.
  7. It is a part of the phallic worship which existed in India from,early Dravidian times.

All the above theories have been put forward from time to time by men whose opinions have been, or are, respected.

The evidence already laid before the reader shows clearly that most of them are quite insufficient to account for the whole institution of dēva-dāsīs, while others, such as Nos. 5 and 6, have already been disproved. No. 4, supported by Frazer and many other scholars, seems to be feasible, although it certainly does not account for everything.

The above theories have been presented by men who made comparisons, and I feel that the fact is often overlooked that the origin of a certain custom in one part of the world may not necessarily be the same as that of a similar custom in another part of the world.

In speaking of sacred prostitution in Western Asia Frazer[50] says:

“The true parallel to these customs is the sacred prostitution which is carried on to this day by dedicated women in India and Africa.”

This is a sweeping statement to make, especially when we bear in mind how scanty is our knowledge of the early Semitic pantheon, the differences of opinion held by some of our greatest Babylonian scholars, and the lack of reliable historic data of the early Yedic period in India.

We must also remember that the religion, ethics and philosophy of India have been ever changing, and nothing is more inapplicable than to speak of the “changeless East” in this respect.

Our knowledge of the early Dravidian religion of India before it was “taken over” by the Aryan invaders is so slight that it is impossible to make any definite statement with regard to the origin of any particular custom of ritual or religious observance.

In order, however, to enable readers to make their own deductions and to follow up any branch of the subject, I shall give a few notes on sacred prostitution in countries other than India.

Religious prostitution in Western Asia is first mentioned in some of the earliest records of Babylonia, and has also been traced in Syria, Phoenicia, Arabia, Egypt, Greece and Rome. Similar cults also occur in the Far East, Central America, West Africa and other localities to be mentioned later.

The subject is a very extensive one, upon which volumes could be written. The following remarks, therefore, merely deal with it in a very general manner. Care, however, has been taken to provide ample references, so that the student can pursue the subject to any length.

As Mesopotamia was the original home of sacred prostitution, I shall deal with the Babylonian evidence more fully than with that from other localities outside India, about which the classical writers of Rome and Greece have already made us sufficiently familiar.



In discussing the “sacred servants,” or hierodouloi, in ancient Babylonia we can conveniently divide the subject under the two following headings: —

  1. The Code of Hammurabi.
  2. The Epic of Gilgamesh.


1. About 2090 B.C., during the first dynasty of Babylon (which corresponds to the twelfth Egyptian dynasty), Hammurabi set up in the temple of Marduk, the city god, at Babylon, a code of laws embodying the decisions of a long series of judges who were already acquainted with a system of laws probably of Sumerian origin. Babylonian law ran in the name of God, and the temple was naturally a very large factor in the life of the people. It formed an intimate connection between their god and themselves, and their ritual tended to emphasise this fact.

Accordingly their god would dine with them at sacrificial feasts, he would intermarry with them, and would be appealed to as an adviser and helper in times of danger or difficulty. The temple was, moreover, the house of the god and thus was the outward sign of human relations with divine powers. It was also the centre of the country’s wealth, the equivalent of the modern bank. Its wealth was derived partly from the land it owned, which was either leased out or used for cattle-breeding, and partly from dues of various kinds.

The Code of Hammurabi[51] affected the whole realm, and the laws therein applied to every temple, no matter what god or goddess happened to be locally enshrined. Although Marduk was worshipped at Babylon, at Larsa or Sippar it was Shamash, at Erech it was Innini or Ishtar the mother-goddess, in Ur it was Nannar the moon-god, and so on. Each temple had a staff, varying with its size, which in most cases included both male and female hierodouloi in its service.

The priestesses and temple women formed several distinct classes which need some detailed description.

The priestesses were of two kinds, the entu (Nin-An) and the naṭitu (Sal-Me). Both classes were held in respect, and the entu (brides of the god) were looked upon as the highest class in the land. It is not clear if they married mortal husbands or not, anyway no mention of a father is made. The naṭitu were much more numerous and were allowed to marry, but were not expected to bear children, a maid being supplied for this purpose. Both the entu and the naṭitu were wealthy and owned property.

They could either live in the gagum (cloister) adjoining the temple or in their own houses. If they chose the latter they were forbidden, on pain of being burned alive, to own or enter a wine-shop, so great was the prestige the class had to maintain.

A study of the contract-literature of the period seems to make it clear that just as an ordinary well-to-do citizen could have a chief wife and many inferior ones as well as concubines, so also the god would have his chief wife (entu), his many inferior ones (naṭitu) and his concubines (zikru).

This latter class of consecrated women known as zikru or zermashitu came immediately after the two varieties of priestesses already mentioned. They, too, were well-to-do and held in respect. The zikru or “ vowed” woman is not mentioned in religious literature, nor is zermashitu (seed-purifying). Both of these temple harlots could marry and bear children. The zikru appears to be slightly superior to the zermashitu owing to the fact that in the laws relating to the inheritance of property it is stated that if the father of a zikru died and nothing was left her in his will she was to inherit equally with her brothers, but if she was a zermashitu or a ḳadishtu (to be discussed shortly) she received only one-third of a brother’s share.

The ḳadishtu, although classed with the zermashitu as regards the inheriting of property, clearly occupied a subordinate position. Her name means “sacred woman” and is the same as the Biblical ĕdēṣāh (Deut. xxiii, 18). There is no record of her marriage, and her speciality, outside her temple duties, was suckling the children of Babylonian ladies, for which service she received payment, together with a clay tablet recording the contract. Several examples of such tablets can be seen in the British Museum.[52]

Apart from the various temple women already mentioned there were others who were more especially connected with the worship of Ishtar. In the time of Hammurabi the centre of this cult was at Erech, although she had a shrine in the temple of Marduk in Babylon, where, under the name of Sarpanit, she appears in later texts as the wife of Marduk. it is undoubtedly Sarpanit to whom Herodotus refers in his well-known account of the enforced temporary prostitution of every Babylonian woman (i, 199).

In order to understand the cult of the great mother-goddess throughout Western Asia it is necessary to say a few words on the origin of Ishtar. Recent evidence[53] seems to show that Ishtar was not of Semitic Babylonian or even of Sumerian creation, but was a primitive Semitic divinity personifying the force of nature which showed itself in the giving and taking of life. The various functions of Sumerian local goddesses became by absorption merely fresh attributes of Ishtar, the original name sometimes remaining.

Thus we find different cities sacred to different goddesses which are all certain aspects of Ishtar, the great mother-goddess. It follows, therefore, that the characteristics of Ishtar were numerous, for besides being connected with creation of animal and vegetable life and the goddess of sexual love, marriage and maternity, she was also the storm and war goddess and the destroyer of life. It is interesting to compare similar attributes in the male-female (Ardha-nārīśvara) form of Śiva, who was both a creator and a destroyer.

In Erech Ishtar was known as Innini, Innanna or Nanā, and as many hymns originally addressed to Innini are appropriated by Ishtar, she bears, among others, the titles of “Queen of Eanna,” “Queen of the land of Erech.”[54] Her cult extended to all cities of importance in Babylonia and Assyria, and it is in her capacity as goddess of sexual love that she concerns us here.

Her character is clearly represented in numerous hymns, where she is described as “the languid-eyed,” “goddess of desire,” “goddess of sighing,” and refers to herself as “a loving courtesan” and “temple-harlot.”

In one hymn she says:

“I turn the male to the female, I turn the female to the male, I am she who adorneth the male for the female, I am she who adorneth the female for the male.”[55]

In art she is depicted as naked with her sexual features emphasised, or as lifting her robe to disclose her charms.[56] Several statues represent her as offering her breasts; some have been found outside Babylonia— e.g. in Northern Syria and Carchemish.[57]

The names given to the licentious ministrants at the Ishtar temple at Erech were kizrēti (harlot), shamkhāti (joy-maiden), and kharimāti (devoted one). If they differed from the zermashitu and ḳadishtu it is impossible to say exactly what the difference was.

They are thus described in the Legend of Girra:

“ Of Erech, home of Anu and of Ishtar,
The town of harlots, strumpets and hetæræ,
Whose (hire) men pay Ishtar, and they yield their hand.”

We will now pass on to the Epic of Gilgamesh, where further data can be obtained.


2. The Epic of Gilgamesh is one of the most important literary products of Babylonia, and sheds considerable light on the cult of Ishtar. It consists of a number of myths of different ages—some dating back to 2000 B.C. or even earlier —which have all been gathered round the name of Gilgamesh, an early Sumerian ruler of about 4500 B.C.

The Epic is known to us chiefly from a collection of twelve sets of fragments found in the library of Assur-bani-pal, King of Assyria (668-626 B.C.). In the first tablet the goddess Aruru creates a kind of “wild man of the woods,” by name Engidu,[58] to act as a rival to Gilgamesh, whose power and tyranny had begun to be a burden to the people. In order to get Engidu away from his desert home and his beasts, a shamkhāt from Ishtar’s temple is taken to him.

“This woman, when they approached Engidu, opened wide her garments, exposing her charms, yielded herself to his embrace, and for six days and seven nights gratified his desire, until he was won from his wild life.”[59]

In the second tablet the harlot takes him back to Erech, where she clothes and generally looks after him.

He finally meets Gilgamesh, and the next three tablets relate their friendship, quarrels and adventures. The sixth tablet is especially interesting, for here we get a reference to the Ishtar-Tammuz myth which is so inseparable from the great mother-goddess.

After overcoming an enemy named Khumbaba the two friends returned to Erech in triumph. Ishtar asks Gilgamesh to be her husband and promises him all manner of riches and power. He refuses, reminding her of the numerous lovers she has had in the past and what ill luck befell them. In particular he refers to Tammuz, the lover of her youth, whose death she bewails every year. This is, of course, the youthful solar God of the Springtime, who was wooed by the Goddess of Fertility, Ishtar. Each year that Tammuz died Ishtar went to Hades (Sheol) in search of him. The myth has been detailed by many scholars and does not in itself concern us here.[60]

The effects of Ishtar’s descent to Sheol in search of her youthful lover have, however, direct bearing upon our inquiry.

As soon as Ishtar had gone on her annual journey to the underworld, copulation in men and animals ceased. Consequently some remedy had to be sought in order to circumvent such a disastrous state of affairs. Thus arose the necessity for women to play her part as goddess of sexual love and fertility; and to fill this office the “ sacred prostitute 55 was created.

This applies only to the Ishtar cult and not to those cases where priestesses were found in temples dedicated to other deities.

We have seen that in the case of Marduk the god was credited with all human attributes and passions.

To return to Gilgamesh, we find Ishtar very wroth at having her offers of love refused. She sent a bull to kill him, but he destroyed it. Thereupon Ishtar gathered together all her temple women and harlots, and made great outcry and lamentation.[61]

The remaining tablets, containing, among other incidents, the story of the Deluge, do not concern us.

We have seen that at this early period sacred prostitution was fully established and entered into the literature and mythology of the country. Under the male deity the temple harlot plays the part of concubine, while under the female deity she was a kind of “understudy,” always ready to symbolise by her action the purpose of the great mother-goddess.

Without going farther into the cult of Ishtar it will serve our purpose better to move slowly westwards, noting the spread of the worship of a goddess of love and fertility which clearly resembled that of Ishtar. We must not necessarily conclude that whenever we find a mother-goddess it is merely Ishtar transplanted to new soil and given a new name. It seems to be more probable, anyhow in several cases, that local female deities acquired fresh attributes from Ishtar which occasionally became the most prominent features of the cult.


Syria, Phoenicia, Canaan, etc.

In Syria the great mother-goddess was known by the name of Attar or Athar, while at the sacred city of Hierapolis (the modern Membij) in the Lebanon she was called Atar-gatis, a word compounded out of’Atar and’Ate, two well-known Syrian deities. The full etymology of these names has been discussed by L. B. Paton,[62] who gives a large number of useful references.

Our information on the worship at Hierapolis is mainly derived from Lucian’s De Dea Syria, which is considered one of his earliest works, probably written about A.D. 150. Recent researches in Asia Minor and Northern Syria, largely numismatic, show that at the height of the Hittite domination in the fourteenth century B.C. the chief religious cult was very similar to that described by Lucian. There were, however, certain differences. The Hittites worshipped a mated pair, a bull god and a lion goddess, while in later days it was the mother-goddess who became prominent, representing fertility, and (in Phoenicia) the goddess who presided over human birth. Religion in the East adapted itself to changing conditions and the immediate needs of the community.

Thus in Syria the climate and temperament of the people tended to develop the sensuous aspect of the goddess. As the cult became more popular, the rites and festivals became more orgiastic in character. The phallic nature of some of the rites at Hierapolis is described by Lucian (28), where he speaks of two huge phalli, thirty fathoms high, which stood at the door of the temple. Twice every year a man (probably one of the castrated Galli) climbed to the summit from the inside, where he was supposed to hold converse with the gods to ensure the prosperity and fertility of the land.

Speaking of the temple at Byblos, Lucian states that after the termination of the mourning for the loss of Adonis (cf. the Tammuz myth) the men shave their heads and the women who refuse to submit to a similar treatment have to prostitute themselves for a whole day in the temple. The proceeds of their hire paid for a sacrifice to the mother-goddess. The fact that the women were only allowed to be hired by strangers forms a curious relic of the system of exogamy.

Evidence seems to make it practically certain that there was a permanent, besides a temporary, system of religious prostitution at the temples, and Eusebius tells us that matrons as well as maids served the goddess in this manner. Lucian shows that the system of enforced temporary prostitution had been modified, and that a modest woman might substitute a portion of her hair instead of her person. This fact is interesting as showing the belief in the hair possessing a large and important percentage of the owner’s personality. Readers will remember the care with which the savage hides or destroys his hair, nail-clippings, etc., lest an enemy get possession of them and work him harm through their means.

By this passage in Lucian we see that at Byblos (Gebal) the sacrifice of chastity was looked upon as the most personal, and therefore most important, offering a woman could make. If she did not give this, then the next best thing—her hair— would be accepted. No such substitution, however, appears to have been allowed in former days— i.e. before Lucian’s time.

The name given to the great mother-goddess in Phoenicia, Canaan, Paphos, Cyprus, etc., was Ashtart, Ashtoreth or Astarte. Her attributes closely resemble those of Ishtar, for we find her represented as a goddess of sexual love, maternity, fertility and war. Both the Greeks and Phoenicians identified her with Aphrodite, thus showing evidence of her sexual character. As is only natural, the Phoenicians carried this worship into their colonies, and so we read in Herodotus (i, 199), Clement of Alexandria (Protrept, ii), Justin (xviii, 5, 4) and Athenæus (xii, 2) of sacred prostitution closely resembling that in Syria. Special mention is made of male prostitutes at the temple of Kition in Cyprus. They are the same as the ḳādhēsh of Deut. xxiii, 18, 19.

Phoenician inscriptions give evidence of a temple of Ashtart at Eryx in Sicily, while along the coast of North Africa the Semitic mother-goddess became very popular under the names of Ashtart and Tanith.

St Augustine (De Civ. Dei, ii, 4) gives some account of the worship which, when stripped of its oratorical vagueness, points to a system of temporary hierodouloi, very similar to that described by Lucian.

In Arabia the mother-goddess was Al-Lāt or Al-‘Uzzā, whose worship was accompanied by the temporary practice of sacred prostitution. It would be superfluous to magnify examples.

We have seen that the practice spread all over Western Asia and into Europe and Africa. Egypt we have not discussed, but the numerous references given by G. A. Barton in his article, “Hierodouloi,” in Hastings’ Ency. Rel. Eth., vol. vi, pp. 675-676, show that the system can be clearly traced, especially at Thebes.

To sum up our evidence from Western Asia, there appear to be several reasons to which the institution of sacred prostitution owes its origin:

1. The male deity needed concubines like any mortal, thus women imitated at the temples their divine duties.

2. The female deity, being a goddess of fertility, had under her special care the fruitfulness of vegetation as well as of the animal world. Thus she endeavours to hasten on the return of spring. It is only natural that at her temples women should assist in this great work of procreation, chiefly by imitating the functions necessary to procreate. When the goddess was absent in search of spring, the whole duties of the cult would fall on her mortal votaries.

3. Sacrifices of as important and personal nature as possible would be acceptable to such a goddess, and the hopes of prosperity in the land would be increased.

When human passions enter so largely into a ritual, and when the worshippers and ministrants of the goddess are of an excitable and highly temperamental nature, and finally when one takes into account such factors as climate and environment, it is not surprising that at times the religious side of the ritual would play but a minor part. This happened in India and also in Western Asia, and evidence shows the same thing to have occurred both in ancient Central America and Western Africa.


West Africa

Before comparing the above with our Indian data, reference might suitably be made to the sacred men and women in West Africa.

Among the Ewe-speaking peoples of the Slave Coast and the Tshi-speaking peoples of the Gold Coast is to be found a system of sacred prostitution very similar to that which we have already considered. The subject was mentioned by Burton[63] and has since been fully discussed by Ellis,[64] and as Frazer has quoted so largely from him,[65] it will not be necessary to give any detailed description here.

Two quotations will be sufficient:

“Young people of either sex, dedicated or affiliated to a god, are termed kosio, from kono, ‘unfruitful,’ because a child dedicated to a god passes into his service and is practically lost to his parents, and si, ‘to run away.’ As the females become the ‘wives’ of the god to whom they are dedicated, the termination si in võdu-si has been translated ‘wife’ by some Europeans; but it is never used in the general acceptation of that term, being entirely restricted to persons consecrated to the gods. The chief business of the female kosi is prostitution, and in every town there is at least one institution in which the best-looking girls, between ten and twelve years of age, are received. Here they remain for three years, learning the chants and dances peculiar to the worship of the gods, and prostituting themselves to the priests and inmates of the male seminaries; and at the termination of their novitiate they become public prostitutes. This condition, however, is not regarded as one for reproach; they are considered to be married to the god, and their excesses are supposed to be caused and directed by him. Properly speaking, their libertinage should be confined to the male worshippers at the temple of the god, but practically it is indiscriminate. Children who are born from such unions belong to the god.”

Just as in India, these women are not allowed to marry a mortal husband. On page 148 of the same work Ellis says:

“The female kosio of Dañh-gbi, or Dañh-sio, that is, the wives, priestesses, and temple prostitutes of Dañh-gbi, the python-god, have their own organisation. Generally they live together in a group of houses or huts inclosed by a fence, and in these inclosures the novices undergo their three years of initiation. Most new members are obtained by the affiliation of young girls; but any woman whatever, married or single, slave or free, by publicly simulating possession, and uttering the conventional cries recognised as indicative of possession by the god, can at once join the body, and be admitted to the habitations of the order. The person of a woman who was joined in this manner is inviolable, and during the period of her novitiate she is forbidden, if single, to enter the house of her parents, and, if married, that of her husband. This inviolability, while it gives women opportunities of gratifying an illicit passion, at the same time serves occasionally to save the persecuted slave, or neglected wife, from the ill treatment of the lord and master; for she has only to go through the conventional form of possession and an asylum is assured.”

The reader will, I think, notice a closer relationship to the customs of West Africa in India than in Western Asia, but we must remember that we have much more evidence on such customs in India and Africa than in Babylonia, Syria and Phoenicia. In Western Asia we have no account of the initiation and duties taught to the new votary, so we cannot make sufficiently close comparisons.

There are undoubtedly instances of the past existence of somewhat similar institutions to those we have been considering in other parts of the world—such as Peru, Mexico, Borneo, Japan, etc. The evidence has been collected, and references given, by John Main in “his” Religious Chastity, New York, 1913, pp. 136-181.

Now that we have considered our subject in countries other than India we feel in a better position to theorise as to the origin of the institution of the dēva-dāsī.

The basis on which all such systems rest seems to be the natural desire to ensure fertility in both the animal and vegetable kingdoms. Environment, changing sentiment, temperament and religious feeling account for the particular channel into which such a system, toucing the human passions so closely, has run.

Different conditions may produce quite different schools of thought in exactly the same place. Old customs may be followed by modern people with little idea of why they follow them.

In India the system of caste, the status of women, suttee, śrāddha and numerous other customs already mentioned in the Ocean of Story have all left their mark on such an institution as that of the dēva-dāsī.

More than this it is impossible to say. Much research still remains to be done on this highly important anthropological problem.

Footnotes and references:


See F. S. Growse, Mathurā: A District Memoir, 2nd edition, 1880. Published by the N.W. Provinces & Oudh Government Press.


See R. Pischel and K. F. Geldner, Vedische Studien. Stuttgart, 1888-1889, I, xxv, pp. 196, 275, 309 et seq.; ii, p. 120; also A. A. Macdonell and A. B. Keith, A Vedic Index of Names and Subjects, London, 1912 , i, p. 395; ii, p. 480 et seq.


See index volume to the English translation of the Jātaka stories under the word “courtesan.” Cambridge, 1913.


See the English translation by R. Shama Sastri in Mysore Review, 1906- 1909, Books I-IV, and Indian Antiquary, 1909-1910, Books V-XV; also list of modern articles, etc., on the Arthaśāstra on pp. 679,680 of vol. i of the Cambridge History of India, 1922. Both author and date are, however, still doubtful.


See J. J. Meyer, Kāvyasaingraha: erotische und exoterische Lieder. Metrische Ūbersetzungen aus indischen und anderen Sprachen. Leipzig [1903]. Das Weib im altindischen Epos. Ein Beitrag zur indischen und zur vergleichenden Kulturgeschiclite. Leipzig, 1915. Also R. Schmidt, Beiträge zur Indischen Erotik;das Liebesleben des Sanskritvolkes nach den Quellen dargestellt. Leipzig, 1902; Berlin, 1911.


See Kāma Ṣāstra Society (R. F. Burton and F. F. Arbuthnot) edition, 1883, and that by K. R. Iyeṅgar, Mysore, 1921. Details of various articles on the Kāma Sūtra and its author will be found in my Bibliography of Sir Richard F. Burton, London, 1923, pp. 166-171.


Thurston in his Castes and Tribes of Southern India, vol. ii, p. 125, says that old Hindu works give seven classes of dēva-dāsī, but gives no reference.


Edited by H. H. Wilson, G. Bühler and P. Peterson, and freely translated by P. W. Jacob.


Apart from the earlier European translations see that by A. W. Ryder, issued in 1905 by the Harvard University. It forms vol. ix of the Harvard Oriental Series.


The extract is from p. 76 of Early Ideas:A Group of Hindoo Stories, 1881, by “Anaryan”—that is to say, by F. F. Arbuthnot. He was helped in his translations by Edward Rehatsek, who assisted both Burton and Arbuthnot in the Kāma Ṣāstra Society publications.


See the German translation by J. J. Meyer, 1903 [Altindische Schelmen-bücher, ii], and Les Legons de l' Entremetteuse, by Louis de Langle, Bibliothèque des Curieux, Paris, 1920, p. 127 to end.


For the English translation see the edition of the Kāma Ṣāstra Society (Burton and Arbuthnot), 1885. Further details will be found in my Burton Bibliography, 1923, pp. 171-173.


Translated by J. J. Meyer, 1903 [ Altindische Schelmenbücher, i].


Le Bréviaire de la Courtisane, Louis de Langle, Bibliothèque des Curieux, Paris, 1920, pp. 1-126.


Ā’īn-i- Akbarī, Abū-1-Fazl, Blochmann and Jarrett, Biblio. Indica. Calcutta, 1873, 1891, 1894 (3 vols.).


Bholanāth Candra, Travels, ii, 68 et seq. W. H. Sleeman, Rambles and Recollections of an Indian Official, ii, 333 et seq. 1844. A. Cunningham, Archaeological Reports, ii, 370; xxi, 110.


Manucci, Storia do Mogor, edited by W. Irvine. Indian Text Series. London, 1907. See vol. ii, p. 9.


Muntakhabu-l-lubāb (H. Elliot, History of India. London, 1867-1877, vol. vii, p. 283).


W. Crooke, The Tribes and Castes of the North-Western Provinces and Oudh, 4 vols., Calcutta, 1896. See vol. i, p. 245; vol. ii, p. 379 et seq.; and vol. iv, p. 364 et seq.


Hastings’ Encyclopaedia of Religion and Ethics, vol. x, 1918. See article on “ Prostitution,” by W. Crooke, p. 406 et seq.


“The Gypsies of Bengal,” Memoirs read before the Anthropological Society of London, vol. iii, pp. 120-133.


The Tribes and Castes of the North-Western Provinces and Oudh, vol. i,. p. 245; vol. iv, pp. 56-80.


Translated from the Chinese and annotated by Hirth and Rockhill, St Petersburg Printing Office of the Imperial Academy of Sciences, 1911.


See Henri Cordier’s Marco Polo,Notes and Addenda, 1920 , pp. 115, 11 6 .


Travels of Tavernier, translated by V. Ball, 2 vols,. 1889. See vol. i, pp. 157, 158.


A View of the History, Literature and Religion of the Hindoos, 2nd edition, Serampore, 1815-1818. See vol. ii, p. 327.


Orissa, 2 vols., 1872, and District Gazetteer of Puri, 1908. See also p. 355 et seq. of Yule and Burnell’s Hobson Jobson, London, 1886.


R. V. Russell, The Tribes and Castes of the Central Provinces of India, 4 vols., London, 1916 . See under the word “ Kasbi,” vol. iii, p. 373.


The number 108 is mystical among both Brāhmans and Buddhists. Thus at Gautama’s birth the number of Brāhmans summoned to foretell his destiny was 108; there are 108 shrines of special sanctity in India; there are 108 Upaniṣads; 108 rupees is a usual sum for a generous temple or other donation. In Tibet and China we also find 108 occurring as a sacred or mystic number in connection with architecture, ritual and literature. See Yule’s Marco Polo, vol. ii, p. 347, London, 1903.


Sir H. Risley, Tribes and Castes of Bengal, art. “Vaiṣṇava,” Calcutta, 1891.


Op. cit., vol. iii, p. 383.


Syed Siraj Ul Hassan, The Tribes and Castes of H.E.H. the Nizams Dominions [Hyderabad], Bombay, 1920. See vol. i, p. 91 et seq.


In the Central Provinces we saw that this worship was made in the spring, not the autumn.


Siraj Ul Hassan, op. cit., p. 94.


See the Ethnographical Survey of Bombay, monograph 60, Bhāvins and Devlis, 190.9; and monograph 92, Dāsa, 1907. Reference should also be made to Kennedy’s Criminal Classes of Bombay, 1908, pp. 13, 122, 274 and 283, and to R. E. Enthoven’s Tribes and Castes of Bombay, 3 vols., 1920 .


E. Hultzsch, South Indian Inscriptions, vol. ii, Part III, pp. 259-303, Archæological Survey of India, Madras, 1895.


26,755 square metres.


Yule and Cordier, The Book of Ser Marco Polo, 1903, vol. ii, pp. 345-346. See also p. 335 for identification of the places visited by Polo.


(a) Nicolo Conti (1420). See his account in India in the Fifteenth Century, (Part II, p. 23), R. H. Major: No. 22 of Series 1 of the Hakluyt Society publications, 1858. (b) ‘Abdu-r Razzāq (1443). See Elliot’s History of India, vol. iv, p. 89 et seq.; also first section of Major’s work quoted above, (c) Domingos Paes (1522). See A Forgotten Empire, R. Sewell, 1900, p. 236 etseq. (d) Fernäo Nuniz (1537). See A Forgotten Empire, p. 291 et seq.


For further information on Vijayanagar see S. K. Ayyaṅgar, Sources of Vijayanagar History, Madras University Series, 1919. Also see the various articles, etc., quoted by V. A. Smith in his Oxford History of India, 1919, pp. 319, 320.


Details of these travellers’ works with reference to the dēva-dāsīs can be found in Hobson Jobson, Yule and Burnell, 1886. See under “ dancing-girl,” dēva-dāsī, bayadere, “ nautch-girl,” cuñcurree.


From the third edition, with notes by Henry K. Beauchamp, Oxford, 1906, pp. 585-587.


A Journey from Madras through the Countries of Mysore, Canara and Malabar, 3 vols., London, 1807.


Memoirs read before the Anthropological Society of London, 1867-1869, vol. iii, London, 1870, pp. 182-194. The word bayadère is merely a French form of the Portuguese bailadeira, from bailar = to dance.


“Basivis: Women who through Dedication to a Deity assume Masculine Privileges,” Journal of the Anthropological Society of Bombay, vol. ii, 1892, pp. 322-345. This is followed by a note on the same subject by Dr W. Dymock (pp. 345, 346) and an appendix (pp. 346-353).


Five is a mystical number. It consists of 2 + 3, the first even and first odd numbers— i.e. if unity is God alone, 2 = diversity, while 3 = 1 +2 = unity and diversity. Thus the two principles of nature are represented.

Mankind has five senses. The Brāhmans worship the five products of the cow. Śiva has five aspects. The Dravidians recognise five divine foods, the Assamese five essentials for worship, and the Avestan doctrine five divisions of human personality. Five wards off the evil eye among the Mohammedans, and, being considered lucky by the Romans, entered into their wedding ceremonies.


This plant, which is used in India as a substitute for saffron and other yellow dyes, always plays an important part in marriage ceremonies—not only in India, but also in ancient Greece. It has a distinct erotic significance and has magical properties ascribed to it. See the paper by Dr W. Dymock on “The Use of Turmeric in Hindoo Ceremonial” in the Journal of the Anthropological Society of Bombay, p. 441 et seq. of the volume quoted in note 1 above.


In the order given they form Nos. 22, 17, 13, 11, 3, 1 and 20 of a series of short pamphlets issued by the Ethnographical Survey of Mysore, Baṅgalore, 1906-1911.


Castes and Tribes of Southern India, by Edgar Thurston and K. Raṅgachari, Madras, 1909, vol. ii, pp. 125-153. See also Ethnographic Notes in Southern India, by Thurston, Madras, 1906, pp. 35-41.


Golden Bough, Adonis, Attis and Osiris, vol. i, p. 61.


For further details of the Code see the articles on Babylonian law by C. H. W. Johns in Ency. Brit., vol. iii, p. 115 et seq., and Ency. Rel. Eth., vol. vii, p. 817 et seq. Special reference should be made to J. Kohler and A. Ungnad, Hammurabi’s Gesetz, Leipzig, 1909, and finally the Bibliography of p. 651 of the Cambridge Ancient History, vol. i, 1923.


See D. G. Lyon, "The Consecrated Women of the Hammurabi Code,” in Studies in the History of Religions, presented to C. H. Toy, New York, 1912 , pp. 341-360. Both Lyon and Johns (Amer. Journ. Sem. Lang., vol. xix, 1902, pp. 96-107) tried to show that the temple women were chaste. This view has, however, been proved untenable by G. A. Barton (art. " Hierodouloi,” Hastings’ Ency. Rel. Eth., vol. vi, p. 672 et seq.) and D. D. Luckenbill ("The Temple Women of the Code of Hammurabi,” in Amer. Journ. Sem. Lang., vol. xxxiv, 1917, pp. 1-12).

I am indebted to Mr R. Campbell Thomson for drawing my attention to the above papers, and to his own excellent chapter on " The Golden Age of Hammurabi” in the Cambridge Ancient History, vol. i, 1923, pp. 494-551, which has been of the greatest help in this appendix.


See note on page 270.


Proc. Soc. Bibl. Arch., vol. xxxi, p. 60 .


Op. cit., vol. xxxi, pp. 22, 34.


W. H. Ward, The Seal Cylinders of Western Asia, Washington, 1910, pp. l6l et seq., 296, 380, 387.


D. G. Hogarth, Liverpool Ann. Arch., ii, 1909, p. 170, fig. 1.


Engidu is now considered a more correct reading than Eabani.


Schrader, Keilinschriftliche Bibliothek, 1878, vol. vi, p. 127.


See Frazers Golden Bough, Attis, Adonis and Osiris, and the numerous articles in Hastings’ Ency. Rel. Eth., under such headings as “Babylonians and Assyrians,” “Heroes and Hero Gods,” “Tammuz” “Ishtar,” etc.


Schrader, Keilins. Biblio., vol. vi, p. 86 et seq.


Hastings’ Ency. Bel. Eth., vol. ii, p. 164 et seq., art. “ Atargatis.”


A Mission to Gelele, vol. ii, p. 155.


A. B. Ellis, The Ewe-speaking Peoples of the Slave Coast of West Africa, London, 1890, p. 140 et seq.; and The Tshi-speaking Peoples of the Gold Coast of West Africa, London, 1887, pp. 120-138.


Golden Bough, Attis, Adonis and Osiris, vol. i, pp. 65-70.

Like what you read? Consider supporting this website: