It is said in the Uttara Khanda of the Padma, that the Puranas, as well as other works, are divided into three classes, according to the qualities which prevail in them. Thus the Vishnu, Naradiya, Bhagavata, Garuda, Padma, and Varaha Puranas, are Satwika, or pure, from the predominance in them of the Satwa quality, or that of goodness and purity. They are, in fact, Vaishnava Puranas.

The Matsya, Kurma, Linga, Siva, Skanda, and Agni Puranas, are Tamasa, or Puranas of darkness, from the prevalence of the quality of Tamas, ‘ignorance,’ ‘gloom.’ They are indisputably Saiva Puranas.

The third series, comprising the Brahmanda, Brahma-vaivartta, Markandeya, Bhavishya, Vamana, and Brahma Puranas, are designated as Rajasa, ‘passionate,’ from Rajas, the property of passion, which they are supposed to represent.

The Matsya does not specify which are the Puranas that come under these designations, but remarks that those in which the Mahatmya of Hari or Vishnu prevails are Satwika; those in which the legends of Agni or Siva predominate are Tamasa; and those which dwell most on the stories of Brahma are Rajasa. I have elsewhere stated, that I considered the Rajasa Puranas to lean to the Sakta division of the Hindus, the worshippers of Sakti, or the female principle; founding this opinion on the character of the legends which some of them contain, such as the Durga Mahatmya, or celebrated legend on which the worship of Durga or Kali is especially founded, which is a principal episode of the Markandeya.

The Brahma-vaivartta also devotes the greatest portion of its chapters to the celebration of Radha, the mistress of Krishna, and other female divinities. Col. Vans Kennedy, however, objects to the application of the term Sakta to this last division of the Puranas, the worship of Sakti being the especial object of a different class of works, the tantras, and no such form of worship being particularly inculcated in the Brahma Purana. This last argument is of weight in regard to the particular instance specified, and the designation of Sakti may not be correctly applicable to the whole class, although it is to some of the series; for there is no incompatibility in the advocacy of a Tantrika modification of the Hindu religion by any Purana, and it has unquestionably been practised in works known as Upa-puranas.

The proper appropriation of the third class of the Puranas, according to the Padma Purana, appears to be to the worship of Krishna, not in the character in which he is represented in the Vishnu and Bhagavata Puranas, in which the incidents of his boyhood are only a portion of his biography, and in which the human character largely participates, at least in his riper years, but as the infant Krishna, Govinda, Bala Gopala, the sojourner in Vrindavan, the companion of the cowherds and milkmaids, the lover of Radha, or as the juvenile master of the universe, Jagannatha. The term Rajasa, implying the animation of passion, and enjoyment of sensual delights, is applicable, not only to the character of the youthful divinity, but to those with whom his adoration in these forms seems to have originated, the Gosains of Gokul and Bengal, the followers and descendants of Vallabha and Chaitanya, the priests and proprietors of Jagannath and Srinath-dwar, who lead a life of affluence and indulgence, and vindicate, both by precept and practice, the reasonableness of the Rajasa property, and the congruity of temporal enjoyment with the duties of religion.

The Puranas are uniformly stated to be eighteen in number. It is said that there are also eighteen Upa-puranas, or minor Puranas; but the names of only a few of these are specified in the least exceptionable authorities, and the greater number of the works is not procurable. With regard to the eighteen Puranas, there is a peculiarity in their specification, which is proof of an interference with the integrity of the text, in some of them at least; for each of them specifies the names of the whole eighteen. Now the list could not have been complete whilst the work that gives it was unfinished, and in one only therefore, the last of the series, have we a right to look for it. As however there are more last words than one, it is evident that the names must have been inserted in all except one after the whole were completed: which of the eighteen is the exception, and truly the last, there is no clue to discover, and the specification is probably an interpolation in most, if not in all.

The names that are specified are commonly the same, and are as follows: 1. Brahma, 2. Padma, 3. Vaishnava, 4. Saiva, 5. Bhagavata, 6. Narada, 7. Markanda, 8. Ágneya, 9. Bhavishya, 10. Brahma-vaivartta, 11. Lainga, 12. Varaha, 13. Skanda, 14. Vamana, 15. Kaurma, 16. Matsya, 17. Garuda, 18. Brahmanda. This is from the twelfth book of the Bhagavata, and is the same as occurs in the Vishnu.

In other authorities there are a few variations. The list of the K.urma P. omits the Agni Purana, and substitutes the Vayu. The Agni leaves out the Saiva, and inserts the Vayu. The Varaha omits the Garuda and Brahmanda, and inserts the Vayu and Narasinha: in this last it is singular. The Markandeya agrees with the Vishnu and Bhagavata in omitting the Vayu. The Matsya, like the Agni, leaves out the Saiva.

Some of the Puranas, as the Agni, Matsya, Bhagavata, and Padma, also particularize the number of stanzas which each of the eighteen contains. In one or two instances they disagree, but in general they concur. The aggregate is stated at 400,000 slokas, or 1,600,000 lines. These are fabled to be but an abridgment, the whole amount being a krore, or ten millions of stanzas, or even a thousand millions. If all the fragmentary portions claiming in various parts of India to belong to the Puranas were admitted, their extent would much exceed the lesser, though it would not reach the larger enumeration. The former is, however, as I have elsewhere stated, a quantity that an individual European scholar could scarcely expect to peruse with due care and attention, unless his whole time were devoted exclusively for many years to the task. Yet without some such labour being achieved, it was clear, from the crudity and inexactness of all that had been hitherto published on the subject, with one exception, that sound views on the subject of Hindu mythology and tradition were not to be expected. Circumstances, which I have already explained in the paper in the Journal of the Royal Asiatic Society referred to above, enabled me to avail myself of competent assistance, by which I made a minute abstract of most of the Puranas. In course of time I hope to place a tolerably copious and connected analysis of the whole eighteen before Oriental scholars, and in the mean while offer a brief notice of their several contents.

In general the enumeration of the Puranas is a simple nomenclature, with the addition in some cases of the number of verses; but to these the Matsya Purana joins the mention of one or two circumstances peculiar to each, which, although scanty, are of value, as offering means of identifying the copies of the Puranas now found with those to which the Matsya refers, or of discovering a difference between the present and the past. I shall therefore prefix the passage descriptive of each Purana from the Matsya. It is necessary to remark, however, that in the comparison instituted between that description and the Purana as it exists, I necessarily refer to the copy or copies which I employed for the purpose of examination and analysis, and which were procured with some trouble and cost in Benares and Calcutta. In some instances my manuscripts have been collated with others from different parts of India, and the result has shewn, that, with regard at least to the Brahma, Vishnu, Vayu, Matsya, Padma, Bhagavata, and Kurma Puranas, the same works, in all essential respects, are generally current under the same appellations. Whether this is invariably the case may be doubted, and farther inquiry may possibly shew that I have been obliged to content myself with mutilated or unauthentic works. It is with this reservation, therefore, that I must be understood to speak of the concurrence or disagreement of any Purana with the notice of it which the Matsya Purana has preserved.

The Brahma Purana

“That, the whole of which was formerly repeated by Brahma to Marichi, is called the Brahma Purana, and contains ten thousand stanzas.” In all the lists of the Puranas, the Brahma is placed at the head of the series, and is thence sometimes also entitled the Adi or ‘first’ Purana. It is also designated as the Saura, as it is in great part appropriated to the worship of Surya, ‘the sun.’ There are, however, works bearing these names which belong to the class of Upa-puranas, and which are not to be confounded with the Brahma. It is usually said, as above, to contain ten thousand slokas; but the number actually occurring is between seven and eight thousand. There is a supplementary or concluding section called the Brahmottara Purana, and which is different from a portion of the Skanda called the Brahmottara Khanda, which contains about three thousand stanzas more; but there is every reason to conclude that this is a distinct and unconnected work.

The immediate narrator of the Brahma Purana is Lomaharshana, who communicates it to the Rishis or sages assembled at Naimisharanya, as it was originally revealed by Brahma, not to Marichi, as the Matsya affirms, but to Daksha, another of the patriarchs: hence its denomination of the Brahma Purana.

The early chapters of this work give a description of the creation, an account of the Manwantaras, and the history of the solar and lunar dynasties to the time of Krishna, in a summary manner, and in words which are common to it and several other Puranas: a brief description of the universe succeeds; and then come a number of chapters relating to the holiness of Orissa, with its temples and sacred groves dedicated to the sun, to Siva, and Jagannath, the latter especially. These chapters are characteristic of this Purana, and shew its main object to be the promotion of the worship of Krishna as Jagannath. To these particulars succeeds a life of Krishna, which is word for word the same as that of the Vishnu Purana; and the compilation terminates with a particular detail of the mode in which Yoga, or contemplative devotion, the object of which is still Vishnu, is to be performed. There is little in this which corresponds with the definition of a Pancha-lakshana Purana; and the mention of the temples of Orissa, the date of the original construction of which is recorded, shows that it could not have been compiled earlier than the thirteenth or fourteenth century.

The Uttara Khanda of the Brahma P. bears still more entirely the character of a Mahatmya, or local legend, being intended to celebrate the sanctity of the Balaja river, conjectured to be the same as the Banas in Marwar. There is no clue to its date, but it is clearly modern, grafting personages and fictions of its own invention on a few hints from older authorities.

The Padma Purana

“That which contains an account of the period when the world was a golden lotus (padma), and of all the occurrences of that time, is therefore called the Padma by the wise: it contains fifty-five thousand stanzas.” The second Purana in the usual lists is always the Padma, a very voluminous work, containing, according to its own statement, as well as that of other authorities, fifty-five thousand slokas; an amount not far from the truth. These are divided amongst five books, or Khandas; 1. the Srishti Khanda, or section on creation; 2. the Bhumi Khanda, description of the earth; 3. the Swarga Khanda, chapter on heaven; 4. Patala Khanda, chapter on the regions below the earth; and 5. the Uttara Khanda, last or supplementary chapter. There is also current a sixth division, the Kriya Yoga Sara, a treatise on the practice of devotion.

The denominations of these divisions of the Padma P. convey but an imperfect and partial notion of their contents. In the first, or section which treats of creation, the narrator is Ugrasravas the Suta, the son of Lomaharshana, who is sent by his father to the Rishis at Naimisharanya to communicate to them the Purana, which, from its containing an account of the lotus (padma), in which Brahma appeared at creation, is termed the Padma or Padma Purana. The Suta repeats what was originally communicated by Brahma to Pulastya, and by him to Bhishma. The early chapters narrate the cosmogony, and the genealogy of the patriarchal families, much in the same style, and often in the same words, as the Vishnu; and short accounts of the Manwantaras and regal dynasties: but these, which are legitimate Pauranik matters, soon make way for new and unauthentic inventions, illustrative of the virtues of the lake of Pushkara, or Pokher in Ajmir, as a place of pilgrimage.

The Bhumi Khanda, or section of the earth, defers any description of the earth until near its close, filling up one hundred and twenty-seven chapters with legends of a very mixed description, some ancient and common to other Puranas, but the greater part peculiar to itself, illustrative of Tirthas either figuratively so termed–as a wife, a parent, or a Guru, considered as a sacred object–or places to which actual pilgrimage should be performed.

The Swarga Khanda describes in the first chapters the relative positions of the Lokas or spheres above the earth, placing above all Vaikuntha, the sphere of Vishnu; an addition which is not warranted by what appears to be the oldest cosmology. Miscellaneous notices of some of the most celebrated princes then succeed, conformably to the usual narratives; and these are followed by rules of conduct for the several castes, and at different stages of life. The rest of the book is occupied by legends of a diversified description, introduced without much method or contrivance; a few of which, as Daksha’s sacrifice, are of ancient date, but of which the most are original and modern.

The Patala Khanda devotes a brief introduction to the description of Patala, the regions of the snake-gods; but the name of Rama having been mentioned, Sesha, who has succeeded Pulastya as spokesman, proceeds to narrate the history of Rama, his descent and his posterity; in which the compiler seems to have taken the poem of Kalidasa, the Raghu Vansa, for his chief authority. An originality of addition may be suspected, however, in the adventures of the horse destined by Rama for an Aswamedha, which form the subject of a great many chapters. When about to be sacrificed, the horse turns out to be a Brahman, condemned by an imprecation of Durvasas, a sage, to assume the equine nature, and who, by having been sanctified by connexion with Rama, is released from his metamorphosis, and dispatched as a spirit of light to heaven. This piece of Vaishnava fiction is followed by praises of the Sri Bhagavata, an account of Krishna’s juvenilities, and the merits of worshipping Vishnu. These accounts are communicated through a machinery borrowed from the Tantras: they are told by Sadasiva to Parvati, the ordinary interlocutors of Tantrika compositions.

The Uttara Khanda is a most voluminous aggregation of very heterogeneous matters, but it is consistent in adopting a decidedly Vaishnava tone, and admitting no compromise with any other form of faith. The chief subjects are first discussed in a dialogue between king Dilipa and the Muni Vasishtha; such as the merits of bathing in the month of Magha, and the potency of the mantra or prayer addressed to Lakshmi Narayana. But the nature of Bhakti, faith in Vishnu–the use of Vaishnava marks on the body–the legends of Vishnu’s Avataras, and especially of Rama–and the construction of images of Vishnu–are too important to be left to mortal discretion: they are explained by Siva to Parvati, and wound up by the adoration of Vishnu by those divinities. The dialogue then reverts to the king and the sage; and the latter states why Vishnu is the only one of the triad entitled to respect; Siva being licentious, Brahma arrogant, and Vishnu alone pure. Vasishtha then repeats, after Siva, the Mahatmya of the Bhagavad Gita; the merit of each book of which is illustrated by legends of the good consequences to individuals from perusing or hearing it. Other Vaishnava Mahatmyas occupy considerable portions of this Khanda, especially the Kartika Mahatmya, or holiness of the month Kartika, illustrated as usual by stories, a few of which are of an early origin, but the greater part modern, and peculiar to this Purana.

The Kriya Yoga Sara is repeated by Suta to the Rishis, after Vyasa’s communication of it to Jaimini, in answer to an inquiry how religious merit might be secured in the Kali age, in which men have become incapable of the penances and abstraction by which final liberation was formerly to be attained. The answer is, of course, that which is intimated in the last hook of the Vishnu Purana–personal devotion to Vishnu: thinking of him, repeating his names, wearing his marks, worshipping in his temples, are a full substitute for all other acts of moral or devotional or contemplative merit.

The different portions of the Padma Purana are in all probability as many different works, neither of which approaches to the original definition of a Purana. There may be some connexion between the three first portions, at least as to time; but there is no reason to consider them as of high antiquity. They specify the Jains both by name and practices.; they talk of Mlechchhas, ‘barbarians,’ flourishing in India; they commend the use of the frontal and other Vaishnava marks; and they notice other subjects which, like these, are of no remote origin. The Patala Khanda dwells copiously upon the Bhagavata, and is consequently posterior to it. The Uttara Khanda is intolerantly Vaishnava, and is therefore unquestionably modern. It enjoins the veneration of the Salagram stone and Tulasi plant, the use of the Tapta-mudra, or stamping with a hot iron the name of Vishnu on the skin, and a variety of practices and observances undoubtedly no part of the original system. It speaks of the shrines of Sri-rangam and Venkatadri in the Dekhin, temples that have no pretension to remote antiquity; and it names Haripur on the Tungabhadra, which is in all likelihood the city of Vijayanagar, founded in the middle of the fourteenth century. The Kriya Yoga Sara is equally a modern, and apparently a Bengali composition. No portion of the Padma Purana is probably older than the twelfth century, and the last parts may be as recent as the fifteenth or sixteenth.

The Vishnu Purana

“That in which Parasara, beginning with the events of the Varaha Kalpa, expounds all duties, is called the Vaishnava; and the learned know its extent to be twenty-three thousand stanzas.” The third Purana of the lists is that which has been selected for translation, the Vishnu. It is unnecessary therefore to offer any general summary of its contents, and it will be convenient to reserve any remarks upon its character and probable antiquity for a subsequent page. It may here be observed, however, that the actual number of verses contained in it falls far short of the enumeration of the Matsya, with which the Bhagavata concurs. Its actual contents are not seven thousand stanzas. All the copies, and in this instance they are not fewer than seven in number, procured both in the east and in the west of India, agree; and there is no appearance of any part being wanting. There is a beginning, a middle, and an end, in both text and comment; and the work as it stands is incontestably entire. How is the discrepancy to be explained?

The Vayaviya Purana

“The Purana in which Vayu has declared the laws of duty, in connexion with the Sweta Kalpa, and which comprises the Mahatmya of Rudra, is the Vayaviya Purana: it contains twenty-four thousand verses.” The Siva or Saiva Purana is, as above remarked, omitted in some of the lists; and in general, when that is the case, it is replaced by the Vayu or Vayaviya. When the Siva is specified, as in the Bhagavata, then the Vayu is omitted; intimating the possible identity of these two works. This indeed is confirmed by the Matsya, which describes the Vayaviya Purana as characterised by its account of the greatness of Rudra or Siva; and Balambhatta mentions that the Vayaviya is also called the Saiva, though, according to some, the latter is the name of an Upa-purana. Col. Vans Kennedy observes, that in the west of India the Saiva is commonly considered to be an Upa or ‘minor’ Purana.

Another proof that the same work is intended by the authorities here followed, the Bhagavata and Matsya, under different appellations, is their concurrence in the extent of the work, each specifying its verses to be twenty-four thousand. A copy of the Siva Purana, of which an index and analysis have been prepared, does not contain more than about seven thousand: it cannot therefore be the Siva Purana of the Bhagavata; and we may safely consider that to be the same as the Vayaviya of the Matsya.

The Vayu Purana is narrated by Suta to the Rishis at Naimisharanya, as it was formerly told at the same place to similar persons by Vayu; a repetition of circumstances not uncharacteristic of the inartificial style of this Purana. It is divided into four Padas, termed severally Prakriya, Upodghata, Anushanga, and Upasanhara; a classification peculiar to this work. These are preceded by an index, or heads of chapters, in the manner of the Mahabharata and Ramayana; another peculiarity.

The Prakriya portion contains but a few chapters, and treats chiefly of elemental creation, and the first evolutions of beings, to the same purport as the Vishnu, but in a more obscure and unmethodical style. The Upodghata then continues the subject of creation, and describes the various Kalpas or periods during which the world has existed; a greater number of which is specified by the Saiva than by the Vaishnava Puranas. Thirty-three are here described, the last of which is the Sweta or ‘white’ Kalpa, from Siva’s being born in it of a white complexion. The genealogies of the patriarchs, the description of the universe, and the incidents of the first six Manwantaras, are all treated of in this part of the work; but they are intermixed with legends and praises of Siva, as the sacrifice of Daksha, the Maheswara Mahatmya, the Nilakantha Stotra, and others. The genealogies, although in the main the same as those in the Vaishnava Puranas, present some variations. A long account of the Pitris or progenitors is also peculiar to this Purana; as are stories of some of the most celebrated Rishis, who were engaged in the distribution of the Vedas.

The third division commences with an account of the seven Rishis and their descendants, and describes the origin of the different classes of creatures from the daughters of Daksha, with a profuse copiousness of nomenclature, not found in any other Purana. With exception of the greater minuteness of detail, the particulars agree with those of the Vishnu P. A chapter then occurs on the worship of the Pitris; another on Tirthas, or places sacred to them; and several on the performance of Sraddhas, constituting the Sraddha Kalpa. After this, comes a full account of the solar and lunar dynasties, forming a parallel to that in the following pages, with this difference, that it is throughout in verse, whilst that of our text, as noticed in its place, is chiefly in prose. It is extended also by the insertion of detailed accounts of various incidents, briefly noticed in the Vishnu, though derived apparently from a common original. The section terminates with similar accounts of future kings, and the same chronological calculations, that are found in the Vishnu.

The last portion, the Upasanhara, describes briefly the future Manwantaras, the measures of space and time, the end of the world, the efficacy of Yoga, and the glories of Siva-pura, or the dwelling of Siva, with whom the Yogi is to be united. The manuscript concludes with a different history of the successive teachers of the Vayu Purana, tracing them from Brahma to Vayu, from Vayu to Vrihaspati, and from him, through various deities and sages, to Dwaipayana and Suta.

The account given of this Purana in the Journal of the Asiatic Society of Bengal was limited to something less than half the work, as I had not then been able to procure a larger portion. I have now a more complete one of my own, and there are several copies in the East India Company’s library of the like extent. One, presented by His Highness the Guicowar, is dated Samvat 1540, or A. D. 1483, and is evidently as old as it professes to be. The examination I have made of the work confirms the view I formerly took of it; and from the internal evidence it affords, it may perhaps be regarded as one of the oldest and most authentic specimens extant of a primitive Purana.

It appears, however, that we have not yet a copy of the entire Vayu Purana. The extent of it, as mentioned above, should be twenty-four thousand verses. The Guicowar MS. has but twelve thousand, and is denominated the Purvarddha, or first portion. My copy is of the like extent. The index also spews that several subjects remain untold; as, subsequently to the description of the sphere of Siva, and the periodical dissolution of the world, the work is said to contain an account of a succeeding creation, and of various events that occurred in it, as the birth of several celebrated Rishis, including that of Vyasa, and a description of his distribution of the Vedas; an account of the enmity between Vasishtha and Viswamitra; and a Naimisharanya Mahatmya. These topics are, however, of minor importance, and can scarcely carry the Purana to the whole extent of the verses which it is said to contain. If the number is accurate, the index must still omit a considerable portion of the subsequent contents.

The Bhagavata Purana

“That in which ample details of duty are described, and which opens with (an extract from) the Gayatri; that in which the death of the Asura Vritra is told, and in which the mortals and immortals of the Saraswata Kalpa, with the events that then happened to them in the world, are related; that, is celebrated as the Bhagavata, and consists of eighteen thousand verses.” The Bhagavata is a work of great celebrity in India, and exercises a more direct and powerful influence upon the opinions and feelings of the people than perhaps any other of the Puranas. It is placed the fifth in all the lists; but the Padma Purana ranks it as the eighteenth, as the extracted substance of all the rest. According to the usual specification, it consists of eighteen thousand slokas, distributed amongst three hundred and thirty-two chapters, divided into twelve Skandhas or books. It is named Bhagavata from its being dedicated to the glorification of Bhagavat or Vishnu.

The Bhagavata is communicated to the Rishis at Naimisharanya by Suta, as usual; but he only repeats what was narrated by Suka, the son of Vyasa, to Parikshit, the king of Hastinapura, the grandson of Arjuna. Having incurred the imprecation of a hermit, by which he was sentenced to die of the bite of a venomous snake, at the expiration of seven days; the king, in preparation for this event, repairs to the banks of the Ganges; whither also come the gods and sages, to witness his death. Amongst the latter is Suka; and it is in reply to Parikshit’s question, what a man should do who is about to die, that he narrates the Bhagavata, as he had heard it from Vyasa; for nothing secures final happiness so certainly, as to die whilst the thoughts are wholly engrossed by Vishnu.

The course of the narration opens with a cosmogony, which, although in most respects similar to that of other Puranas, is more largely intermixed with allegory and mysticism, and derives its tone more from the Vedanta than the Sankhya philosophy. The doctrine of active creation by the Supreme, as one with Vasudeva, is more distinctly asserted, with a more decided enunciation of the effects being resolvable into Maya, or illusion. There are also doctrinal peculiarities, highly characteristic of this Purana; amongst which is the assertion that it was originally communicated by Brahma to Narada, that all men whatsoever, Hindus of every caste, and even Mlechchhas, outcastes or barbarians, might learn to have faith in Vasudeva.

In the third book the interlocutors are changed to Maitreya and Vidura; the former of whom is the disciple in the Vishnu Purana, the latter was the half-brother of the Kuru princes. Maitreya, again, gives an account of the Srishti-lila, or sport of creation, in a strain partly common to the Puranas, partly peculiar; although he declares he learned it from his teacher Parasara, at the desire of Pulastya; referring thus to the fabulous origin of the Vishnu Purana, and furnishing evidence of its priority. Again, however, the authority is changed, and the narrative is said to have been that which was communicated by Sesha to the Nagas. The creation of Brahma is then described, and the divisions of time are explained. A very long and peculiar account is given of the Varaha incarnation of Vishnu, which is followed by the creation of the Prajapatis and Swayambhuva, whose daughter Devahuti is married to Karddama Rishi; an incident peculiar to this work, as is that which follows of the Avatara of Vishnu as Kapila the son of Karddama and Devahuti, the author of the Sankhya philosophy, which he expounds, after a Vaishnava fashion, to his mother, in the last nine chapters of this section.

The Manwantara of Swayambhuva, and the multiplication of the patriarchal families, are next described with some peculiarities of nomenclature, which are pointed out in the notes to the parallel passages of the Vishnu Purana. The traditions of Dhruva, Vena, Prithu, and other princes of this period, are the other subjects of the fourth Skandha, and are continued in the fifth to that of the Bharata who obtained emancipation. The details generally conform to those of the Vishnu Purana, and the same words are often employed, so that it would he difficult to determine which work had the best right to them, had not the Bhagavata itself indicated its obligations to the Vishnu. The remainder of the fifth book is occupied with the description of the universe, and the same conformity with the Vishnu continues.

This is only partially the case with the sixth book, which contains a variety of legends of a miscellaneous description, intended to illustrate the merit of worshipping Vishnu: some of them belong to the early stock, but some are apparently novel. The seventh book is mostly occupied with the legend of Prahlada. In the eighth we have an account of the remaining Manwantaras; in which, as happening in the course of them, a variety of ancient legends are repeated, as the battle between the king of the elephants and an alligator, the churning of the ocean, and the dwarf and fish Avataras. The ninth book narrates the dynasties of the Vaivaswata Manwantara, or the princes of the solar and lunar races to the time of Krishna. The particulars conform generally with those recorded in the Vishnu.

The tenth book is the characteristic part of this Purana, and the portion upon which its popularity is founded. It is appropriated entirely to the history of Krishna, which it narrates much in the same manner as the Vishnu, but in more detail; holding a middle place, however, between it and the extravagant prolixity with which the Hari Vansa repeats the story. It is not necessary to particularize it farther. It has been translated into perhaps all the languages of India, and is a favourite work with all descriptions of people.

The eleventh book describes the destruction of the Yadavas, and death of Krishna. Previous to the latter event, Krishna instructs Uddhava in the performance of the Yoga; a subject consigned by the Vishnu to the concluding passages. The narrative is much the same, but something more summary than that of the Vishnu. The twelfth book continues the lines of the kings of the Kali age prophetically to a similar period as the Vishnu, and gives a like account of the deterioration of all things, and their final dissolution. Consistently with the subject of the Purana, the serpent Takshaka bites Parikshit, and he expires, and the work should terminate; or the close might be extended to the subsequent sacrifice of Janamejaya for the destruction of the whole serpent race. There is a rather awkwardly introduced description, however, of the arrangement of the Vedas and Puranas by Vyasa, and the legend of Markandeya’s interview with the infant Krishna, during a period of worldly dissolution. We then come to the end of the Bhagavata, in a series of encomiastic commendations of its own sanctity, and efficacy to salvation.

Mr. Colebrooke observes of the Bhagavata Purana, “I am inclined to adopt an opinion supported by many learned Hindus, who consider the celebrated Sri Bhagavata as the work of a grammarian (Vopadeva), supposed to have lived six hundred years ago.” Col. Vans Kennedy considers this an incautious admission, because “it is unquestionable that the number of the Puranas has been always held to be eighteen; but in most of the Puranas the names of the eighteen are enumerated, amongst which the Bhagavata is invariably included; and consequently if it were composed only six hundred years ago, the others must be of an equally modern date.” Some of them are no doubt more recent; but, as already remarked, no weight can be attached to the specification of the eighteen names, for they are always complete; each Purana enumerates all. Which is the last? which had the opportunity of naming its seventeen predecessors, and adding itself? The argument proves too much. There can be little doubt that the list has been inserted upon the authority of tradition, either by some improving transcriber, or by the compiler of a work more recent than the eighteen genuine Puranas. The objection is also rebutted by the assertion, that there was another Purana to which the name applies, and which is still to be met with, the Devi Bhagavata.

For, the authenticity of the Bhagavata is one of the few questions affecting their sacred literature which Hindu writers have ventured to discuss. The occasion is furnished by the text itself. In the fourth chapter of the first book it is said that Vyasa arranged the Vedas, and divided them into four; and that he then compiled the Itihasa and Puranas, as a fifth Veda. The Vedas he gave to Paila and the rest; the Itihasa and Puranas to Lomaharshana, the father of Suta. Then reflecting that these works may not be accessible to women, Sudras, and mixed castes, he composed the Bharata, for the purpose of placing religious knowledge within their reach. Still he felt dissatisfied, and wandered in much perplexity along the banks of the Saraswati, where his hermitage was situated, when Narada paid him a visit.

Having confided to him his secret and seemingly causeless dissatisfaction, Narada suggested that it arose from his not having sufficiently dwelt, in the works he had finished, upon the merit of worshipping Vasudeva. Vyasa at once admitted its truth, and found a remedy for his uneasiness in the composition of the Bhagavata, which he taught to Suka his son. Here therefore is the most positive assertion that the Bhagavata was composed subsequently to the Puranas, and given to a different pupil, and was not therefore one of the eighteen of which Romaharshana the Seta was, according to all concurrent testimonies, the depositary. Still the Bhagavata is named amongst the eighteen Puranas by the inspired authorities; and how can these incongruities be reconciled?

The principal point in dispute seems to have been started by an expression of Sridhara Swamin, a commentator on the Bhagavata, who somewhat incautiously made the remark that there was no reason to suspect that by the term Bhagavata any other work than the subject of his labours was intended. This was therefore an admission that some suspicions had been entertained of the correctness of the nomenclature, and that an opinion had been expressed that the term belonged, not to the Sri Bhagavata, but to the Devi Bhagavata; to a Saiva, not a Vaishnava, composition. With whom doubts prevailed prior to Sridhara Swamin, or by whom they were urged, does not appear; for, as far as we are aware, no works, anterior to his date, in which they are advanced have been met with. Subsequently, various tracts have been written on the subject. There are three in the library of the East India Company; the Durjana Mukha Chapetika, ‘A slap of the face for the vile,’ by Ramasrama; the Durjana Mukha Maha Chapetika, ‘A great slap of the face for the wicked,’ by Kasinath Bhatta; and the Durjana Mukha Padma Paduka, ‘A slipper’ for the same part of the same persons, by a nameless disputant. The first maintains the authenticity of the Bhagavata; the second asserts that the Devi Bhagavata is the genuine Purana; and the third replies to the arguments of the first. There is also a work by Purushottama, entitled ‘Thirteen arguments for dispelling all doubts of the character of the Bhagavata’ (Bhagavata swarupa vihsaya sanka nirasa trayodasa); whilst Balambhatta, a commentator on the Mitakshara, indulging in a dissertation on the meaning of the word Purana, adduces reasons for questioning the inspired origin of this Purana.

The chief arguments in favour of the authenticity of this Purana are the absence of any reason why Vopadeva, to whom it is attributed, should not have put his own name to it; its being included in all lists of the Puranas, sometimes with circumstances that belong to no other Purana; and its being admitted to be a Purana, and cited as authority, or made the subject of comment, by writers of established reputation, of whom Sankara acharya is one, and he lived long before Vopadeva. The reply to the first argument is rather feeble, the controversialists being unwilling perhaps to admit the real object, the promotion of new doctrines. It is therefore said that Vyasa was an incarnation of Narayana, and the purpose was to propitiate his favour.

The insertion of a Bhagavata amongst the eighteen Puranas is acknowledged; but this, it is said, can be the Devi Bhagavata alone, for the circumstances apply more correctly to it than to the Vaishnava Bhagavata. Thus a text is quoted by Kasinath from a Purana–he does not state which–that says of the Bhagavata that it contains eighteen thousand verses, twelve books, and three hundred and thirty-two chapters. Kasinath asserts that the chapters of the Sri Bhagavata are three hundred and thirty-five, and that the numbers apply throughout only to the Devi Bhagavata. It is also said that the Bhagavata contains an account of the acquirement of holy knowledge by Hayagriva; the particulars of the Saraswata Kalpa; a dialogue between Ambarisha and Suka; and that it commences with the Gayatri, or at least a citation of it. These all apply to the Devi Bhagavata alone, except the last; but it also is more true of the Saiva than of the Vaishnava work, for the latter has only one word of the Gayatri, dhimahi, ‘we meditate;’ whilst the former to dhimahi adds, Ya nah prachodayat, ‘who may enlighten us.’

To the third argument it is in the first place objected, that the citation of the Bhagavata by modern writers is no test of its authenticity; and with regard to the more ancient commentary of Sankara Acharya, it is asked, “Where is it?” Those who advocate the sanctity of the Bhagavata reply, “It was written in a difficult style, and became obsolete, and is lost.” “A very unsatisfactory plea,” retort their opponents, “for we still have the works of Sankara, several of which are quite as difficult as any in the Sanscrit language.” The existence of this comment, too, rests upon the authority of Madhwa or Madhava, who in a commentary of his own asserts that he has consulted eight others. Now amongst these is one by the monkey Hanuman; and although a Hindu disputant may believe in the reality of such a composition, yet we may receive its citation as a proof that Madhwa was not very scrupulous in the verification of his authorities.

There are other topics urged in this controversy on both sides, some of which are simple enough, some are ingenious: but the statement of the text is of itself sufficient to shew that according to the received opinion of all the authorities of the priority of the eighteen Puranas to the Bharata, it is impossible that the Sri Bhagavata, which is subsequent to the Bharata, should be of the number; and the evidence of style, the superiority of which to that of the Puranas in general is admitted by the disputants, is also proof that it is the work of a different hand. Whether the Devi Bhagavata have a better title to be considered as an original composition of Vyasa, is equally questionable; but it cannot be doubted that the Sri Bhagavata is the product of uninspired erudition. There does not seem to be any other ground than tradition for ascribing it to Vopadeva the grammarian; but there is no reason to call the tradition in question. Vopadeva flourished at the court of Hemadri, Raja of Devagiri, Deogur or Dowlutabad, and must consequently have lived prior to the conquest of that principality by the Mohammedans in the fourteenth century. The date of the twelfth century, commonly assigned to him, is probably correct, and is that of the Bhagavata Purana.

The Naradiya Purana

“Where Narada has described the duties which were observed in the Vrihat Kalpa, that, is called the Naradiya, having twenty-five thousand stanzas.” If the number of verses be here correctly stated, the Purana has not fallen into my hands. The copy I have analysed contains not many more than three thousand slokas. There is another work, which might be expected to be of greater extent, the Vrihat Naradiya, or great Narada Purana; but this, according to the concurrence of three copies in my possession, and of five others in the Company’s library, contains but about three thousand five hundred verses. It may be doubted, therefore, if the Narada Purana of the Matsya exists.

According to the Matsya, the Narada Purana is related by Narada, and gives an account of the Vrihat Kalpa. The Naradiya Purana is communicated by Narada to the Rishis at Naimisharanya, on the Gomati river. The Vrihannaradiya is related to the same persons, at the same place, by Suta, as it was told by Narada to Sanatkumara. Possibly the term Vrihat may have been suggested by the specification which is given in the Matsya; but there is no description in it of any particular Kalpa, or day of Brahma.

From a cursory examination of these Puranas, it is very evident that they have no conformity to the definition of a Purana, and that both are sectarial and modern compilations, intended to support the doctrine of Bhakti, or faith in Vishnu. With this view they have collected a variety of prayers addressed to one or other form of that divinity; a number of observances and holidays connected with his adoration; and different legends, some perhaps of an early, others of a more recent date, illustrative of the efficacy of devotion to Hari. Thus in the Narada we have the stories of Dhruva and Prahlada; the latter told in the words of the Vishnu: whilst the second portion of it is occupied with a legend of Mohini, the will-born daughter of a king called Rukmangada: beguiled by whom, the king offers to perform for her whatever she may desire. She calls upon him either to violate the rule of fasting on the eleventh day of the fortnight, a day sacred to Vishnu, or to put his son to death; and he kills his son, as the lesser sin of the two. This shews the spirit of the work. Its date may also be inferred from its tenor, as such monstrous extravagancies in praise of Bhakti are certainly of modern origin. One limit it furnishes itself, for it refers to Suka and Parikshit, the interlocutors of the Bhagavata, and it is consequently subsequent to the date of that Purana: it is probably considerably later, for it affords evidence that it was written after India was in the hands of the Mohammedans. In the concluding passage it is said, “Let not this Purana be repeated in the presence of the ‘killers of cows’ and contemners of the gods.” It is possibly a compilation of the sixteenth or seventeenth century.

The Vrihannaradiya is a work of the same tenor and time. It contains little else than panegyrical prayers addressed to Vishnu, and injunctions to observe various rites, and keep holy certain seasons, in honour of him. The earlier legends introduced are the birth of Markandeya, the destruction of Sagara’s sons, and the dwarf Avatara; but they are subservient to the design of the whole, and are rendered occasions for praising Narayana: others, illustrating the efficacy of certain Vaishnava observances, are puerile inventions, wholly foreign to the more ancient system of Pauranik fiction. There is no attempt at cosmogony, or patriarchal or regal genealogy. It is possible that these topics may be treated of in the missing stanzas; but it seems more likely that the Narada Purana of the lists has little in common with the works to which its name is applied in Bengal and Hindustan.

The Markandeya Purana

“That Purana in which, commencing with the story of the birds that were acquainted with right and wrong, every thing is narrated fully by Markandeya, as it was explained by holy sages in reply to the question of the Muni, is called the Markandeya, containing nine thousand verses.” This is so called from its being in the first instance narrated by Markandeya Muni, and in the second place by certain fabulous birds; thus far agreeing with the account given of it in the Matsya. That, as well as other authorities, specify its containing nine thousand stanzas; but my copy closes with a verse affirming that the number of verses recited by the Muni was six thousand nine hundred; and a copy in the East India Company’s library has a similar specification. The termination is, however, somewhat abrupt, and there is no reason why the subject with which it ends should not have been carried on farther. One copy in the Company’s library, indeed, belonging to the Guicowar’s collection, states at the close that it is the end of the first Khanda, or section. If the Purana was ever completed, the remaining portion of it appears to be lost.

Jaimini, the pupil of Vyasa, applies to Markandeya to be made acquainted with the nature of Vasudeva, and for an explanation of some of the incidents described in the Mahabharata; with the ambrosia of which divine poem, Vyasa he declares has watered the whole world: a reference which establishes the priority of the Bharata to the Markandeya Purana, however incompatible this may be with the tradition, that having finished the Puranas, Vyasa wrote the poem.

Markandeya excuses himself, saying he has a religious rite to perform; and he refers Jaimini to some very sapient birds, who reside in the Vindhya mountains; birds of a celestial origin, found, when just born, by the Muni Samika, on the field of Kurukshetra, and brought up by him along with his scholars: in consequence of which, and by virtue of their heavenly descent, they became profoundly versed in the Vedas, and a knowledge of spiritual truth. This machinery is borrowed from the Mahabharata, with some embellishment. Jaimini accordingly has recourse to the birds, Pingaksha and his brethren, and puts to them the questions he had asked of the Muni. “Why was Vasudeva born as a mortal? How was it that Draupadi was the wife of the five Pandus? Why did Baladeva do penance for Brahmanicide? and why were the children of Draupadi destroyed, when they had Krishna and Arjuna to defend them?” The answers to these inquiries occupy a number of chapters, and form a sort of supplement to the Mahabharata; supplying, partly by invention, perhaps, and partly by reference to equally ancient authorities, the blanks left in some of its narrations.

Legends of Vritrasura’s death, Baladeva’s penance, Harischandra’s elevation to heaven, and the quarrel between Vasishtha and Viswamitra, are followed by a discussion respecting birth, death, and sin; which leads to a more extended description of the different hells than is found in other Puranas. The account of creation which is contained in this work is repeated by the birds after Markandeya’s account of it to Kroshtuki, and is confined to the origin of the Vedas and patriarchal families, amongst whom are new characters, as Duhsaha and his wife Marshti, and their descendants; allegorical personages, representing intolerable iniquity and its consequences. There is then a description of the world, with, as usual to this Purana, several singularities, some of which are noticed in the following pages. This being the state of the world in the Swayambhuva Manwantara, an account of the other Manwantaras succeeds, in which the births of the Manus, and a number of other particulars, are peculiar to this work. The present or Vaivaswata Manwantara is very briefly passed over; but the next, the first of the future Manwantaras, contains the long episodical narrative of the actions of the goddess Durga, which is the especial boast of this Purana, and is the text-book of the worshippers of Kali, Chandi, or Durga, in Bengal. It is the Chandi Patha, or Durga Mahatmya, in which the victories of the goddess over different evil beings, or Asuras, are detailed with considerable power and spirit. It is read daily in the temples of Durga, and furnishes the pomp and circumstance of the great festival of Bengal, the Durga puja, or public worship of that goddess.

After the account of the Manwantaras is completed, there follows a series of legends, some new, some old, relating to the sun and his posterity; continued to Vaivaswata Manu and his sons, and their immediate descendants; terminating with Dama, the son of Narishyanta. Of most of the persons noticed, the work narrates particulars not found elsewhere.

This Purana has a character different from that of all the others. It has nothing of a sectarial spirit, little of a religious tone, rarely inserting prayers and invocations to any deity, and such as are inserted are brief and moderate. It deals little in precepts, ceremonial or moral. Its leading feature is narrative, and it presents an uninterrupted succession of legends, most of which, when ancient, are embellished with new circumstances; and when new, partake so far of the spirit of the old, that they are disinterested creations of the imagination, having no particular motive; being designed to recommend no special doctrine or observance. Whether they are derived from any other source, or whether they are original inventions, it is not possible to ascertain. They are most probably, for the greater part at least, original; and the whole has been narrated in the compiler’s own manner, a manner superior to that of the Puranas in general, with exception of the Bhagavata.

It is not easy to conjecture a date for this Purana: it is subsequent to the Mahabharata, but how long subsequent is doubtful. It is unquestionably more ancient than such works as the Brahma, Padma, and Naradiya Puranas; and its freedom from sectarial bias is a reason for supposing it anterior to the Bhagavata. At the same time, its partial conformity to the definition of a Purana, and the tenor of the additions which it has made to received legends and traditions, indicate a not very remote age; and, in the absence of any guide to a more positive conclusion, it may conjecturally be placed in the ninth or tenth century.

The Agni Purana

“That Purana which describes the occurrences of the Isana Kalpa, and was related by Agni to Vasishtha, is called the Agneya: it consists of sixteen thousand stanzas.” The Agni or Agneya Purana derives its name from its having being communicated originally by Agni, the deity of fire, to the Muni Vasishtha, for the purpose of instructing him in the twofold knowledge of Brahma. By him it was taught to Vyasa, who imparted it to Suta; and the latter is represented as repeating it to the Rising at Naimisharanya. Its contents are variously specified as sixteen thousand, fifteen thousand, or fourteen thousand stanzas. The two copies which were employed by me contain about fifteen thousand slokas. There are two in the Company’s library, which do not extend beyond twelve thousand verses; but they are in many other respects different from mine: one of them was written at Agra, in the reign of Akbar, in A. D. 1589.

The Agni Purana, in the form in which it has been obtained in Bengal and at Benares, presents a striking contrast to the Markandeya. It may be doubted if a single line of it is original. A very great proportion of it may be traced to other sources; and a more careful collation –if the task was worth the time it would require–would probably discover the remainder.

The early chapters of this Purana describe the Avataras; and in those of Rama and Krishna avowedly follow the Ramayana and Mahabharata. A considerable portion is then appropriated to instructions for the performance of religious ceremonies; many of winch belong to the Tantrika ritual, and are apparently transcribed from the principal authorities of that system. Some belong to mystical forms of Saiva worship, little known in Hindustan, though perhaps still practised in the south. One of these is the Diksha, or initiation of a novice; by which, with numerous ceremonies and invocations, in which the mysterious monosyllables of the Tantras are constantly repeated, the disciple is transformed into a living personation of Siva, and receives in that capacity the homage of his Guru. Interspersed with these, are chapters descriptive of the earth and of the universe, which are the same as those of the Vishnu Purana; and Mahatmyas or legends of holy places, particularly of Gaya.

Chapters on the duties of kings, and on the art of war, then occur, which have the appearance of being extracted from some older work, as is undoubtedly the chapter on judicature, which follows them, and which is the same as the text of the Mitakshara. Subsequent to these, we have an account of the distribution and arrangement of the Vedas and Puranas, which is little else than an abridgment of the Vishnu: and in a chapter on gifts we have a description of the Puranas, which is precisely the same, and in the same situation, as the similar subject in the Matsya Purana. The genealogical chapters are meagre lists, differing in a few respects from those commonly received, as hereafter noticed, but unaccompanied by any particulars, such as those recorded or invented in the Markandeya. The next subject is medicine, compiled avowedly, but injudiciously, from the Sausruta. A series of chapters on the mystic worship of Siva and Devi follows; and the work winds up with treatises on rhetoric, prosody, and grammar, according to the Sutras of Pingala and Panini.

The cyclopædical character of the Agni Purana, as it is now described, excludes it from any legitimate claims to be regarded as a Purana, and proves that its origin cannot be very remote. It is subsequent to the Itihasas; to the chief works on grammar, rhetoric, and medicine; and to the introduction of the Tantrika worship of Devi. When this latter took place is yet far from determined, but there is every probability that it dates long after the beginning of our era. The materials of the Agni, Purana are, however, no doubt of some antiquity. The medicine of Susruta is considerably older than the ninth century; and the grammar of Panini probably precedes Christianity. The chapters on archery and arms, and on regal administration, are also distinguished by an entirely Hindu character, and must have been written long anterior to the Mohammedan invasion. So far the Agni Purana is valuable, as embodying and preserving relics of antiquity, although compiled at a more’ recent date.

Col. Wilford has made great use of a list of kings derived from an appendix to the Agni Purana, which professes to be the sixty-third or last section. As he observes, it is seldom found annexed to the Purana. I have never met with it, and doubt its ever having formed any part of the original compilation. It would appear from Col. Wilford’s remarks, that this list notices Mohammed as the institutor of an era; but his account of this is not very distinct. He mentions explicitly, however, that the list speaks of Salivahana and Vikramaditya; and this is quite sufficient to establish its character. The compilers of the Puranas were not such bunglers as to bring within their chronology so well known a personage as Vikramaditya. There are in all parts of India various compilations ascribed to the Puranas, which never formed any portion of their contents, and which, although offering sometimes useful local information, and valuable as preserving popular traditions, are not in justice to be confounded with the Puranas, so as to cause them to be charged with even more serious errors and anachronisms than those of which they are guilty.

The two copies of this work in the library of the East India Company appropriate the first half to a description of the ordinary and occasional observances of the Hindus, interspersed with a few legends: the latter half treats exclusively of the history of Mina.

The Bhavishya Purana

“The Purana in which Brahma, having described the greatness of the sun, explained to Manu the existence of the world, and the characters of all created things, in the course of the Aghora Kalpa; that, is called the Bhavishya, the stories being for the most part the events of a future period. It contains fourteen thousand five hundred stanzas.” This Purana, as the name implies, should be a book of prophecies, foretelling what will be (bhavishyati), as the Matsya Purana intimates. Whether such a work exists is doubtful. The copies, which appear to be entire, and of which there are three in the library of the East India Company, agreeing in their contents with two in my possession, contain about seven thousand stanzas. There is another work, entitled the Bhavishyottara, as if it was a continuation or supplement of the former, containing also about seven thousand verses; but the subjects of .both these works are but to a very imperfect degree analogous to those to which the Matsya alludes.

The Bhavishya Purana, as I have it, is a work in a hundred and twenty-six short chapters, repeated by Sumantu to Satanika, a king of the Pandu family. He notices, however, its having originated with Swayambhu or Brahma; and describes it as consisting of five parts; four dedicated, it should seem, to as many deities, as they are termed, Brahma, Vaishnava, Saiva, and Twashtra; whilst the fifth is the Pratisarga, or repeated creation. Possibly the first part only may have come into my hands, although it does not so appear by the manuscript.

Whatever it may be, the work in question is not a Purana. The first portion, indeed, treats of creation; but it is little else than a transcript of the words of the first chapter of Manu. The rest is entirely a manual of religious rites and ceremonies. It explains the ten Sanskaras, or initiatory rites; the performance of the Sandhya; the reverence to be shewn to a Guru; the duties of the different Asramas and castes; and enjoins a number of Vratas, or observances of fasting and the like, appropriate to different lunar days. A few legends enliven the series of precepts. That of the sage Chyavana is told at considerable length, taken chiefly from the Mahabharata. The Naga Panchami, or fifth lunation, sacred to the serpent-gods, gives rise to a description of different sorts of snakes. After these, which occupy about one-third of the chapters, the remainder of them conform in subject to one of the topics referred to by the Matsya. They chiefly represent conversations between Krishna, his son Samba, who had become a leper by the curse of Durvasas, Vasishtha, Narada, and Vyasa, upon the power and glory of the sun, and the manner in which he is to be worshipped. There is some curious matter in the last chapters, relating to the Magas, silent worshippers of the sun, from Sakadwipa, as if the compiler had adopted the Persian term Magh, and connected the fire-worshippers of Iran with those of India. This is a subject, however, that requires farther investigation.

The Bhavishyottara is, equally with the preceding, a sort of manual of religious offices, the greater portion being appropriated to Vratas, and the remainder to the forms and circumstances with which gifts are to be presented. Many of the ceremonies are obsolete, or are observed in a different manner, as the Rath-yatra, or car festival; and the Madanotsava, or festival of spring. The descriptions of these throw some light upon the public condition of the Hindu religion at a period probably prior to the Mohammedan conquest. The different ceremonies are illustrated by legends, which are sometimes ancient, as, for instance, the destruction of the god of love by Siva, and his thence becoming Ananga, the disembodied lord of hearts. The work is supposed to be communicated by Krishna to Yudhishthira, at a great assemblage of holy persons at the coronation of the latter, after the conclusion of the great war.

The Brahma-vaivartta Purana

“That Purana which is related by Savarni to Narada, and contains the account of the greatness of Krishna, with the occurrences of the Rathantara Kalpa, where also the story of Brahma-varaha is repeatedly told, is called the Brahma-vaivartta, and contains eighteen thousand stanzas.” The account here given of the Brahma-vaivartta Purana agrees with its present state as to its extent. The copies rather exceed than fall short of eighteen thousand stanzas. It also correctly represents its comprising a Mahatmya or legend of Krishna; but it is very doubtful, nevertheless, if the same work is intended.

The Brahma-vaivartta, as it now exists, is narrated, not by Savarni, but the Rishi Narayana to Narada, by whom it is communicated to Vyasa: he teaches it to Súta, and the latter repeats it to the Rishis at Naimisharanya. It is divided into four Khandas, or books; the Brahma, Prakriti, Ganesa, and Krishna Janma Khandas; dedicated severally to describe the acts of Brahma, Devi, Ganesa, and Krishna; the latter, however, throughout absorbing the interest and importance of the work. In none of these is there any account of the Varaha Avatara of Vishnu, which seems to be intended by the Matsya; nor any reference to a Rathantara Kalpa.

It may also be observed, that, in describing the merit of presenting a copy of this Purana, the Matsya adds, “Whoever makes such gift, is honoured in the Brahma-loka;” a sphere which is of very inferior dignity to that to which a worshipper of Krishna is taught to aspire by this Purana. The character of the work is in truth so decidedly sectarial, and the sect to which it belongs so distinctly marked, that of the worshippers of the juvenile Krishna and Radha, a form of belief of known modern origin, that it can scarcely have found a notice in a work to which, like the Matsya, a much more remote date seems to belong. Although therefore the Matsya may be received in proof of there having been a Brahma-vaivartta Purana at the date of its compilation, dedicated especially to the honour of Krishna, yet we cannot credit the possibility of its being the same we now possess.

Although some of the legends believed to be ancient are scattered through the different portions of this Purana, yet the great mass of it is taken up with tiresome descriptions of Vrindavan and Goloka, the dwellings of Krishna on earth and in heaven; with endless repetitions of prayers and invocations addressed to him; and with insipid descriptions of his person and sports, and the love of the Gopis and of Radha towards him. There are some particulars of the origin of the artificer castes, which is of value because it is cited as authority in matters affecting them, contained in the Brahma Khanda; and in the Prakrita and Ganesa Khandas are legends of those divinities, not wholly, perhaps, modern inventions, but of which the source has not been traced. In the life of Krishna the incidents recorded are the same as those narrated in the Vishnu and the Bhagavata; but the stories, absurd as they are, are much compressed to make room for original matter, still more puerile and tiresome. The Brahma-vaivartta has not the slightest title to be regarded as a Purana.

The Linga Purana

“Where Maheswara, present in the Agni Linga, explained (the objects of life) virtue, wealth, pleasure, and final liberation at the end of the Agni Kalpa, that Purana, consisting of eleven thousand stanzas, was called the Lainga by Brahma himself.”

The Linga Purana conforms accurately enough to this description. The Kalpa is said to be the Ísana, but this is the only difference. It consists of eleven thousand stanzas. It is said to have been originally composed by Brahma; and the primitive Linga is a pillar of radiance, in which Maheswara is present. The work is therefore the same as that referred to by the Matsya.

A short account is given, in the beginning, of elemental and secondary creation, and of the patriarchal families; in which, however, Siva takes the place of Vishnu, as the indescribable cause of all things. Brief accounts of Siva’s incarnations and proceedings in different Kalpas next occur, offering no interest except as characteristic of sectarial notions. The appearance of the great fiery Linga takes place, in the interval of a creation, to separate Vishnu and Brahma, who not only dispute the palm of supremacy, but fight for it; when the Linga suddenly springs up, and puts them both to shame; as, after traveling upwards and downwards for a thousand years in each direction, neither can approach to its termination. Upon the Linga the sacred monosyllable Om is visible, and the Vedas proceed from it, by which Brahms and Vishnu become enlightened, and acknowledge and eulogize the superior might and glory of Siva.

A notice of the creation in the Padma Kalpa then follows, and this leads to praises of Siva by Vishnu and Brahma. Siva repeats the story of his incarnations, twenty-eight in number; intended as a counterpart, no doubt, to the twenty-four Avataras of Vishnu, as described in the Bhagavata; and both being amplifications of the original ten Avataras, and of much less merit as fictions. Another instance of rivalry occurs in the legend of Dadhichi, a Muni and worshipper of Siva. In the Bhagavata there is a story of Ambarisha being defended against Durvasas by the discus of Vishnu, against which that Saiva sage is helpless: here Vishnu hurls his discus at Dadhichi, but it falls blunted to the ground, and a conflict ensues, in which Vishnu and his partisans are all overthrown by the Muni.

A description of the universe, and of the regal dynasties of the Vaivaswata Manwantara to the time of Krishna, runs through a number of chapters, in substance, and very commonly in words, the same as in other Puranas. After which, the work resumes its proper character, narrating legends, and enjoining rites, and reciting prayers, intending to do honour to Siva under various forms. Although, however, the Linga holds a prominent place amongst them, the spirit of the worship is as little influenced by the character of the type as can well be imagined. There is nothing like the phallic orgies of antiquity: it is all mystical and spiritual. The Linga is twofold, external and internal. The ignorant, who need a visible sign, worship Siva through a ‘mark’ or ‘type’–which is the proper meaning of the word ‘Linga’–of wood or stone; but the wise look upon this outward emblem as nothing, and contemplate in their minds the invisible, inscrutable type, which is Siva himself. Whatever may have been the origin of this form of worship in India, the notions upon which it was founded, according to the impure fancies of European writers, are not to be traced in even the Saiva Puranas.

Data for conjecturing the era of this work are defective, but it is more of a ritual than a Purana, and the Pauranik chapters which it has inserted, in order to keep up something of its character, have been evidently borrowed for the purpose. The incarnations of Siva, and their ‘pupils,’ as specified in one place, and the importance attached to the practice of the Yoga, render it possible that under the former are intended those teachers of the Saiva religion who belong to the Yoga school, which seems to have flourished about the eighth or ninth centuries. It is not likely that the work is earlier, it may be considerably later. It has preserved apparently some Saiva legends of an early date, but the greater part is ritual and mysticism of comparatively recent introduction.

The Varaha Purana

“That in which the glory of the great Varaha is predominant, as it was revealed to Earth by Vishnu, in connexion, wise Munis, with the Manava Kalpa, and which contains twenty-four thousand verses, is called the Varaha Purana.”

It may be doubted if the Varaha Purana of the present day is here intended. It is narrated by Vishnu as Varaha, or in the boar incarnation, to the personified Earth. Its extent, however, is not half that specified, little exceeding ten thousand stanzas. It furnishes also itself evidence of the prior currency of some other work, similarly denominated; as, in the description of Mathura contained in it, Sumantu, a Muni, is made to observe, “The divine Varaha in former times expounded a Purana, for the purpose of solving the perplexity of Earth.”

Nor can the Varaha Purana be regarded as a Purana agreeably to the common definition, as it contains but a few scattered and brief allusions to the creation of the world, and the reign of kings: it has no detailed genealogies either of the patriarchal or regal families, and no account of the reigns of the Manus. Like the Linga Purana, it is a religious manual, almost wholly occupied with forms of prayer, and rules for devotional observances, addressed to Vishnu; interspersed with legendary illustrations, most of which are peculiar to itself, though some are taken from the common and ancient stock: many of them, rather incompatibly with the general scope of the compilation, relate to the history of Siva and Durga. A considerable portion of the work is devoted to descriptions of various Tirthas, places of Vaishnava pilgrimage; and one of Mathura enters into a variety of particulars relating to the shrines of that city, constituting the Mathura Mahatmyam.

In the sectarianism of the Varaha Purana there is no leaning to the particular adoration of Krishna, nor are the Rath-yatra and Janmashtami included amongst the observances enjoined. There are other indications of its belonging to an earlier stage of Vaishnava worship, and it may perhaps be referred to the age of Ramanuja, the early part of the twelfth century.

The Skanda Purana

“The Skanda Purana is that in which the six-faced deity (Skanda) has related the events of the Tatpurusha Kalpa, enlarged with many tales, and subservient to the duties taught by Maheswara. It is said to contain eighty-one thousand one hundred stanzas: so it is asserted amongst mankind.”

It is uniformly agreed that the Skanda Purana in a collective form has no existence; and the fragments in the shape of Sanhitas, Khandas, and Mahatmyas, which are affirmed in various parts of India to be portions of the Purana, present a much more formidable mass of stanzas than even the immense number of which it is said to consist. The most celebrated of these portions in Hindustan is the Kasi Khanda, a very minute description of the temples of Siva in or adjacent to Benares, mixed with directions for worshipping Maheswara, and a great variety of legends explanatory of its merits, and of the holiness of Kasi: many of them are puerile and uninteresting, but some are of a higher character.

The story of Agastya records probably, in a legendary style, the propagation of Hinduism in the south of India: and in the history of Divodasa, king of Kasi, we have an embellished tradition of the temporary depression of the worship of Siva, even in its metropolis, before the ascendancy of the followers of Buddha, There is every reason to believe the greater part of the contents of the Kasi Khanda anterior to the first attack upon Benares by Mahmud of Ghizni. The Kasi Khanda alone contains fifteen thousand stanzas.

Another considerable work ascribed in upper India to the Skanda Purana is the Utkala Khanda, giving an account of the holiness of Urissa, and the Kshetra of Purushottama or Jagannatha. The same vicinage is the site of temples, once of great magnificence and extent, dedicated to Siva, as Bhuvaneswara, which forms an excuse for attaching an account of a Vaishnava Tirtha to an eminently Saiva Purana. There can be little doubt, however, that the Utkala Khanda is unwarrantably included amongst the progeny of the parent work. Besides these, there is a Brahmottara Khanda, a Reva Khanda, a Siva Rahasya Khanda, a Himavat Khanda, and others. Of the Sanhitas, the chief are the Súta Sanhita, Sanatkumara Sanhita, Saura Sanhita, and Kapila Sanhita: there are several other works denominated Sanhitas.

The Mahatmyas are more numerous still. According to the Súta Sanhita, as quoted by Col. Vans Kennedy, the Skanda Purana contains six Sanhitas, five hundred Khandas, and five hundred thousand stanzas; more than is even attributed to all the Puranas. He thinks, judging from internal evidence, that all the Khandas and Sanhitas may be admitted to be genuine, though the Mahatmyas have rather a questionable appearance. Now one kind of internal evidence is the quantity; and as no more than eighty-one thousand one hundred stanzas have ever been claimed for it, all in excess above that amount must be questionable. But many of the Khandas, the Kasi Khanda for instance, are quite as local as the Mahatmyas, being legendary stories relating to the erection and sanctity of certain temples or groups of temples, and to certain Lingas; the interested origin of which renders them very reasonably objects of suspicion. In the present state of our acquaintance with the reputed portions of the Skanda Purana, my own views of their authenticity are so opposed to those entertained by Col. Vans Kennedy, that instead of admitting all the Sanhitas and Khandas to be genuine, I doubt if any one of them was ever a part of the Skanda Purana.

The Vamana Purana

“That in which the four-faced Brahma taught the three objects of existence, as subservient to the account of the greatness of Trivikrama, which treats also of the Siva Kalpa, and which consists of ten thousand stanzas, is called the Vamana Purana.”

The Vamana Purana contains an account of the dwarf incarnation of Vishnu; but it is related by Pulastya to Narada, and extends to but about seven thousand stanzas. Its contents can scarcely establish its claim to the character of a Purana.

There is little or no order in the subjects which this work recapitulates, and which arise out of replies made by Pulastya to questions put abruptly and unconnectedly by Narada. The greater part of them relate to the worship of the Linga; a rather strange topic for a Vaishnava Purana, but engrossing the principal part of the compilation. They are however subservient to the object of illustrating the sanctity of certain holy places; so that the Vamana Purana is little else than a succession of Mahatmyas. Thus in the opening almost of the work occurs the story of Daksha’s sacrifice, the object of which is to send Siva to Papamochana tirtha at Benares, where he is released from the sin of Brahmanicide.

Next conies the story of the burning of Kamadeva, for the purpose of illustrating the holiness of a Siva-linga at Kedareswara in the Himalaya, and of Badarikasrama. The larger part of the work consists of the Saro-mahatmya, or legendary exemplifications of the holiness of Sthanu tirtha; that is, of the sanctity of various Lingas and certain pools at Thanesar and Kurukhet, the country north-west from Delhi. There are some stories also relating to the holiness of the Gódavari river; but the general site of the legends is in Hindustan. In the course of these accounts we have a long narrative of the marriage of Siva with Uma, and the birth of Kartikeya.

There are a few brief allusions to creation and the Manwantaras, but they are merely incidental; and all the five characteristics of a Purana are deficient. In noticing the Swarochisha Manwantara, towards the end of the book, the elevation of Bali as monarch of the Daityas, and his subjugation of the universe, the gods included, are described; and this leads to the narration that gives its title to the Purana, the birth of Krishna as a dwarf, for the purpose of humiliating Bali by fraud, as he was invincible by force. The story is told as usual, but the scene is laid at Kurukshetra.

A more minute examination of this work than that which has been given to it might perhaps discover some hint from which to conjecture its date. It is of a more tolerant character than the Puranas, and divides its homage between Siva and Vishnu with tolerable impartiality. It is not connected, therefore, with any sectarial principles, and may have preceded their introduction. It has not, however, the air of any antiquity, and its compilation may have amused the leisure of some Brahman of Benares three or four centuries ago.

The Kurma Purana

“That in which Janarddana, in the form of a tortoise, in the regions under the earth, explained the objects of life–duty, wealth, pleasure, and liberation–in communication with Indradyumna and the Rishis in the proximity of Sakra, which refers to the Lakshmí Kalpa, and contains seventeen thousand stanzas, is the Kurma Purana.”

In the first chapter of the Kurma Purana it gives an account of itself, which does not exactly agree with this description. Suta, who is repeating the narration, is made to say to the Rishis, “This most excellent Kaurma Purana is the fifteenth. Sanhitas are fourfold, from the variety of the collections. The Brahmí, Bhagavatí, Saurí, and Vaishnavi, are well known as the four Sanhitas which confer virtue, wealth, pleasure, and liberation. This is the Brahmi Sanhita, conformable to the four Vedas; in which there are six thousand slokas, and by it the importance of the four objects of life, O great sages, holy knowledge and Parameswara is known.”

There is an irreconcilable difference in this specification of the number of stanzas and that given above. It is not very clear what is meant by a Sanhita as here used. A Sanhita, as observed above (p. xi), is something different from a Purana. It may be an assemblage of prayers and legends, extracted professedly from a Purana, but is not usually applicable to the original. The four Sanhitas here specified refer rather to their religious character than to their connexion with any specific work, and in fact the same terms are applied to what are called Sanhitas of the Skanda. In this sense a Purana might be also a Sanhita; that is, it might be an assemblage of formulæ and legends belonging to a division of the Hindu system; and the work in question, like the Vishnu Purana, does adopt both titles. It says, “This is the excellent Kaurma Purana, the fifteenth (of the series):” and again, “This is the Brahmí Sanhita.” At any rate, no other work has been met with pretending to be the Kurma Purana.

With regard to the other particulars specified by the Matsya, traces of them are to be found. Although in two accounts of the traditional communication of the Purana no mention is made of Vishnu as one of the teachers, yet Suta repeats at the outset a dialogue between Vishnu, as the Kurma, and Indradyumna, at the time of the churning of the ocean; and much of the subsequent narrative is put into the mouth of the former.

The name, being that of an Avatara of Vishnu, might lead us to expect a Vaishnava work; but it is always and correctly classed with the Saiva. Puranas, the greater portion of it inculcating the worship of Siva and Durga. It is divided into two parts, of nearly equal length. In the first part, accounts of the creation, of the Avataras of Vishnu, of the solar and lunar dynasties of the kings to the time of Krishna, of the universe, and of the Manwantaras, are given, in general in a summary manner, but not unfrequently in the words employed in the Vishnu Purana. With these are blended hymns addressed to Maheswara by Brahma and others; the defeat of Andhakasura by Bhairava; the origin of four Saktis, Maheswarí, Siva, Satí, and Haimavatí, from Siva; and other Saiva legends. One chapter gives a more distinct and connected account of the incarnations of Siva in the present age than the Linga; and it wears still more the appearance of an attempt to identify the teachers of the Yoga school with personations of their preferential deity. Several chapters form a Kasí Mahatmya, a legend of Benares. In the second part there are no legends. It is divided into two parts, the Íswara Gíta and Vyasa Gita. In the former the knowledge of god, that is, of Siva, through contemplative devotion, is taught. In the latter the same object is enjoined through works, or observance of the ceremonies and precepts of the Vedas.

The date of the Kurma Purana cannot be very remote, for it is avowedly posterior to the establishment of the Tantrika, the Sakta, and the Jain sects. In the twelfth chapter it is said, “The Bhairava, Vama, Arhata, and Yamala Sastras are intended for delusion.” There is no reason to believe that the Bhairava and Yamala Tantras are very ancient works, or that the practices of the left-hand Saktas, or the doctrines of Arhat or Jina were known in the early centuries of our era.

The Matsya Purana

“That in which, for the sake of promulgating the Vedas, Vishnu, in the beginning of a Kalpa, related to Manu the story of Narasinha and the events of seven Kalpas, that, O sages, know to be the Matsya Purana, containing twenty thousand stanzas.”

We might, it is to be supposed, admit the description which the Matsya gives of itself to be correct, and yet as regards the number of verses there seems to be a misstatement. Three very good copies, one in my possession, one in the Company’s library, and one in the Radcliffe library, concur in all respects, and in containing no more than between fourteen and fifteen thousand stanzas: in this case the Bhagavata is nearer the truth, when it assigns to it fourteen thousand. We may conclude, therefore, that the reading of the passage is in this respect erroneous. It is correctly said that the subjects of the Purana were communicated by Vishnu, in the form of a fish, to Manu.

The Purana, after the usual prologue of Suta and the Rishis, opens with the account of the Matsya or ‘fish’ Avatara of Vishnu, in which he preserves a king named Manu, with the seeds of all things, in an ark, from the waters of that inundation which in the season of a Pralaya overspreads the world. This story is told in the Mahabharata, with reference to the Matsya as its authority; from which it might be inferred that the Purana was prior to the poem. This of course is consistent with the tradition that the Puranas were first composed by Vyasa; but there can be no doubt that the greater part of the Mahabharata is much older than any extant Purana. The present instance is itself a proof; for the primitive simplicity with which the story of the fish Avatara is told in the Mahabharata is of a much more antique complexion than the mysticism and extravagance of the actual Matsya Purana. In the former, Manu collects the seeds of existing things in the ark, it is not said how: in the latter, he brings them all together by the power of Yoga.

In the latter, the great serpents come to the king, to serve as cords wherewith to fasten the ark to the horn of the fish: in the former, a cable made of ropes is more intelligibly employed for the purpose.

Whilst the ark floats, fastened to the fish, Manu enters into conversation with him; and his questions, and the replies of Vishnu, form the main substance of the compilation. The first subject is the creation, which is that of Brahma and the patriarchs. Some of the details are the usual ones; others are peculiar, especially those relating to the Pitris, or progenitors. The regal dynasties are next described; and then follow chapters on the duties of the different orders. It is in relating those of the householder, in which the duty of making gifts to Brahmans is comprehended, that we have the specification of the extent and subjects of the Puranas. It is meritorious to have copies made of them, and to give these away on particular occasions.

Thus it is said of the Matsya; “Whoever gives it away at either equinox, along with a golden fish and a milch cow, gives away the whole earth;” that is, he reaps a like reward in his next migration. Special duties of the householder–Vratas, or occasional acts of piety–are then described at considerable length, with legendary illustrations. The account of the universe is given in the usual strain. Saiva legends ensue; as, the destruction of Tripurasura; the war of the gods with Taraka and the Daityas, and the consequent birth of Kartikeya, with the various circumstances of Uma’s birth and marriage, the burning of Kamadeva, and other events involved in that narrative; the destruction of the Asuras Maya and Andhaka; the origin of the Matris, and the like; interspersed with the Vaishnava legends of the Avataras. Some Mahatmyas are also introduced; one of which, the Narmada Mahatmya, contains some interesting particulars. There are various chapters on law and morals; and one which furnishes directions for building houses, and making images. We then have an account of the kings of future periods; and the Purana concludes with a chapter on gifts.

The Matsya Purana, it will be seen even from this brief sketch of its contents, is a miscellaneous compilation, but including in its contents the elements of a genuine Purana. At the same time it is of too mixed a character to be considered as a genuine work of the Pauranik class; and upon examining it carefully, it may be suspected that it is indebted to various works, not only for its matter, but for its words. The genealogical and historical chapters are much the same as those of the Vishnu; and many chapters, as those on the Pitris and Sraddhas, are precisely the same as those of the Srishti Khanda of the Padma Purana. It has drawn largely also from the Mahabharata: amongst other instances, it is sufficient to quote the story of Savitrí, the devoted wife of Satyavat, which is given in the Matsya in the same manner, but considerably abridged.

The Garuda Purana

“That which Vishnu recited in the Garuda Kalpa, relating chiefly to the birth of Garuda from Vinata, is here called the Garuda Purana; and in it there are read nineteen thousand verses.”

The Garuda Purana which has been the subject of my examination corresponds in no respect with this description, and is probably a different work, though entitled the Garuda Purana. It is identical, however, with two copies in the Company’s library. It consists of no more than about seven thousand stanzas; it is repeated by Brahma to Indra; and it contains no account of the birth of Garuda. There is a brief notice of the creation; but the greater part is occupied with the description of Vratas, or religious observances, of holidays, of sacred places dedicated to the sun, and with prayers from the Tantrika ritual, addressed to the sun, to Siva, and to Vishnu. It contains also treatises on astrology, palmistry, and precious stones; and one, still more extensive, on medicine.

The latter portion, called the Preta Kalpa, is taken up with directions for the performance of obsequial rites. There is nothing in all this to justify the application of the name. Whether a genuine Garuda Purana exists is doubtful. The description given in the Matsya is less particular than even the brief notices of the other Puranas, and might have easily been written without any knowledge of the book itself, being, with exception of the number of stanzas, confined to circumstances that the title alone indicates.

The Brahmanda Purana

“That which has declared, in twelve thousand two hundred verses, the magnificence of the egg of Brahma, and in which an account of the future Kalpas is contained, is called the Brahmanda Purana, and was revealed by Brahma.”

The Brahmanda Purana is usually considered to be in much the same predicament as the Skanda, no longer procurable in a collective body, but represented by a variety of Khandas and Mahatmyas, professing to be derived from it. The facility with which any tract may be thus attached to the non-existent original, and the advantage that has been taken of its absence to compile a variety of unauthentic fragments, have given to the Brahmanda, Skanda, and Padma, according to Col. Wilford, the character of being the Puranas of thieves or impostors. This is not applicable to the Padma, which, as above shewn, occurs entire and the same in various parts of India. The imposition of which the other two are made the vehicles can deceive no one, as the purpose of the particular legend is always too obvious to leave any doubt of its origin.

Copies of what profess to be the entire Brahmanda Purana are sometimes, though rarely, procurable. I met with one in two portions, the former containing, one hundred and twenty-four chapters, the latter seventy-eight; and the whole containing about the number of stanzas assigned to the Purana. The first and largest portion, however, proved to be the same as the Vayu Purana, with a passage occasionally slightly varied, and at the end of each chapter the common phrase ‘Iti Brahmanda Purane’ substituted for ‘Iti Vayu Purane.’ I do not think there was any intended fraud in the substitution. The last section of the first part of the Vayu Purana is termed the Brahmanda section, giving an account of the dissolution of the universe; and a careless or ignorant transcriber might have taken this for the title of the whole. The checks to the identity of the work have been honestly preserved, both in the index and the frequent specification of Vayu as the teacher or narrator of it.

The second portion of this Brahmanda is not any part of the Vayu; it is probably current in the Dakhin as a Sanhita or Khanda. Agastya is represented as going to the city Kanchí (Conjeveram), where Vishnu, as Hayagríva, appears to him, and, in answer to his inquiries, imparts to him the means of salvation, the worship of Parasaktí. In illustration of the efficacy of this form of adoration, the main subject of the work is an account of the exploits of Lalita Deví, a form of Durga, and her destruction of the demon Bhandasura. Rules for her worship are also given, which are decidedly of a Sakta or Tantrika description; and this work cannot be admitted, therefore, to be part of a genuine Purana.

The Upa-puranas

The Upa-puranas, in the few instances which are known, differ little in extent or subject from some of those to which the title of Purana is ascribed. The Matsya enumerates but four; but the Deví Bhagavata has a more complete list, and specifies eighteen. They are, 1. The Sanatkumara, 2. Narasinha, 3. Naradíya, 4. Siva, 5. Durvasasa, g. Kapila, 7. Manava, 8. Ausanasa, 9. Varuna, 10. Kalika, 11. Samba, 12. Nandi, 13. Saura, 14. Parasara, 15. Aditya, 16. Maheswara, 17. Bhagavata, 18. Vasishtha.

The Matsya observes of the second, that it is named in the Padma Purana, and contains eighteen thousand verses. The Nandi it calls Nanda, and says that Kartikeya tells in it the story of Nanda. A rather different list is given in the Reva Khanda; or, 1. Sanatkumara, 2. Narasinha, 3. Nanda, 4. Sivadharma, 5. Durvasasa, 6. Bhavishya, related by Narada or Naradíya, 7. Kapila, 8. Manava, 9. Ausanasa, 10. Brahmanda, 11. Varuna, 12. Kalika, 13. Maheswara, 14. Samba, 15. Saura, 16. Parasara, 17. Bhagavata, 18. Kaurma. These authorities, however, are of questionable weight, having in view, no doubt, the pretensions of the Deví Bhagavata to be considered as the authentic Bhagavata.

Of these Upa-puranas few are to be procured. Those in my possession are the Siva, considered as distinct from the Vayu; the Kalika, and perhaps one of the Naradíyas, as noticed above. I have also three of the Skandhas of the Deví Bhagavata, which most undoubtedly is not the real Bhagavata, supposing that any Purana so named preceded the work of Vopadeva. There can be no doubt that in any authentic list the name of Bhagavata does not occur amongst the Upa-puranas: it has been put there to prove that there are two works so entitled, of which the Purana is the Deví Bhagavata, the Upa-purana the Srí Bhagavata. The true reading should be Bhargava, the Purana of Bhrigu; and the Deví Bhagavata is not even an Upa-purana. It is very questionable if the entire work, which as far as it extends is eminently a Sakta composition, ever had existence.

The Siva Upa-purana contains about six thousand stanzas, distributed into two parts. It is related by Sanatkumara to Vyasa and the Rishis at Naimisharanya, and its character may be judged of from the questions to which it is a reply. “Teach us,” said the Rishis, “the rules of worshipping the Linga, and of the god of gods adored under that type; describe to us his various forms, the places sanctified by him, and the prayers with which he is to be addressed.” In answer, Sanatkumara repeats the Siva Purana, containing the birth of Vishnu and Brahma; the creation and divisions of the universe; the origin of all things from the Linga; the rules of worshipping it and Siva; the sanctity of times, places, and things, dedicated to him; the delusion of Brahma and Vishnu by the Linga; the rewards of offering flowers and the like to a Linga; rules for various observances in honour of Mahadeva; the mode of practising the Yoga; the glory of Benares and other Saiva Tírthas; and the perfection of the objects of life by union with Maheswara. These subjects are illustrated in the first part with very few legends; but the second is made up almost wholly of Saiva stories, as the defeat of Tripurasura; the sacrifice of Daksha; the births of Kartikeya and Ganesa the sons of Siva, and Nandi and Bhringaríti his attendants and others; together with descriptions of Benares and other places of pilgrimage, and rules for observing such festivals as the Sivaratri. This work is a Saiva manual, not a Purana.

The Kalika Purana contains about nine thousand stanzas in ninety-eight chapters, and is the only work of the series dedicated to recommend the worship of the bride of Siva, in one or other of her manifold forms, as Girija, Deví, Bhadrakalí, Kalí, Mahamaya. It belongs therefore to the Sakta modification of Hindu belief, or the worship of the female powers of the deities. The influence of this worship spews itself in the very first pages of the work, which relate the incestuous passion of Brahma for his daughter Sandhya, in a strain that has nothing analogous to it in the Vayu, Linga, or Siva Puranas.

The marriage of Siva and Parvati is a subject early described, with the sacrifice of Daksha, and the death of Sati: and this work is authority for Siva’s carrying the dead body about the world, and the origin of the Píthasthanas, or places where the different members of it were scattered, and where Lingas were consequently erected. A legend follows of the births of Bhairava and Vetala, whose devotion to different forms of Deví furnishes occasion to describe in great detail the rites and formulæ of which her worship consists, including the chapters on sanguinary sacrifices, translated in the Asiatic Researches. Another peculiarity in this work is afforded by very prolix descriptions of a number of rivers and mountains at Kamarupa-tírtha in Asam, and rendered holy ground by the celebrated temple of Durga in that country, as Kamakshí or Kamakhya. It is a singular, and yet uninvestigated circumstance, that Asam, or at least the north-east of Bengal, seems to have been in a great degree the source from which the Tantrika and Sakta corruptions of the religion of the Vedas and Puranas proceeded.

The specification of the Upa-puranas, whilst it names several of which the existence is problematical, omits other works, bearing the same designation, which are sometimes met with. Thus in the collection of Col. Mackenzie we have a portion of the Bhargava, and a Mudgala Purana, which is probably the same with the Ganesa Upa-purana, cited by Col. Vans Kennedy. I have also a copy of the Ganesa Purana, which seems to agree with that of which he speaks; the second portion being entitled the Krída Khanda, in which the pastimes of Ganesa, including a variety of legendary matters, are described. The main subject of the work is the greatness of Ganesa, and prayers and formulæ appropriate to him are abundantly detailed. It appears to be a work originating with the Ganapatya sect, or worshippers of Ganesa. There is also a minor Purana called Adi, or ‘first,’ not included in the list. This is a work, however, of no great extent or importance, and is confined to a detail of the sports of the juvenile Krishna.