Part 04 of 16Guptas & Huns

The Guptas and the Huns

Four hundred years in which India had a golden age and Punjab was not invited to it — invaded from the passes wave after wave, until a king almost nobody remembers finally broke the worst man in this book.

1The Golden Age That Happened Somewhere Else

Every Indian schoolbook calls this the classical age — poetry, mathematics, the zero, Kalidasa, temples. Almost none of it happened in Punjab. Punjab spent these four hundred years being fought over.

There is a habit in Indian history writing that this part has to break before it can start.

The period from about 320 to 550 CE is normally told as the story of the Gupta empire — a settled, wealthy,

confident state on the Ganges, with a court that produced some of the finest work in Sanskrit and some of the best

mathematics on earth. That is all true. It is also a story about a different part of India. The Guptas never really

ruled Punjab. They touched it, listed it, taxed parts of it on paper, and fought on its edge. They did not govern it

the way they governed Magadha or Malwa.

What actually happened on the five rivers between 250 and 650 CE is a completely different story, and a harder

one. It goes like this. The Kushan empire that ended Part Three dies. For about a hundred years, the old armed

republics of the plain govern themselves again — the only long stretch of self-rule Punjab gets between Alexander

and Ranjit Singh. Then the passes open, and they keep opening: the Kidarites, then the Alchon Huns under

Toramana and Mihirakula, then a long slow fragmentation. By the time the period ends, the Buddhist civilisation

of Taxila and Gandhara — one of the great cultures of the ancient world, four centuries in the building — has been

physically destroyed and never comes back.

Four things you need to hold in your head

  • “Huna” is not one people. Indian sources use the single word Huna for at least two, probably three, different groups arriving over 150 years — Kidarites, then Alchon, with Hephthalites next door. They were not the Huns of Attila, though they were relatives and rivals of the same broad Central Asian world. When a Sanskrit inscription says Huna, treat it like the word “foreigner”: accurate about the direction they came from, useless about who exactly they were.
  • Punjab in this period is written about almost entirely by outsiders. Our sources are a Persian king’s cliff inscription, Chinese pilgrims and envoys, victory pillars set up in Madhya Pradesh, and a Kashmiri poet writing six hundred years later. What survives from inside Punjab is a handful of stones — the most important of them a monastery dedication in the Salt Range, dated by the reign of a Hun king, which describes no event at all. Not one narrative account written in Punjab about any of these wars survives. That absence is the single biggest fact in this part.
  • Coins do a lot of the work here. When there are no chronicles, money becomes the record. Who struck it, in whose name, in what metal, and how badly — that is how we date the arrival and departure of whole dynasties. You will see coins used as evidence on nearly every page.
  • This is the period Gandhara dies. Taxila, the great university city of Part Two and Part Three, is a ruin by the end of it, and stays a ruin. That is not a footnote. It is the largest single loss in the whole book.
THE HONESTY LABELS, AGAIN

Same three as before. CERTAIN means contemporary evidence — an inscription carved at the time, a coin, or an

eyewitness. LIKELY means it is reported by good sources writing later and most historians accept it. STORY

means legend or a much later tradition, which may have a real war underneath it.

Part Four is unusually lopsided. Several of these wars are known from a single carved stone, which makes the fact of the

war certain and everything else guesswork. Others are known only from a poet writing after six hundred years. You will

see both labels used honestly, sometimes in the same war.

2Persia Takes Peshawar — The Quiet End of the Kushans LIKELY

A superpower swallowed the north-west and left one sentence about it, carved on a wall in Iran.

Won by: Sasanian Persia — in a war nobody described

When: Roughly 230–270 CE, in two pushes a generation apart.

Where: Bactria, the Kabul valley, Gandhara, and the Indus — the country just west and north of the Punjab plain.

Who: Ardashir I and his son Shapur I of Persia, against the last strong emperors of the Kushan house.

The evidence: One enormous royal inscription, and several thousand coins. No battle account of any kind.

Why it happened

Part Three ended with the Kushan empire cracking — beaten in the north-west by the Han Chinese, and then

having its Punjab provinces pulled away by the armed republics of the plain. An empire in that condition is a

target.

At exactly that moment, Iran changed hands. In 224 CE a provincial nobleman named Ardashir destroyed the

Parthian empire that had ruled Iran for four centuries and founded a new dynasty, the Sasanians (the ruling

family of Persia from 224 to 651 CE). New dynasties need victories, urgently, because a new dynasty is just a

successful rebellion until it wins a war somebody else recognises.

Ardashir looked east and saw a rich, tired empire sitting on the road to India. The Kushans had spent two

hundred years advertising their wealth in gold coinage. That advertisement worked exactly as well as

you would expect.

What happened, as far as we can tell

1. Ardashir moves east, some time in the 230s. He takes Sistan, Merv, Balkh and Khwarezm — the whole

  • northern approach. The Arab historian Tabari, writing about seven hundred years later, says that after this the kings of the Kushans, of Turan and of Makran sent envoys to him offering submission. Seven hundred years is a long gap. Treat it as a memory, not a report.
  • Shapur I finishes the job, some time between 240 and 270. Shapur was one of the most successful soldiers of the ancient world — the man who captured a Roman emperor alive. His eastern campaign is not described anywhere. We only know its result.
  • The result is carved on the Ka‘ba-ye Zartosht. On the wall of a stone tower at Naqsh-e Rustam in Iran, Shapur had a long inscription cut in three languages, listing what he ruled. The list includes Hindestan, and “Kushanshahr up to Peshawar”, and on to the borders of Kashgar, Sogdia and Tashkent. This is a
  • contemporary document written on the winner’s orders. On its own terms it is CERTAIN.
  • Persia installs governors and calls them Kushan kings. The men Persia put in charge of the east took the title Kushanshah — “King of the Kushans” — and struck coins in the old Kushan style, same shape, same script, same standing king at a fire altar, with the Persian god Ahura Mazda quietly swapped in. Modern historians call them the Kushano-Sasanians.
  • A rump Kushan state survives in the east. Around Mathura, and probably in parts of the Punjab plain, later Kushan kings — Vasudeva II, Shaka, Kipunada — keep minting into the middle of the fourth century, almost certainly as somebody’s dependants. Then they stop, and the Kushan name disappears from India for good.
WHAT “UP TO PESHAWAR” DOES AND DOES NOT MEAN

Shapur’s inscription is a claim of overlordship, not a map of occupied territory, and the phrase stops at Peshawar. There

is no evidence at all that a Sasanian army fought on the Punjab plain, crossed the Jhelum, or garrisoned anything east

of the Indus. What Persia took was the mountain approach and the river frontier — the doorway, not the house.

That distinction matters for the rest of this part. Whoever holds Gandhara and the Kabul valley can enter Punjab

whenever they like. From 250 CE onwards, that doorway is almost always in foreign hands.

WHY PERSIA WON
  • They picked their moment. They attacked an empire that had already lost its Chinese war and its Indian provinces. Nobody in Iran had to beat the Kushans at their best.
  • They came from the strong side. Persia’s power base in Iran sat directly against Bactria, the Kushan empire’s richest and most exposed half.
  • They took the passes first. Sistan, Merv, Balkh, Khwarezm — the whole northern arc — before touching Gandhara. Standard practice for anyone competent, and almost nobody in this book does it.
  • They did not try to govern India. They installed governors under a local-sounding title, kept the existing coinage and left the machinery alone. Cheap, and it worked for a century.
WHY THE KUSHANS LOST
  • Two fronts, one army. They were being pulled apart by Persia in the west and by the Yaudheyas and their neighbours in the east at the same time.
  • Their empire was a trade route, not a country. Kushan wealth came from the Silk Road caravans. Cut the road in Bactria and the whole structure starves.
  • Their capital was in the wrong place. Purushapura — Peshawar — sits right at the mouth of the passes. Excellent for an empire that faces both ways. Fatal when the west turns hostile.
  • Nobody in India came to help. The Kushans were foreign rulers who had become Indian. When the crisis came, they were still foreign enough that no Indian power fought for them.
THE STORY WORTH REMEMBERINGYou Can Watch an Empire Die in the Metal

We have no account of the Kushan collapse. What we have instead is their money, and it tells the story

better than a chronicle would.

Start with a gold coin of Kanishka, around 130 CE. It is heavy, it is nearly pure gold, and the design is

confident: the emperor standing in a heavy coat and felt boots, sacrificing at an altar, with a god on the

back and clean lettering all around.

Now put next to it a gold coin of Kipunada, one of the last Kushan kings, around 350 CE. It is the same

design. It is meant to be the same coin. But the gold has been cut with so much base metal that it is pale, the

weight has dropped, the die has been copied from a copy from a copy, and the king’s face has degraded into

a few scratched lines. The lettering has become a pattern that imitates lettering. The mint no longer

employed anyone who could read it.

Nobody wrote down the fall of the Kushan empire. Nobody needed to. Two hundred years of decline sit in a

coin tray, in order, in metal you can weigh — an empire that kept making the same promise on the same

coin long after it had stopped being able to keep it.

3The Free Century — When Punjab Governed Itself LIKELY

The longest stretch of self-rule between Alexander and Ranjit Singh, and it has no chronicle, no king list, and no name.

The republics of the plain — Yaudheyas, Madrakas, Kunindas, Audumbaras, WON BY

Trigartas

WHEN
  • Roughly 250–350 CE, between the fall of the Kushans and the arrival of the Guptas.
WHERE
  • The whole plain: Rohtak and the Sutlej– Yamuna country, Sialkot, the Kangra hills, the Beas–Ravi doab, Bahawalpur.
WHO
  • Half a dozen gana-sanghas (tribal republics — states run by an assembly of clan heads, not a king).
THE EVIDENCE
  • Almost entirely coins and clay seals. No history of these states was ever written.

Why this counts as a war

Because you cannot get from “Kushan province” to “independent republic minting its own money” without

fighting, and Part Three ended with exactly that fight — the revolt that threw the Kushans out of the plain.

What we have here is the aftermath, and the aftermath lasted about a hundred years. For roughly four

generations, no empire ruled Punjab. The land of the five rivers was governed by its own people, in its own

way, using a political system that had already been old when Alexander met it: the republic, run by an assembly,

electing a war-leader when it needed one.

And we know almost nothing about it, because the men who governed it were not in the habit of carving their

achievements into rock.

Who these people were

1. The Yaudheyas. The biggest and most militant. Their country ran from around Rohtak across the Sutlej–

  • Yamuna plain and down into Bahawalpur. Their coins carry the legend Yaudheya-ganasya jaya — “victory to the Yaudheya nation” — and the god Karttikeya, the god of war, on the front. A clay sealing found at Bijaygarh calls them “the Yaudheyas who hold the spell of victory.” 2. The Madrakas. The people of Madra, around Sakala — modern Sialkot. You met their city in Part Three as the capital of the Greek king Menander. It is about to become the capital of a Hun.
  • The Kunindas. The Sivalik foothills and the upper Yamuna–Sutlej valleys, minting handsome silver coins in the name of a ruler called Amoghabhuti.
  • The Audumbaras in the Beas–Ravi country around Pathankot, the Trigartas in the Jalandhar doab and Kangra, and the Arjunayanas and Malavas to the south-east. Several of these names are still on the map, if you know how to look — Trigarta is Kangra, Jalandhara is Jalandhar.
WHY COINS ARE THE WHOLE ARGUMENT

A state that mints coins in its own name is claiming sovereignty. It is a public statement, in metal, that says: we tax here,

we decide value here, nobody above us. When you find Kushan coins in the Punjab plain up to about 250 CE, then

Yaudheya coins in the same soil for the next century, then Gupta coins after that, you are reading a change of

government even though nobody wrote one down.

The limits are real too. Coins do not give you dates, battles, leaders, or borders. They tell you that Punjab was

independent. They cannot tell you how it fought for it.

WHY THE REPUBLICS WON
  • They only had to beat a corpse. The Kushan state in Punjab had been cut off from Bactria by the Persians. Its garrisons were on their own.
  • They were already armed. These were societies where the clan and the militia were the same thing. There was no army to raise — the population was the army.
  • Nobody else wanted the job. Persia stopped at Peshawar; the Guptas had not been invented yet. For a century there was simply no empire able to come.
  • They had beaten empires before. The Yaudheyas had survived being crushed by Rudradaman in the second century, as his own inscription boasts. Survival is a habit.
WHY THEY LOST IT AGAIN
  • Six republics, six armies. They never merged. Each was strong enough to defend itself and none was strong enough to defend the region.
  • No permanent command. An assembly that elects a war-leader for a crisis is excellent at repelling a raid and hopeless at fighting a ten-year campaign.
  • They did not fortify the passes. Not one of these states put a garrison in the Khyber or the Kabul valley. They defended their own fields and left the doorway open.
  • They wrote nothing down. Which cost them nothing militarily and cost them everything historically. We remember the men who beat them.
THE STORY WORTH REMEMBERINGA Nation That Signed Its Name on Money

There is no Yaudheya chronicle. No king list. No capital we can point to with confidence. No temple with

their name on it, no poem, no epic, no tomb.

What there is, is copper. Thousands of small copper coins, turned up by ploughs and canal-diggers across

Rohtak, Ludhiana, Sonepat and Bahawalpur for two hundred years, each one carrying six words of Brahmi

script: Yaudheya-ganasya jaya — victory to the Yaudheya nation.

Not victory to the king. There was no king. Victory to the nation — to the gana, the assembly, the people

who voted. In an age when every other state in India was announcing the personal glory of a named man,

the Yaudheyas were stamping a collective noun onto their money.

They did it for four hundred years, through the Kushans, through the Guptas, until the Huns came and the

coins stop. They are one of the longest-lived political communities in Indian history, and we do not know

the name of a single one of their leaders.

They were not silent. They just wrote on the smallest surface available, in the largest possible

number of copies, and trusted the ground to keep it. It worked.

4Samudragupta Comes to the Frontier — And Stops CERTAIN

The greatest conqueror of the age reached the edge of Punjab, took its submission on paper, and went home.

Won by: Samudragupta — on parchment, possibly without a battle

When: Somewhere in c. 350–375 CE. The inscription carries no date, which is itself annoying.

Where: The south-eastern edge of the Punjab plain and the desert margin — Yaudheya, Madraka, Arjunayana and Malava country.

Who: Samudragupta of the Gupta empire, against the republics; and, at a distance, the Kushan and Saka rulers of the far north-west.

The evidence: One inscription, composed by the king’s own war minister, carved on a pillar Ashoka had put up six hundred years earlier.

Why it happened

Samudragupta, who ruled from about 335 to 375 CE, is the man who turned the Guptas from a regional dynasty

into an empire. His method, described in his own record, was blunt and clever at the same time: in the Ganges

plain he destroyed kings and annexed their land; everywhere else he beat people and gave the land

back.

That second policy is the whole explanation for what happens in Punjab. The north-west was not worth occupying.

It was far away, it was full of armed farmers who had just spent a century proving they could throw out an empire,

and beyond it sat a frontier that would need permanent defending. What Samudragupta wanted from Punjab

was a quiet flank and a tribute payment. He got both, and the record says so.

What the pillar says

The Allahabad pillar inscription — the prashasti (a formal praise-poem in verse and prose, carved in stone) —

was composed by a court official named Harishena, whose titles included Sandhivigrahika, minister of peace and

war. He was, in modern terms, the foreign minister writing the press release.

Its Punjab passage lists the frontier peoples who submitted. Among them: the Malavas, the Arjunayanas, the

Yaudheyas, the Madrakas, the Abhiras, the Prarjunas, the Sanakanikas, the Kakas and the Kharaparikas. What

they did, in the inscription’s own words, was give “all kinds of taxes, obey orders, and perform obeisance.”

Then it goes further out, to the powers beyond the Indus. The Daivaputra-Shahi-Shahanushahi — that is the

old Kushan royal title, devaputra shahi shahanushahi, “son of heaven, king, king of kings”, so it means the Kushan

or Kushano-Sasanian ruler of the north-west — and the Saka lords are said to have offered “self-surrender,

offering their own daughters in marriage, and a request for the administration of their own districts and provinces.”

READ THAT LAST SENTENCE AGAIN

Strip the flattery and it says: foreign rulers sent embassies, sent marriage alliances, and asked to keep governing their

own territory. That is a description of diplomacy. A court poet is entitled to call an embassy a surrender, and Harishena

does, and we should not.

Notice also what is not claimed. No siege. No named battle. No city taken. No governor appointed. For the north-west,

Harishena lists titles and gestures — because titles and gestures were what there was.

Did any fighting actually happen?

Probably some, and probably not much. The honest position is this:

1. The Yaudheyas kept minting. Their coins continue after Samudragupta. A conquered and annexed people

  • does not usually keep issuing money in its own name.
  • Gupta gold appears in the region, but Gupta administration does not. We have no Gupta land grants, no Gupta governors and no Gupta officials recorded for the Punjab plain in this reign.
  • Later Gupta kings had to deal with the same peoples again. If they had been absorbed in 360 CE, they would not still be there as independent actors afterwards.
  • The most likely reality is a large armed demonstration. An imperial army marches to the frontier. The republics count it, calculate, and buy peace with tribute and formal submission — which costs them money and dignity but leaves their assemblies, their armies and their coinage intact. Both sides get what they actually want. Only one side writes it down.
ONE CAUTION ABOUT THE PILLAR

Even the pillar’s address is disputed. Several scholars think the column originally stood at Kausambi, about fifty

kilometres away, and was moved to Allahabad later — possibly by Akbar when he built the fort. If so, its Ashokan edicts

were addressed to a different city than the one they now stand in. It is a small thing, but it is a useful reminder: in this

period, even the objects move.

WHY SAMUDRAGUPTA WON
  • He knew what he was buying. He wanted the flank quiet and the tribute flowing, not a province to garrison. He priced the campaign correctly.
  • Overwhelming, visible force. The empire that had just eaten the Ganges plain arriving on your border is an argument that does not need to be made twice.
  • He offered terms worth taking. Submit and keep your government. Against that, fighting is a bad trade even if you might win.
  • He controlled the record. His minister wrote the only surviving account, and wrote it as total victory. Two thousand years later, that is still the version in the textbooks.
WHY THE REPUBLICS GAVE WAY
  • They could not combine, again. Nine separate peoples are listed. Nine separate decisions were made. Not one alliance appears anywhere in the record.
  • Their strength was defensive. A militia can hold its own fields against a raid. It cannot threaten an empire, so it has no way to make the empire go away.
  • Submission was survivable. The same brutal logic as under Persia in Part Two: paying is cheaper than dying, and a republic that survives can outlast the man taxing it.
  • And in the end it did not save them. They bought a century. Then the Huns came, and there was nothing to buy them off with.
THE STORY WORTH REMEMBERING — THE IRONIC ONEThree Empires on One Stone

The pillar Samudragupta’s minister used was not a new one. It was a polished sandstone column already

six hundred years old, put up by Ashoka in the third century BCE, and it already had writing on it.

Ashoka’s pillar edicts are about restraint. They are instructions to his officials and his people about self-

control, about care for prisoners, about not injuring living creatures, about the emperor’s duty to do good

rather than to be feared. They are the most famous experiment in non-violent government in the ancient

world, carved by a man who took up that policy after a war he could not forgive himself for.

Harishena carved a war record on top of them. Kings destroyed, kings captured, kings restored as

tributaries, tribes made to pay, a list of conquests running from Bengal to the Deccan, on the same stone, in

the same beautiful script, a few inches below an emperor’s public promise to stop conquering.

And it does not end there. In 1605 the Mughal emperor Jahangir had his own accession record cut into the

same column, in Persian, straight across the older texts.

Three empires, eighteen hundred years, one pillar — and each of them treated the previous one’s

words as blank space. If you want the shortest possible summary of how power in India talks to its own

past, it is standing in Allahabad fort.

5The Seven Mouths of the Indus LIKELY

A seven-metre iron post standing in Delhi says a king crossed the rivers of Punjab and beat somebody. Every other word of that sentence is argued over.

Won by: A king called Chandra — almost certainly Chandragupta II

When: Probably c. 380–410 CE, if the king is Chandragupta II. Undated.

Where: The Indus and its river system — either the delta in Sindh, or the five rivers themselves.

Who: “Chandra”, against a people the inscription calls the Vahlikas.

The evidence: Six lines of Sanskrit on the Mehrauli iron pillar, now standing in the Qutb complex in Delhi.

What the pillar actually says

The inscription is short, and it is one of the very few Gupta-period records that describes fighting in the north-

west at all. In the standard translation, the relevant lines run: the king, in battle in the Vanga countries, turned

back with his own chest the enemies who came against him united; and “having crossed in warfare the seven

mouths of the Sindhu, the Vahlikas were conquered by him”; and the southern ocean is still perfumed by the

breezes of his valour.

That is it. One clause about Punjab, or Sindh, in a poem about a dead king. And around that one clause there is a

hundred and fifty years of argument.

The four things nobody agrees on

1. Which Chandra? The inscription names only “Chandra”. The majority view is Chandragupta II (ruled

  • roughly 375–415 CE), on the grounds of script style, the mention of the Vangas, and the sheer scale of the claims. But serious scholars have argued for Chandragupta I, and for a completely different king, Chandravarman of Pushkarana, whose name turns up on an inscription in Bengal.
  • Who were the Vahlikas? Two readings. Either Bahlika means Bactria — the country around Balkh, north of Afghanistan — which would make this an enormous overseas campaign. Or it means the Bahlikas of the
  • Punjab, a people named repeatedly in the Mahabharata as living in the land of the five rivers. The second reading is geographically far easier and would put this war squarely in this book.
  • Where are the seven mouths? Either the mouths of the Indus in the delta, which suits a Bactrian campaign badly and a Sindh campaign well. Or the rivers of the Punjab system, which the ancients often counted as a bundle. Sapta Sindhu — the seven rivers — is one of the oldest names in this book. It appears in the Rigveda, in Part One.
  • Where was the pillar originally? Not Delhi. It was moved there in the medieval period, and its first home is unknown — Udayagiri in central India is the most-argued candidate, on the strength of the Gupta connection. A victory monument whose location we have lost is a victory monument that has lost half its meaning.
HANDLE WITH CARE

This is the most honest way to put it. Something happened, and a Gupta-era king crossed a river system in force

and won. Beyond that, this part will not pretend. If the Vahlikas were the Punjab Bahlikas, this is a major Gupta

campaign on the five rivers and belongs at the centre of the chapter. If they were Bactrians, it is a claim about a war

fought a thousand kilometres away that may never have happened at all.

Be suspicious of anyone — in a textbook, in a documentary, in a comment section — who tells you which one it is

without telling you there is a question.

What we can say about Chandragupta II with more confidence

Quite a lot, actually, and none of it from the pillar.

  • He finished the Saka war in the west. Chandragupta II destroyed the Western Kshatrapas, the Saka dynasty that had ruled Gujarat and Malwa since Part Three, and took their territory. We know this because he took over their silver coinage as well, restriking their distinctive thin silver coins with Gupta types. Coinage again, doing the work of a chronicle.
  • That war gave the Guptas the sea. With Gujarat came the western ports and the trade with Rome and Persia. It is the single most profitable thing any Gupta king ever did.
  • He married the frontier rather than fighting it. Gupta policy under him and his father runs on marriage alliances — with the Vakatakas in the Deccan, with the Nagas. It is the same logic as Samudragupta’s: buy the flanks, spend the army in one direction.
WHY THE GUPTAS WON IN THIS GENERATION
  • They fought one war at a time. West first, into Gujarat and the sea trade. The north-west got demonstrations and diplomacy, not campaigns.
  • Money, then soldiers. The Saka conquest paid for everything that came after, including the army that would meet the Huns two generations later.
  • They absorbed rather than destroyed. Taking over the enemy’s own currency system is a conqueror telling merchants that nothing has changed. It keeps the tax base alive.
WHY IT DID NOT LAST
  • Nothing was built in the north-west. No fort line, no permanent frontier army, no standing command in Punjab. The Guptas treated the passes as somebody else’s problem for a hundred years.
  • Prosperity in the east, exposure in the west. The empire’s wealth and its weak point were at opposite ends of it, with a very long march in between.
  • The Punjab republics were left as a buffer. Buffers work until something arrives that they cannot stop, and then the empire discovers it has no second line.
THE STORY WORTH REMEMBERINGSixteen Hundred Years Standing in the Rain

Forget the inscription for a moment and look at the object.

The Mehrauli pillar is a single shaft of wrought iron, about seven metres tall and over six tonnes,

made some time around 400 CE by forge-welding lumps of hot iron together, blow by blow, by hand. Nobody

in Europe could make an iron object at that scale for another thousand years or more. It is, by itself, one of

the finest pieces of metalwork surviving from the ancient world.

And it has stood outdoors, in the open Delhi air, through sixteen centuries of monsoons — and it has barely

rusted.

That is not magic and it is not a lost secret. Modern metallurgical work, most thoroughly by R.

Balasubramaniam, traced it to the iron’s unusually high phosphorus content, which under Delhi’s

wetting-and-drying cycles forms a thin, dense protective film over the metal — a layer that seals the

surface instead of flaking off the way ordinary rust does. The ancient smiths did not know the chemistry.

They knew their ore and their process, and the process happened to be right.

There is a sad modern footnote. For generations, visitors stood with their back to the pillar and tried to link

their hands around it behind them — the story was that if you could, your wish came true. Millions of arms

wore a bright polished band into a protective layer that had taken centuries to form. The pillar survived

every empire in this book and was being worn away by affection. It is fenced now.

6Kidara Crosses the Great Mountains — The First Wave LIKELY

The day India learned a new word for the enemy.

Won by: The Kidarites — the first of the peoples India would call Huna

When: Roughly 380–430 CE. The dating rests on coins and is argued over by decades.

Where: Bactria first, then Gandhara, Taxila, Kashmir and the western Punjab.

Who: Kidara and his successors, against the Kushano-Sasanian governors and whatever local powers stood in the way.

The evidence: Chinese dynastic histories, Kidarite coins, and a Byzantine diplomat’s report. Nothing Indian.

Why it happened

Something enormous was moving across the steppe in the fourth century. In Europe, it produced the Huns of Attila

and pushed Gothic nations across the Danube into the Roman empire. In Central Asia, it produced a series of

confederations — the Chionites, the Kidarites, the Hephthalites, the Alchon — that came down out of the grass in

wave after wave over roughly two hundred years.

We do not know what started it. Drought, population pressure, a collapse further east, one successful war-leader

— all have been argued. What we can see clearly is the effect: for two centuries, every settled empire from

Rome to China had a horse-archer problem, and they all had it at once.

Persia had already weakened the Kushan structure in the north-west and replaced it with governors. Governors

are easier to overthrow than dynasties. When the Kidarites arrived, the doorway to India was held by a caretaker.

What happened

1. The Kidarites take Bactria from the Kushano-Sasanians, some time in the middle or later fourth century, and

  • start minting there.
  • They come south through the passes. The Chinese Wei shu, compiled in the sixth century from earlier records, tells it plainly: the king of the Great Yuezhi, named Jiduoluo — Kidara — was brave and warlike, sent his troops south, crossed the great mountains, and subdued the kingdoms north of Gandhara.
  • They take the Kushan brand with them. Kidarite coins copy Kushan and Sasanian designs and use the old titles. In Kashmir and Gandhara their money imitates the coinage of Vasudeva so closely that numismatists spent a long time untangling it. The new masters advertised themselves as the old ones.
  • They push into the western Punjab. How far east is genuinely unclear. Their coins are common in Gandhara and Kashmir, thinner on the plain. What is not in doubt is that from about 400 CE onward, the country between the Indus and the Jhelum was under a Central Asian dynasty again.
  • They fight Persia for eighty years. The Sasanian kings Bahram V and Yazdegerd II campaigned against them repeatedly. The Byzantine diplomat Priscus records that around 456 CE the Kidarite king refused to keep paying tribute to the Persian king Peroz, and the war restarted. Persia only finished them — in Bactria — with Hephthalite help.
THREE NAMES, ONE WORD

This is where the confusion in Indian sources begins, so it is worth being precise once. Kidarites arrive first, roughly

380–450. Hephthalites — the “White Huns” of Byzantine writers — dominate Bactria and Iran’s eastern frontier from

roughly 450 to 560, and mostly stay north of the mountains. Alchon — the group of Toramana and Mihirakula — come

into Gandhara and India from around 460 and are the ones who do the real damage here.

Indian inscriptions call all of them Huna. Chinese sources use different names again. Older history books merge the lot

into “the White Huns”, which is why so much popular writing on this period is a muddle. When you read “the Hunas

invaded India”, always ask which ones, and in which decade.

WHY THE KIDARITES WON
  • Horse archers against garrison troops. Mounted archery gives you the choice of whether to fight at all. An army that can decline battle and reappear a hundred kilometres away cannot be pinned down by infantry.
  • They arrived at a caretaker frontier. Gandhara was held by Persian-appointed governors with no local roots and no reason to die for the job.
  • They took the passes, which nobody was holding. Same doorway, fourth time in this book.
  • They dressed as the previous regime. Old titles, old coin designs, old scripts. Merchants and tax-collectors carried on. Continuity is cheaper than conquest.
WHY THE NORTH-WEST LOST
  • Nobody owned the frontier. Persia had stopped at Peshawar. The Guptas had stopped a thousand kilometres short of it. The republics had never gone there. It was, in practice, unclaimed.
  • Gandhara was rich, exposed and unwarlike. Centuries of monastery-building had made it one of the wealthiest religious landscapes in Asia and one of the least militarised.
  • Cities cannot run away. Taxila, Peshawar and the great monasteries were fixed, visible and full of portable wealth.
  • The warning went unheard. India had fifty years’ notice that a new power was on the other side of the mountains, and used it to do nothing.
THE STORY WORTH REMEMBERINGThe Day India Learned a New Word

Languages record invasions. Sanskrit records this one.

Before the late fourth century, the word Huna is essentially absent from Indian writing. After it, the word is

everywhere — in inscriptions, in the Puranas, in the lists of northern peoples that Indian texts liked to

compile, in poetry as a byword for a distant and dangerous edge of the world. It becomes a fixed term, the

way “Mughal” or “Angrez” later would.

What makes it worth pausing over is the reach of it. The same shockwave that put a new word into Sanskrit

put related words into Latin and Greek at almost the same time, as Gothic refugees crossed the Danube and

Roman writers reached for a name for whatever was behind them. Two civilisations at opposite ends of

the world, neither with any clear idea of the other, both suddenly needing a new noun in the

same century.

And here is the part that should stay with you. India got its warning first. The Kidarites were in Gandhara

for two full generations before the really destructive wave arrived. Two generations is enough time to fortify

passes, raise a frontier army, garrison Taxila, agree on a joint command.

The word entered the language decades before anyone acted on what it meant.

7Skandagupta and the Whirlpool CERTAIN

The one clear Indian victory in this entire part — and it cost so much that the empire never recovered.

Won by: Skandagupta of the Guptas — the man who bought India seventy years

When: Around 455–460 CE, near the start of his reign. Precise dating is impossible.

Where: Not stated. Almost certainly the western approaches — the country between the Indus and the upper Ganges plain.

Who: Skandagupta against the Hunas — most probably Kidarites, though this is argued.

The evidence: The Bhitari pillar inscription, carved for Skandagupta himself, and the Junagadh rock inscription of 457 CE.

Why it happened

Skandagupta came to the throne in about 455 CE in the middle of a crisis, and possibly in the middle of a disputed

succession. His own inscription says he restored the ruined fortunes of his family, which is not the sort of thing a

king says if everything was fine.

Two enemies are named on the Bhitari pillar. One is a people called the Pushyamitras, who had grown powerful

and against whom he fought a hard battle. The other is the Hunas. The empire was being attacked in two

directions at the accession of a new and possibly contested king. That is the worst combination in politics.

What happened

1. He fights the Pushyamitras first and wins — the inscription describes him joining that fight in person and

  • creating a disturbance like a terrible whirlpool, and says the earth shook under his arms. Court language, but the emphasis on personal combat is consistent and unusual.
  • Then the Hunas come. The inscription records the clash and records that he prevailed. It does not give a place, a date, a number, a season or a name for the opposing commander.
  • He wins decisively enough to stop them for a generation. This is the strongest argument, and it is not from the inscription — it is from the silence afterwards. There is no Huna power inside India again until the 470s at the earliest, and no Huna king ruling Indian territory until about 490. Something stopped them, hard, for thirty years.
  • He then rules for about twelve more years, repairs the great Sudarshana dam in Gujarat after a flood — recorded in the Junagadh inscription of 457 CE, which is where we get his date — and dies around 467.
  • And the empire visibly weakens under him. Gupta gold coinage in his reign gets worse: more base metal, uneven weights, hurried dies. The classic explanation is the cost of the wars. It is a reasonable inference rather than a proven one, but the coins are real and the direction is not in doubt.
THREE GENUINE DISPUTES, STATED PLAINLY

Which Hunas? Older books say Hephthalites or the Huns of Toramana. On the dates, that is difficult: Toramana’s

people arrive later. Most current scholarship makes Skandagupta’s enemy the Kidarites, the first wave. The inscription

just says Huna and leaves us to argue.

Who were the Pushyamitras? Nobody knows. They are not otherwise recorded as a power. Some scholars have even

proposed a different reading of the damaged text. A war we are certain about, against a people we cannot identify.

How decisive was it? The inscription is a praise-poem written for the winner. Read alone, it proves the fight happened

and that Skandagupta was pleased with the result. The thirty-year gap that follows is the real evidence, and it is

circumstantial.

WHY SKANDAGUPTA WON
  • He fought them before they were established. He met the Hunas while they were still raiding from outside, not ruling from inside. Every later Indian ruler in this part has to fight them as an occupying power.
  • The Gupta army was still intact. He inherited the fully funded military machine of a century of successful reigns, and he used it immediately.
  • He led from the front. Three separate passages stress his personal presence in the fighting. For an army facing an unfamiliar and terrifying enemy, that is not decoration; it is the difference between a line that holds and a line that runs.
  • He dealt with the internal enemy first. Pushyamitras, then Hunas. In order, not at once. This is the correct answer to a two-front problem and it is depressingly rare in this book.
WHY THE VICTORY DID NOT HOLD
  • He beat the army, not the source. The steppe kept producing new confederations. Nothing Skandagupta could do in India addressed that.
  • The frontier was still nobody’s job. After the victory, no permanent north-western command was created, no fortress line built, no standing frontier force stationed. The next wave found the same open door.
  • The war broke the money. Debased gold is the visible symptom of a state spending more than it has.
  • The succession collapsed. After Skandagupta died the Gupta house fell into competing branches and rival claimants. Within twenty-five years of the great victory, there was no longer a single Gupta empire to defend anything.
THE STORY WORTH REMEMBERING — THE HUMAN ONEHe Went to Tell His Mother

The Bhitari pillar is a formal royal inscription in high Sanskrit, and most of it does what such things do: it

lists ancestors, piles up compliments, and compares the king to gods.

Then, in the middle of it, comes one of the most unexpected passages in Indian epigraphy. Having beaten

his enemies and restored his family’s ruined fortunes, the new king — in the words carved for him —

cried “the victory has been achieved,” and betook himself to his mother, whose eyes were full of tears from

joy, just as Krishna, when he had slain his enemies, betook himself to his mother Devaki.

Bhitari pillar inscription of Skandagupta, in Fleet’s translation

Consider what is being described. A man in his twenties or thirties, who has just inherited a crumbling

empire in a contested succession, has fought a hard war on two fronts and won it. He has the resources of

an empire to celebrate with.

What he does — or what he chose to have carved in stone for all future time, which tells you the

same thing about what he valued — is go and find his mother and say: we are all right.

The Krishna comparison is not accidental. In the story, Devaki had spent years imprisoned and had lost

child after child before Krishna came back and freed her. Whoever wrote this line was saying that this

mother had also been through something.

Everything else on that pillar is about the throne. That one sentence is about a family that had been

frightened.

8Toramana Takes the Punjab CERTAIN

For the first time since Alexander, a foreign conqueror did not stop at the five rivers. He kept going, and reached the middle of India.

Toramana of the Alchon Huns — who ruled from the Salt Range to Madhya WON BY

Pradesh

WHEN
  • Roughly 490–515 CE. His first regnal year is recorded; the calendar it belongs to is not.
WHERE
  • Gandhara, the Salt Range, the whole Punjab plain, Kashmir, and on into Malwa and central India.
WHO
  • Toramana, king of the Alchon, against the fragmenting Gupta empire and everyone in between.
THE EVIDENCE
  • Two inscriptions in his own name — one of them in Punjab — plus copper plates and a large coinage.

Why it happened

Because there was nothing left to stop it. In the thirty years after Skandagupta’s death the Gupta empire came

apart into competing branches. The Punjab republics were long past their strength. The Kidarites had been broken

in Bactria by Persia and the Hephthalites, which removed the only organised power between the Alchon and the

passes.

Every barrier that had held from 455 to 480 was gone by 490. Toramana did not defeat India’s defences. He

arrived after they had dissolved.

What happened

1. The Alchon establish themselves in Gandhara and Kabul from around 460 onward under earlier kings, of

  • whom the best-attested is Khingila. They mint heavily. Their coins carry a distinctive mark — a tamga, a clan symbol used like a brand — and the name Alchono in Bactrian script.
  • Toramana comes down onto the plain some time around 490. There is no account of the campaign. The evidence for it is that he ends up in charge. 3. He is recorded ruling in the Punjab. The Kura inscription, found at Kura in the Salt Range in the Jhelum district, is dated by the reign of the great king Toramana Shahi Jauvla. It is the single most important document of Hun rule in Punjab, and we will come back to what it actually records, because it is not what you would expect.
  • He reaches central India. The Eran boar inscription, cut at Eran in what is now Madhya Pradesh, opens by dating itself to the first year of Maharajadhiraja Toramana, ruling the earth in great splendour. Maharajadhiraja means great king over kings — the full imperial title, used without apology. Copper plates from Sanjeli in Gujarat also record his overlordship.
  • He is still being resisted in 510. The Eran pillar inscription of that year records a great battle involving a Gupta prince named Bhanugupta and a warrior called Goparaja, who died in it. The inscription does not say who the enemy was. The date, the place and the context make the Huns the obvious candidate — obvious, but not stated.
  • He dies around 515 and is succeeded by his son. That son is the reason this part exists.
THE ARCHAEOLOGY, WHICH IS GRIMMER THAN THE INSCRIPTIONS

At Taxila, the excavations of Sir John Marshall found the late city and its surrounding monasteries ending in

destruction — burning, collapse, coin hoards buried and never recovered, and no meaningful rebuilding afterwards.

Similar endings show up at monastic sites across Gandhara in roughly the same window.

Assigning that destruction to a specific king or year is beyond what the evidence supports, and honest archaeologists

say so. What the evidence does support is blunt and terrible: in the late fifth and early sixth centuries, the

monastic civilisation of the north-west stopped, and it never restarted. When the Chinese pilgrim Xuanzang

came through a century later, he described monastery after monastery standing empty and ruined, with a handful of

monks where there had been thousands.

WHY TORAMANA WON
  • He came into a vacuum. No Gupta central authority, no republic strong enough, no frontier power left in Bactria. There was, for the first time in a century, nobody whose job it was to say no.
  • Mounted archers on open plains. The Punjab doab country is close to ideal cavalry ground — flat, dry for much of the year, and cut only by rivers that a mobile army can choose when to cross.
  • Two generations of local knowledge. The Alchon had been sitting in Gandhara since the 460s. They did not invade a country they had never seen; they annexed the one next door.
  • He governed rather than raided. Dated regnal years, imperial titles, an issued coinage, land grants confirmed. Toramana set up a state. That is why his rule stuck and his son inherited an empire.
WHY INDIA LOST
  • The succession war did what no enemy could. The Guptas spent the years 470 to 500 fighting each other. Toramana simply walked into the result.
  • No frontier institution ever existed. Not under the Kushans, not under the Guptas, not under the republics. Two hundred years of warnings produced not one permanent garrison at the passes.
  • The rivers were not used. Punjab’s five rivers are the best defensive line in northern India, and in this war nobody held one.
  • Everyone defended their own ground. Republics defended their fields, Gupta claimants defended their claims, cities defended their walls, monasteries defended nothing. Nobody defended the region.
THE STORY WORTH REMEMBERING — THE COMPLICATED ONEThe Monastery in the Salt Range

Here is the awkward document mentioned above, and it deserves to be awkward.

The one inscription we have from Punjab that names Toramana as its ruling king — the Kura record from

the Salt Range — is not a war memorial. It is not a proclamation. It is a dedication recording the

building of a Buddhist monastery, made by a local man, dated in the ordinary way by the reign of the

ruling sovereign: the great king, king of kings, Toramana Shahi Jauvla.

Think about how that document came to exist. A subject of the Hun king, in Hun-ruled Punjab, spent his

own money founding a Buddhist establishment, recorded it publicly on stone, and dated it by his king’s

name — exactly as he would have done under any Indian ruler. Nobody stopped him. The stone survived.

That does not cancel Taxila. The burnt monasteries are also real, and the wave that Toramana led is the

wave under which Gandharan Buddhism died. Both things are in the record, and a book that tells you only

one of them is selling you something.

The most likely truth is the ordinary, unsatisfying one that historians keep arriving at. Conquest was

catastrophic where armies passed and unremarkable a few valleys away. In the Salt Range, life went

on well enough that a man could plan a monastery and expect it to still be standing when he was old.

Remember that in the next war, because the next king is a monster, and it will be tempting to make his

father one too. The evidence does not allow it.

9Mihirakula of Sialkot — Three Years of War With Kashmir LIKELY

Every source that mentions him hates him. That is unusual enough to be worth investigating.

Won by: Mihirakula — at first, and everywhere

When: Roughly 515–528 CE for the height of his power.

Where: Ruled from Sakala — Sialkot — over Punjab and Gandhara, campaigning into Kashmir and eastward.

Who: Mihirakula, son of Toramana, against Kashmir, the Buddhist establishments of the north-west, and eventually everyone.

The evidence: His own inscription at Gwalior; a Chinese envoy who was there; two later accounts that loathe him.

Who he was

Mihirakula inherited his father’s empire around 515 CE and ruled it from Sakala — the city we now call Sialkot.

Readers of Part Three will recognise that address. Sakala was the capital of the Indo-Greek king Menander six

centuries earlier, the setting of the Milindapanha, one of the great cosmopolitan cities of the ancient Punjab. The

same city, twice over: first the seat of a Greek king who is remembered for a philosophical dialogue with

a monk, then the seat of a Hun king who is remembered for burning monasteries.

His own inscription survives. At Gwalior, a record dated to the fifteenth year of Mihirakula commemorates the

building of a sun temple by a man named Matricheta, and calls the king’s father Toramana a ruler of great fame.

So we have him in his own words, or at least in his subjects’ words about him, doing perfectly normal royal things.

The eyewitness

Around 520 CE a Chinese envoy named Song Yun travelled through Gandhara on his way to fetch Buddhist texts,

and wrote a report. He is the only person in this entire part who saw the Hun court in Punjab and left a

description of it. What he says:

1. The country had been conquered by the Huns two generations earlier, and the current king was a hard, violent

  • ruler with no interest in Buddhism, who preferred to worship demons. 2. The king had been at war with Kashmir for three years and had no intention of stopping, keeping his army permanently in the field.
  • He kept a great force of war elephants — Song Yun gives a figure of seven hundred, each carrying armed men, with blades fastened to the elephants’ trunks.
  • The population was taxed heavily to pay for it, and the people were miserable.
ON THAT ELEPHANT FIGURE

Seven hundred war elephants is a very large number — comparable to the biggest elephant forces claimed anywhere in

Indian history — and it comes from a foreign envoy who was shown what a king wanted him to see. Treat it as “an

impressive number of elephants, displayed deliberately to an ambassador.”

The detail about swords on the trunks is repeated widely and may well be accurate; elephants were trained to use fitted

blades. But note who is telling us: a visitor, being given a military parade. Every ruler in history has done this to every

ambassador in history.

The three-year war

The Kashmir war Song Yun describes is one of the few concrete military facts we have about Mihirakula, and it fits

everything else. Kashmir was the one neighbouring region that was hard to take: a valley behind mountains,

reachable through a handful of passes that a defender can hold. It was also wealthy and it was on the flank of his

Punjab territory.

The war lasted at least three years and was still going when Song Yun passed through. That, as much as

any inscription, tells you what kind of ruler he was: the kind who commits an army to a mountain campaign for

years, keeps it in the field through winters, and taxes his own country flat to pay for it.

It also tells you something about his empire’s condition. Long wars in bad country are how large states exhaust

themselves. Within a few years of Song Yun’s visit, Mihirakula would lose everything.

WHY MIHIRAKULA WAS HARD TO BEAT
  • He inherited, not conquered. The army, the treasury, the territory and the administration were handed to him intact. He started at full strength.
  • Combined arms. Steppe cavalry from his own people plus Indian elephants and infantry from his subject territory. That is a genuinely dangerous combination, and it is what happens when invaders stay long enough to become local.
  • A central position. From Sialkot he could reach Kashmir, Gandhara, the Punjab plain and the road east. It is one of the best strategic addresses in northern India.
  • Terror as policy. Whether the stories are exaggerated or not, a reputation for atrocity makes cities open their gates. It works. It also, eventually, unites everyone against you.
THE WEAKNESSES THAT WOULD END HIM
  • He fought everyone at once. Kashmir for three years, the Buddhist establishment at home, and expansion eastward, simultaneously.
  • He made enemies of his own subjects. A king who persecutes the wealthiest institutions in his own territory has converted a tax base into an opposition.
  • He gave the Indian kingdoms a reason to combine. The one thing this book shows to be nearly impossible was an Indian coalition. He achieved it, against himself.
  • His empire was one generation deep. Loyalty in his territories was to a name, not an institution. When he was beaten once, it evaporated.
THE STORY WORTH REMEMBERING — THE CRUEL ONEThe Elephant at Hastivanj

This story comes from the Rajatarangini, the chronicle of the kings of Kashmir written by the poet

Kalhana in about 1148 CE — six hundred years after the events. It is the most famous thing ever written

about Mihirakula, and its label is Story, and it should stay that way.

Kalhana says that the king was known by the sound that travelled ahead of his army: not trumpets, but the

wailing of the women of the districts he was about to reach. He calls him a man of violent deeds, and

compares him to Death itself.

Then he tells this. Mihirakula was marching his army through the mountains when one of his elephants

lost its footing on a steep path and fell into the gorge below. As it fell, it screamed.

The king found the sound delightful. He ordered a hundred more elephants driven over the edge so that he

could hear it again.

Kalhana says the place kept the name — Hastivanja, the place of the elephants — and that people still

pointed it out in his own day.

Now the honest part. Kalhana wrote six centuries later. He was a Kashmiri Brahmin writing about a foreign

invader who had attacked Kashmir, which is not a neutral position. He was a poet, and this is a poet’s image

— the man who enjoys the scream. And in the very same chronicle he also records Mihirakula founding a

shrine, establishing a town and granting villages to Brahmins, which is what he records of every king.

The story is almost certainly not true as told. It survived because it is the shortest possible way to

say what an entire region remembered about being ruled by this man. That memory is the historical

fact. The elephants are how it was carried.

10Sondani — The Alliance That Broke Mihirakula CERTAIN

The one time in this book that India’s kings combined in time, won, and then vanished.

Yashodharman — a king who appears from nowhere and disappears within a WON BY

decade

WHEN
  • Around 528 CE. His related inscription at Mandsaur is dated to 532 CE.
WHERE
  • Sondani, near Mandsaur in Malwa — a very long way from Punjab, which is the point.
WHO
  • Yashodharman of Dashapura, probably in alliance with Gupta and other Indian forces, against Mihirakula.
THE EVIDENCE
  • Two identical victory pillars raised on the spot by the winner, and still standing.

Why it happened here and not in Punjab

Because Mihirakula had pushed too far east and south. His empire already held Punjab and Gandhara; his

expansion was aimed down the old road into Malwa and central India, the same corridor his father had reached at

Eran.

That expansion did the one thing his enemies could never do for themselves. It made him a threat to kingdoms

that had no other reason to cooperate. A Hun empire on the Ravi is a Punjabi problem. A Hun empire on the

Chambal is everybody’s problem.

What we know, which is more than usual

1. Yashodharman ruled at Dashapura — Mandsaur, in western Madhya Pradesh. Before this war he is

  • unrecorded. We do not know how he came to power or how he assembled the force he used.
  • He beat Mihirakula in the field and set up two matching stone pillars at Sondani to say so. Both survive.
  • The pillars claim two specific things, and their specificity is what makes them valuable. First, that he holds countries which even the Guptas never held and which even the Hunas could not command. Second, and far more pointedly, that Mihirakula himself paid homage to him — that the head which had never bowed to any human being was made to bow. 4. He calls his own achievement out by name. Naming your defeated enemy on a monument is not standard practice in Indian inscriptions, which usually prefer vague categories of vanquished foes. Yashodharman named him, twice, on two pillars, at the battlefield.
  • And then he disappears. Yashodharman is known from three inscriptions inside about five years. Then there is nothing. No successor, no dynasty, no later mention. The empire that broke the Huns lasted less than a decade.
WAS IT REALLY AN ALLIANCE?

The inscriptions credit Yashodharman alone, as victory inscriptions do. But a local ruler from Malwa defeating the

emperor of the north-west is a very large step, and there are two other threads. The Chinese pilgrim Xuanzang, a

century later, credits the defeat of Mihirakula to a king of Magadha he calls Baladitya — a name associated with the

later Gupta king Narasimhagupta. And the Guptas, though weakened, were still in the field in this period.

The standard reconstruction — that Yashodharman won the decisive battle with Gupta and allied support, and took the

whole credit on his own pillars — is plausible and unproven. What is certain is that Mihirakula was beaten, and

that after 528 he never again ruled an empire in India.

WHY YASHODHARMAN WON
  • He fought at the end of the enemy’s reach. Mihirakula’s power base was Punjab and Gandhara. At Sondani he was a very long way from his own country, in someone else’s.
  • He fought a tired empire. Three years of mountain war in Kashmir, a long eastward expansion, and a persecuted domestic population had already spent Mihirakula’s strength.
  • He assembled a combination — probably. Everything about the scale of the result suggests more than one kingdom’s army in the field.
  • He understood the value of a monument. He built the memory as deliberately as he built the victory, on the ground, in stone, with the loser named. It is why we know about him at all.
WHY MIHIRAKULA LOST
  • Overreach. He was fighting Kashmir, holding Punjab and Gandhara, and campaigning into central India at the same time.
  • He had no friends anywhere. No subject population with a reason to hold, no allied kingdom, no faction inside India that preferred him to the alternative.
  • Cavalry far from home. Horse armies need remounts, grazing and space. Deep in central India, at the end of a supply line, they get none of the three.
  • He had made himself the point of the empire. Once he was beaten in person, there was no institution underneath to keep the thing standing.
THE STORY WORTH REMEMBERING — THE IRONIC ONEThe Victory Pillar of a Man With No Kingdom

Go to Sondani today and the pillars are still there. What they say is one of the boldest claims any Indian

inscription makes: I rule lands the Guptas never ruled and the Hunas could not take, and the king

whose head had never bowed to anyone bowed to me.

Every word of that is very likely true. He did beat the most feared man in India. He did end an empire that

had held the north-west for forty years. The Hun threat to India effectively stops here.

And within about five years, Yashodharman is gone. No coins of his are known. No son. No successor. No

dynasty. Not one later inscription mentions him. The kingdom he claims on the pillar leaves no other trace

anywhere in India, and the historians who reconstructed his existence had to do it almost entirely from the

monuments he built for himself.

He is the only man in this book who beat the invader outright, took the credit in stone, and then

vanished more completely than the enemy he defeated.

There is a hard lesson under the irony, and it runs through the whole of Punjab’s history. Coalitions in this

book form against a threat and dissolve with it. The alliance that broke Mihirakula did not become a state,

an army or an institution. It won, and then it went home, and when the next invader came through the

passes there was nothing there but the pillars.

11Mihirakula’s Revenge — Kashmir and the Killing of Gandhara STORY

A defeated king with one army left, and nothing to lose, went and did the worst thing in this part of the book.

Won by: Mihirakula — the most destructive victory in these four hundred years

When: Roughly 528–535 CE, in the years after Sondani.

Where: Kashmir, and then Gandhara — Taxila, Peshawar and the monastery country.

Who: Mihirakula, against the king who gave him shelter and then against the Buddhist establishment of the north-west.

The evidence: Xuanzang, writing a century later, and Kalhana, six centuries later. No contemporary record at all.

What the sources say happened

This is the section of Part Four with the weakest evidence and the heaviest consequences, and both those facts

need stating at the top.

1. He is defeated and loses his empire. Xuanzang’s version has him captured by Baladitya of Magadha and

  • released on the intervention of the victor’s mother, which is a story shaped exactly like a moral lesson and should be read as one.
  • He takes refuge in Kashmir. The king of Kashmir gives shelter to the fallen conqueror.
  • He kills his host and seizes the throne. Both Xuanzang and Kalhana have some version of this. A man with a personal following, no territory and no future, taken in by a smaller kingdom, takes it over.
  • Then he goes back down the mountains into Gandhara. This is the part that matters. Xuanzang says he destroyed the royal house, overthrew stupas, wrecked monasteries, and killed on a scale he gives in the hundreds of thousands.
  • And after that, nothing. There is no reliable account of how or when he died. The coinage tails off. The Alchon fragment into small kingdoms in Kashmir, Gandhara and the Punjab hills, and never threaten India again.
INTERROGATING THE NUMBERS, AND THE SOURCE

Xuanzang’s casualty figures are not usable. He was a Buddhist monk writing a century after the events about the

destruction of Buddhist institutions, working from what monks in ruined monasteries told him. Hundreds of

thousands killed is a number of the kind ancient and medieval writers reach for when they mean beyond counting. Do

not repeat the figure. It is a feeling, not a census.

And yet. Xuanzang walked through Gandhara and Taxila in person around 630 CE and described what he saw with the

calm detail of a man taking notes: monasteries deserted, stupas broken, foundations robbed, a handful of monks in

buildings meant for hundreds. That part is not rhetoric — it is observation, and it is corroborated in the ground by

excavation.

So: the number is worthless, and the devastation is real. Both statements are true at once, and holding both is the

whole skill of reading a source like this.

WHY HE WAS ABLE TO DO IT
  • He still had an army when he had nothing else. Steppe followings were loyal to a leader, not a state. Losing the empire did not dissolve the war-band.
  • Kashmir was a small kingdom with a big guest. Taking in a defeated conqueror with his own troops is an act of charity that has almost never worked out in history.
  • Gandhara was undefended and wealthy. The monasteries held gold, relics and reliquaries and had no military capacity whatsoever. As targets go, there was nothing safer.
  • Nobody came. The Guptas were finished, Yashodharman had evaporated, the Punjab republics were long gone. The victors of Sondani never followed up.
WHY THE NORTH-WEST COULD NOT STOP HIM
  • It had already been ruled by his family for forty years. Whatever independent power once existed there had been dismantled by his own father.
  • Its wealth was in buildings. You cannot evacuate a stupa. Fixed, famous, valuable and unarmed is the worst possible combination.
  • Its protector had always been an outside empire. Kushan, Persian, Kidarite, Alchon — Gandhara had not defended itself in centuries and no longer knew how.
  • Nobody outside considered it their responsibility. That is the oldest pattern in this book, and here it costs an entire civilisation.
THE STORY WORTH REMEMBERING — THE ONE THAT COSTS THE MOSTThe City That Never Came Back

Taxila has been in every part of this book so far.

In Part Two it was the city of Ambhi, the king who invited Alexander in — already old, already famous,

already the place people came to learn. In Part Three it revolted twice against the Mauryan empire, was

governed by two future emperors, and grew into one of the great meeting points of the ancient world: Greek

sculpture, Buddhist philosophy, Persian administration and Indian learning stacked in the same streets. It

was one of the very few places on earth where all four of those things were in one place at one time.

And then, in the sixth century, it stops.

Not declines. Stops. The excavations show burning, collapse and abandonment. Money buried in panic and

never dug up. Monasteries emptied and not reoccupied. No later great city on the site — no medieval

Taxila, no Mughal Taxila. The road moved, the schools scattered east and north, and what had been a centre

of the world became a set of mounds that farmers ploughed around for thirteen hundred years until

archaeologists arrived to dig them up.

When Xuanzang passed through in about 630 CE, he was walking through the wreckage with a guidebook

in his head. He knew every story attached to every ruin, because Buddhists in China had been reading

about these places for centuries. He recorded what he had come to see and then recorded that it was gone.

This is the single largest loss in the whole of The Wars of Punjab, and unlike a battle it has no

date. Nobody can tell you the day Gandhara ended. The most any honest account can say is that it was alive

when Toramana came down the passes, and it was dead when the Arabs reached Multan in Part Five.

12The War That Freed Punjab Was Fought a Thousand Miles Away LIKELY

Two empires with no interest whatsoever in the five rivers destroyed the power that had been ruling them.

Won by: Sasanian Persia and the Turks — jointly, and briefly

When: Around 557–561 CE, with the long build-up starting in 484.

Where: Bactria and Sogdia — the country around the Oxus, near Bukhara. Not India.

Who: Khosrow I of Persia and the Turk khagan Istami, against the Hephthalite empire.

Why it is in this book: Because it removed the last Central Asian power capable of feeding invasions into Punjab, and nobody in Punjab had anything to do with it.

The seventy years before it

To understand why this mattered, you need one earlier disaster.

In 484 CE, the Persian king Peroz I led a full imperial army east against the Hephthalites, who had been beating

and humiliating him for years. The accounts differ in detail but agree on the shape of it: the Hephthalites let him

come, drew his army into ground of their choosing, and destroyed it. Several versions describe a concealed trench

dug across the line of the Persian advance and hidden, with the Hephthalite army waiting behind it.

Peroz died in that battle, and Persia paid tribute afterwards. A shah of Iran, at the head of the empire that

had conquered Gandhara in War One of this part, was killed in the field by the same broad group of peoples who

were about to take Punjab. It is the best single measure of how dangerous these confederations were, and it

explains why nobody in India managed a permanent answer to them.

What happened in 557–561

1. A new power appears behind the Hephthalites. The Turks — the first empire in history to use that name

  • for itself — came out of Mongolia in the 550s and expanded west across the steppe with extraordinary speed under a khagan named Istami. 2. Persia had spent eighty years waiting for exactly this. Khosrow I, one of the ablest rulers Persia ever had, made an alliance with the Turks against the enemy in between.
  • The Hephthalites were crushed between them, in fighting around Bukhara at the end of the 550s. Their empire ended.
  • The victors then divided the corpse and immediately fell out. The Oxus became the border: Persia south of it, the Turks north. Within a decade the two allies were enemies, which is what usually happens after a partition.
  • And south of the mountains, the pressure stopped. The Alchon in India had already been broken at Sondani. With the Hephthalite power in Bactria gone as well, there was no longer a reservoir behind the passes to send new waves down them. The Huna period of Indian history ends here, in a war India did not fight.
WHAT HAPPENED TO THE HUNS IN INDIA

They stayed. That is the short answer, and it is the one most often left out.

The Alchon and their descendants fragmented into small kingdoms in Kashmir, the Punjab hills, Gandhara and the

Kabul valley, and their coins keep appearing for generations. The dynasties that will hold the north-western door in

Part Five — the Shahi kings of Kabul and Gandhara — grew out of that wreckage.

Beyond the ruling families, there is a much larger and much more argued question: how far the Huna and related

populations settled into the general population of north-western India, and which later clan and caste names carry

their memory. Some of the connections proposed in the nineteenth century were confident, colonial and

evidence-free, and a lot of them are still repeated. The careful version is this: the invaders of the fifth and sixth

centuries did not evaporate, they merged, and the details are genuinely uncertain. Anyone who tells you exactly which

modern community descends from whom is telling you about politics, not history.

WHY PERSIA AND THE TURKS WON
  • Two fronts, one enemy. The Hephthalites were attacked from north and south at the same time — the exact situation that had destroyed the Kushans in War One.
  • Patience. Khosrow waited eighty years and paid tribute in the meantime rather than repeat Peroz’s mistake. Then he moved once, with an ally.
  • The Turks were faster and larger. A steppe empire beaten by a bigger steppe empire, which is how most of them end.
WHY THE HEPHTHALITES LOST
  • Success made them the middle. They had made enemies of everyone on both sides of them, which is fatal for a power that sits between two civilisations.
  • No depth. Their strength was a confederation of war-bands. Beat the central army badly enough and there is nothing behind it.
  • Their Indian branch was already gone. Sondani had taken away the wealth of the south a generation before the northern collapse.
THE STORY WORTH REMEMBERINGNobody in Punjab Was Consulted

Sit with this for a moment, because it is the most uncomfortable point in the whole part.

For roughly seventy years the people of the five rivers had lived under foreign rulers who came through the

passes. They had been taxed for a three-year mountain war they had no interest in. Their cities had been

burnt, their monasteries emptied and their most famous university town reduced to mounds.

And the thing that finally ended it was a battle near Bukhara, fought between a Persian emperor in

Ctesiphon and a Turkish khagan from Mongolia, about who should control the silk trade across the Oxus.

Neither of them had ever been to Punjab. Neither of them was thinking about it. Punjab was not a

party to the war, not a beneficiary anyone considered, and not informed of the outcome.

This is what it means to be a frontier rather than a power. Your fate is decided in rooms you are not in, by

people arguing about something else, and the news reaches you afterwards as a change in who does not

arrive at your gates.

It happens again in this book. It happens when the Mongols turn away from India because of a succession

crisis in Karakorum. It happens when the fate of Punjab in the 1840s is settled by decisions taken in

Calcutta and London. Remember that the relief of 561 CE was not a victory. It was a vacancy.

13A Lion to the Huna Deer — The Vardhanas of Thanesar LIKELY

The last Huna war in Indian history was won by a family from the Punjab plain, and it produced the last great empire before the Muslim conquests.

Won by: Prabhakaravardhana and his sons — the house of Thanesar

When: c. 580–648 CE. Prabhakaravardhana dies in 605; Harsha rules 606–647.

Where: From Sthanvishvara — Thanesar, in the Kurukshetra country of the eastern Punjab plain.

Who: The Pushyabhuti or Vardhana family, against the Hunas, the ruler of Sindh, the Gurjaras, Malwa and Gandhara.

The evidence: The Harshacharita of Bana, a court biography written for the family, and Xuanzang, who met Harsha in person.

Where they came from

Thanesar sits in the Kurukshetra plain, the country of the Mahabharata, at the eastern end of the Punjab plain on

the great road between the Yamuna and the Sutlej. It is exactly where you would expect a defensive power to grow

after two centuries of invasion: far enough east to be safe from a first blow, close enough west to have to

care.

The Vardhanas were minor rulers who became major ones in one generation, under Prabhakaravardhana, who

took the imperial title and spent his reign fighting his neighbours in every direction.

What Bana says about him

The Harshacharita is a court biography, written by the family’s own poet, and it should be read with all the

scepticism that implies. It is also the best piece of prose in this part and it contains the single most quoted

description of a king in early medieval India. Prabhakaravardhana, it says, was:

a lion to the Huna deer, a burning fever to the king of the Indus land, a troubler of the sleep of Gurjara, a bilious

plague to that scent-elephant the lord of Gandhara, a looter of the lawlessness of the Latas, an axe to the creeper

of the glory of Malwa.

Bana, Harshacharita, on Prabhakaravardhana

Strip the poetry and it is a foreign policy summary. Punjab’s neighbours, listed in order, each with the injury

done to them. The Hunas are first, which tells you which one the family was proudest of.

What happened

1. Prabhakaravardhana fights the Hunas repeatedly and wins often enough for the boast to stick. These are no longer imperial Hunas; they are the fragments left after Sondani, holding the hills and the north-west. The

  • threat is real but reduced.
  • He sends his elder son Rajyavardhana north against them in about 604 or 605, in what is the last recorded Huna campaign in Indian history. The prince is winning when a messenger reaches him: his father is dying.
  • The family collapses in a single year. Prabhakaravardhana dies. His queen, Yashomati, follows him into the fire, in the practice these sources record without comment and which this book will meet again. Rajyavardhana comes home to a funeral.
  • Then it gets worse. The king of Malwa attacks and kills the Vardhanas’ ally; Rajyavardhana marches out, defeats him — and is then killed, Bana says by treachery, at the hands of Shashanka, the king of Gauda in Bengal, while apparently in negotiation.
  • The younger brother inherits everything at about sixteen. This is Harsha. He takes the throne in 606 CE, vows revenge, and then spends thirty years building the last great north Indian empire before the Muslim conquests — from the Punjab to Bengal, ruled from Kanauj.
  • And in 647 he dies without an heir, and it all comes apart within months. Northern India returns to being a collection of kingdoms, and stays that way for four hundred years.
WHAT PUNJAB LOOKED LIKE AT THE END OF THIS PART

We have an eyewitness, and he is unusually reliable. Xuanzang travelled through the Punjab in the 630s, and what he

describes is a plain full of separate kingdoms rather than a province of anything: a large territory he calls the country

of the Takkas, covering much of the plain, along with Jalandhara, Kashmir and others.

He also describes Sakala — Sialkot, and this is the detail worth keeping. He records the ruins of the old city, and

remembers it as the seat of Mihirakula, who had ruled there a hundred years before. The Greek capital of Part Three,

the Hun capital of Part Four, standing in pieces, being pointed out to a foreign traveller as the place where the cruel king

used to live.

WHY THE VARDHANAS WON
  • They fought a broken enemy. The Hunas of 600 CE were the wreckage of Sondani, not the empire of Toramana. Timing was most of the victory.
  • They were based in the right place. Thanesar had depth behind it and the great road in front of it — a base you can defend and campaign from.
  • Three capable rulers in a row. Father and two sons, all of them soldiers. In a hereditary system that is the whole game, and it is largely luck.
  • They campaigned in every direction and consolidated behind them. Unlike Yashodharman, Harsha turned victory into an administration that lasted forty years.
WHY IT ENDED ANYWAY
  • One man, no heir. Harsha’s empire was a personal achievement with no institution underneath it. He died, and it stopped.
  • The centre had moved east for good. Harsha ruled from Kanauj, not from Thanesar. Punjab was again a frontier province of somebody else’s state — the pattern set in Part One and unbroken since.
  • Nothing was fixed at the passes. For the fourth time in this part: the door was never garrisoned. Sixty years after Harsha, an Arab army walks through a different door into Sindh, and Part Five begins.
THE STORY WORTH REMEMBERING — THE LAST WORD ON THE HUNASThe Deer That Was Once the Terror of the World

Read Bana’s line again, and notice what it takes for granted.

A lion to the Huna deer. Deer. Not wolves, not tigers, not the storm out of the mountains — deer. Something

you hunt for sport. Something that runs.

Two lifetimes before Bana wrote that sentence, the Hunas held everything from the Kabul valley to central

India. They had killed a Persian emperor in the field, ended the Gupta empire’s reach in the north-west,

taxed the whole Punjab plain, and destroyed the greatest Buddhist civilisation in the world. Mihirakula

ruled from Sialkot and marched wherever he liked.

And by 600 CE a court poet in Thanesar can use them as the small animal in a compliment, in a list, third

from the top, without any need to explain the joke to his readers. The most feared name in India had

become a figure of speech.

That is what the end of an era actually looks like from the inside. Not a treaty, not a final battle, not a date

anybody records. Just a change in the way people talk — the moment a word that used to make a room go

quiet turns into the sort of thing you put in a poem, because everyone in the room already knows it cannot

hurt them any more.

14The Whole Chapter on One Page

c. 230–270 CESasanian Persia takes Gandhara and Peshawar. The Kushan empire ends in the north-west. No battle is recorded. Won by Persia.
c. 250–350 CEThe free century. The Yaudheyas, Madrakas, Kunindas, Audumbaras and Trigartas govern themselves and mint their own coins. Won by the republics of Punjab.
c. 350–375 CESamudragupta reaches the frontier. Nine peoples pay tribute and submit on paper; the Kushan and Saka rulers send embassies. Won by Samudragupta — on parchment.
c. 380–410 CEThe Mehrauli iron pillar. A king called Chandra crosses the seven mouths of the Sindhu and beats the Vahlikas. Won by Chandra — probably Chandragupta II.
c. 380–430 CEKidara crosses the great mountains. The first Huna wave takes Bactria, Gandhara and Kashmir. India learns a new word. Won by the Kidarites.
c. 455–460 CESkandagupta and the whirlpool. The Pushyamitras beaten, then the Hunas stopped for a generation. Won by Skandagupta.
484 CEThe Hephthalites destroy a Persian imperial army and kill the shah Peroz I. Won by the Hephthalites.
c. 490–515 CEToramana takes the Punjab and reaches central India. His name dates a monastery record in the Salt Range. Won by the Alchon Huns.
510 CEThe battle at Eran. Goparaja dies beside the Gupta prince Bhanugupta; his wife follows him into the fire, in the first dated record of the practice in India.
c. 515–528 CEMihirakula rules from Sialkot, fights Kashmir for three years, and persecutes the Buddhist establishment. Won by Mihirakula.
c. 528 CESondani. Yashodharman breaks Mihirakula and raises two victory pillars naming him. Won by Yashodharman — who then vanishes.
c. 528–535 CEMihirakula’s revenge. He seizes Kashmir and devastates Gandhara. Taxila does not recover. Won by Mihirakula — the costliest victory in this part.
c. 557–561 CEPersia and the Turks destroy the Hephthalite empire near Bukhara. The pressure on the passes stops. Won by people who had never been to Punjab.
c. 580–605 CEPrabhakaravardhana of Thanesar — “a lion to the Huna deer.” Won by the Vardhanas.
605–606 CEThe last Huna campaign. Rajyavardhana is recalled from it, and is murdered soon after by Shashanka of Gauda.
606–647 CEHarsha builds the last great north Indian empire and rules Punjab from Kanauj. He dies without an heir and it dissolves.
712 CEAn Arab army reaches Multan. Part Five begins here.

Five patterns to carry into Part Five

1. Nobody ever garrisons the door. This is the defining failure of Part Four and you should be angry about it by

  • now. Four separate powers — Kushan, Gupta, republic, Vardhana — knew that every invasion of India for eight hundred years had come through the same three or four passes, and not one of them stationed a permanent army there. Punjab is invaded repeatedly not because it is weak but because the entrance to it is always somebody else’s problem. Watch what Jayapala and the Shahis try to do about this in Part Five, and what it costs them.
  • Conquering Punjab is easy; holding it is not — and this part adds a crueller version. Toramana took the plain in a season. His son lost everything within thirteen years of inheriting it. The Alchon empire in India lasted about forty years in total. Every foreign power in this part rules Punjab quickly and briefly, and none of them build anything that outlives the man in charge.
  • The coalition forms too late and dissolves too early. Part One had ten kings who could not combine; Part Two had the Malavas and Kshudrakas who combined and then could not agree on a commander. Part Four finally produces a coalition that works — and it works only after the invader has ruled for forty years, and it disbands the moment it wins. Yashodharman is the proof that India could beat these invaders whenever it chose to combine, and the proof that it almost never chose to.
  • The invader becomes the neighbour. The Kushans became an Indian dynasty. The Alchon fragments became the hill kingdoms and the ancestors of the Shahis, who will spend Part Five defending India against the next wave from exactly the same passes they themselves came through. In this book, today’s invader is tomorrow’s frontier guard. It happens again with the Ghaznavids in Lahore, and again after that.
  • Punjab’s history in this period was written by everyone except Punjab. A Persian cliff inscription, two Chinese travellers, victory pillars in Madhya Pradesh, a court poet in Thanesar and a Kashmiri writing six hundred years later. Not one contemporary Punjabi account of any of these twelve wars exists. The winner is whoever gets written about, and in Part Four almost nobody local was writing at all.
COMING NEXT — PART FIVE

Arabs, Hindu Shahis and Mahmud of Ghazni, 712–1030. Muhammad bin Qasim reaches Multan in 712 and stops.

For nearly three hundred years the Shahi kings of Kabul and the Punjab hold the passes — the first dynasty in this

book to actually try — and then the Ghaznavids come. Peshawar in 1001 and the self-immolation of Jayapala. Waihind

in 1008, and the panicking elephant that lost a battle. The sack of Multan, Nandana in 1014, and the long end of

Trilochanapala. And at the finish, something new: Lahore becomes the capital of a foreign empire, and Punjab acquires

a Muslim ruling class that never leaves.

15Words Used in This Part

Sasanian

The Persian imperial dynasty founded by Ardashir I in 224 CE and destroyed by the Arabs in 651. The last pre-Islamic Persian empire.

Kushanshah

“King of the Kushans.” The title used by Persian-appointed governors of the east after the Kushan collapse. Historians call them the Kushano-Sasanians.

Gana / gana-sangha

A republic — a state governed by an assembly of clan heads rather than a hereditary king. The Yaudheyas were one. The word survives in modern Indian usage for “republic”.

Prashasti

A formal praise-inscription. Always written by the winner’s own court. Essential evidence and never neutral.

Maharajadhiraja

“Great king over kings” — the full imperial title. Its use by Toramana is how we know he considered himself an emperor of India, not a raider.

Pratyanta

“Frontier.” In the Allahabad inscription, the pratyanta kings are the border rulers who submit without being annexed.

Huna

The Sanskrit word used for all the Central Asian invaders of this period, regardless of who they actually were. Related to the word behind “Hun”, but not the same people as Attila’s.

Kidarite

The first of the Huna groups to enter India, from around 380 CE. Named after their king Kidara.

Hephthalite

The “White Huns” of Byzantine writers. Dominated Bactria and Iran’s eastern frontier from about 450 to 560, mostly staying north of the mountains.

Alchon

The group of Toramana and Mihirakula, named from the word that appears on their coins in Bactrian script. These are the Hunas who actually ruled Punjab.

Tamga

A clan or dynastic symbol stamped on coins and seals, used across the steppe like a brand. The Alchon tamga is how numismatists identify their coinage at a glance.

Sakala

Ancient Sialkot. Capital of the Madrakas, then of the Indo-Greek king Menander in Part Three, then of Mihirakula in this part.

Sthanvishvara

Ancient Thanesar, in the Kurukshetra country. Capital of the Vardhana family.

Vihara

A Buddhist monastery. The Kura inscription in the Salt Range records the founding of one under Toramana’s rule.

Stupa

A solid domed monument built over a relic. The characteristic religious building of Gandhara, and the main target of the destruction described in War Ten.

Gandhara

The country around Peshawar and Taxila. In this part it is a Buddhist civilisation of monasteries, sculpture and learning — and it does not survive the period.

Uttarapatha

“The northern road.” The ancient trunk route from the Ganges plain through Punjab to the north-west. Every army in this part uses it.

Rajatarangini

“The river of kings.” The chronicle of Kashmir written by Kalhana in about 1148 CE. The source of the Mihirakula elephant story, and the only sustained historical chronicle produced in early India.

Harshacharita

“The deeds of Harsha.” A court biography of the Vardhana family by the poet Bana, written in the seventh century. Brilliant prose, entirely partisan.

Sources used for this part

  • The inscription of Shapur I on the Ka‘ba-ye Zartosht at Naqsh-e Rustam (the Res Gestae Divi Saporis), for the Sasanian conquest of
  • the east
  • The Allahabad pillar inscription of Samudragupta, composed by Harishena — in J. F. Fleet, Corpus Inscriptionum Indicarum, vol.
  • III
  • The Mehrauli iron pillar inscription of Chandra; the Bhitari pillar and Junagadh rock inscriptions of Skandagupta; the Eran boar inscription and the Kura inscription of Toramana; the Gwalior inscription of Mihirakula; the Sondani and Mandsaur inscriptions
  • of Yashodharman; the Eran pillar inscription of 510 CE — all in Fleet, Corpus Inscriptionum Indicarum, vol. III
  • Xuanzang, Records of the Western Regions (Si-Yu-Ki), in Beal’s translation
  • Song Yun’s account of his journey through Gandhara, c. 518–522 CE
  • The Wei shu and related Chinese dynastic histories, for Kidara and the Yuezhi
  • Priscus of Panium, fragments, for the Kidarite–Sasanian wars
  • Procopius, Wars, Book I, and al-Tabari, History, for Peroz I and the Hephthalites
  • Kalhana, Rajatarangini, Book I, in Stein’s translation
  • Bana, Harshacharita, in the Cowell and Thomas translation
  • John Marshall, Taxila: An Illustrated Account of Archaeological Excavations (1951)
  • Modern scholarship on the Alchon and Kidarite coinages, particularly the work of Michael Alram and colleagues at the Kunsthistorisches Museum, Vienna
  • Hans Bakker, The Alkhan: A Hunnic People in South Asia (2020)
  • R. Balasubramaniam, on the corrosion resistance of the Delhi iron pillar
  • Standard reference accounts of the Gupta period, the Huna invasions, the Yaudheya and Kuninda coinages, and the reign of
  • Harsha
A NOTE ON DISAGREEMENT

The open questions in this part are unusually large. They include: how far east Sasanian authority ever reached;

whether Samudragupta fought anything at all in the north-west; the identity of “Chandra” on the iron pillar and of the

Vahlikas he defeated; whether Skandagupta’s Hunas were Kidarites or Alchon; who the Pushyamitras were; whether

Yashodharman fought alone or as part of a coalition with the later Guptas; and how much of the destruction of

Gandharan Buddhism can honestly be laid at Mihirakula’s door rather than spread across a century of disorder. Where

scholars differ, this part follows the mainstream reading and says so on the page.

Quotations from inscriptions follow the standard published translations, chiefly Fleet’s. Those translations are more

than a century old and bracket a good deal of wording that is not on the stone; where a reading is disputed, it is flagged

in the text rather than smoothed over.

And the warning from Part Two applies twice as hard here. Every word we have about Punjab in these four

hundred years was written somewhere else, by someone with an interest — a Persian king listing his possessions, a

Gupta minister praising his employer, Chinese Buddhists describing the ruin of Buddhist places, a victorious rival

naming his defeated enemy on a pillar, a Kashmiri poet six centuries later. Nobody who lived through Toramana’s

conquest left an account of it. When you read that Mihirakula was a monster, remember that his own subjects’ only

surviving word about him is a building record — and that we would think very differently about him if the people who

hated him had not been the only ones taking notes.

Download The complete book · 15.7 MB