Between Krishna and Kalki: Why the Avatara Lists Disagree About the Ninth Place

Ask which descent comes after Krishna and the tradition gives two answers, both old, both used in worship. The standard list of ten puts the Buddha ninth, and the Buddha is placed after Krishna, inside Kali Yuga. Sri Vaishnava lists and Jayadeva's twelfth-century hymn give the place to Balarama instead. So the familiar claim that nothing divine descends between Krishna and Kalki is not what the core texts say. What they leave is not an empty interval. It is a disputed one.

The difference is not a technicality. If the slot is empty, the question for this age is what carries dharma across the remaining hundreds of thousands of years. If the slot is contested, the question becomes who has standing to fill it, and there the sources are far less decisive than either side of the argument usually lets on. The Bhagavata Purana lists twenty-two descents and then says the number is uncountable. That one verse does more work than any list.

Where Kalki Appears, and How Late the Fullest Source Is

The oldest substantial account is not Puranic. It sits in the Markandeya dialogue of the Vana Parva in the Mahabharata, where the sage describes the collapse of the age and then a brahmana named Kalki Vishnuyasha, born in a village called Shambhala, who destroys the robber-kings and turns the ages over. Chapter numbering for that passage differs between the Calcutta recension that Ganguli translated and the Poona critical edition, by several chapters, so a citation to it is only as good as the edition named beside it.

What Markandeya supplies is short and mostly negative space. Rulers plunder instead of protect, lifespans and harvests shorten, and the restoration arrives from outside the political order rather than from within it. There is no biography.

The Bhagavata Purana repeats the same two proper nouns twice. At 1.3.25 it says that at the junction of the yugas, when rulers have mostly become plunderers, the lord of the world will be born as the son of Vishnuyasha, named Kalki. At 12.2.18 and the verses following, it gives the fuller version: born in Shambhala village in the house of the brahmana Vishnuyasha, mounting a swift horse called Devadatta, sword in hand, riding across the earth. Then Satya Yuga begins.

In the major texts, that is close to the whole of it. A father's name, a village, a horse, a sword. The names hold steady across sources. The narrative is thin.

The Kalki Purana is where the thin narrative becomes a story: a full life, a marriage, campaigns, a cast of named characters. It is also the latest of the sources by a wide margin, and honesty about that is part of using it well. Most of what circulates as Kalki detail, the personal history and the order of the battles, comes from here.

The Kalki Purana is an Upapurana, not one of the eighteen Mahapuranas. Its surviving manuscripts come from Bengal, in Sanskrit written in Bengali script, and their colophons place them in the eighteenth century; no manuscript earlier than the seventeenth is known. Wendy Doniger dates the composition somewhere between about 1500 and 1700 CE. It is a real text with real standing in the traditions that read it. It is not evidence of what anyone believed in the first millennium.

Then there is the arithmetic. On the reckoning standard in Puranic and siddhantic literature, Kali Yuga runs 432,000 years and began a little over five thousand years ago. The tradition's own numbers therefore put Kalki something over four hundred thousand years out. No reading that respects those numbers can place the descent anywhere near the present.

Balarama or the Buddha: Two Lists, Both Traditional

The Bhagavata's long enumeration in Canto 1, chapter 3, runs to twenty-two, opening with the cosmic purusha and closing with Kalki. At 1.3.23 it names the nineteenth and twentieth together, Balarama and Krishna, born among the Vrishnis. At 1.3.24 comes the Buddha. At 1.3.25, Kalki. The Bhagavata does not choose between Balarama and the Buddha; it has room for both. And it places the Buddha after Krishna, which means that by its own account the age we are living in already contains a descent.

The familiar ten is a later compression: Matsya, Kurma, Varaha, Narasimha, Vamana, Parashurama, Rama, Krishna, Buddha, Kalki. Versions of that sequence appear across the Agni, Padma, Garuda, Linga, Narada, Skanda and Varaha Puranas, with variation between them and sometimes within a single text. Ten does work that a list of twenty-two cannot. It is memorable, it fits a temple wall and a hymn metre, and it can be recited by someone who has never opened a Purana.

Sri Vaishnava lists commonly run differently. Balarama takes eighth place, Krishna ninth, Kalki tenth, and the Buddha is absent. The Yatindramatadipika, a seventeenth-century Sanskrit digest of Sri Vaishnava doctrine, records that arrangement. The omission is doctrinal rather than careless. A school unwilling to accept a descent whose stated purpose was to mislead has to put someone else in the slot, and Balarama is already standing there in the Bhagavata's own count.

Jayadeva solves the same problem a third way. The hymn that opens the Gita Govinda gives ten forms: fish, tortoise, boar, man-lion, dwarf, Parashurama, Rama, then Balarama with the plough, then the Buddha, then Kalki. Krishna is not in the ten, because Krishna is the one being addressed. Every verse closes on the same refrain, keshava dhrita-form-sharira jaya jagadisha hare, addressing Keshava as the one who took that particular body. Taking Krishna out of the list is not a demotion. It makes him the person the list is sung to.

These are not three counts of one fixed set with two of them wrong. They are three answers to a prior question the lists never bother to state: what is the number ten for? For the Bhagavata it abbreviates something larger. For the Sri Vaishnava digests it marks a doctrinal boundary. For Jayadeva it is liturgy, and liturgy needs a second person to sing to.

The Bhagavata then removes most of the pressure to decide.

avatārā hy asaṅkhyeyā hareḥ sattva-nidher dvijāḥ

yathāvidāsinaḥ kulyāḥ sarasaḥ syuḥ sahasraśaḥ

"The descents of Hari, the reservoir of goodness, are countless, brahmanas, as thousands of streams flow from an inexhaustible lake."

Bhagavata Purana 1.3.26

A closed list of ten, sitting inside a text that says the number cannot be counted, is a teaching device rather than a census.

The Polemical Frame, and the Generous Reading

There is a passage in the Vishnu Purana, book three, chapters seventeen and eighteen, that gets cited constantly in this argument and usually cited loosely. The gods, losing to the Daityas, appeal to Vishnu, who emits from his own body an illusory form, mayamoha, and hands it to them. Mayamoha approaches the Daityas twice. First as a naked mendicant with a shaven head, carrying a bunch of peacock feathers, teaching a doctrine in which a thing might be and might not be. Then in red garments, benevolent in aspect and soft in speech, telling them to desist from the slaughter of animals in sacrifice. The Daityas abandon Vedic practice, and once stripped of what the text treats as the protection their religion gave them, they lose the next war.

Two things about that passage usually get dropped. The narrative does not enumerate mayamoha among Vishnu's descents; it calls the figure an illusory form produced for a purpose and handed to the gods to use. And the identification of that figure with the Buddha owes a great deal to H. H. Wilson's 1840 English translation, whose editorial chapter heading supplies the name where the narrative gives only a deluder in two costumes.

The first costume is not Buddhist. Nudity and a whisk of shed peacock feathers, the picchika, are Digambara Jain monastic equipment, and a doctrine in which a proposition both is and is not is the Jain syadvada. The red robe of the second appearance is the Buddhist marker. One passage, two rival non-Vedic schools, and a single English chapter heading that has been read back into the whole of it for the better part of two centuries.

The Bhagavata's own line is harder to soften, and hiding it would be worse than quoting it. At 1.3.24 the text says that at the onset of Kali the lord will appear as the Buddha, sammohaya sura-dvisham, for the bewilderment of those hostile to the devas, in the Kikata country, described as anjana-sutah, son of Anjana.

That last word is where the verse complicates itself. Siddhartha Gautama's mother was Maya and his father Shuddhodana. Anjana is neither. The Sanskrit commentators noticed. Sridhara Svamin and Vishvanatha Chakravarti both record the alternative reading ajina-sutah, son of Ajina, and a line of later Vaishnava writers drew the conclusion that the Bhagavata's Buddha and the teacher of Magadha are not the same person. Whether that reconciliation convinces anyone is a separate question, and it looks convenient. What it shows is that the tradition's own commentators found the verse difficult and put the difficulty in writing rather than papering over it.

Jayadeva, writing in the twelfth century, took the ninth place somewhere else entirely. His ninth verse praises the figure for precisely what the Puranic frame condemned: out of a compassionate heart, it says, you decried the slaughter of animals laid down in the rules of sacrifice. Deception in one text and mercy in another, the same slot, the same name, the same broad devotional world.

The tradition never resolved this. It kept both readings in circulation, one in Purana and one in hymn, and in practice it is the hymn that gets sung at festivals. Anyone claiming the sources speak with one voice about the ninth place has not read both of them.

The Traditions That Say the Interval Is Not Empty

Canto 11, chapter 5 of the Bhagavata has the sage Karabhajana explaining to King Nimi that the lord takes a distinct form in each yuga, with a distinct complexion, distinct names and a distinct method of worship. At 11.5.21 the Satya form is white and four-armed, matted-haired, in bark cloth, carrying the staff and waterpot of a brahmacharin. At 11.5.24 the Treta form is red and four-armed, marked with the implements of sacrifice. At 11.5.27 the Dvapara form is dark, in yellow garments, bearing the familiar signs and weapons. Then verse 32 describes the form for Kali Yuga as krishna-varnam tvishakrishnam, accompanied by associates, limbs, weapons and companions, worshipped by sacrifices consisting chiefly of sankirtana, by those of good intelligence.

That phrase has carried an enormous amount of theological weight, because in Sanskrit it is genuinely ambiguous.

Gaudiya Vaishnavism reads it as a description of Chaitanya. Krishna-varnam is taken as one who repeats the name Krishna, and tvisha akrishnam as not dark in complexion, therefore golden. Jiva Gosvamin argues the reading in the Krama-sandarbha, and Krishnadasa Kaviraja quotes the verse in the Chaitanya-charitamrita at Adi-lila 3.52 to establish the point. Inside that sampradaya the interval is not empty in the slightest. Its central descent arrived in 1486 and its whole practice follows from him.

Other commentators read the same words as a description of Krishna himself, worshipped in Kali through the chanting of his names, with no second figure implied at all.

Both readings are defended with grammar, with commentary and with lineage. This article reports them and adjudicates neither, deliberately. A general article is not the venue where a question internal to a sampradaya gets closed.

The Madhva tradition makes a claim that often gets flattened into this discussion and should not be. Madhvacharya identified himself, in his own writings, as the third descent of Vayu, following Hanuman and Bhima. Vayu, not Vishnu. In Dvaita theology that distinction carries load: Vayu is the highest of the dependent beings and never identical with the one independent being. Describing Madhva as an avatara without the qualification misstates the doctrine in exactly the direction the doctrine cares about most.

Comparable claims attach to other acharyas in their biographical literature. The Shankara-vijaya texts present Shankara as a portion of Shiva; Sri Vaishnava tradition treats Ramanuja as a descent of Adishesha. These are hagiographical claims about partial or associated descent, not proposals to alter the list of ten.

What Carries Dharma If Nothing Descends

Set the disputes aside for a moment. On the practical question the texts are consistent, and their answer is not a person.

The first part of it is shastra. The Bhagavata makes the claim about itself, and it makes it in the same chapter as the list of descents. At 1.3.43, eighteen verses after the enumeration closes, Suta says that when Krishna departed to his own abode, taking dharma and knowledge with him, this Purana rose like a sun for those who had lost their sight in Kali. Text outlives teacher. That is why so much energy went into fixing recensions, into commentary, into the machinery of copying and reciting. A tradition expecting a rescuer shortly does not invest that heavily in preservation.

The second is sampradaya. The function of an acharya in the classical model is not to descend but to conserve and to explain: to receive a reading, defend it, and hand it on intact. The whole architecture of guru-parampara exists because the descent is not expected.

The third is the name.

kaler doṣa-nidhe rājann asti hy eko mahān guṇaḥ

kīrtanād eva kṛṣṇasya mukta-saṅgaḥ paraṁ vrajet

"In this ocean of faults that is Kali, king, there is one great virtue: by the chanting of Krishna alone, one is released from attachment and reaches the highest."

Bhagavata Purana 12.3.51

Read the design of that verse rather than its sentiment. It requires no priest, no fire, no initiation into a ritual technique, no qualification by birth, no fixed time or place. Every barrier that the sacrificial and temple systems put between a person and the practice is removed at once.

The next verse, 12.3.52, states the equivalence directly: what meditation on Vishnu gave in Krita, sacrifice in Treta and ritual service in Dvapara, Kali gives through the kirtana of Hari. A verse to nearly the same effect is standardly cited from Vishnu Purana 6.2.17, so the idea was not the property of one school. That is not a consolation prize written into a bad age. It is a lowered barrier written into a bad age, which is a different and more interesting design decision.

What This Piece Does Not Settle

Three things, and the refusal is deliberate in each case.

It names no living person and assesses no contemporary claim to descent. The texts supply criteria, a named village, a named father, a position at the junction of the yugas, and on the tradition's own arithmetic that junction is not now. Beyond restating those criteria there is nothing a reading of the sources can decide. Any assessment of a living figure is a claim about the present, not a claim about the texts, and this site does not make it.

It does not tell Gaudiya readers that Chaitanya is not a descent, nor Madhva readers that Madhva is not the third form of Vayu. Those positions rest on verses, commentaries and unbroken lines of teachers, and they are set out here as what those traditions hold rather than as claims to be scored.

It does not choose between Balarama and the Buddha. Both lists are old, both are chanted in living worship, and the Bhagavata's own statement at 1.3.26 that the descents cannot be counted makes the choice considerably less urgent than the arguing about it suggests.

One detail of arrangement, with no moral attached. The Kalki chapter is Canto 12, chapter 2. The chanting verse comes later, at 12.3.51, after the rescue has been narrated in full. That is not the order assumed by anyone who treats the rescue as the last thing the text has to say.