Pramāṇa · Search · About · API

Indian Epigraphy c15: Chapter VII. Dating and the Eras

Indian Epigraphy c15

Language
English

Contents

Chapter VII. Dating and the Eras

c15.2

DATES IN INDIAN EPIGRAPHS

1. Nature of the Dates

c15.t1.1

As will be seen below, the dates in the earliest Indian epigraphs as well as a large number of later documents quote only a particular year of the reign of a king, the years of various eras being quoted in some records from about the beginning of the Christian era. In the earlier instances of the specification of a day in the dates, the number of the day is mentioned as belonging to one of the 8 fortnights of a season, of which three were regarded as constituting a year.¹ We have generally a reference to the month instead of the fortnight in the dates of the Kuṣāṇa epigraphs. In later instances, the day is quoted as belonging to one of the 12 solar months of the year or to either of the two fortnights (e.g. the dark and the bright) of one of the 12 lunar months. The mention of the name of the week-day became gradually popular.² After the mention of the day, sometimes the date was referred to as the *pūrvā* which really means ‘the above’ (in the feminine gender) and apparently indicates the *tithi* or lunar mansion on the day in question,³ although the *vāra* or week-day was reckoned from one sun-rise to another. The week-day of a tithi is generally the day on which the tithi ends. In some inscriptions, the nakṣatra, yoga, karaṇa, lagna and many other details are quoted in the dates besides the above.¹ Some epigraphs exhibit the mention of several eras in the date.²

c15.t1.2

¹ These were *grīṣma* (summer), *varṣā* (the rainy season) and *hemanta* (winter). See below, pp. 236-37.

c15.t1.3

² The earliest mention of the week-day by name is found in the Eran inscription of the time of Budhagupta, dated 484 A.D. (*Select Inscriptions*, p. 326). But the word *vāra* probably referring to a week-day seems to be used in the Andhau inscriptions of 130 A.D. (ibid., pp. 167 ff.).

c15.t1.4

³ The expressions used are *etasyām pūrvāyām*, *asyām divasa-pūrvāyām*, *asyām saṃvatsara-māsa-divasa-pūrvāyām*, etc., in the earlier records (cf. Bhandarkar’s List, Nos. 2, 1261, 1287, etc.; *Select Inscriptions*, pp. 120, 279, etc.) and, in some late cases (cf. Bhandarkar’s List, Nos. 241, 438,451, etc.), *asyām saṃvatsara-māsa-pakṣa-divasa-pūrvāyām tithau*. Sometimes the mention of the *tithi* is followed by such expressions (cf. ibid., No. 122). See also expressions like *daśamī-pūrvakam* (ibid., No. 1077). Instead of *pūrvā*, the words *anupūrvā*, *kṣaṇa*, *divasa*, *divasa-kṣaṇa*, *ghaṭikā*, etc., are used in some cases (cf. *Select Inscriptions*, pp. 127, 135, 136, 143, 149, 153, etc.). In one case, the expression used in an inscription dated in the Vikrama era is *asyām Mālava-pūrvāyām* (Bhandarkar’s List, No. 5).

c15.t1.5

¹ See, e.g., Bhandarkar’s List, No. 27.

c15.t1.6

² Cf. ibid., No. 565.

c15.t1.7

A peculiarity of the dating of Indian records is the use of abbreviations. Indeed, the use of contractions is first noticed in the epigraphs in connection with dates. The words varṣa, vatsara, saṁvatsara, etc., meaning ‘the year’, was often contracted into va, saṁ, saṁva, saṁvat, etc. The word saṁvat and its numerous corrupt variants were later used in the sense of ‘an era’. A few early records suggest a similar contraction of the word kāla meaning ‘an era’ into kā.³ Of the seasons mentioned in early inscriptions, hemanta (i.e. winter) was often indicated by he or hema, varṣā (i.e. the rainy season) by va or vā (for Prakrit vāsa), and grīṣma as grī or gi (for Prakrit gimha). Likewise di or diva often stood for divasa and ti for tithi. In the commonly found expression śu-di or su-di, śu stands for śuddha (bright) and su seems to be its corrupt spelling. Similarly, in ba-di or va-di, ba stands for bahula (dark) and va is either a wrong spelling or stands for vadya. In the Kashmir region, śu-ti or su-ti and ba-ti or va-ti (i.e. śudha-tithi and bahula-tithi) are generally found for śu-di and ba-di. The names of the months were often abbreviated; e.g. Phā for Phālguna. Sometimes the expression rājya-saṁvatsara is contracted into rā-saṁ.⁴ In the medieval records of South India, very often the names of the week-days are likewise contracted; e.g. So standing for Soma-vāra, Ma for Maṅgala-vāra, etc.⁵

c15.t1.8

³ Cf. Select Inscriptions, p. 131.

c15.t1.9

⁴ Cf. Bhandarkar’s List, No. 1787.

c15.t1.10

⁵ See Ep. Ind., Vol. XXXIII, pp. 108, 112, 115. In Telugu-Kannaḍa epigraphs, Saturday was often called Bṛiha-vāra or Vaḍḍa-vāra (ibid., Vol. XXXI, p. 100; Vol. XXXIII, pp. 125-26). Both the names were sometimes also applied to Thursday.

c15.t1.11

The numbers of the years, etc., in the earliest historical records of India, e.g. the inscriptions of Aśoka, were written in words. In later times, the practice of writing them by numerical figures became gradually popular.¹ In many cases, they are expressed by both words and figures. In some inscriptions from Dwarahat in the Almora District, U. P., the months and *tithis* of the lunar months are given in numbers; e.g. *māsa* 10 (i.e. the tenth month of the year).¹

c15.t1.12

¹ Cf. *A. R. Ep.*, 1958-59, No. C 382 (V.S. 1081, Month 11, *tithi* 10, Friday=20th August 1025 A. D., the year being Kārttikādi and the month *Pūrṇimānta*); No. C 384 (*Śaka* 1136, month 11, Māgha su 7, Thursday= 8th January 1215 A. D., the year being Caitrādi).

c15.t1.13

The numerals were written by symbols in the early inscriptions in Brāhmī and Kharoṣṭhī, the signs of the Kharoṣṭhī system being borrowed from the Aramaic alphabet with slight modifications. The decimal system of writing numbers with the figures for 1 to 9 and the cypher with the application of the principle of space value, which is the most convenient system of writing numbers and is now popular all over the world, originated in India, probably in West India amongst the astronomers of the Ujjayinī school, about the 5th century A. D.²

c15.t1.14

² It was borrowed by the Arabs from the Indians and gradually spread to the rest of the world. See our *Indian Palaeography*, the section on Numerals.

c15.t1.15

There was also a convention according to which ordinary words of the Sanskrit language, that can be associated with particualr numbers, were used to indicate those numbers. Thus words meaning 'the hand' (i.e. *kara*, etc.) came to be used to imply 2 since man has only two hands. Such words indicating numbers were arranged according to the principle of *vāma-gati*, i.e. 'the turning round of the sum', i.e. reading the figures from right to left.³ This system of expressing numbers is often noticed in the dates of inscriptions especially of the medieval age.⁴ Another system of expressing numbers by particular letters of the alphabet is known since 499 A.D. when Āryabhaṭa composed his *Āryabhaṭīya*. Although Āryabhaṭa's system is not noticed in inscriptions, a similar system called Kaṭapayādi, which developed in South India, is sometimes found in South Indian epigraphs of the medieval period. According to this system, the 10 letters from *ka* to *ña* and from *ṭa* to *na* have the values respectively of the numbers 1 to 9 and 0, while the 5 letters from *pa* to *ma* indicate the numbers 1 to 5 and the values of the numbers 1 to 9 are assigned respectively to the letters *ya* and *ḷa*. In the formation of chronograms, vowels, without numerical value, are added to the consonants and the last consonants in the conjuncts used have alone numerical values, while the principle of *vāma-gati* is also observed. Thus the Kandiyur inscription¹ of Kodavarman is dated in the 1,511,564th day of the Kali-yuga era expressed by the Kaṭapayādi chronogram *viṣamampuṇyamekam* in which *va*=4, *ṣa*=6, *ma*=5, *pa*=1, *ya*=1, *ma*=5 and *ka*=1 and they are to be read from right to left. Similar other known systems, such as the Nannādi developed in Kerala, do not appear in inscriptions.

c15.t1.16

³ In a few cases, the *vāma-gati* principle is found to have been ignored. Cf. Bhandarkar's List, No. 620.

c15.t1.17

⁴ See, e.g., *veda-aṅk-ābdh-indu-varṣa* (Bhandarkar's List, No. 775) in which *veda*=4, *aṅka*=9, *abdhi*=4 and *indu*=1 and, since the words are arranged according to the *vāma-gati* principle, they give the year as 1494 [ of the Vikrama-saṃvat ]. See below, pp. 228 ff.

c15.t1.18

¹ *Trav. Arch. Ser.*, Vol. I, p. 294. See also the Srirangam inscription (*Ep. Ind.*, Vol. XXIX, pp. 71 ff.) of Sadāśivarāya in which the *Śaka* year 1467 is expressed by the chronogram *setu-vandya* which gives *sa*=7, *ta*=6 *va*=4 and *ya*=1, i.e. 1467. See below, pp. 233 ff.

c15.t1.19

The dates in the earlier inscriptions generally do not give details enough for finding out their equivalents in known calendars. Mention of the week-day together with the name of the month or that of the month and the fortnight as well as of the year of an era with known epoch is very useful for this purpose. Such details are often found in later inscriptions. Gifts were generally made on auspicious occasions such as the eclipses of the sun and the moon, the full-moon and new-moon days, the sun's movement from the south to the north or vice versa, the vernal and autumnal equinoxes, the holy *tithis* and particular positions of the planets, etc., and these are often clearly indicated in the inscriptions recording royal grants. For the purpose of verification of dates, one requires to have some knowledge about the nature of the Indian years, months, etc. The beginning of the years, which were either solar or lunar, was counted variously from the months of Kārttika, Caitra, Āṣāḍha or Śrāvaṇa while the months were counted differently as solar or lunar.

c15.t1.20

The ecliptic is divided into 12 *rāśis* or signs of the Zodiac, viz. (1) Meṣa, (2) Vṛṣa, (3) Mithuna, (4) Karkaṭa, (5) Siṃha, (6) Kanyā, (7) Tulā, (8) Vṛścika, (9) Dhanus, (10) Makara, (11) Kumbha, and (12) Mīna. The sun's passage from the beginning of Meṣa to the end of Mīna is called a *saura* (solar) year (*varṣa, saṃvatsara, vatsara, etc.*). Generally the length of the solar year is regarded as 365 days, 15 *ghaṭikās* or *daṇḍas*, 31 *palas* and 30 *vipalas*,¹ although there is some slight difference of opinion among astronomers on this point. The sun's passage from the beginning to the end of a *rāśi* is called a *saura-māsa* (solar month) and its entry into a new *rāśi* a *saṅkrānti* or *saṅkramaṇa*. The day on which the *saṅkrānti* falls is counted as the first day of the solar month in the Punjab while, if it falls in the night, the following day becomes the first day of the month. The system followed in the Tamil region is similar to that followed in the Punjab. But, in Bengal, the day following the *saṅkrānti* day is regarded as the first day of the month. This leads to considerable difference in the almanacs of various parts of the country.

c15.t1.21

¹ The *ghaṭikā* or *daṇḍa* is the one-sixtieth part of the day, i.e. about 24 minutes, while the *pala* is its one-sixtieth part and the *vipala* is one-sixtieth of the *pala*.

c15.t1.22

The solar months are usually named after the particular *rāśi* through which the sun passes at the time, although the names of the lunar months are also often applied to them. The solar years are prevalent in Bengal and the Punjab and their neighbourhood as well as in the Tamil- and Malayāḷam-speaking areas. The solar month of Meṣa, with which the solar year begins, is called Vaiśākha in Bengal but Caitra (Cittirai) in the Tamil-speaking area.

c15.t1.23

There are generally between 29 and 32 days in a solar month and they are counted from 1 to 29, 30, 31 or 32 without reference to a *tithi* or lunar day (i.e. one-thirtieth part of the whole lunation) of the dark or bright fortnight (*pakṣa*), even though the *tithis* are also counted for the observance of particular functions in the areas where the solar months are prevalent.¹

c15.t1.24

¹ The names of the solar months in Tamil and Bengali are as follows: (1) Meṣa—Cittirai, Vaiśākha; (2) Vṛṣabha—Vaikāsi, Jyaiṣṭha; (3) Mithuna—Āni, Āṣāḍha; (4) Karkaṭa—Āḍi, Śrāvaṇa: (5) Simha—Avani, Bhādra: (6) Kanyā—Puraṭṭāśi (Puraṭṭādi), Āśvina; (7) Tulā—Aippaśi, Kārttika; (8) Vṛścika—Kārttigai, Mārgaśīrṣa; (9) Dhanus—Mārgaḷi, Pauṣa; (10) Makara—Tai, Māgha; (11) Kumbha—Māśi, Phālguna; (12) Mīna—Paṅguni, Caitra.

c15.t1.25

A lunar month consists of two fortnights or *pakṣas*, viz. dark and bright. The fifteen *tithis* or lunar days of either of the dark or bright fortnight, called respectively the *kṛṣṇa* or *bahula* and the *śukla*, are counted from 1 to 15. In North India, the lunar month is generally counted as *Pūrṇimānta*, i.e. ‘ending in the full-moon (i.e. sudi 15)’ or lasting from *Kṛṣṇa* (i.e. badi) 1 to *Śukla* (i.e. sudi ) 15, while, in the land to the south of the Narmadā, the lunar month is *Amānta*, i.e. it begins on *Śukla* (sudi) 1 and ends with the new-moon (i.e. badi 15).²

c15.t1.26

² *Amānta* | *Pakṣa* | *Pūrṇimānta*

c15.t1.27

The twelve lunar months are named after the *nakṣatras* or lunar mansions on which the full-moon occurs in each one

c15.t1.28

| Phālguna ... ... ... | {Śukla Kṛṣṇa} ... ... ... Caitra |
| Caitra ... ... ... | {Śukla Kṛṣṇa} ... ... ... Vaiśākha |
| Vaiśākha ... ... ... | {Śukla Kṛṣṇa} ... ... ... Jyaiṣṭha |
| Jyaiṣṭha ... ... ... | {Śukla Kṛṣṇa} ... ... ... Āṣāḍha |
| Āṣāḍha ... ... ... | {Śukla Kṛṣṇa} ... ... ... Śrāvaṇa |
| Śrāvaṇa ... ... ... | {Śukla Kṛṣṇa} ... ... ... Bhādrapada |
| Bhādrapada ... ... ... | {Śukla Kṛṣṇa} ... ... ... Āśvina |
| Āśvina ... ... ... | {Śnkla Kṛṣṇa} ... ... ... Kārttika |
| Kārttika ... ... ... | {Śukla Kṛṣṇa} ... ... ... Mārgaśīrṣa |
| Mārgaśīrṣa ... ... ... | {Śukla Kṛṣṇa} ... ... ... Pauṣa |
| Pauṣa ... ... ... | {Śukla Kṛṣṇa} ... ... ... Māgha |
| Māgha ... ... ... | {Śukla Kṛṣṇa} Śukla ... ... ... Phālguna |

c15.t1.29

of them.¹ These are : (1) Caitra, (2) Vaiśākha, (3) Jyaiṣṭha, (4) Āṣāḍha, (5) Śrāvaṇa, (6) Bhādrapada, (7) Āśvina (also called Āśvayuja), (8) Kārttika, (9) Mārgaśīrṣa (also called Agrahāyaṇa probably because the beginning of the year was once counted from it in some parts of the country), (10) Pauṣa, (11) Māgha, and (12) Phālguna.

c15.t1.30

¹ The nakṣatras are 27 in number, viz. (1) Aśvinī, (2) Bharaṇī, (3) Kṛttikā, (4) Rohiṇī, (5) Mṛgaśīrā, (6) Ārdrā, (7) Punarvasu, (8) Puṣyā, (9) Aśleṣā, (10) Maghā, (11) Pūrva-Phalgunī, (12) Uttara-Phalgunī, (13) Hastā, (14) Citrā, (15) Svātī, (16) Viśākhā, (17) Anurādhā, (18) Jyeṣṭhā, (19) Mūlā, (20) Pūrv-Āṣāḍhā (21) Uttar-Āṣāḍhā, (22) Śravaṇā, (23) Dhaniṣṭhā, (24) Śatabhiṣā, (25) Pūrva-Bhādrapadā, (26) Uttara-Bhādrapadā, and (27) Revatī.

c15.t1.31

The length of the lunar year made of twelve such months is regarded as 354 days, 22 ghaṭikās or daṇḍas, 1 pala and 24 vipalas, the year therefore being shorter than the solar year by 10 days, 56 ghaṭikās or daṇḍas, 30 palas and 6 vipalas. Thus 32 lunar months become roughly equal to 31 solar months. Since the tithis have considerable importance in the religious life of the people throughout the country, the lunar year is made equal to the solar year in the areas following the solar system by what may be called the luni-solar reckoning. According to this, the lunar month, in which the saṅkrānti (i.e. the sun's entry into a rāśi) does not occur, is called the Adhika or Mala month, while the lunar month witnessing two saṅkrāntis is regarded as the Kṣaya or Nija month. Medieval inscriptions sometimes mention yogas and karaṇas in the dates. The yogas (the word yoga meaning 'conjunction' used with reference to the astronomical association of the movements of the sun and the moon and noticed in some records)² are 27 in number; e. g. (1) Viṣkambha, (2) Prīti, (3) Āyuṣmat, (4) Saubhāgya, (5) Śobhaṇa, (6) Atigaṇḍa, (7) Sukarman, (8) Dhṛti, (9) Śūla, (10) Gaṇḍa, (11) Vṛddhi, (12) Dhruva, (13) Vyāghāta; (14) Harṣaṇa, (15) Vajra, (16) Siddhi or Asṛja, (17) Vyātīpāta, (18) Varīyas, (19) Parigha, (20) Śiva, (21) Siddha, (22) Sādhya, (23) Śubha, (24) Śukla, (25) Brahman, (26) Aindra and (27) Vaidhṛti. The karaṇas, each of which indicate half of a tithi, are noticed in medieval inscriptions like the Ajmer inscription of Vikrama 1210.¹ The karaṇas called (1) Bava, (2) Bālava, (3) Kaulava, (4) Taitila, (5) Gara, (6) Vaṇija and (7) Viṣṭi (Bhadrā or Kalyāṇī) make a cycle which are repeated for eight times covering 56 half tithis from the second half of śudi 1 to the first half of badi 14. The remaining four half tithis from the second half of badi 14 to the first half of śudi 1 are covered by the karaṇas : (8) Śakuni, (9) Catuṣpada, (10) Kintughna and (11) Nāga.

c15.t1.32

² See, e. g., Bhandarkar's List, No. 27 dated Vikrama 898.

c15.t1.33

¹ See Bhandarkar's List, No. 289.

c15.t1.34

The conception of a week of seven days does not appear in very early Indian literature. As indicated above, the name of the week-day first occurs in an Indian epigraph of the last quarter of the 5th century A.D., and it is generally believed that the conception was borrowed by the Indians from Greek astronomers. As in the West, the Indian week-days are named after seven out of the nine grahas : (1) the sun (called Ravi, Āditya, etc.), (2) the moon (called Soma, Candra, etc.), (3) Mars (called Bhauma, Maṅgala, etc.), (4) Mercury (called Budha, etc.), (5) Jupiter (called Bṛhaspati, Suraguru, etc.), (6) Venus (called Śukra, Daityaguru, etc.), and (7) Saturn (called Śani, Śanaiścara, etc.). Thus Sunday is called Ravi-vāra, Āditya-vāra, etc.

c15.t1.35

From the nature of the Indian year and month discussed above, it will be seen that the verification of the dates of some records is beset with difficulties. By way of illustration, it may be pointed out that it is often difficult to determine whether the years quoted, even when they are referred to an era, are Kārttikādi, Caitrādi, Āṣāḍhādi or Śrāvaṇādi and whether the month is Amānta or Pūrṇimānta. It is also often unknown whether the regnal years of kings were counted exactly from the date of his accession or they were adjusted with the regular years of the almanac. Another difficulty is that the years were sometimes regarded as current though generally they were counted as expired.² But there is a greater difficulty which is the possibility of errors in the enumeration of the dates. There is no doubt that such errors account for a large number of irregular dates in the inscriptions of various ages in different parts of the country.

c15.t1.36

² Sometimes the records use the words *pravartamāna* and *atīta* to indicate respectively the current and expired years of a reckoning.

c15.t1.37

Some of the errors in the dates of Indian records have to be ascribed to the astrologer at the villages and courts.¹ With the innumerable copies of printed almanacs, called Pañjī, Pañjikā or Pañcāṅga, of various types in each one of the numerous regional languages, available to the literate people in all parts of India, it is impossible in these days of advanced literacy in the country to realise the great importance of the astrologer in Indian social life before the 19th century.² There is epigraphic evidence to show that the office of the village astrologer was an institution recognised by the State during Hindu rule.³ That the astrologers at the royal courts of ancient and medieval India were often not competent astronomers is also clearly shown by the Sarkho plates⁴ of Kalacuri Ratnadeva II, dated 1128 A.D.

c15.t1.38

¹ See IHQ, Vol. XXVIII, pp. 342 ff.

c15.t1.39

² See L. B. Day, Bengal Peasant Life, Chapter VIII; Dubois, Hindu Manners and Customs, 3rd ed., pp. 135-36. Cf. Kane, History of Dharmaśāstra, Vol. III, p. 126.

c15.t1.40

³ Cf. T. V. Mahalingam, Administration and Social Life under Vijayanagara, pp. 218-19. The Brahmanical family name Jośi (Sanskrit Jyotiṣin), found in various parts of India, seems to speak of royal recognition for the families of the astrologers in question. The family name Praharāja (Sanskrit Praharāja), found among Oriya Brāhmaṇas, was originally a title conferred on an astrologer by the king.

c15.t1.41

⁴ Ep. Ind., Vol. XXII, pp. 159 ff. It is stated that the astronomers at the court of the Kalacuri king were following wrong methods of calculation, as a result of which their predictions relating to eclipses hardly came true. A good astronomer named Padmanābha discovered the mistakes in their methods, made the required bīja-saṃskāra and correctly calculated the exact time of a lunar eclipse. Padmanābha was rewarded by the king for his success.

c15.t1.42

There is a tendency among a section of epigraphists to view the genuineness of an inscription bearing an irregular date with suspicion. But it seems to be a wrong approach to the problem since the irregularity in the dates may be due to factors absolutely unconnected with the question of the genuineness of the documents. Many cases of irregularity were no doubt due to mistakes in the calculation of incompetent astrologers and the erroneous method of calculation followed in the preparation of the almanacs.¹ Varying local traditions regarding the astrological character of particular days, *tithis* or moments may have also contributed to the difference referred to. Another cause of irregularity in the dates of royal documents seems to have been the fact that, in some cases at least, the court astrologer had reason to fabricate an auspicious moment when really there was none. This was hardly difficult for him to do since his royal master and most of the latter’s courtiers were blessedly ignorant of astronomy and had to depend entirely on the astrologer for the determination of an appropriate moment for a particular undertaking. Of course there may be other possible reasons for an irregularity creeping into the date of a royal charter.

c15.t1.43

¹ The almanacs now published in different parts of India or even in the same part of the country do not exactly tally with one another owing principally to the difference in the method of calculation followed or in the approach to particular astronomical problems. If the details of a date do not tally with those quoted by modern astronomers (e.g. L. D. Swamikannu Pillai in his *Indian Ephimeris*), it may be irregular technically speaking, even though it was quite regular according to a particular almanac from which the details were quoted in the inscription.

2. Numbers expressed by Words

c15.t2.1

The Indian students of astronomy, mathematics and many other subjects generally composed their works in verse. The nature of the works in some of these subjects necessitated the repeated mention of numbers, e.g., in the setting and solution of the problems of an arithmetical sum. There was, however, generally only one word in Sanskrit to indicate a particular number. Since the repeated mention of a word in the same sense was regarded as a bad style of composition, and such a word would often not fit in the metre of a stanza, the authors required a large number of words for indicating the same number. A convention therefore gradually developed, according to which ordinary words that can somehow be associated with particular numbers were used to imply those numbers. Thus all the Sanskrit words meaning ‘the hand’ came to be used to indicate 2, since man has only two hands. Similarly words meaning ‘the arrow’ were employed in the sense of 5 because Kāma, the Indian god of love, is represented in tradition as armed with five arrows of flowers. When there were more than one tradition regarding the number associated with a particular object, only one of them was adopted. Thus, even though traditionally the Vedas are regarded either as three or as four and the seas either as four or as seven in number, the early Indian astronomers and mathematicians used the words indicating both the Vedas and the sea to indicate 4 only. But uncritical authors of the medieval age are sometimes found to have violated this old convention and introduced an amount of confusion in the system. Thus, in some cases, we notice the use of a word for ‘the sea’ to imply both the numbers 4 and 7.¹ The name of a metre was often used to indicate a number according to the number of syllables in a foot of a stanza in that metre. Thus the word anuṣṭubh, the name of a metre with eight syllables in a foot, implied 8. But there was want of complete uniformity since in some cases the number of syllables in all the four feet of a stanza was counted; e.g., Jagatī, a metre with 12 syllables in a foot, was used by late authors in the sense of both 12 and 48. When such a word-numeral was a compound of two words, often either the first or the second member of the compounds was used to indicate the number in question.

c15.t2.2

¹ The polygraphic (multiscriptal) inscription of king Jaya-Pratāpamalla of Nepal is dated in *Nepāl-ābde yug-ābdhau muni-gaṇa-sahite* given in figures as Saṃva (Saṃvat) 774, thus equating *abdhi* (sea) to 7. The exact date of the record is quoted as Friday, Māgha-sudi 1 (13th January 1654 A.D.). See *Journ. As.*, Tome CCXLV, 1957, p. 338 and Plate. Cf. Śaka 1748=Bengal San 1233, the date of the composition of Jagaddurlabha’s *Uddhavacamatkāra* given in the colophon of the work (*Bhāratavarṣa*, Bengali, Śrāvaṇa, 1365 B.S., p. 173) respectively as *aṣṭa-sāgara-payonidhi-candra* and *netra-Rāma-yuga-candra*. The word *netra* = 3 according to the Bengali convention and *yuga* = 2 (not 4 as usual). Indeed it would have been difficult to determine the date of the composition of the work, in spite of the years in both the eras being quoted, unless it was separately stated that the work was completed on the 12th Caitra, 1233 B. S. (24th March 1827 A.D.). The Bengal San has been referred to as *yāvanika* and *ādhunika*. See below, pp. 230 ff.

c15.t2.3

In the early astronomical and arithmetical works, single such words are generally used to indicate particular number. Piṅgala, Varāhamihira and later authors use such words with numerical meanings as grouped in Dvandva compounds. But, while Piṅgala’s *ved-artu-samudrāḥ* means ‘4 or 6 or 4’, Varāhamihira’s *kha-kh-aṣṭi-yamāḥ* would mean ‘0 and 0 and 16 and 2’ to be read actually from right to left as 21600.¹ This principle of *aṅkānāṁ vāmato gatiḥ*, ‘the movement of the numerals from right to left’, is generally followed in the dates expressed in inscriptions in word-numerals although there are a few exceptions especially in late records.² Such numerical words are sometimes found also in the ordinary enumeration of a number in words; cf. *Śaka* 1172 indicated by *dvisaptaty-adhika-Śiva-śata*,³ in which *Śiva* means ‘eleven.’

c15.t2.4

¹ Cf. Bühler, op. cit., p. 86.

c15.t2.5

² See, e.g., Bhandarkar’s List, No. 620.

c15.t2.6

³ A.R. Ep., May 1890, p. 2.

c15.t2.7

We give below an indication as to the words that were used to express particular numbers.⁴

c15.t2.8

⁴ See Bühler, op. cit., pp. 84 ff.; Ojha, op. cit., p. 120; Filliozat, op. cit., pp. 708-09; Sewell, Eclipses of the Moon in India, pp. 7-8. A few words have been added from other sources, especially epigraphic.

c15.t2.9

0 = *śūnya*, ‘the empty space of the sky’ (*kha, gagana, ambara, abhra, viyat, nabhas, antarikṣa, Viṣṇupada, ākāśa, ananta, vyoman, dyu, etc.*), *pūrṇa, vindu*, etc. The word *randhra*, which generally means 9, is used to imply 0 (cypher) in a few cases by late authors.

c15.t2.10

1 = *ādi*, ‘the moon’ (*śaśin, indu, vidhu, candra, śītāṁśu, śītaraśmi, soma, śaśāṅka, sudhāṁśu, kṣapeśvara, Atrinayanaja, mṛgāṅka, candramas, etc.*), ‘the earth’ (*bhū, bhūmi, kṣiti, dharā, urvarā, go, vasundharā, pṛthvī, kṣmā, dharaṇī vasudhā, ilā, ku, mahī, dhātrī, jagatī kṣoṇi, kṣauṇī, avanī, etc.*), *ajba, raśmi, rūpa, Pitāmaha* (but cf. *Brahman* meaning 9), *nāyaka, tanu, jana, Airāvata* (sometimes wrongly *hastin*, etc., which really mean 8), *Uccaiśravas* (sometimes wrongly *aśva*, etc., which really mean 7), ‘the single eye of Śukra’ (*Śukra-netra*, etc.), *Dhruva, ātman, Parabrahman, Mūlaprakṛti*, etc. *Jagatī* is sometimes used to mean 12 and 48 and *go* 9.

c15.t2.11

2 = *yama, yamala, yugala, dvaya, yugma, dvandva, Aśvin* (sometimes also *Aśvina*), *Nāsatya, Dasra*, ‘the eye’ (*locana, netra, akṣi, dṛṣṭi, cakṣus, nayana, īkṣaṇa, ambaka, etc.*), the words meaning other limbs which are in pairs such as bāhu, kara, karṇa, kuca, oṣṭha, gulpha, jānu, etc., pakṣa, kuṭumba, ravi-candra, grahaṇa, nadī-kūla, Rāma-nandana, viṣuvat, ayana, etc. In the Bengal region, the word netra, ‘the eye’, is generally taken to stand for ‘Śiva’s eyes’ and therefore in the sense of 3. Sometimes yuga=2 instead of 4, and pakṣa=15.

c15.t2.12

3=Rāma, guṇa, tri-guṇa, ‘the world’ (loka, jagat, tri-jagat, bhuvana, etc., some of which are sometimes also used for 14), ‘the fire’ (agni, vahni, pāvaka, vaiśvānara, dahana, tapana, udarcis, hutāśana, jvalana, śikhin, kṛśānu, hotṛ, anala, etc.), kāla, tri-kāla, tri-gata, ‘Śiva’s eyes’ (trinetra, Hara-nayana Śaṅkar-ākṣi, Īśa-dṛś, etc.), sahodara, doṣa, nāḍī, śūla, gaṅgā-mārga, viṣṭapa, mūrti, vacana, ‘Daśaratha’s wives’, etc. For śakti=3, see Ep. Ind., Vol. XIX, p. 166, and guṇa=6 below. Veda rarely indicates 3, though it really stands for 4. Tapana in the sense of ‘the sun’ would mean ‘twelve’.

c15.t2.13

4=‘the Veda’ (veda, śruti, etc.), ‘the sea’ (sāgara, arṇava, amburāśi, abdhi, jaladhi, jalanidhi, ambhonidhi, vārinidhi, ambudhi, jalāśaya, sindhu, vāridhi, etc., later used also in the sense of 7), kendra, varṇa, āśrama, yuga (rarely 2), turya, kṛta, aya, āya, bandhu, koṣṭha, ‘Brahman’s faces’ (Brahm-āsya, Caturānana-vadana, etc.), ‘Viṣṇu’s arms’ (Hari-bāhu, etc.), vyūha, sen-āṅga, khaṭvā-pāda, go-stana, puruṣ-ārtha, etc. For gati in the sense of 4, cf. Ep. Ind., Vol. XXV, p. 336. The words diś, diśā, etc., meaning 10, are sometimes found in the sense of 4. Jala (for jaladhi), dadhi (for udadhi) and vanadhi (mistake for vāridhi ?) are also counted in the same sense.

c15.t2.14

5=‘the arrow’ (bāṇa, śara, sāyaka, iṣu, viśikha, kalamba, mārgaṇa, etc.), bhūta, mahābhūta, parvan, prāṇa, ‘Pāṇḍu’s sons’ (Pāṇḍava, etc.), artha, viṣaya, indriya, ratna (also 9 or 14), ‘Śiva’s faces’ (Rudr-āsya, Tryakṣa-mukha, etc.), ‘the sins’ (pātaka, mahāpāpa, etc.), Purāṇa-lakṣaṇa, pallava, gavya, tanmātra, kṣāra, lavaṇa, ‘the sacrifices’ (mahāyajña, etc.), tattva (generally used to indicate 25), etc. Words meaning ‘a son’ (suta, putra, etc., probably meaning Pāṇḍu’s sons) and tata (for tattva ?) are also counted in the same sense.

c15.t2.15

6=rasa, aṅga, kāya, darśana (meaning the 6 darśanas and not dṛṣṭi, ‘the eye’, meaning 2), rāga, ‘the enemy’ (ari, ripu, etc.), śastra, mala, tarka, kāraka, ṛtu, māsārdha (cf. māsa=12; but pakṣa=2), ‘Karttikeya’s faces’ (Kumār-āsya, etc.), etc. In late medieval records, the word guṇa is sometimes used to indicate 6 instead of 3. Cf. A. R. Ep., 1957-58, No. B 68.

c15.t2.16

7 = ‘the hill or mountain’ (naga, aga, bhūbhṛt, parvata, śaila, adri, giri, bhūdhara, mahīdhara, etc.), ṛṣi, muni, Atri, vāra, svara, dhātu, ‘a horse’ (aśva, turaga, turaṅgama, vājin, etc.), chandas, dhī, kalatra, dvīpa, etc. Loka, pātāla, sāgara and Sagara are also counted in the same sense.

c15.t2.17

8 = Vasu, ‘an elephant’ confused with ‘a snake’ (ahi, nāga, gaja, dantin, diggaja, hastin, mātaṅga, kuñjara, dvīpa, sarpa, etc.), takṣan, siddhi, bhūti, Anuṣṭubh, maṅgala, Dikpāla, tanu, yāma, etc. Naga (really 7), mūrti (cf. rūpa=1) and tanu (really 1) are soemtimes also counted in the same sense. Naga is used in the sense of ‘eight’ in Ep. Ind., Vol. XXXV, p. 140.

c15.t2.18

9 = aṅka, Nanda, nidhi, nidhāna, randhra, chidra, dvāra, ‘the planet’ (graha, etc.), go (also 1), pavana, Bṛhatī, varṣa, khaṇḍa, ‘the god Brahman’ (Aja, Abjagarbha, etc.), ratna (also 5 or 14), etc.

c15.t2.19

10 = ‘the quarters’ (diś, diśā, āśā, kakubh, etc.), aṅguli (also used to indicate 20), paṅkti, ‘Rāvaṇa’s heads’ (Rāvaṇa-śiras, etc.) avatāra, karman, la-kāra, etc.

c15.t2.20

11 = ‘the god Rudra’ (Rudra, Śiva, Īśvara, Hara, Īśa, Bhava, Bharga, Śūlin, Mahādeva, etc.), akṣauhiṇī, lābha, Triṣṭubh, etc.

c15.t2.21

12 = ‘the sun’ (ravi, sūrya, sahasrāṃśu, arka, mārtaṇḍa, dyu-maṇi, bhānu, āditya, divākara, etc.), māsa, rāśi, vyaya, saṅkrānti, Jagatī (sometimes used also in the senses of 1 and 48), ‘the eyes or arms of Kārttikeya’ (Senānī-netra, Ṣaṇ-mukha-bāhu, etc.), etc.

c15.t2.22

13 = Viśvedevāḥ (sometimes contracted to Viśva), Kāma, Manmatha, Atijagatī, aghoṣa, etc.

c15.t2.23

14 = Manu, vidyā, ‘the god Indra’ (Indra, Śakra, etc.), loka (sometimes also used to indicate 3), etc. Pūrva [of the Jains] and ratna (also 5 or 9) are sometimes counted in the list of words of the same sense.

c15.t2.24

15 = tithi, ghasra, dina, ahan, pakṣa (sometimes also used to imply 2), etc.

c15.t2.25

16 = ‘a king’ (*nṛpa*, *bhūpa*, etc.), *Aṣṭi*, *kalā*, etc. 17 = *Atyaṣṭi*. 18 = *Dhṛti*. 19 = *Atidhṛti*. 20 = *nakha*, *aṅguli* (also used to indicate 10), *Kṛti*, ‘Rāvaṇa’s arms’ (*Rāvaṇa-bhuja*, etc.), etc. 21 = *Prakṛti*, *svarga*, *Utkṛti* (generally used in the sense of 26). 22 = *Akṛti*, *jāti*, *Kṛtin*. 23 = *Vikṛti*. 24 = *Gāyatrī*, *Jina*, *Arhat*, *Siddha*, etc. 25 = *tattva* (rarely used to indicate 5). 26 = *Utkṛti* (rarely used to indicate 21). 27 = ‘the stars’ (*nakṣatra*, *uda*, *bha*, etc.). 32 = ‘the teeth’ (*danta*, *rada*, etc.). 33 = ‘the gods’ (*amara*, *deva*, *sura*, *tridaśa*, etc.). 34 = *tāna* (generally used in the sense of 49). 40 = *naraka*. 48 = *Jagatī* (sometimes used to indicate 1 and 12). 49 = *tāna* (cf. its use in the sense of 34), *vāyu*. 100 = *Dhārtarāṣṭra*, *puruṣ-āyus*, *abja-dala*, *Śakra-yajña*, etc. 1000 = ‘the mouths of the Ganges’ (*Jāhnavī-vaktra*, etc.) ‘Śeṣa’s heads’ (*Śeṣa-śīrṣa*, etc.), ‘the Sun’s rays’ (*Ravi-bāṇa*, etc.), ‘Arjuna’s arms’ (*Arjuna-kara*, etc.), ‘Indra’s eyes’ (*Indra-dṛṣṭi*, etc.), etc.

3. Numbers expressed by Letters

c15.t3.1

The celebrated astronomer Āryabhaṭa, who wrote his Āryabhaṭīya at Pāṭaliputra in the Kaliyuga year 3600 (499 A.D.), used a system of representing numbers by letters of the alphabet.¹ The 25 consonants from *ka* to *ma* indicated respectively the numerals 1 to 25. The 8 letters from *ya* to *ha* similarly indicated the numbers 30, 40, 50, 60, 70, 80, 90 and 100 respectively. The vowels *i*, *u*, *ṛ* and *ḷ* (short or long) and the diphthongs *e*, *ai*, *o* and *au* were added to the letters to indicate the multiplication of a number by 10² to 10¹⁶, i.e. the addition of two, four, six, eight, ten, twelve, fourteen and sixteen zeroes after a number indicated by a consonant. Thus *ga*=3, *gi*=300, *gu*=30000, *ya*=30, *yi*=3000, *yu*=300000, *ri*=4000, *giri* (or *gri*)=4300 (300+4000), etc.

c15.t3.2

¹ Cf. Filliozat, op. cit., p. 188. See above, p. 222.

c15.t3.3

A system developed in South India was called Kaṭapayādi. In this, the 10 letters from *ka* to *ña* had the value respectively of the numbers 1 to 9 and 0. The same was the case with the following 10 letters from *ṭa* to *na*. The five letters from *pa* to *ma* indicated the numbers 1 to 5, and the 9 letters from *ya* to *la* had the value of the numbers 1 to 9 respectively. In the formation of chronograms, vowels were added to the consonants without affecting the numerical value of the letters and two or more consonants could be combined in conjuncts in which only the last consonant of a ligature had the numerical value. But the principle of *vāma-gati*, i.e. the turning round of the sum, was observed in the determination of the value of a word. Thus *ga*=3 and *ja*=8; but the word *gaja*=83 and not 38. The chronogram *khago=ntyān=meṣam=āpa* would give *kha*=2, *ga*=3, *ya*=1, *ma*=5, *ṣa*=6, *ma*=5 and *pa*=1, i.e. 1565132.¹

c15.t3.4

¹ See Bühler, op. cit., p. 87; Ojha, op. cit., 122 ff. See above, p. 222.

c15.t3.5

The Nannādi system popular in Kerala was based on the following lines : *na-nna-nya-ṣka-jhra-hā(ha)-gra-pra-dre-ma* | *tha-li(la)-pia-ba-tra(tru)-tri-ca-ṇa-ña* ||²

c15.t3.6

² Cf. *JRAS*, 1896, p. 790; *Trav. Arch. Ser.*, Vol. II, pp. 132. ff.

c15.t3.7

The 10 letters of the first line had the value respectively of the numbers from 1 to 10 while the others in the second line represented 20, 30, 40, 50, 60, 70, 80, 90 and 100. This system was generally employed in paginating manuscripts.

c15.t3.8

In a simpler system prevalent in South India, the letters of the alphabet from *ka* to *ḷa* represented respectively the numbers 1 to 34, *kā* to *ḷā* the numbers 35 to 68, *ki* to *ḷi* the numbers 69 to 102, and so on.

c15.t3.9

There was a similar system of paginating manuscripts prevalent in Ceylon. According to this *ka*, *kā*, *ki*, *kī* *ku*, kū, kṛ, kṛ, kl, kl, ke, kai, ko, kau, kaṃ and kaḥ had the value respectively of 1 to 16. Similarly 16 akṣaras formed by kha khā, etc., had the value of the numbers from 17 to 32. The 16 akṣaras associated with ḷa, the last consonant of the alphabet, thus represented the numbers 529 to 544. Numbers beyond 544 could be expressed by beginning again as 2 ka, 2 kā, and so on. In Burma, the same system was used with the omission of ṛ, ṝ, ḷ and ḹ, so that ka to kaḥ would indicate the numbers 1 to 12.

4. Regnal Reckoning and Connected Problems

c15.t4.1

¹ See *Vikrama Volume*, ed. R. K. Mookerji, pp. 557 ff.; *Ep. Ind.*, Vol. XXX, pp. 5-6; above, p. 219.

c15.t4.2

The earliest documents of the indigenous kings of India bear no trace of the use of any era. Many of them contain no date at all. But some of the official records of the independent rulers as well as those belonging to their officials and subordinates and persons residing in their dominions are dated. The date is, however, always expressed in the king’s regnal years. A few of the numerous instances may be cited.

c15.t4.3

(1) Rock Edict III of Aśoka (c. 272-32 B.C.) —dbādasa-vās-ābhisitena mayā idaṃ āñapitaṃ (Girnar text),² “I have ordered this when 12 years [have passed after] my installation [on the throne].”

c15.t4.4

² *Select Inscriptions*, p. 19. Cf. the same date in Rock Edict IV (ibid., p. 22) and other dates similarly expressed in Rock Edict V (ibid., p. 23: Mansehra—treḍaśa-vaṣ-abhisitena maya dhrama-mahamatra kaṭa); Rock Edict VIII (ibid., p. 28 : Girnar—Devānampriyo Piyadasi rājā dasa-vas-ābhisito saṃto ayāya Saṃbodhiṃ); Rock Edict XIII (ibid., pp. 35-36: Shahbazgarhi—aṭha-vaṣa-abhisitasa Devanapriasa Priadrasisa Kaliga vijita); etc.

c15.t4.5

(2) Besnagar inscription of the time of Bhāgabhadra (c. close of the 2nd century B.C.)—raṇo Kāsīputrasa Bhāgabhadrasa trātārasa vasena catudasena rājena vadhamānasa,³ ‘when king Kāsīputra Bhāgabhadra, the saviour, was prospering with his sovereignty (or, realm) for 14 years (i.e., when 14 years of his reign had passed).’

c15.t4.6

³ Ibid., p. 91.

c15.t4.7

(3) Pabhosa inscription of the time of Udāka (c. close of the 1st century B.C.)—Ūdākasa dasama-savacchare,¹ ‘in the tenth [regnal] year of Ūdākā.’

c15.t4.8

¹ Ibid., p. 98.

c15.t4.9

(4) Hāthigumphā inscription of Khāravela (c. close of the 1st century B.C.)—abhisitamato ca padhame vase,² ‘during the first regnal year of the installed [monarch].’

c15.t4.10

² Ibid., p. 207.

c15.t4.11

(5) Nasik inscription of Gautamīputra Śātakarṇi (c. 106-30 A.D.)—savachare 18 vāsa-pakhe 2 divase 1,³ ‘on the 1st day of the 2nd fortnight of the rainy season in the 18th [regnal] year.’

c15.t4.12

³ Ibid., p.

c15.t4.13

(6) Nagarjunikonda inscription of Ehuvula Śāntamūla (c. 280-335 A.D.)—Ehuvula-Caṃtamūlasa saṃvacharaṃ bitiyaṃ gimha-pakkhaṃ chaṭhaṃ 6 divasoṃ dasamaṃ 10,⁴ ‘10th day of the 6th fortnight of summer in the second regnal year of Ehuvula Śāntamūla.’

c15.t4.14

⁴ Ibid., p.

c15.t4.15

A study of the above and numerous other instances of the kind would suggest that originally the Indian custom was to refer to the past or current regnal year of the king without specification of the day. A little later, the day is found mentioned in some parts of India with reference to the specific fortnight of a season. The year seems to have been divided into 3 seasons, viz., summer (grīṣma), the rains (varṣā) and winter (hemanta) each consisting of 8 fortnights.⁵ The seasons ended with the Cāturmāsī which was usually identified with the full-moon tithi of the months of Āṣāḍha, Kārttika and Phālguna. According to pupular tradition, summer consisted of the pūrṇimānta months of Caitra, Vaiśākha, Jyaiṣṭha and Āṣāḍha; the rains of Śrāvaṇa, Bhādra, Āśvina and Kārttika; and winter of Mārgaśīrṣa, Pauṣa, Māgha and Phālguna.⁶ A South Indian epigraph of c. 300 A.D.¹ suggests the later inclusion of *śarad*, autumn, in the list of seasons in some parts of the country.

c15.t4.16

¹⁹² Cf. similar other dates in the other records of the time of the Śātavāhanas (ibid., p. 195 : Puḷumāvisa savachare satame 7 gimha-pakhe pacame 5 divase pathame 1; etc.).

c15.t4.17

²²⁹ For other dates, cf. Ep. Ind., Vol. XX, pp. 1 ff.; Vol. XXXV, pp. 1 ff.

c15.t4.18

⁵ Cf., e.g., Lüder’s List, No. 1186: hemantāna pakho 7 divasa 1.

c15.t4.19

⁶ Select Inscriptions, p. 63, p. 119 note, p. 134 note. The seasons are often counted as six as follows (cf. Amarakosa, Kālavarga, v. 21): (1) Vasanta or Madhu=Caitra-Vaiśākha; (2) Grīṣma or Śuci=Jyaiṣṭha-Āṣāḍha; (3) Varṣā or Nabhas=Śrāvaṇa-Bhādrapada; (4) Śarad=Āśvina-Kārttika; (5) Hemanta or Sahas=Mārgaśīrṣa-Pauṣa; (6) Śita, Śiśira or Tapas=Māgha-Phālguna. But sometimes Vasanta is regarded as consisting of Phālguna and Caitra instead of Caitra and Vaiśākha, and the other seasons are counted accordingly.

c15.t4.20

¹ Lüder's List, No. 1196. Cf. *sarada-pakkham bitiyam divasam paḍamam*; cf. also *Successors of the Sātavāhanas*, p. 223.

c15.t4.21

The dates found in the earliest records of the indigenous kings of India thus point to the absence of the custom of dating royal documents in the years of an era and also to the non-existence of any popular era in ancient India. There are, however, some Indian eras which have their epochs falling before the 4th century B.C. These are the Kali-yuga era of 3102 B.C., the Śrī-Harṣa era of 457 B.C. referred to by Al-Bīrūnī and the Parinirvāṇa eras of the Buddhists and the Jains. Purāṇic passages like the one giving the intervening period between the birth of Parikṣit and the accession of Mahāpadma Nanda as 1015 or 1050 or 1500² years no doubt suggest that the chroniclers of ancient Indian historical traditions attempted to base their chronological scheme on such important events as the birth of Parikṣit (placed by some authorities immediately after the battle of Kurukṣetra and at the beginning of the Kali-yuga) and the accession of the great empire-builder Mahāpadma Nanda. But the facts that the chroniclers do not always begin their computation from a point near about the supposed beginning of the Kali-yuga and that there is no uniformity among the traditions go to show that the reckoning from the starting of the Kali age (which itself belongs to the domain of mythology, the *yuga* division being unknown to very early works) was not a regular, far less a popular, era. The contemporaneity of the last Nanda king with Alexander (336-23 B.C.) and the Puranic traditions regarding the predecessors of the Mauryas suggest that Mahāpadma Nanda is not to be assigned to a period much earlier than 400 B.C. Different versions of the Puranic tradition quoted above thus place the birth of Parikṣit (and therefore the beginning of the Kali age) 1015 or 1050 or 1500 years before that date, i.e., about 1415 B.C. or 1450 B.C. or 1900 B.C. According to the evidence of the Aihole inscription¹ (634 A.D.) and certain other sources, however, the first year of the Kali-yuga corresponds to 3102-01 B.C. Another tradition, supported by authorities including Varāhamihira² and Kalhaṇa,³ says that the battle of Kurukṣetra (and therefore the birth of Parikṣit) occurred 653 years later than the beginning of the Kali age, i.e. about 2449 B.C. Thus, whatever may have been the value attached by Purāṇic chroniclers to the event of Parikṣit’s birth, there was apparently no regular and uniform reckoning from that chronological point. As to the Kali-yuga era of 3102 B.C., scholars have satisfactorily demonstrated that it is not a real historical era, but was invented by Indian astronomers for the purposes of their calculations some 35 centuries after that date.⁴ The use of Al-Bīrūnī’s Śrī-Harṣa era of 457 B.C.⁵ is unknown to Indian epigraphic and literary records; but the fact that its epoch is just 400 years earlier than that of the Vikrama-saṃvat may suggest that this reckoning also was probably invented by astronomers for the purposes of calculations, although Al-Bīrūnī’s statement really appears to be based on a wrong information. The existence of a Maurya era starting from the fourth century B.C. was suggested by some scholars on the basis of a wrong reading and interpretation of a passage in the Hāthigumphā inscription⁶ of Khāravela.

c15.t4.22

² Pargiter, *Puranik Texts of the Dynasties of the Kali Age*, p. 58; Raychaudhuri, *Pol. Hist. Anc. Ind.*, 4th ed., pp. 25-26.

c15.t4.23

¹ *Ep. Ind.*, Vol. VI, p. 7.

c15.t4.24

² *Bṛhatsaṃhitā*, XIII. 3.

c15.t4.25

³ *Rājatarangiṇī*, I. 51-56.

c15.t4.26

⁴ Cf. Fleet in *JRAS*, 1911, p. 479.

c15.t4.27

⁵ Sachau, *Alberuni’s India*, II, p. 7.

c15.t4.28

⁶ *Select Inscriptions*, p. 210 and note 15.

c15.t4.29

Al-Bīrūnī does not mention the Buddhist and Jain Parinirvāṇa reckonings apparently because he never heard of them. But there is evidence to show that the Buddhists had a sort of rough reckoning calculated from the Buddha’s death. It is however to be remembered that, in ancient India, the dates of this reckoning were usually referred to not by individual years, but by centuries. A few instances may be noticed.

c15.t4.30

(1) The *Milindapañha*¹ on the date of Milinda usually identified with the Indo-Greek king Menander—*parinibbāṇato paṃca-vassa-sate atikkamte*, ‘at a time when 500 years elapsed from the death of the Buddha,’ i.e. in the sixth century (between 501 and 600 years) after the Buddha’s Parinirvāṇa.

c15.t4.31

¹ Trenckner, *Milindapañho*, p. 3.

c15.t4.32

(2) The *Laṅkāvatārasūtra*² on the date of Vyāsa, the Bhāratas (viz. the Kauravas and Pāṇḍavas), and the Nandas—*mayi nirvr̥te varṣa-śate*, ‘100 years [after] my (the Buddha’s) death,’ i.e. between 101 and 200 years after the Parinirvāṇa.

c15.t4.33

² *JRAS*, 1905, p. 835.

c15.t4.34

(3) Paramārtha’s *Life of Vasubandhu* (6th century A.D.)³ on the date of the Sāṅkhya philosophers Vṛṣagaṇa (or Vārṣagaṇya) and Vindhyavāsa, a rival of Vasubandhu’s teacher Buddhamitra and a contemporary of king Vikramāditya of Ayodhyā—‘in the 900 years’ (i.e. between the years 901 and 1000) after the Buddha’s Parinirvāṇa.

c15.t4.35

³ Ibid., p. 51 and note.

c15.t4.36

(4) The same work⁴ on the date of Aśvaghoṣa—‘in the 500 years’ (i.e. between the years 501 and 600) after the Parinirvāṇa.

c15.t4.37

⁴ Ibid., p. 52.

c15.t4.38

(5) Tradition recorded by Hiuen-tsang⁵ about the date of Kaniṣka—‘400 years (i.e. in the 5th century between the years 401 and 500) after my (the Buddha’s) decease.’

c15.t4.39

⁵ Watters, *On Yuan Chwang’s Travels in India*, Vol. I, p. 203.

c15.t4.40

(6) Tradition recorded by the same authority⁶ about the date of Vasubandhu’s teacher Manoratha and king Vikramāditya of Śrāvastī—‘within the 1000 years (i.e. in the 11th century between the years 1001 and 1100) after the Buddha’s decease.’

c15.t4.41

⁶ Ibid., p. 211.

c15.t4.42

The indication of the dates of the Buddhist reckoning in ancient India only by centuries has to be coupled with the fact that a large number of different epochs of the Parinirvāṇa is recognised in different parts of the world⁷ and with the absurd chronological position created by such traditons as those assigning Kaniṣka to the 5th and Menander to the 6th century after the Buddha’s death. These facts show that the use of the Parinirvāṇa reckoning was limited within the Buddhist church and that it was never used as a popular era in ancient India. The absence of any use of the reckoning in the records of the early Indian Buddhists also points to the same direction.¹ Some scholars, however, believe that the Minor Rock Edict I of Aśoka refers to year 256 of the Buddhist Parinirvāṇa era,² although they are unable to explain why Aśoka used the era only in one of his many records which are usually dated according to his regnal reckoning. The theory is based entirely on misunderstanding. There can hardly be any doubt about the interpretation of the passage in question, if only the different versions of the record are read together. The explanation of the number 256 is clearly given in the Sahasram version where we have duve-sapaṁnā-lāti-satā vivuthā ti 256³ (=dvi-ṣatpañcāśad-rātri-śate vyuṣitaḥ [aham] = iti—256). It will be seen that 256 was the number of nights which, as the word vivāsa of the Rupnath version⁴ proves beyond doubt, Aśoka passed away from the capital apparently in connection with a dharma-yātrā (pilgrimage). Thus the Buddha-Parinirvāṇa era is entirely unknown to the early period of Indian epigraphy. The Jain ecclesiastical reckoning has even less claim than the Buddhist one to be taken as anything like a regular era in the early history of India.⁵

c15.t4.43

⁷ S. C. Vidyabhushan (*Buddhadeva*, p. xi) notices the Ceylonese tradition of 543 B.C. (or 544 B.C.), a Japanese tradition of 947 B.C., a Tibetan tradition of 433 B.C. and a Chinese tradition of 770 B.C. The Cantonese dotted record suggesting 486 B.C. is now usually relied on by scholars.

c15.t4.44

¹ The Chronicles of Ceylon use the reckoning as an era. In India, however, the Nirvāṇa reckoning is used as an era only in the medieval period; cf. the Gayā inscription of Aśokacalla (Bhandarkar’s List, No. 1459).

c15.t4.45

² Cf. Bhandarkar, Aśoka, 2nd ed., p. 372.

c15.t4.46

³ Select Inscriptions, p. 51 note; p. 52 note; Maski Inscription of Aśoka, pp. 29-30.

c15.t4.47

⁴ Select Inscriptions, p. 51.

c15.t4.48

⁵ There is really no mention of this era in the early Brāhmī inscription from Barli. Cf. Ojha, Bhāratīya Prācīn Lipimālā, p. 2; JBRS, Vol. XXXVII, Parts 1-2, pp. 34-38; Vol. XL, Part 1, pp. 8 ff.

c15.t4.49

The absence of any trace of the use of an era in the dated records of the early indigenous kings of India has to be considered along with the fact that the earliest use of the regular era in India is noticed in the epigraphic and numismatic records of foreigners. The two facts together would doubtless point to the extraneous origin of the custom of dating records according to the years of an era. It seems very probable that the use of the era in royal as well as private records was introduced and popularised in India by kings belonging to non-Indian extraction such as the Scytho-Parthians and the Kuṣāṇas.

c15.t4.50

In connection with the question of regnal reckoning, it has to be pointed out that some royal families (e.g. the Coḷas and Pāṇḍyas of South India, the Pālas and Candras of Bengal, the Brahmapāla kings of Assam and many others) dated their records in regnal years even when the use of an era was quite popular in the neighbouring territories. The influence of local conservatism explains the fact that, in spite of the introduction of the Gupta era in Bengal and Assam, later ruling families of those areas resumed the practice of dating charters in the regnal reckoning.¹

c15.t4.51

¹ A few records of the Pāṇḍyas are dated in the Kali-yuga or Śaka eras. See, e.g., *Ep. Ind.*, Vol. VII, pp. 11-12 ff.; Vol. XXXII, p. 335; Vol. VIII, p. 320; *A.R.Ep.*, 1952-53, No. B 143. The Kandiyur inscription of Kodavarman gives the date as the 1,511, 564th day from the beginning of the Kali-yuga (*Trav. Arch. Ser.*, Vol. I, p. 294). Similarly a few private records of the Pāla age are dated in the Vikrama or Śaka era. See Bhandarkar's List, Nos. 114, 370; *Ep. Ind.*, Vol. XXVIII, p. 145; Vol XXXVI, p. 42.

First 129 of 746 passages. Read on

Linked texts (30)