THE CURRENCY SYSTEM 93

professions, or as Indian castes began to take to them by a long process

of adjustment, there arose a sufficiently large supply of free skilled
labour that could dispense with the demand for slave labour. For slaves
in productive work had the obvious disadvantage of lacking the
incentive as well as the insecurity of subsistence, which drove and
haunted free labour.

4 The currency system

Possibly the most decisive evidence of change in the economic life of
northern India after the Muslim expansion is numismatic. Some
evidence of a mild economic revival from the tenth century is provided
by the reappearance, in limited quantities, of a gold coinage based on
older models of a 40 rati standard of which the earliest examples appear
to be of the Chedis or Kalachuris of Madhya Pradesh followed by
dynasties of northern-central India, the Chandelas of Mahoba, and, most
numerous of all, the Gahadavalas of Kanawj in the twelfth century.
After the Muslim conquest of the Gangetic plain, the earliest gold issues
in the name of Mu'izz al-Din Muhammad bin Sam are of this weight
and type, with a conventionalized usage of the goddess Lakshmi on the
obverse and the name of the ruler in Nagari characters on the reverse.
Clearly they have no Muslim ideological content and must represent the
survival of pre-existent arrangement with Indian moneyers to provide
a coinage in the name of the ruler of the day. Monetary assay being a
skill of goldsmiths and sarrdfs or money-changers, who down to the
present day have mostly been members of the Hindu sonar caste, we may
assume that the implementation of the monetary policy of Muslim
sultans was mainly entrusted to these experts. The mint-master at Delhi
during the reign of the Khalji Sultan Qutb al-Din Mubarak, Thakkura
Pheru, was a Hindu or Jain, and left a valuable treatise written in
Apabhramsa regarding the exchange rate and precious-metal content
of coins brought to the royal mint for exchange, the coin-types dating
in some cases from a century before the time at which he was writing.1
Pure silver coinage was extremely scarce in northern India. Recent
evidence suggests that, notwithstanding the thirteenth-century chron-
icler Juzjani's observations, it was current in Bengal, where the silver
came from a more easterly source of supply. The broad-struck bull and
triglyph coins of the Chandra dynasty have now been reassigned from
the seventh to the eleventh and twelfth centuries, and from Arakan to
east Bengal.2 A large find of more than 300 such coins has been reported
1
See: [206], 87-101.
2
D. W. Macdowell, 'Eight coins of Arakan from Sylhet', Numismatic Chronicle, 6th Series, xx
(i960), 229-33: A. H. Dani, 'Coins of the Chandra Kings of East Bengal', JNSI, 24 (1962),
141-2.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 94 NORTHERN INDIA UNDER THE SULTANATE

from Mainamati in a twelfth-century context.1 These Chandra dynasty

coins are important in suggesting a prototype for the broad-struck silver
and gold issues of the Delhi and Bengal sultanates and with regard to
the evident dependence of the Delhi sultanate on an eastern source of
supply from Bengal, to maintain the silver currency in circulation.
Literary sources are entirely lacking regarding the establishment of
a new gold and silver currency, the coin types of which endured in India
through the following seven centuries to the present-day white-metal
rupee. It may be traced to the overrunning of Bengal by the adventurer
'Ala 'al-Dln Bakhtyar Khalji with a small body of horse following the
establishment of Muslim power in Delhi and the Gangetic Doab. The
conquest of the capital of Gaur provided an occasion for the remission
of treasure from Bengal to the Ghurid overlord Mu'izz al-Dln
Muhammad bin Sam in 1205,2 as well as later remittances to the Delhi
Sultan Shams al-Dln Iltutmish. Possibly the original remittance was the
result of the immediate striking of coins from captured treasure, as in
addition to the Arabic titles of the sultan they bear the Sanskrit
inscription gaura vijaye, 'at the victory of Gaur'. The recorded weights
of 170.8 and 172.18 g. are evidence of the adoption of a new standard,
probably of one tola consisting of 96 ratts, {gunja seeds of the abrus
precatorius plant — a light but variable measure in different areas and at
different periods in Indian history, used by jewellers and moneyers).
The device upon the earliest gold and silver tankas remitted from
Bengal was a galloping horseman. Unconvincing parallels have been
drawn with the static and extremely stylized obverse of the north-western
Indian bull-and-horseman coinages. An acquaintaince with the copper
horseman-coins on the part of the Muslim invaders of Bengal (who were
Khaljls from the Afghan borderland) is unlikely. The most probable
explanation is that this was an imaginative commemoration of 'Ala 'al-
Din Bakhtyar's exploit with a very small body of Afghan horse. It is
improbable that those who introduced the new coinage were influenced
either by Indo-Greek tetradrachms or by the Gupta gold coinage of many
centuries earlier. A model for the broader silver pieces may well have
been suggested by the lighter Chandra silver coins still in circulation.
The metrology of the new coinage is firmly Indian with no parallels
in earlier Islamic coinage. Accordingly the decision to mint gold and
silver coins in these weights (though we have no evidence for the silver
coins till fifteen years later) may have been a chance decision, but it was
of paramount importance in later Indian monetary history.
1
M. H. Rashid, ' T h e Mainamati gold coins', Bangladesh Lalit Kala, i: i, pp. 41, 45, 57-8. The
hoard must be assigned to the twelfth century A.D. It contradicts the Muslim chronicler Juzjanl's
assertion that there was no silver coinage current in Bengal in the period immediately before
the Muslim conquest - see Tabaqit-i Nasirf (trans. Raverty), p. 556.
2
Lowick [39;], 196-208; P. L. Gupta, 'Nagari legend on horseman tanka of Muhammad bin
Sam', JNS1, 35 (1973), 210-12.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 THE CURRENCY SYSTEM 95

The earliest issues of gold and silver coins from Delhi itself often have
a commemorative character, reflecting the immediate coinage of hoards
plundered or remitted in tribute. However, the remonetarization of the
economy must have occurred by the middle of the thirteenth century,
for at this period the Suhrawardi shaykhs of Multan left assets of lakhs
of tankas. By the first decade of the fourteenth century, the salary of
a trooper or knight with one horse was calculated at 234 tankas per
annum, with the increment of 78 tankas if he maintained a second
warhorse. As we shall see below, the purchasing power of such a salary
was considerable.
In the monetary system of the Delhi sultanate a firm equation
between gold and silver appears to have been established at 1:10. The
plunder of Hindu kingdoms and religious establishments yielded both
gold and silver. An inflexibility of monetary policy did not lead to the
considerable fluctuations in the rate between the two metals which is
visible in medieval Europe and Mamluk Egypt, but rather to the
disappearance from circulation of the scarcer metal, which was silver.
The prevalent coin of north-western India in the late twelfth and early
thirteenth centuries was a debased variety of the bull-and-horseman
issues of the Hindushahl kings of Kabul and Waihind, which was
current in the later Ghaznavid kingdom of Lahore and had spread
south-eastwards into Rajasthan and northern-central India. It was the
money of account, called by the Muslim conquerors dehliwdl (i.e. Delhi
coin). In the inscription recording the construction of the mosque at
their new capital the cost of the material from twenty-seven Hindu
temples re-employed in its construction is reckoned at 12 million (or
120 lakhs) of dehliwdls. H. Nelson Wright argues that with the standard-
ization of the Delhi coinage by Shams al-Dln Iltutmish (1210—35) the
small silver content of these coins was reduced by nearly a half. The
jital, as the coin came to be called, was of 32 rails weight. The value
of its silver and copper combined equalled 2 ratis of silver; the relative
value of silver to copper was 1:80. The jital changed at i/48th of a silver
tanka in northern India and 1/5oth in the Deccan after the Muslim
conquest of Devgiri; the difference is accountable by a slightly higher
price for copper in southern India, where it was imported from
overseas.1
The existence of smaller moneys of account in the capital city of Delhi
is demonstrated by Baranl's numerous references to dings ('quarters')
1
Wright [549], 393-8, 401. Nelson Wright convincingly refuted the contention of Edward
Thomas that there were b^jitals to the silver tanka, Chronicles of the Patban kings ofDehli, London,
1871, pp. 3-4. Nelson Wright appears also to argue that thejital or dehlival in the early thirteenth
century represented i/i2th of a tanka; but his demonstration that from the end of that century
theyi7a/was i/48th part of a tanka is convincing, see op. cit., pp. 397-8. See also p. 96, note
1, below.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 96 NORTHERN INDIA UNDER THE SULTANATE

and dirams. From a comparison of prices quoted in the different

recensions of his work it is possible to establish the equation: 1 silver
tanka = 4%jitah = \yzda~ngs = 4S0 dirams.1 The dang, calledfiis (pi. fulus)
by the Arab geographer al-'Umarl, may be identified with the 40-rati
weight pure-copper coin which persistently appears in the coinage of
the Delhi sultans and 'dirams', single or multiple, with smaller copper
issues. Similarly, binary multiple jttals or fractions of the tanka are
represented by tiny silver issues, mainly in the thirteenth century, or
by mixed silver and copper issues (billon), whose denomination could
be distinguished by their silvery appearance at the time of minting and
by assay.
It is clear that the establishment of a trimetallic coinage in northern
India in the thirteenth century was heavily dependent on the remittance
of gold and silver from Bengal, some of the gold as well as the silver
issues having been minted there. The remittances from Bengal were
irregular, dependent on the degree of fear in which the local governor
or sultan stood towards the power of the Delhi sultan. Equally coinages
represented a 'dethesaurization' (E. Ashtor's term) or release of
treasure from hoards into monetary circulation, derived from the
plunder of the treasure of local rulers or religious establishments. Here
we note two forces in operation. In hoarding, the more precious metal,
gold, is more likely to be withdrawn from circulation than the less
precious silver. Hence when hoards are released into circulation, the
ratio of exchange between gold and silver is likely to be under pressure.
Secondly, while most of the gold hoarded in India must have come from
the western world, though there was the maritime import of gold to
India from China during the period, there is also evidence of a
continuing import into Bengal of silver which represents the surplus
of a trade balance probably from an inland source.
With the successive plunder of the Deccan kingdoms at the end of
the thirteenth and beginning of the fourteenth centuries, which were
more prosperous in the preceding centuries than northern India, a great
quantity of precious metal came into the hands of the Delhi sultans,
though less of silver than of gold. One may perhaps accept the
information of the sixteenth and seventeenth-century historian Farishta
that the indemnity extracted by 'Ala 'al-DIn Khaljl from Ramadeva of
Deogiri in his raiding expedition before his usurpation of the throne
amounted to roughly 7.7 metric tonnes of gold and 12.8 metric tonnes
of silver, and we note the shorter supply of silver. Towards the end
of the reign, Malik Kafur's plundering expedition to the Pandya
1
The equation is established in a forthcoming article which compares the price lists in the standard
recension of BaranI [140], 319-20, with a variant represented by S. Digby collection MS 57,
ff. 117-18.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 THE CURRENCY SYSTEM 97

kingdom is said by the fourteenth-century historian BaranI to have

brought back 96,000 man of gold, which would correspond to 241 metric
tonnes; it is likely that this is an exaggeration. Subsequent expeditions
in the reigns of Qutb al-Dln Mubarak, Ghiyath al-Dln Tughluq, and
Muhammad bin Tughluq extracted the remaining available surplus.
The consequence of this acquisition of wealth is visible in the coinage
of the Delhi sultanate. Both the gold and silver issues of 'Ala 'al-Dln
Khaljl are plentiful, and the brighter appearance of the silver issues, due
to the absence of lead, distinguish these issues from those coined from
the eastern Indian source. The gold coined in this reign provided a vast
surplus, not wholly exhausted by the policy and largesse of Muhammad
bin Tughluq. When, at the close of the fourteenth century, Timur
plundered Delhi, his officers found stores of tankas coined in 'Ala
'al-Din's name.
The predominance of gold over silver now released into circulation
placed a strain on the monetary relation between the two metals of 10:1,
with some evidence for an unofficial exchange-rate of 7:1. The coinage
of the Delhi sultanate was more efficiently controlled than any Middle
Eastern or European currency of the period, with the fixed ratio
between the two precious metals remaining constant; this may reflect
the exact calculations of Indian money-changers.
The disturbance of the gold: silver rate becomes apparent after the
accession of Muhammad bin Tughluq in 1325, probably made visible
by this monarch's keenness for monetary innovation. Very shortly gold
coins one-sixth above and one-sixth below the 196-ra// standard were
issued, followed in his third year by a mixed-metal tanka of 80 ratis
weight, one-sixth reduced in weight from the silver coin, with a silver
content of about 45 grains, a little more than a quarter of what was found
in the earlier coins. Three years later the sultan tried a more desperate
expedient, the issue of a token coinage of brass and copper to replace
the silver coinage. It is clear from BaranI that he was led to this by
accounts of Chinese token currency {chad) in the form of silk or paper
notes of credit.1 The disruption of commerce which this token currency
caused led to its withdrawal two years later.
It is clear from the scale of Muhammad bin Tughluq's subsequent
military operations, from the plentiful gold coins issued later in his reign
and accounts of his largesse to foreign visitors, as well as from evidence
cited elsewhere that the accumulated treasure of the Delhi sultanate was
not exhausted, but the problem was one of the relative scarcity of silver
in a cash economy with urban inflation. This was soon accentuated by
the loss of political control of Bengal.
1
BaranI, variant recension (see preceding note), f. 167V.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 98 NORTHERN INDIA UNDER THE SULTANATE

The sources of the silver supply of Bengal remain obscure. It can

hardly have been a Far Eastern maritime source, for this would also
have reached Coromandel, Kerala, and Gujarat and have been reflected
in currencies. We have earlier noted the eleventh-and twelfth-century
silver coinages of the Chandra dynasty; during the Mughal dynasty
strong silver coinages are characteristic of Assam, Tipperah, Nepal, and
other border states of north-east India. The silver coinage of the sultans
of Bengal continues uninterrupted and plentiful from the middle of the
fourteenth to the middle of the sixteenth century; and the revival of
silver rupee coinage throughout northern India immediately follows
Sher Shah's conquest of Bengal. Further evidence against a Far Eastern
maritime source of supply in the early sixteenth century is provided by
the Portuguese, Pires, who mentions the profit of 25 per cent on its
import from Bengal to Malacca.1 As no significant quantity of silver
is found in Bengal, an overland source of supply is implied. From
Tavernier's seventeenth-century evidence, this may have been from
Yunnan. The argentiferous lead galena in Burma is another possible
source and there is later eighteenth-century evidence of silver brought
from Tibet to Nepal, which may have been conveyed to eastern India.
Literary evidence confirms the demand of the Delhi sultans for silver
from eastern sources. Muhammad bin Tughluq's governor Qadar Khan
was preocuppied with gathering revenue in the form of silver tankas;
the more he gathered the better, he used to say.2 The silver tankas were
gathered, but his unpaid soldiery turned to the local pretender Fakhr
al-Din before the hoard was dispatched to Delhi.
Surviving examples of the pure-silver coinage of Muhammad bin
Tughluq's reign are much scarcer than the gold issues. The situation
grew more acute in the reign of his successor Firuz Shah, from which
only four silver 96 ratt pure-silver tankas have been recorded.3 The quest
for silver is indicated by the realization of arrears of revenue from Hindu
chiefs in the region of Gorakhpur at the outset of Firuz Shah's
expedition against Ilyas Shah of Bengal in 1359; the arrears were
realized in silver tankas and bags of silver were distributed to the Muslim
religious establishments of Delhi. In the Broach hoard of the late-
fourteenth century in western India, about 1,200 silver coins from
western Asia were recorded, but none from the Delhi sultanate which
ruled over this area.
The monetary system initiated by the Muslim expansion in the early
thirteenth century was in decay by the middle of the fourteenth century.
The debased 80 rati billon tanka replaced the relatively pure silver tanka
around 1330, though Ibn Battuta clearly reckoned prices in terms of
1
Pires, Suma oriental, n , 93.
2 3
SIhrindl [145], i°4~5- Wright [549], 160, 163-4.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 THE CURRENCY SYSTEM 99

the pure silver tanka in the following decade. The silver tanka —
supported by denominations of cowries, not copper - remained as a
coinage of commercial transactions in Bengal down to the revival of
a coinage in precious metals throughout northern India in the middle
of the sixteenth century. Elsewhere in Muslim-dominated India in the
fifteenth century, as in the earliest days of the Delhi sultanate, gold and
silver coinage may have been merely ceremonial, intended not as a
currency of trade but as a proclamation of sovereignty or for ceremonial
distributions. Gold issues of the later Tughluq sultans of Delhi, some
posthumous and dating from the early fifteenth century, and of the
Sayyid sultans of Delhi are rare. The fifteenth-century sultans of
Kashmir issued rare gold coins of tanka weight, with almost illegible
legends, which must be regarded as commemorating accession. Both
gold and silver tankas survive from the fifteenth-century sultanates of
the Deccan, Jawnpur, Malwa, and Gujarat which must be regarded as
mainly commemorative or celebratory issues rather than currencies of
trade. Apart from Bengal, where the pure-silver tanka continued to
circulate, in fifteenth-century Gujarat smaller coins of moderately
debased silver, which must be regarded as a trading currency, were
minted and circulated as a result of the import of the silver coinages
of the Middle East in exchange for commodities across the Arabian Sea.
In the whole of inland Muslim India the fifteenth century is a period
of reversion to mixed-metal (billon) currencies, of copper with a scanty
admixture of silver, as the standard currency of trade. The Lodi sultans
of Delhi, less concerned with traditional Muslim concepts of sovereignty
than their contemporary rivals, appear not to have issued gold or silver
coins.
Wassaf, an early fourteenth-century Persian historian, saw India as the
drain of gold which it had been in earlier and later centuries.1 It is clear
that in spite of Chinese imports, most of the gold, whether it had been
concealed as treasure centuries before or had been recently imported,
came from the principal Eurasian source, the goldmines of west Africa.
From medieval Ghana great quantities of gold passed not only to
western Europe but also through the Mamluk kingdom, which were
circulated in eastern trade.
The clearest evidence regarding the import of precious metals into
the territories of the Delhi sultanate from the west and of the balances
of trade is found in the Broach hoard.2 The greatest quantity of gold
and silver coins is from the Mamluk kingdom of Egypt and Syria. A
considerable number of coins are from the Rasulid kingdom of the
Yemen, probably reflecting a balance on the export of foodstuffs,
1
Wassaf [138], joo.
2
O. Codrington, 'On a hoard of coins found at Broach', JBBRAS, xv (1882-3), PP- 339~7°-

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 IOO N O R T H E R N I N D I A U N D E R THE S U L T A N A T E

particularly rice, and cotton cloth to south Arabia. The European coins,
mainly Venetian, show growing European participation in the spice
trade. Coins from the Persian Gulf or Persia are rare. This reinforces
other indications that this was a region where, mainly on account of
the import of warhorses, the balance of trade involved an export from
India.
The drain of Indian riches towards the Middle East, particularly
through the Persian Gulf, in Muhammad bin Tughluq's benefactions
and trading advances is vividly attested by Ibn Battuta. Some con-
firmatory evidence is found in a fifteenth-century Persian revenue
manual, which reproduces documents which must date from the reign
of the Ilkhan Ghazan (694-703) or Oljeytu (703-17), more probably the
latter in view of the Indian context. The gold at the royal treasury (at
Tabriz) which came from India and was coined in tankas (therefore from
the Delhi sultanate), exceeded that from any other source. The gold
from India was recorded as 122,000 mithqals, as against 64,000 mithqals
of Egyptian gold and 24,000 of' western' (Maghribiyya) gold and 110,000
in the local coinage of Ghazan Khan.1
The overland trade through the Kipchak desert to northern India
probably led to an export of specie for the same reason as that from
the Persian Gulf to India. Finds of medieval Indian coins in European
Russia are exclusively from the southern steppes bordering on the Sea
of Azov and the Black Sea. This is in contrast to finds of Sasanian and
'Abbasid treasure and coins in north Russia or Sweden, which reflect
the importance in those periods of the trade in furs and in slaves.
Though there was a taste for furs among the nobility of the Indian
sultanates, this strongly suggests that the coins of the Delhi sultans
found their way there in payment for the thousands of warhorses
dispatched from this area to the frontier of the Delhi sultan noticed by
Ibn Battuta.
With the exception of the brief disordered reign of the usurper Nasir
al-Dln Khusraw, gold tankas of every reign from that of 'Ala 'al-Dln
Khaljl (1296-1316) to that of Flriiz Shah Tughluq (13 5 2-88) have been
recorded, but no coins of the later Delhi sultanate or the provincial
sultanates of the fifteenth century. Though the number of gold coins
recovered by Russian institutions is not impressive - forty-two gold
coins in all - they are from no less than seventeen different find-spots.2
The Russian numismatist Bykov on one occasion mentions that they
were part of a larger hoard of similar gold tankas which was dispersed,
but nowhere records that they were found in association with other
Islamic coins. His evidence fits well with our general thesis of the
1
' A b d A l l a h al-Mazandaranl, Risala-yi falakiyya [ I I J ] , 229.
2
B y k o v [ 2 4 7 ] , 146—56.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008
 THE CURRENCY SYSTEM IOI

politico-economic pattern of the Delhi sultanate, viz. that it put into

circulation the hoarded treasure captured by its military expeditions and,
apart from profligate displays of conspicuous consumption, spent much
of this capital upon military materiel, the import of war-horses and
fighting-men.

Downloaded from https://www.cambridge.org/core. University of Western Ontario, on 12 Sep 2018 at 16:06:51, subject to the Cambridge Core
Cambridge Histories Online https://doi.org/10.1017/CHOL9780521226929.007
terms of use, available at https://www.cambridge.org/core/terms. © Cambridge University Press, 2008