Pramāṇa · Search · About · API

The Colas c31: Chapter XXIII. Coins, Weights and Measures

The Cōḷas c31

Language
English

Contents

Chapter XXIII. Coins, Weights and Measures

c31.2

The absence of prominent land-marks in the numismatic history of Southern India, and the small proportion of inscribed specimens of coins discovered so far have stood in the way of a scientific treatment of the coinage of the South. At the same time the relative richness of epigraphical material has contributed to make the study of South Indian history largely independent of the always difficult and none too certain conclusions of numismatics.

c31.3

Two weight systems are traceable from the ancient coins of the South. 'The gold *gadyāṇa* coin of the Deccan averages 58 grains, the heaviest reaching 60·1 grains'; this was the standard unit called *gadyāṇa* or *kaḻañju* in the Tamil country.¹ If the weight of the lost gold coin of Uttama Cōḻan figured by Elliot² is correctly recorded as 50 to 60 grains, this coin must have followed the old *gadyāṇa* standard and must have remained in circulation late in the tenth century. The survival to late times of a small cess called *Kumārakaccāṇam*³ may be accepted as confirmation of the same fact. But in the Cōḻa period the more usual standard was the *kaḻañju* of twenty *mañjāḍis* equal in theory to 72 grains, but sometimes going up to 80.⁴ It is apparently this unit of bullion weight that is employed in an inscription of the thirtieth year of Parāntaka I which equates the *kaḻañju* with the *niṣka* (Sanskrit).⁵ When exactly the Cōḻa currency was brought into line with this weight standard does not admit of precise determination.

c31.4

By the side of several inscriptions which seem to employ the *kaḻañju* in recording payments by weight, there are some which mention the *pon* and equate it to the *kaḻañju*, implying thereby that the *pon* was coined gold of the full weight of one *kaḻañju*.⁶ This coin was also called *madhurāntakadēvan-māḍai*, served as the standard of fineness for testing gold,⁷ and yielded the same interest as one *kaḻañju* of fine gold. This coin is mentioned in the thirty-first year of Rājarājadēva,⁸ and if this king was Rājarāja I, as seems likely, its issue must have been commenced under his predecessor Madhurāntaka Uttama Cōḷa.

c31.5

Exactly half of this *māḍai* was the Rājarājan *kāśu*,⁹ issued apparently by Rājarāja I. But the *kāśu* of this weight and fineness was certainly in use earlier than the time of Rājarāja I. A record¹⁰ of the fourth year of Āditya II mentions that twenty *kāśus* were equal to ten *kaḻañjus*. It was in vogue after Rājarāja also, as the same relation between the *kāśu* and *kaḻañju* re-appears in some of the later inscriptions.¹¹ In fact both the *māḍai* and the *kāśu*, the standard gold coins of the realm, were issued by each one of the Cōḷa kings of the period before 1070—the different issues being distinguished in epigraphy by the names of the kings being prefixed to those of the coins in question: thus we have expressions like the *māḍai* of Rājarāja,¹² the *māḍai* of Rājēndraśōḷan,¹³ a phrase which by the way makes it more likely that the *madhurāntakadēvan-māḍai* was issued by Rājarāja's predecessor, *anṟāḍu-naṟkāśu* meaning 'good current *kāśu*', and *paḻangāsu* ('old *kāśu*')¹⁴ and sometimes even *anṟāḍu*-(nar)-*paḻangāśu*, current (good) old *kāśu*, in later inscriptions.¹⁵

c31.6

The *madhurāntakan māḍai* was still current in the reign of Kulōttunga I;¹⁶ and it is said to have been equivalent to the *kaḻañju* of the fineness of 9½ *māṟi* or two *kāśus*.¹⁷ That this high standard was not maintained at all times in the actual currency goes without saying, and the inscriptions give valuable evidence on the extent of the departure from the standard at different times and places.

c31.7

An accidental discovery in 1946 in the village of Dhavalēśvaram (E. Godavari Dt.) brought to light a fine hoard of which only 127 gold coins could be recovered. Here is a genuine collection of ancient gold coins which have elsewhere mostly disappeared in the goldsmith's crucible. The coins are all of fine gold—thin round discs bearing a *lāñchana* in the middle, and letters round the margin, all punched on one side of the disc, the other side being left blank. There are 49 coins of Rājarāja I, the Eastern Cāḷukya king (Nos. 5 and 6 on the plate) which call for no remark; they bear the regnal years 33, 34 and 35 corresponding to A.D. 1055-57 as Rājarāja is known to have celebrated his coronation in 1022 with the aid of Rājēndra I Cōḷa. The remaining coins fall into two sets with different inscriptions in Tamil-grantha but one and the same *lāñchana* in the centre of the field which is clearly the Cōḷa mark. The inscriptions on one set (Nos. 1 and 2) read :

c31.8

kam-gai-ko-ṇḍa-cō-ḷan

c31.9

and bear regnal years ranging from 28 to 33; these are obviously issues of Rājēndra I and the regnal years correspond to A.D. 1040-45. There are other figures above the regnal year such as 4000 and 11 of which the significance is not clear; so also there are some letters with the *lāñchana* in the centre which await explanation. The second set of 46 coins bears the inscription:

c31.10

ma-la (i) -nā-ḍu-ko-ṇḍa-cō-ḷan

c31.11

and bears regnal years ranging from 34 to 36. From their association these coins are easily attributed to Rājakēsarī Rājādhirāja I who was recognised as heir apparent by his father Rājēndra I in 1018 and who continued to rule jointly with his father till about 1044 and then for ten years more in his own right till he fell on the battle-field of Koppam. It will be recalled that Rājarāja I Cāḷukya had a troubled reign and had to appeal for Cōḷa aid on many occasions against the aggressions of the Western Cāḷukyas and Vijayāditya VII. Rājādhirāja must have taken the title *malai-nāḍu-koṇḍa* by virtue of the wars he waged for his father in Kēraḷa, and the title must have been chosen for being put on his coins to distinguish them from Rājēndra's issue. It is not known if these coins which closely follow the fabric of Eastern Cāḷukya coins were minted for circulation in the Vēṅgī country only or went into more general use.

c31.12

From the reign of Kulōttunga I, several other types of *māḍais* make their appearance in the inscriptions, and these were undoubtedly issued by local rulers who were feudatories of the Cōḷa empire. Examples of such issues are: the *jayamāḍa* mentioned in a record from Cebrōlu dated Ś. 998;¹⁸ the *uttama-gaṇḍa-māḍa* of a Bāpatla inscription;¹⁹ the *cāmara-māḍa* and the *biruda-* *māḍa* in somewhat later records from the same place;²⁰ the *nakki-māḍa* mentioned in an inscription from Kāñcīpuram, dated in the fourteenth year of Rājādhirāja II,²¹ and recording a gift from a chieftain from Gaṅga-maṇḍalam, who had the title Bhujabalavīran; the *Bhujabala-māḍai*, first mentioned in an inscription of the third year of Kulōttuṅga, perhaps the third of that name, from Nandalūr;²² and in later records from Kāñcīpuram and Tiruppālaivanam;²³ the *paḻam-puḷḷi-māḍai* of a record from Tirumullaivāyil dated 1232;²⁴ and the *gaṇḍa-gōpālan-māḍai* obviously issued by the Telugu-Cōḍa chieftains of Nellore who were contemporary with Rājarāja III and Rājēndra III.²⁵

c31.13

By a dexterous use of two inscriptions of the same time and locality, Codrington seeks to establish that the fineness of the *madhurāntakan-māḍai* was roughly about 8⅜ *māttu*;²⁶ but this contradicts the express statement of the inscription of Kulōttuṅga's reign cited above, and it is not improbable that two different *kāśus* are meant by the two records though they stand so close to each other, and that the standard of fineness adopted in both cases was the same, in other words, that *śembon* of one of these records was of the same fineness as the *madhurāntakan māḍai* of 9½ *māttu*. Codrington adds: 'Now none of the Cōḷa coins even approach this standard, and it seems possible that this *māḍai* was a unit of account equal to a *kaḻañju* of gold of the touch mentioned, and had ceased to have any connection with the progressively debased coinage.' This conclusion can hardly be accepted as correct, if it implies that the *kāśu* of proper fineness was not actually issued from the royal mints for circulation. Considering that as late as the thirty-fifth year of Rājādhirāja I²⁷, A.D. 1053, the standard *kāśu* was still current and that a heavier *kāśu*, equal to ¾ *kaḻañju* and 3 *mañjāḍis*, is sometimes mentioned in the records,²⁸ while a lighter *kāśu* of seven *mañjāḍis* was also known,²⁹ the real explanation for the prevalence of different units of currency must be sought elsewhere. As Codrington has himself observed 'it seems probable that each province of the empire retained its local currency' and its relation to the standard currency was a matter for calculation on each separate occasion. That few coins of the high standard of fineness of the *māḍai* have reached us is perhaps due to no other reason than this, that no old coins that were at all fit for the gold-smith's brazier had any chance of survival through many centuries; it is probable that a good part of the Dhavalēśvaram hoard mentioned above has also disappeared in this manner.

c31.14

The standard *kāśu*, moreover, derived ultimately from Ceylon which had a more ancient and continuous currency tradition than the Cōḷa kingdom which came up in the ninth and tenth centuries. The *īḻak-kāśu*, Ceylon *kāśu*, which was also half-a-*kaḻañju* of the fineness of the *māḍai*³⁰, was current in the island as early as the seventh and eighth centuries; it is also mentioned in Cōḷa inscriptions from about A.D. 937 in the reign of Parāntaka I.³¹ There was close contact between Ceylon and the Madura country for centuries before Parāntaka's conquest of Madura, and the introduction of this coin into Cōḷa currency must have been the result of Parāntaka's invasion of Madura and the island of Ceylon.³² From the actual specimens of coins known, the Ceylon type 'with a rude human figure standing on the obverse, and seated on the reverse' and the traditional Cōḷa type with the seated tiger, fish and bow emblems, both may be seen to have existed side by side almost from the beginning, the Ceylon type being specially suited for circulation in the Pāṇḍyan country where it may have been long known before. However that may be, the main thing was the adoption of the Ceylon standard, and this took place apparently much earlier than the time of Rājarāja.³³

c31.15

Besides the gold coins *māḍai* and *kāśu* and others of local provenance and uncertain standard, the inscriptions mention '*Karungāśu*' or '*Īḻakkarungāśu*', the black *kāśu* of Ceylon.³⁴ This coin is also apparently traceable to Ceylon, where the '*nīla kahāpaṇa*', a silver coin of the *gadyāṇa* standard in weight, is known to have been in circulation from very early times.³⁵ Some of the impure silver coins of the Cōḷas now known may be assigned to this series, full-weight or half, as the case may be.³⁶ Copper coins issued by the Cōḷas conforming more to the standard of the *gadyāṇa* than of the *kaḻañju* are figured by Elliot and other writers.

c31.16

In an inscription of the thirty-third year of Parāntaka I from the Madura district,³⁷ we are told that the *Īḻak-kāśu* was the equivalent of 7½ new *akkams*. What sort of a coin the *akkam* was is not easily decided. It is again mentioned in a record³⁸ of Āditya II, but without any relation to other units. In the days of Rājarāja, however, the *akkam* was definitely a twelfth of the *kāśu*,³⁹ and therefore a twenty-fourth of the *Kalañju*. It is obvious that the term *akkam* is applied to coins of different value in the Madura and the Tanjore inscriptions of Parāntaka and Rājarāja; for there is much support, as we have seen, from epigraphy for holding that the *Īḻak-kāśu* of Parāntaka's time and Rājarāja's *kāśu* were of the same value and fineness.⁴⁰

c31.17

With the growth of the Cōḻa empire as a result of the conquests of Rājarāja I, the currency system of the Cōḻas was extended over the whole empire, including its feudatory provinces. The new standard was adopted for instance in the Vēṅgī country about A.D. 1000;⁴¹ 'the weight of Rājarāja's coin is found also beyond the limits of his empire, for example, in the gold pieces of the Western Cāḷukya Jagadēkamalla and of the Kadambas of Goa.⁴² The *āḍavalam gadyāṇa* of the Koṅgāḷvas⁴³ in the eleventh century as opposed to *gadyāṇa* perhaps refers to the Cōḻa reformed gold coin.'⁴⁴ On the other hand, the local varieties of other areas are found intruding on occasions into the heart of the Cōḻa empire; about A.D. 1049, an Eastern Cāḷukya prince presented to the temple at Tiruvaiyāṟū a sum of 300 Rājarājamāḍas equal to 337½ *kalañjus* by the *kuḍiñaikal*.⁴⁵

c31.18

An inscription of the tenth year of Kulōttunga I from the Chingleput district gives the ratio between the standard *kāśu* and silver bullion by stating that 433 *kalañjus* of silver formed the equivalent of 100 *kāśus*, so that 1 *kalañju* of gold, probably of the standard fineness of 9½ *māttu*, was equal to 8·66 *kalañjus* of silver.⁴⁶

c31.19

Examples of various types of the gold *kāśu* deviating more or less from the standard have been given already. The gold content of the *kāśu* is clearly stated in several inscriptions, and it seems probable that in such statements, the standard of fineness adopted was that of the *māḍai*; this is, however, expressly mentioned only in some cases. About A.D. 1063, inscriptions from the North Arcot district seem to mention two varieties of *kāśu* containing 8·356 and 7 *mañjāḍis* respectively of standard gold;⁴⁷ a *kāśu* of a little over 6 *mañjāḍis* occurs in an inscription of 1077 from Tiruvorriyūr,⁴⁸ of 6·813 *mañjāḍis* in a record 1111 from Tiruvāḍuturai;⁴⁹ and of exactly 6½ *mañjāḍis* in 1122 at Tiruvārūr.⁵⁰ Another inscription from the Tanjore district, A.D. 1133, mentions a *kāśu* of the weight of three-fourths of a *kaḷañju*,⁵¹ a coin which seems to have still kept up the weight of the old *māḍai* of the *gadyāṇa* standard. Even so late as 1152 a *kāśu* of the gold content of a third of a *kaḷañju* is mentioned in an inscription from Ālanguḍi.

c31.20

These coins with a substantial gold content are, however, exceptional instances of the survival of old issues or very limited new issues minted from time to time; for the general history of Cōḷa currency seems to have been marked by a progressive deterioration. While the 'Ceylon type' issues of Rājarāja I are usually of fairly good gold, all the available specimens of later monarchs beginning from Rājādhirāja I are of very base gold, 'or rather silver washed with gold.'⁵²

c31.21

From the time of Kulōttunga III, the term *kāśu* comes to be applied to a copper coin of rather low value and its content must have varied greatly from place to place, or, what seems equally likely, with each issue. While two to three *paḷangāśus* were quite enough to endow a lamp in this period,⁵³ 1100 current *kāśus* were needed for a lamp being maintained during services in a temple, and 9000 for a perpetual lamp.⁵⁴ In another case, 200 new *kāśus* sufficed for a lamp.⁵⁵ An inscription of the twentieth year of Rājēndra III states that the *kaḷañju* of gold was then equal to 411 7|13 *kāśus*.⁵⁶ And a Pāṇḍya inscription of the thirteenth century from the Salem district (Āragaḷūr) states that one hundred *Sōḷiya kāśus* went to the *faṇam*.⁵⁷ Coppers of various sizes usually with the legend Rājarāja on them can still be picked up easily in the bazaars of South Indian cities, and they, no doubt, are the *kāśus* of these later Cōḷa inscriptions. Of the copper issues of the later monarchs, Codrington says⁵⁸ that they are 'of the same design as the coins of Rājarāja I, but the human figure is more degraded and the legend roughly executed, the *Ja* sometimes being reversed. In the degradation of the type various stages are visible, and it is probable that the coins were issued by various later kings following the old model without alteration of name. The better executed and presumably older specimens with a flat blank about ·76 inch in width, and a weight of some 63 grains, may be the new *kāśu* or rather its double.’

c31.22

The *tiramam* is occasionally found mentioned in inscriptions; it is not clear if it was a coin or only a smaller unit of reckoning. At Kāñcīpuram, in A.D. 1076, six *tiramams* went to the *kāśu*,⁵⁹ while more than forty years later, the *kāśu* was the equal of seven *tiramams*⁶⁰ in the Ramnad district.

c31.23

Though the weight and fineness of the standard gold coin of the realm were fairly determinate, there still existed a large variety of systems of weighing and assaying. Without stopping to give an exhaustive account of these systems, a few examples may be given to illustrate the general conditions that prevailed. The mention of *Viḍēlviḍugu-kal* used for weighing gold at Kumaravayalur in the Trichinopoly area in an inscription of Āditya I may be some indication that continuity in the system of weights and measures was ordinarily sought to be maintained irrespective of dynastic changes. In the Tanjore inscriptions of Rājarāja, two different weight units are found used, the *āḍavallān* for weighing gold and the *dakṣiṇamēruviṭaṅkan* for jewels. Outside Tanjore may be noticed the ‘stone used for weighing Rājarājan-kāśu’ at Tiruvārūr,⁶¹ the *vaiyagattār-kal* at Tirumalavāḍi,⁶² the *viḍēlviḍugu-kal* at Tiruccenduṟai⁶³ and Paḻuvūr⁶⁴ and the *kempōnāgarasu-niṟai* at Taḍi-malingi,⁶⁵ or simply the stone used in this village of Śrīkaṇṭha-caturvēdi-maṅgalam (Tiruverumbūr).⁶⁶ The absolute weights of these different units are not easily determined, but they all used the same denominations of *kalañju* divided into 20 *mañjāḍis*, each *mañjāḍi* being the equivalent of two *kunṟis*. The fineness of gold is likewise

c31.24

expressed in different ways. The fineness of the *māḍai* and that of *śembon* (red-gold) have already been mentioned. Of these, the term ‘red gold’ was applied to gold purified and tested according to certain prescribed methods;⁶⁷ apparently it was the same as *tuḷai-pon* or *tuḷai-niṟai-pon*. Sometimes, the fineness is explained in terms of touch—as *niṉe māttu* by the *kāśu-niṟai-kal*.⁶⁸ 'Of the fineness of *paḻaṅgāśu*' and 'of the fineness prescribed by the (royal) treasury' (*tālaċċemmai*)⁶⁹ are other expressions that are employed in the early inscriptions of the period. There was thus no uniformity either in the actual units of gold currency in circulation or in the standard of weight and fineness adopted in different areas for the regulation of local transactions. The need for 'gold committees' which were set up by different village-assemblies for testing gold is thus clear.

c31.25

The same diversity of usage marks the systems of land, liquid and grain measures adopted in different areas. The *nilam* or *vēli*, called *vāṭikā* in Sanskrit in the Anbil plates, was the unit measure of land.⁷⁰ It was divided into 1/2, 1/4, 1/8, 1/20, 1/80, and 1/320, a secondary series (the first *kīḻ*) of 1/320 of this primary series, and a tertiary (1/320)² (the second *kīḻ*) of the same;⁷¹ even a third *kīḻ* series (1/320)³ seems to be occasionally employed.⁷²

c31.26

How exactly the measurement of land was carried out to such minute accuracy is not known. But the system of measurement by *vēli* which was an ancient one in the Cōḻa country spread over the other parts of Southern India with the Cōḻa empire and apparently disappeared with it. Cōḻa inscriptions in the Pāṇḍya country for instance used this system, which is not found there either before or after the Cōḻas. For all the minute care bestowed in the inscriptions from Tanjore and elsewhere on the fractional sub-divisions of the *vēli*, the extent of the *vēli* seems itself to have been somewhat indeterminate. The length of the rod (*kōl*) used is gathered from expressions like the 16-span-*kōl*,⁷³ the *kaḍigaikkaḷattu-kōl*,⁷⁴ *śrī-pāda-kōl*,⁷⁵ the *māḷigai-kōl*,⁷⁶ and so on. Not only did the extent of the *kuḻi* differ with the length of the measuring rod used, but the number of *kuḻis* that made up a *mā* or *śeru*,⁷⁷ the twentieth of the *vēli*, also varied considerably. In Kiḻūr in South Arcot, the *mā* contained 256 *kuḻis* by the 16 span rod in the sixth year of Rājēndra I;⁷⁸ in the same year in the Tanjore district (Tiruvāḍuturai) one hundred *kuḻis* by the *māḷigaikōl* made up the *mā*.⁷⁹ About twelve years later, in Tiruvāmāttūr not far from Kiḻūr, 200 *kuḻis* by the 16 span rod were reckoned to the *mā*.⁸⁰ In land-measures, as in currency, therefore, the same tendency to standardisation on the part of the central government can be traced by the side of the persistence of local usage. The *māḷigai-kōl* (the rod of the palace), and the 100-*kuḷi*-*mā* as well as the spread of the *vēli*-unit outside the Cōḷa country proper may be taken as evidence of the attempt to impose a uniform system on the provinces. In the temple at Tiruvālangāḍu in the Tanjore district the length of the standard *kōl* (rod) is marked on the stone wall and on the *gōpuram*, to conform to the *kōl* found engraved on the wall of the Tanjore temple.⁸¹

c31.27

How little these efforts at standardisation succeeded is clear from the chaos of measuring standards revealed by the later Cōḷa inscriptions. At Tiruvorṟiyūr, in A.D. 1072, the *vēli* comprised 2000 *kuḷis* by the sixteen-span rod;⁸² while in 1204 it is equated to 6⅙ *pāḍagams* in Uttaramērūr.⁸³ The *mā* was made up of 138 *kuḷis* in 1097 at Tirukkaḍaiyūr,⁸⁴ and of 128 *kuḷis* by the standard of the *sabhā* of Tiruvāḍutuṟai in 1110,⁸⁵ the same *sabhā* employing in the very next year another measure by which the *mā* was 100 *kuḷis*.⁸⁶ Again, in the South Arcot district, the *mā* was reckoned at 512 *kuḷis* by the fourteen span rod in 1138;⁸⁷ and the same number of *kuḷis* to the *mā* is found in a record from Muniyūr (Tanjore district) in 1220, the length of the rod, however, not being specified in this case.⁸⁸ And a *mā* of 513 *kuḷis* is found in Vaḷuvūr in the same district in the reign of Rājādhirāja II.⁸⁹ This list is by no means exhaustive; it is seldom that we have the data for converting one unit into another as is found in some records of the *sabhā* of Tiruvāḍutuṟai which equate 4½ *mās* of their measure to six *mās* of the survey, and six *mās* and a *kāṇi* to eight *mās* according to the general survey which took 100 *mās* to the *kuḷi*.⁹⁰

c31.28

The same characteristic marks the liquid and grain measures used in different places, and numerous kinds of *nāḷi* and *marakkāl* find mention in the inscriptions. From the Tanjore inscriptions it becomes clear that the *Āḍavallān* equal to *Rājakēsari* was adopted as the standard in the accounts of the capital city and perhaps of the empire in the time of Rājarāja I. Possibly, the same measure is referred to as *Arumoḷidēvan* in the reigns of Rājēndra and Rājādhirāja.⁹¹ though the name *Rājakēsari* also continued in use.⁹² The distinction between this standard measure *Rājakēsari* and another called *Vīdiviḍangan* is clearly brought out in a record of the twenty-sixth year of Rājarāja I from Tiruvāmāttūr stating that a surplus was left as a result of measuring paddy by the *Vīdiviḍangan* instead of by the *Rājakēsari*.⁹³ *Kōmaḷam* (Covelong) and *Kacci-ppēḍu-ninṟān* were other measures used in Chingleput.⁹⁴ It would seem, however, that better success attended the effort to standardise liquid and grain measures than in other directions; at any rate, after the reign of Rājarāja I the inscriptions seem to mention fewer varieties of these measures than of land measures or gold weights.

c31.29

Another standardised unit of value was, strangely enough, the *āḍu* (lit. sheep) which was the equivalent of a sum enough to yield one measure of *ghī* per annum as interest; in the sixteenth year of Rājēndra I, 25 *kāśus* were equal to 22½ such *āḍus*,⁹⁵ giving as many measures of *ghī* in the year. The fractional *āḍus* (sheep) mentioned in inscriptions⁹⁶ can be understood only if they are taken to be fractions of a monetary unit, and not of the live animal.

c31.30

The year is reckoned generally at 360 days, but instances also occur where calculations are based on a year of 365 days.⁹⁷ It is well known that Hindu astronomical works like the *Sūryasiddhānta* and the *Siddhānta Śirōmaṇi* mention, among others, both these reckonings and prescribe the particular purposes for which each may be used. In the Cōḷa inscriptions, however, we find both reckonings used for the same purpose, *e.g.*, for calculating the quantity of *ghī* necessary for the maintenance of lamps.

Tables of Main Weights and Measures in the Inscriptions

c31.s1.1

1. Liquid and grain measure:—

c31.s1.2

| 2 śeviḍu | = 1 piḍi |
| 5 śeviḍu | = 1 āḻakku |
| 2 āḻakku | = 1 uḻakku |
| 2 uḻakku | = 1 uri |
| 2 uri | = 1 nāḻi |
| 8 nāḻi | = 1 kuṟuṇi |
| 2 kuṟuṇi | = 1 padakku |
| 2 padakku | = 1 tūṇi or kāḍi⁹⁸ |
| 3 tūṇi | = 1 kalam |

c31.s1.3

2. Weight of gold:—

c31.s1.4

One mañjāḍi = 2 kunṟi = 10 mā = 40 kāṇi.
Twenty mañjāḍi = one kaḻañju (about 68 to 72 grains).

c31.s1.5

Note.—The *mā* and *kāṇi* are usually 1|20 and 1|80 when applied to a *vēli* of land. Hultzsch conjectures (*SII.* ii, p. 65 n.) very plausibly that originally the *mā* and *kāṇi* had the same fractional value even here, and were subdivisions of a unit of 2 *mañjāḍis* or a tenth of the *kaḻañju*; coins called *faṇams* were usually a tenth of the standard *kaḻañju* in weight; and the *paṇa-tūkkam* (paṇam-weight) has always been a tenth of the *kaḻañju*. In 273 of 1927 (Vik. 3, Śivāpuram) the *mā* is really 1|20 *kāśu*.

c31.s1.6

3. Linear measure used in measuring icons: —

c31.s1.7

| 8 tōrai (rice-corn) | = one viral (finger). |
| 12 viral | = one śāṇ (span). |
| 2 śāṇ | = one muḻam (cubit). |

c31.s1.8

1. Codrington—*Ceylon Coins and Currency*, pp. 6-7.
2. *CSI*. No. 151.
3. *TAS*. i, p. 165; 182 of 1915.
4. Codrington, pp. 3 and 7.
5. *SII*. iii, 104; also 181 of 1912.
6. 49 of 1888, 54 of 1893, etc. *contra* Codrington:—'The early mediaeval *pon* seems to be a *kalañju* of gold, and not necessarily a coin'. (p. 52).
7. 140 of 1912; *ARE*. 1913, II, 22.
8. 252 of 1915.
9. 141 of 1912. Mentioned also in 484 of 1925, A.D. 1054.
10. 241 of 1923.
11. 203 of 1925; 228 of 1923.
12. 104 of 1925. This *mādai* is used as the standard of fineness in 671 of 1919.
13. 203 of 1925.
14. 629 of 1916; 484 of 1925.
15. 71 of 1926; 217 of 1901; 329 of 1929.
16. 17 of 1893; 180 of 1911. *ARE*. 1936-7 II. 27.
17. 90 of 1928.
18. 151 of 1897.
19. 236 of 1897.
20. 210 of 1897; 176 of 1897.
21. 48 of 1893.
22. 586 of 1907.
23. 360 of 1919, 311 of 1929. *ARE*. 1929, II, 38.
24. 674 of 1904.
25. 266 of 1921, and several others.
26. op. cit. p. 86.
27. 228 of 1923.
28. 105 of 1925; also 571 of 1904 for a *kāśu* of slightly less value. Both of Rājarāja I.
29. 5 of 1890.
30. 25 and 156 of 1895; also 252 of 1915. The word *Īḻam* came to mean 'gold' in Tamil; and this sense is known to the *Divākaram*; but I have not come across any early use of the word in this sense. Can it be that the meaning was derived from the fact that the *Īḻak-kāśu* was a gold coin?
31. 435 of 1904.
32. Codrington, p. 73.
33. *Contra* Codrington, p. 84. It is perhaps worth noting that the so-called Laṅkēśvara type of Rājarāja I (Desikachari, *South Indian Coins*, p. 183) is not a Cōḻa coin, but most probably belongs to the Ceylonese *Kahāvaṇu* series and that the legend on it is now read as 'Śrī Laṅka Vibhu'. Codrington. op. cit. p. 54. Unfortunately, Desikachari does not give the weight of his specimen of this type. The copper '*Kōdaṇḍarāma*' coins (Desikachari, p. 66) are also clearly Pāṇḍyan, not Cōḻa.
34. *SII*. iii, 120; 242 of 1907; 238 and 266 of 1923.
35. Codrington, pp. 13-4.

c31.s1.9

36. Elliot, No. 152 is full weight; so also Nos. 26 and 27 described by Hultzsch at *IA.* 1896, p. 321 which from the title Cōḷa-nārāyaṇa may now be ascribed to Rājarāja who gets the same title in one of the Mysore inscriptions. One coin at p. 317 *IA.* 1896 is half weight, 30 gr. Another, in the same place, of 51½ grains, is near Elliot No. 153 which is 52·2 grains.

c31.s1.10

37. *SII.* iii, 106. This *kāśu* is also mentioned in 435 of 1904, year 30.

c31.s1.11

38. 275 of 1923. The *Aka* was again a fraction piece of the Ceylon *Kahāvaṇuva*, and, like the *kāśu*, may have been derived from there. Codrington, p. 58.

c31.s1.12

39. *SII.* ii, 7.

c31.s1.13

40. Codrington has demonstrated conclusively (pages 71-74) that the Cōḷas borrowed their standard from Ceylon where it had a long and continuous history before the Cōḷa occupation. The older view that 'its use was established in Ceylon, as a result of the Cōḷa occupation of the island' (Rapson) is therefore the exact reverse of the truth. See also Smith (*Catalogue of Coins in the Indian Museum*) IMC. I, pp. 327-8.

c31.s1.14

A careful study of the epigraphical evidence, however, does not seem to support Codrington's theory of the reform of the coinage by Rājarāja I, (sect. 23, p. 7) which rests so far as I am able to see on two facts: (a) the mention of *paḻangāśu* in 629 of 1916 of the 27th year of the king and (b) the weight of the lost gold coin (Elliot No. 151) with the *grantha* legend Uttama-śōḷan (pages 7 and 74). These facts are susceptible of other explanations, and the fixed relation of the Cōḷa kāśu=*Īḻakkāśu*=½ kaḻañju is anterior to Rājarāja's time as is seen from 25 of 1895 of the 24th year of Parāntaka I, 156 of 1895 and 241 of 1923, of the time Parāntaka II (Sundara Cōḷa) and Āditya II respectively.

c31.s1.15

The *nāgari* legends on the Cōḷa coins seem to appear first under Rājarāja I replacing *grantha* legends of the earlier period, cf. the woodcut of the Uttama-Cōḷa gold coin in *CSI.* It is not possible to decide if this marks a further stage in the growth of Ceylonese influence on Cōḷa coinage, or, what is equally, if not more, probable, the result of North Indian Śaiva influences beginning to operate in the Cōḷa court. On the latter hypothesis, the *nāgari* script must have first been employed on the mainland, and thence spread to Ceylon.

c31.s1.16

41. *IA.* xxv, p. 321.

c31.s1.17

42. IMC. I, pp. 313-4.

c31.s1.18

43. *EC.* i, 49.

c31.s1.19

44. Codrington, p. 8.

c31.s1.20

45. 221 of 1894. The Cōḷa *māḍa* was just one *kaḻañju* by this *kal.* 144 of 1925.

c31.s1.21

46. 211 of 1922.

c31.s1.22

47. 157 of 1916; 5 of 1890.

c31.s1.23

48. 401 of 1896.

c31.s1.24

49. 150 of 1925.

c31.s1.25

50. 563 of 1904.

c31.s1.26

51. 509 of 1912; Codrington, op. cit. p. 85.

c31.s1.27

51a. 521 of 1920.

c31.s1.28

⁵² Codrington, op. cit. p. 73.

c31.s1.29

56. 522 of 1922.
57. 439 of 1913.
58. *op. cit.* p. 85.
59. 1 of 1893.
60. 284 of 1923. 60a. 141 of 1936-7, ARE. II. 20.
61. 680 of 1919.
62. 1 of 1920.
63. 316 of 1903.
64. 353 of 1918.
65. 491 of 1911.
66. 100 of 1892.
67. *SII.* iii, p. 229 n. 5.
68. *TAS.* iv, pp. 139-41.
69. 50 of 1925; 356 of 1924.
70. The Tanjore *vēli* today is 6·7 acres. Possibly the ancient unit was not different.
71. *SII.* ii, p. 48 *nn.*
72. *SII.* ii, p. 64, paragraph 15.
73. 261 of 1902; 344 of 1912; 18 of 1922; *SII.* iii. 64. Fourteen spans in 229 of 1910, twenty in 413 of 1922; twelve in 104 of 1928.
74. 160 and 172 of 1921.
75. 87 of 1900.
76. 99 of 1914; 102 of 1925.
77. 250 of 1902.
78. 261 of 1902.
79. 102 of 1925.
80. 18 of 1922.
81. 93, 97 of 1926.
82. *SII.* iii, 64.
83. 76 of 1898.
84. 243 of 1925.
85. 155 of 1925. This record says that 4½ *mās* of this measure were equal to six *mās* by the general survey; is this only an approximation?
86. 150 of 1925.
87. 179-81 of 1918.
88. 607 of 1902.
89. 428 of 1912.
90. 155 and 144 of 1925.
91. 401 of 1921; 262 of 1921.
92. 140 of 1912.
93. 21 of 1922.
94. *ARE.* 1911, II, 21.
95. 78 of 1895.
96. 40 of 1888.
97. 556 of 1904, 731 of 1909, 504 of 1918.
98. 219 of 1921 (Rājarāja I, 22).

Linked texts (6)