Chapter XVIII. Calendar, Eras, Various Reckonings about Years, Months etc.

Volume V: Part I pp. 641–685

Chapter XVIII. Calendar, Eras, Various Reckonings about Years, Months etc.

Calendar, eras, various reckonings about years, months &c.

In order to be able to perform vratas and celebrate utsavas at the proper times and in order to find out the appropriate times for religious rites like sacrifices, upanayana and marriage people require a calendar or almanac. A calendar is a methodical combination or book about days, months and years for the purpose of registering in advance religious festivals and astronomical or astrological phenomena and for purposes of civil life. In India about thirty calendars are in use by Christians, Parsis, Moslems and Hinduṣ Confining oneself to Hindus there is a bewildering variety of calendars in use at present. Some are based on the data of the Sūryasiddhānta and some on those of the Āryasiddhānta, while others follow comparatively later manuals like the Grahālaghava; some begin the year with Caitra śukla pratipad (or pratipadā), others with Kārtika śukla pratipad and there are small tracts like the Halar⁹⁸⁵ prānt (in Kathiawad) where the year begins with the 1st day of the bright half of Āṣāḍha. Gujarat and North India except Bengal use the Vikrama era, while the Deccan and South India gene- rally employ the Śaka year, but Kashmir follows the Laukika reckoning. In some parts (North India and Telanganā) and communities the months are pūrnimānta (i. e. end with Full Moon), while other parts (Bengal, Mahārāstra and south India) follow the amānta (month ending with amāvasyā) reckoning. The result is that some fasts and celebrations universally observed throughout India such as the Ekādaśī and Śivarātri fasts and the celebrations of the birth of Śrīkrsna are observed on two different days in different parts of the country and by different sects and there is a difference of one month between the days of some observances i. e. while on the pūrnimānta reckoning a cer- tain festival may be observed in Āśvina dark half, the same month may be called Bhãdrapada dark half ( by amãnta reckoning ) and the same festival may be celebrated one month later. The con- fusion is worse confounded in modern times by some calendars being what are called Drk or Drkpratyaya based on the data of the Nautical Almanac in order that such happenings as eclipses predicted in the pañcãnga may be found to tally exactly with what is actually seen by the people. It appears that in South India there are rather too many pañcãngas ( calendars ). In Tamilnad there are two kinds of pañcãngas, one is based on drk-ganita, the other is based on the Vãkya method (that follows medieval calculations based on Ãryabhata which produce less accurate results) There is the Pudukottai pañcãnga (a Vãkya one) issued under the authority of the rãjas of the principality of that name. The Srirangam pañcãnga (a Vãkya one) is used by the Srivaishnavas, while followers of another Vaishnava sect, the Mãdhvas, have a pañcãnga of their owṇ The Kanjanur smãrtas who do not use the drk-ganita pañcãnga issued under the authority of the Sankarãcãrya. The Telugus follow the Siddhãnta-candrã ² pañcãnga based on the Grahalãghava of Ganesa composed in sãke 1442 ( 1520 A. D. ). In Malabar the people have a drk-pañcãnga, but it is based on a revision of the old Malabar system called Parahita and is not the drk-pañcãnga followed by the Tamilṣ The Telugus follow the Candra reckon- ing and begin their year called Yugãdi from Caitra sukla, while the Tamils follow the Saura reckoning, start their Caitra with Mesa Visu, though, their vratas and religious ceremonies being based on tithis, they have to follow the Candra-mãna. Bengalis use solar months and lunar days with triennial adjustment by the insertion of an intercalary month.

There are three siddhãntas in use, viz. the Süryasiddhãnta ( used throughout India on account of its accuracy ), the Ãrya- siddhãnta ( relied on in Travancore, Malabar, and by the Mãdhvas in Karnãtaka and in Tamil Districts of Madras State ) and the Brãhma-siddhãnta ( followed in Gujarat and Rajputana ). The last, however, is losing ground in favour of the first In the Siddhãntas calculations are made from the epoch of the mahāyuga and are so elaborate that pañcāṅgas can hardly be made from them direct. Therefore, works called karaṇas based on the Siddhāntas are used to prepare pañcāṅgas, such as Makaranda in Bengal and the Grahālaghava of Gaṇeśa; the tables of the latter called Tithicintāmaṇi are used in the Deccan, Central India and some other parts of India. There are two important points of difference between the Siddhāntas among themselves viz the length of the year ( but the difference is only of a few vrpalas ) and the second is the number of revolutions of the Moon and the planets in the Kalpa or Mahāyuga or Yuga. The Siddhāntaśiromaṇi ( I. 6 ) specifies the matters with which a siddhānta must deal, the very first of which is the enumeration of the units of time from truth, up to the end of pralaya.

It should not be supposed that India is singular in this respect. The present European calendar is also very unsatis- factory. Originally it was Julius Caesar who introduced a revised calendar in 46 B C with the provision that every 4th year was to be a leap year. But his calculations were not very accurate and the result was that by 1582 A. D. the spring equinox occurred on 10th March instead of on 21st March. Pope Gregory XIII declared that 4th October 1582 was to be followed by 15th October ( i. e. ten days were dropped ). He further provided that century years were not to have a leap year unless they could be divided by 400 ( so that years 1700, 1800, 1900 of the christian era have no additional day, but 2000 would have one ) Even so, there is still a slight error, but only after more than 33 centuries one day would have to be droppeḍ The Gregorian year is longer by 26 seconds than the correct length of the year arrived at by modern astronomical calculationṣ Protestant England did not follow Pope Gregory's reform till 1750 A. D. when Acts were passed whereby 2nd of September 1752 was to be followed not by 3rd September but by 14th September ( dropping eleven days ) Vide ‘the Calendar New Style Act’ 1750 ( 24 Geo. 2, chapter 23 ) and the Calendar Act 1751 ( 25 Geo. 2, chapter 30 ), which changed the year beginning from 25th March to 1st January and embodied the reforms made by Pope Gregory XIII. The European calendar is still very inconvenient Under it a month may have from 28 to 31 days, the quarters of a year consist of 90 to 92 days, the two halves of the year ( January to June and July to December ) contain respectively 181 ( or 182 ) and 184 days, working days per month vary from 24 to 27 days and years and months begin on different week-days Easter, the Queen of feasts, could fall even after 1751 on 35 different dates from 22nd March to 25th April, as Easter is first Sunday after the Full Moon occurring on or after 21st March.

It has been stated above (p. 480) that this work will not deal with pure astronomical questions; therefore, the author proposes to avoid going into details of astronomical computationṣ But something would have to be said about them as occasion requireṣ Those who are interested in ancient Indian astronomy and its later ramifications may consult and study the following works and articles; Warren's Kālaśaṅkalita; the Sūryasiddhānta translated and annotated by Whitney; Indian Metrology by J. B Jervis, pp 174–259, the Pañca-siddhāntikā of Varāhamihira ( text and translation by Thibaut and M. M. Sudhākar Dwivedi ), S. B. Dīkṣita's 'Bhāratīya Jyotiṣaśāstra ' (in Marathi a masterly work ); Sewell and Dīkṣit's 'Indian Calendar ' (1896); Sewell's 'Indian Chronography ' (1912); Sewell's 'Siddhāntas and Indian Calendar '; 'Vedic Chronology and Vedāṅga-Jyotiṣa ' by B. G. Tilak (1925): Divan Bahadur Swāmikannu Pillai's 'Indian Ephemeris ', 7 volumes ( vol. I part 1 being general and the rest being Tables ); V B. Ketkar's Jyotirgaṇitam, Ketaki, Vaijayanti, Grahagaṇita and 'Indian and foreign chronology ' (Extra Number LXXV–A of the JBBRAS); Jacobi's papers in Epigraphia Indica vol I pp. 403–460 (on computation of Hindu dates) vol. II. pp. 487–498 ( tables for calculating Hindu dates in true local time), vol. XII. pp. 47 ff (for planetary tables), vol. XII pp 158 ff ( special Tables for Hindu dates), Sewell's papers in E. I. vol 14 pp. 1ff, 24 ff, vol 15 pp. 159 ff, vol 16 pp 100– 221 ( on Ārya-siddhānta and Tables), vol. 17 pp 17 ff (on first Ārya-siddhānta), pp 173ff (on Brāhmasiddhānta of Brahmagupta) and pp. 205 ff; Indian Historical Quarterly, vol. IV.pp 483–511, vol X pp. 332–336 (for Jain calendar ); Nautical Almanac for 1935, explanations by J. K. Fotheringham pp. 754ff; 'Ancient Indian Chronology' by Prof. Sen-Gupta (1947, University of Calcutta); Karaṇa-kalpalatā (in Sanskrit) and Bhāratīya-Jyotiḥśāstra-nirīkṣaṇa (in Marathi) of Dr. K. L Daftari, Dr. M. N. Saha's paper on 'Reform of the Indian calendar ' in 'Science and culture ' ( Calcutta), 1952 pp. 57–68, 109–123 and Report of the Calendar Reform Committee (published by the Government of India 1955, a very exhaustive and useful work )

Among all nations the fundamental periods of time are the same, viz. the day, the month and the year (consisting of seasons). The years are the constituents of eras that are vital for chronology and history. Though the fundamental periods of time are the same, there are variations in the arrangement of days to form months and years, about the sub-divisions of the day, about the commencement of the day, about the divisions of the year among seasons and months, about the number of days in each month and in the year and about various kinds of monthṣ The great time measurers are the Sun and the Mooṇ The day is due to the revolution of the earth round its axiṣ The month is mainly a lunar phenomenon and the year is due to the apparent motion of the Sun (but in reality it is due to the revolution of the earth round the sun). The tropical year is the time of the passage of the Sun from one vernal equinox to the next. The latter is shorter than the sidereal year (i. e. the time between the two successive arrivals of the Sun at the same fixed star) by about 20 minutes as the vernal equinox point shifts to the west at the rate of about 50 seconds per year.³ A modern calendar mentions the current year (by some era), the month and the day of the month and generally the week-day, besides several other matters of religious and social interest. The eras and accurate knowledge of the length of the year and the month came to man rather late. The synodic month is a little over 29¾ days, while the tropical year is a little less than 365¼ dayṣ These are incommensurable periodṣ For ordinary life and calendars whole days are required; besides, the starting of the year and the month must be properly defined and must correspond to seasons and some era-must be used These are the requirements of a calendar which has to be used for civil and religious purposes The complexity of calendars is due mainly to the incommensurability of the above two astronomical periodṣ The Moslems solved it by ignoring the length of the tropical year and holding fast by the Moon as the measurer of time. They had a purely lunar year. The result was that the Moslems’ year was of 354 days and that in about 33 years all their festivals travelled through all the months of the year. On the other hand, the ancient Egyptians ignored the moon as the measurer of time and their year was of 365 days (12 months of 30 days each plus five epagomenal, i. e. additional days). Their priests stuck to this system for 3000 years, they had no leap years or intercalary monthṣ Most ancient peoples including Indians followed a luni-solar calendar and tried to adjust the lunar months to the solar year by the system of intercalary monthṣ It has been seen above (p. 489 note 710) that even the Ṛgveda speaks of an additional month (in I. 25 8) but how it was arrived at and where in the scheme of months it was introduced we do not know. We know that the Vedāṅgajyotiṣa⁴ added two months in five yearṣ The people of those ancient times had to solve the problem of adjusting the reckoning of the months by the moon and the years by the suṇ People wanted to know in advance when the Full Moon or New Moon on which many ancient festivals were celebrated may be expected, when to expect monsoon or winter, when to prepare ground for sowing and when to sow. Sacrifices had to be in spring and other New Moon and Full Mooṇ The lunar year of 354 days was less by about 11 days than the solar year. Therefore, the seasons would shift back if the lunar year alone were observeḍ Hence intercalary months were introduced by several nationṣ The Greeks had the ‘Octaeteris’ (eight years cycle ), that contained 99 months of which three were inter- calary, viz. in the 3rd, 5th and 8th yearṣ Then was introduced the Metonic cycle of 19 years in which occurred seven inter- calary months (19 × 12 plus 7 = 235). Olmstead (in American Journal of Semitic Languages, vol. 55 for 1938) says at p. 116 that the cycle of intercalation in Babylon was of eight years and that it was followed by the Greekṣ Fotherigham (in Journal of Hellenistic studies, vol. 39 p. 179) says that intercalation in Babylon was irregular till 528 B. C. and in Greece it was un- systematic in the 5th and 4th centuries B. C. Vide also C. R. C. Report pp. 175-176. The use of eras in India for calendrical purposes is not much older than about two thousand yearṣ Continuous era reckoning first began to be employed in the records of the Indo-Scythian kings that ruled over modern Afganistan and North Western India between about 100 B. C. and 100 A. D. This is not peculiar to India, most of the ancient civilizations such as those of Egypt, Babylon, Greece and Rome did not employ a continuously running era till late in their careerṣ In the Jyotirvidā- bharana ( which is a later fabrication, though it professes that it was finished in gatakali 3068 i. e. 33 years before the christian era ) mentions the names of six persons as the founders of eras in Kaliyuga, viz Yudhisthira, Vikrama, Śālīvāhana, Vijayabhi- nandana, Nāgārjuna, Kalkin⁵ and states that their eras respectively last for 3044, 135, 18000, 10000, 400000, 821 yearṣ In ancient⁶ countries no continuous era was used, but only regnal years were employed In India also Asoka employs only regnal years in his edicts ( e. g the 4th pillar edict at Laurīya- nandana-garh and the 5th pillar edict at Rāmpurvā, C.I.I. vol I pp. 147, 151, were engraved in the 26th year of his kingship ). Kauṭilya⁷ also, when setting forth the business of the collector of revenue specifies the times with which the collector would be concerned as the royal year, the month, the fortnight, the day &c. This practice continued even under the Kusanas and Sātavāhanas; vide Luder's list in E.I. X, No 22, 32 for Kanişka, No.1024 (Kanheri cave inscription of Gotamīputa-sāmī-Sirīyaña Sātakanī), No 1100 and 1122 ( Karle and Nasik cave inscrip- tions of Vāsiṭhīputa Sirī-Puļumāvi ) The Kalinga king Khāra- vela (generally assigned to the 2nd century B. C. ) uses only regnal years and no era ( vide E. I. vol. 20 p 71 at pp 79–80 ).

A variety of eras have been used in India for hundreds of years; this has created a great deal of confusion in chronology and history. For the lists of eras now in use or met with in inscriptions and Sanskrit literature, vide Cunningham on ‘Indian eras, ‘Indian Ephemeris’ by Swamikannu Pillai vol. I part 1 pp 53–55, V. B. Ketkar’s ‘Indian and foreign chronology’ pp. 171–172; P. O. Sen-Gupta in ‘Ancient Indian eras’ (pp. 222–238) and Dr Saha’s paper in ‘Science and Culture’ for 1952 (Calcutta) at p 116 that gives a table of eras with the starting year of the eras, their current years, year beginnings (solar and luni-solar), lunar month end, provenance &c. Vide also p. 258 of the C.R.C. Report (Table 27 ) for a similar long list of eraṣ In this work only a few eras of importance will be briefly dealt with. Alberuni (Sachau vol II. p. 5 ) names five eras, viz of Śrī Harsha, Vikramāditya, Śaka, Valabha and Gupta He gives two different versions about the first and leaves the question undecideḍ

Differing views were held even in ancient times, as to when the Kaliyuga⁹⁹² started and what event the Kaliyuga era commemorateṣ At present the Kaliyuga era is supposed to have started in 3102 B. C and a past Kaliyuga year is to be arrived at by adding 3101 to the A. D. yearṣ There are four principal views about the starting point, viz. (1) the starting point is when Yudhisthira ascended the throne, (2) it is 36 years later when Yudhisthira crowned as king Pariksit, the grandson of Arjuna; ⁽³⁾ according to the Purāṇas ⁹ Kaliyuga started when Kṛṣṇa passed away ( vide Viṣṇupurāṇa IV. 24, 108 to 113); ⁽⁴⁾ According to ¹⁰ to Varāhamihira, 2426 years prior to the starting of Śaka-kāla, Yudhisthira era started i. e. 653 years after the start of Kaliyuga according to the 2nd view. The Aihole inscription appears to follow the 2nd view, since it equates ¹¹ 555 Śaka year with 3735 Kaliyuga year (past). One of the earliest references to the Kaliyuga era is that by Aryabhaṭṭa who states that he was twenty-three years old when 3600 years of Kaliyuga had expired (i. e. he was born in 476 A. D.). ¹⁴ There is an early Cola record dated 4044 of the Kaliyuga era (i. e. 943 A. D.) Vide JRAS for 1911 pp. 689-694 for several instances of the employment of Kaliyuga era in inscriptionṣ Medieval Indian astronomers believed that all planets (including the sun and the moon) were together at sunrise on Sunday of Caitra śukla pratipad at the beginning of Kaliyuga and also of Kalpa. ¹⁴ It is suggested by modern writers like Burgess and Dr. Saha that this epoch was arrived at by astronomical calculations carried back-ward at a late period and is a mere fiction ( vide p. 253 of the C. R. C. Report). It is rather difficult to hold that this was all pure imagination on the part of the ancient Siddhānta writers; it is more probable that they had a persistent tradition of long standing before theṃ¹⁴

In the Sankalpa (declaratory formula) for every religious ceremony the performer has to mention the grand divisions and sub-divisions of time beginning with Śvetavārāhakalpa, such as Vaivasvata manvantara, Kaliyuga and its first quarter together with the geographical position of the place of performance in Bharatavarṣa, and the Zodiacal signs occupied by the Sun, Jupiter and other planets, the era, name of year (Jovian), the month, the fortnight, the tithi, nakṣatra, Yoga and Karaṇa. A verse of Devala states that if a man performing a religious rite does not mention the month, the fortnight, the tithi and the occasion (of the rite) he will not reap the merit of the rite.¹⁵ That is the importance of eras, years and its sub-divisions in the religious life of our people. Therefore a calendar (pañcāṅga) is absolutely necessary for every religious Hindu.

The origin and use of the Vikrama era are as much shrouded in doubts and mystery as is the case with the śaka era.¹⁶ Doubts have been expressed as to whether there was any king called Vikramāditya about 57 B. C. the supposed starting point being¹⁷ Nov. 58 B C in Gujarat where the year begins with Kārttika śukla pratipad and April 58 B C in Northern India where the year begins with Caitra kṛṣṇ pratipad The expired Vikrama year is equal to A. D. year plus 57. In certain rather early inscriptions, years are referred to as Kṛta For example, in Nandśa Yūpa Inscrip-tion the Kṛta year 282 is mentioned and in three Maukhari In-scriptions on Yūpas (sacrificial posts ) the year 295 Kṛta is found¹⁸ The Bijayagad pillar Inscription is dated in Kṛta 428, the Mandasor Inscription in Kṛta 461 and the Gangadhar Inscrip-tion in Kṛta 480¹⁹ Scholars are generally agreed that the Kṛta era is a precursor of the Vikrama era. But the name Kṛta applied to the years has not been satisfactorily explained²⁰ Some inscriptions are also dated in the era of the Malava-gana (Malava tribe), as e g the Mandasor²⁰ Inscription of Naravarmaṇ The Kṛta and Malava era are held to be identical and the inscriptions so far discovered as dated in these two eras are found in eastern Rajasthan and Western Malva. It is further to be noted that, whereas we have inscriptions in 282 and 295 of the Kṛta year, no inscription of the Malava era has yet been found that is dated so early. It is possible that Kṛta is the earlier name and when the Malavas adopted the same era it came to be called ‘Mālava-gaṇāmnāta’ or ‘Mālava-gaṇa-sthiti.’ But it may be remarked that if Kṛta and Mālava both refer to what later became the Vikrama era, they were both being used concurrently for at least one hundred years, as we have the Kṛta era

( Continued from last page ) Brahmāṇḍa III 74. 225 no doubt read एकाराशो. The three Matsya-purāṇa verses he quotes are not from the same chapter. The first verse simply says that Viṣṇu created Yavas on 3rd tithi of Vaiśākha bright half and that he started Kṛtayuga on that tithi. It is impossible to draw any definite date from this verse by itself. The other two verses are Matsya 65. 2-3. They describe a vrata and fast on 3rd tithi of the bright half of Vaiśākha, which yield inexhaustible fruits to the performer, equal to all good deedṣ If on that tithi when joined to Kṛttikā one does special worship, whatever is donat-ed or offered in the fire and whatever mantras are recited become inexhaustible. These two verses contain not a word about Kṛtayuga and the first says nothing about Kṛttikā. It is, to say the least, a strange method to put to-gether verses apart from their context and aver that a particular date has been establisheḍ 480 and also the Mālava era year 461. The author does not think it possible that the word Krta was applied to the era because it was believed at its foundation that the Krta-yuga had starteḍ It is possible that Krta has the same meaning as siddha (just as krtānta means siddhānta ) and it indicates that it was establi- shed by agreement among certain people. The Vikrama era is specifically mentioned by name only from about 8th or 9th century A. D ²². Besides, the Vikrama era has hardly ever been employed in Sanskrit astronomical works and it is generally desiganted saṃvat to distinguish it from the Śāka era. The Yedārāve Inscription of Cālukya Vikramāditya VI shows that that king superseded the Śāka era and started a Cālukya Vikrama era the first year of which was 1076-77 A. D. Vide Bombay Gazatteer vol. 1 part 2 p 447 and 477 (and I. A. VII pp. 61-66 for reference to 400 śakākāla 1478 A. D. ).

Almost all astronomical works in Sanskrit from about 500 A. D employ the Śaka era Why the Śaka era is so called, whether it was to commemorate the coming of the famous Kuṣāṇa king Kaniṣka to the throne or when and where it came to be first used are among the most difficult and unsolved problems of Indian History and Chronology. In one of the earliest inscriptions expressly dated in the Śaka era, viz that of Cālukya king Maṅgalésvara in 500 śaka (expired, i,e, 578 A.D. ) it is cited as the era of the crowning of the Śaka ²³ king. Varahạmihira speaks of it as Śakākāla (in the Pāñca- siddhāntika ²⁴ and Brhat-saṃhitā 13. 3) and as Śakendrakāla or Śakabhūpākāla (in Bṛhat-saṃhitā VIII 20-21) Utpala (about 966 A. D.) while commenting on Bṛhat-saṃhitā (VIII. 20) remarks that the Śaka era started from the time when the Śaka king was killed by Vikramāditya. Its years begin with Caitra for luni-solar reckoning and Meṣa for solar reckoning. Its years are generally expired ones and it begins with vernal equinox of 78 A. D. The earliest inscription so far discovered in which the Śaka era is expressly mentioned is that of Cālukya Vallabheśvara dated in 465 śaka (i. e. 543 A. D.)²⁵ The inscriptions of the Kṣatrapa kings that bear the number of the year but no era are generally ascribed to the Śaka era, e.g. those of Rudradāman in 52 and 72, of Rudrasimha in 103, of Rudrasena in 122 (Luder's list Noṣ 965, 963, 962). Emperor Kaniṣka (of the Kuśāna race) is also connected by some with the foundation of the śaka era.²⁶ One of the earliest date for the śaka era in literature is mentioned as śaka 380 in the Jain work Lokavibhāga of Simhasūri.²⁷ As the Pañcasiddhāntikā (XII. 2) takes the 2nd year of the śaka era as the epoch of the Pitāmahāsiddhānta, it is not too much to assume that the original composed according to him in the first quarter of the first century of the śaka era 1008. In later medieval and modern times ( as also in the Jyotirvidābharaṇa ) the śaka era goes by the name of Śālivāhana. But the form Śālivāhana as that of the era occurs only in the inscriptions of the 13th or 14th century A. D. e g. the two earliest inscriptions so far discovered men-tioning the word Śālivāhana are those recorded in the plate of Devagiri Yādava king Kṛṣṇa dated in Śālivāhana era 1172 (1251 A. D.)²⁹ and the plate of Bukka I dated in śaka 1276, current (1354 A. D.). Vide No 455 in Kielhorn's list in E. I. VII. It is possible that the name Sātavāhana ( mentioned in Harsa-carita as the composer of the Gāthāsaptaśatī ) was changed into Śālavāhana, then into Śālivāhana and this conjecture is support-ed by the fact that in the Gāthāsaptaśatī ( V. 67 ) king Sālahana (a Prakrit form of Sātavāhana ) is mentioned; ' only two know how to raise high distressed families ( by śleśa, the family of Aparnā ), viz. Śiva beloved of the heart of Gauri or the king Sālahana ( Sātavāhana )' Vide C. R. C. Report pp. 244–256 for the śaka era and Ludwig Bachhofer's paper on 'Greeks and Śakas in India' in J. A. O. S. vol 61 ( 1941 ) pp. 223–250, where he attacks some of the downright and very emphatic assertions of Dr. W. W. Tarn in ' the Greeks in Bactria and India '.

Another era, which is current even now in Kashmir, is the Saptarṣi era, also called Laukika era. According to the Rājataraṅgiṇī the Laukika year 24 was equal to śaka 1070 (expired).³⁰ In the use of that era the centuries are generally omitteḍ It is a luni-solar era supposed to have begun in April 3076 B. C on the first day of Caitra-śukla. Vide Cunningham's ‘Indian eras’ pp. 6–17 and Stein's Introduction to translation of the Rājataraṅgiṇī pp. 58 ff The Bṛhatsamhitā³¹ notes a tradition that the Saptarṣis (Ursa Major) occupy each nakṣatra for a hundred years and that they occupied the constellation of Maghā when king Yudhisthira reigned Probably this is the origin of the Kashmir era in cycles of one hundred yearṣ For Inscriptions dated in the Laukika era, vide Bhandarkar's list No. 1431-1443 in E. I. vol. 20 Appendix and No. 1444-45 for Saptarsi era.

Several other eras such as the Vardhamana and Buddha-nīrvāṇa eras, Gupta, Cedi, Harsa, Laksamana-sena (in Bengal), Kollam or Parasurama era (prevalent in Malabar), most of which were once in great vogue (at least in civil life) but are now practically unused, are passed over here, as of only academic interest in these dayṣ For Vardhamana nirvana era, vide I. A. vol 25 p 346 where the year 2493 after the nirvana of Vardha-mana is said to be equal to Vikrama year 1888 and Saka year 1752 (1830 A. D.) and Buddha-nīrvāṇa era in JRAS for 1875 at p 358 (a Ceylonese Inscription of Sahasa-malla s time about 1191 A. D.), where mention is made that the king was crowned 1743 years, three months and 27 days after the nirvana of Buddha. Great controversies have raged over the Gupta era (which Fleet holds began in 319 A. D.) There are some who hold that it is really the Vikrama era (J. I. H. vol. 17 p. 293 ff and J. I. H. special number of March and April 1941 pp 71–84). For Harsa era, vide E. I. 19 p 52 and Noṣ 1385-1421 in Bhandarkar's list in vol. 20 Appendix of E. I. From Alberuni (Sachau vol II. p 5) it appears that Harsa era started in 606 A.D. There has been some conflict of views on this; vide IHQ vol 27 pp. 321 ff, vol. 28 p 280 ff (by R. C. Majumdar) and vol 29 pp 72–79 (Dr D. C. Sarkar). The Kollam era is said to have started in 1176 B. C.

It has already been shown how in Vedic works the year was denoted by several synonyms such as saṁvatsara, samā, varṣa. In the Nārada-saṃhitā³² it is said that there are nine kinds of reckonings of time, viz. Brāhma ( of Brahmā ), Daiva ( of the gods ), Mānusa ( human ), pitrya ( of the Pitrs ), Saura, Sāvana, Cāndra, Nākṣatra and Bārhaspatya, but that in ordinary worldly matters only five of these are employeḍ The Vedāṅga-jyotiṣa appears to refer to four when it says that in a yuga ( of five years ) there are 61 sāvana months, 62 lunar months, 67 nākṣatra monthṣ Hemādri ( on Kāla, p. 9 ) dwells on only three of these year reckonings, viz. cāndra, saura, and sāvana, while Mādhava ( in Kālanirṇaya-kārikās 11–12 ) speaks of two more, viz Nākṣatra and Bārhaspatya. The Viṣṇudharmottara speaks of four ( not mentioning Bārhaspatya ). The three mentioned by Hemādri are those most employed in religious and civil matters, and dharmaśāstra works are specially concerned with them, while the other two are rarely referred to in the latter. A cāndra month ³³ is a time period from the end of one amāvāsyā ( New Moon ) to the next amāvāsyā and twelve such months make the cāndra year of 354 days It is also called a lunatioṇ The length of a cāndra month varies from 29. 246 to 29. 817 days owing to the eccentri- city of the moon's orbit and other causes and its average period is 29.53059 dayṣ A saura month is the time taken by the sun to move through one sign of the zodiac; twelve such months con- stitute the solar year, ³⁴ and the first day of a solar year is the first day of the Saura month, Meṣa. If the entry of the sun in- to the sign is in the day time then that day is the first day of the month. If the entry is in the night, the next day is the first day of the month. The time of entry of the sun into a rāśi differs from almanac to almanac and the result is that the entry may be before sunset in one almanac and after sunset in another Therefore, there may be the difference of one day as to the first day of the month. Owing to the adoption of different ayanām- śas and the difference in the length of the year, the Dṛk, Vākya and Siddhānta almanacs may differ about the first day of the year, about Pongal and other festivalṣ A Sāvana year is of 12 months of 30 days each, the day being counted from one sunrise to the next. A nākṣatra month is the time taken by the moon to pass through the 27 nakṣatras; the Bārhaspatya year is the time taken by Jupiter to pass through one sign of the zodiac (i e about 361 days). According to modern astronomy, Jupiter takes 11. 86 years (for its revolution round the sun). All these four or five divisions are not mentioned in early works and even in later days all these four divisions do not appear to have been fully utilised, though mentioned in astronomical and dharmaśāstra workṣ Kauṭilya provides that³⁵ the month for workmen is of thirty days and nights, the Saura month is one half day larger i. e. (30½ days in a month), the lunar month a half day less (i. e. 29½ days), the nākṣatra month is of 27 days, the intercalary month is of 32 days (or occurs in the 32nd month?), the month is of 35 days (for the wages) of those who tend horses and of 40 days for those who are keepers of elephantṣ The month for workmen was of 30 days, but it need not have coincided with Full Moon or New Moon, while lunar months end with amāvāsyā or pūrṇimā. The Brāhmasphuṭasiddhānta quoted by Utpala on Bṛhat-saṃhitā II 4 p 40 provides that from the saura reckoning one derives the extent of Yuga, the year, Viṣuvat, Ayana, seasons, the increase and decrease in the length of the day and night, from the cāndra the details about tithis, karanas, intercalary month or kṣaya month, all acts to be per- formed at night; from sāvana reckoning sacrifices, the savanas (the three soma libations) the motions of planets, fasts, impurities on birth and death, medical treatment, expiations and other religious rites Vide Viṣṇudharmottara I. 72. 26-27 for similar verseṣ

At present the year begins either in caitra or kārtika in different parts of India. But in ancient times it began in different months in different countries and for different purposeṣ Many Vedic³⁶ passages show that the reck- mānta, that the year began after the Full moon of Phālguna and that Vasanta was the first season of the year. It is due to this that Mādhava says in his Kāla-nirṇaya p 63 (Calcutta eḍ) that Śruti (Veda) is insistent on the pūrṇimānta month (paurnamāsyantatve śruteh katakso bhūyān) The Sm C. (on śrāddha p 377) notes that the amānta reckoning is followed in Deccan the (Dakṣiṇa-patha) and pūrṇimānta in North India (Uttara-patha.) Vide Mundakhede plate of Sendraka Jayasakti on 10th of Phālguna dark half of year 602 (must be śaka) for indi- cating that in north Mahārāṣṭra there was amānta reckoning then (E I. vol 29 part V pp 116-117). The Vedāngajyotiṣa ³⁷ says that the first year of the cycle of five years began in Māgha- śukla 1. e. at the winter solstice Alberuni (Sachau II. pp. 8-9) mentions that the year began in Caitra, Bhādrapada, Kārtika, Mārgaśīrṣa in different parts of India. Kauṭilya ³⁸ provided that the year for work in the Government office of Accounts was a cāndra one and ended on the Full Moon of Āṣāḍha. From Vanaparva ³⁹ it appears that the year began in Caitra It is possible to suggest that the year began with Mārgaśīrṣa because Anuśāsana (106. 17-30) deals with the results of practising ekabhakta vrata from Mārgaśīrṣa to Kārtika. The Kṛtyaratnākara (p. 452) quoting Brahmapurāṇa says that the year began with the pratipad of Mārgaśīrṣa in Kṛtayuga. From Gītā X. 35 ( māsānām mārgaśīrṣoham ) it does not, I humbly think, follow that Mārgaśīrṣa was the first month of the year, since most of the vibhūtis with which Lord Kṛṣṇa identifies himself are the best in that particular category and as even among Pāṇḍavas Arjuna is mentioned, not as first but as the best. Only a few such as akshara or Dvandva are the first of their respective categorieṣ The Varṣakriyā-kaumudi and Nirṇayasindhu p 79 say that the Brahmapurāṇa⁴⁰ generally accepts the pūrṇimānta reckoning and that it (N. S.) would deal with tithi-nirṇaya by generally following Brāhmapurāṇa. The Bhaviṣyapurāṇa when speaking of Holikā festival on Phālguna Full Moon states that it was the end of Phālguna (Uttara-parva 132. 1). Matsya 159 3-6 say that Skanda and Viśākha are said to have been born on the 15th of dark half of caitra in a clump of reeds and it is added that the two were made into one by Indra on the 5th of the bright half of caitra and on 6th crowned as commander. It follows that Matsya thinks the month to have been pūrṇimānta.

Something must be said about the 60 year cycle called Bārhaspatya (Jovian). The Viṣṇu-dharmottara⁴¹ says that the first year called Prabhava of the sixty year cycle began on Māghaśukla when the sun and the moon occupied the nakṣatra Dhaniṣṭhā and were in conjunction with Jupiter. In Bṛhatsaṁhitā 8. 27-52 the phalas of the 60 years from Prabhava to Kṣaya are described, as also in Agnipurāṇa chap. 139, while Jyotiṣitattva pp. 692-697 enumerate all the 60 by name with their phalaṣ Each year of the Bārhaspatya cycle has the word 'Samvatsara' added to it. As yuga means five years, the sixty samvatsaras of Jupiter are twelve times five and are so described in Br. S. 8. 23-26 and in Viṣṇudharmottara I. 92. 9 and later writers suppose that the samvatsaras are affiliated to the lunar year alone¹⁰²¹. In the Deccan now (and in older times also) it appears that at the beginning of each year the Jovian name is changeḍ But in northern India Vijaya is the first year instead of Prabhava. The length of Jupiter's year is 361.9267 days and is shorter than the sidereal year by 4.23 dayṣ The result is that in 85 sidereal years there are 86 Jovian (or samvatsara) years and therefore there is the suppression (kṣaya) of one year after 85 years (which is done only in Northern India). The Jovian cycle must be an ancient one. some Purāṇas and Varāhamihira speak of it as an established institutioṇ Vide Cunningham's 'eras' pp.11-25 (60 year cycle) and pp. 26-30 (12 year cycle), Warren's 'Kālasankalita' (pp. 199-216), V. B. Katkar's work on 'Indian and foreign chronology' pp. 73-82, I. A. vol. 18 pp. 193-209 (Kielhorn) and pp. 221-224 (Fleet) and Gupta Inscriptions, Introduction pp. 169-176 (a note on 12 year cycle and 60 year cycle of Jupiter by Sh. B. Dikshit) for origin and information about the Bārhaspatya cycle and tableṣ In Malabar and Travancore the Jovian cycle is of twelve years, but it is passed over here for reasons of space. The names of the 60 Bārhaspatya years are (with some differences such as that Br. S. 8. 35 puts Nata after Subhānu and then brings in Tāraṇa and omits Pārthiva): 1. Prabhava 2. Vibhava 3. Śukla 4. Pramoda 5. Prajāpati 6. Aṅgiras 7. Śrīmukha 8. Bhāva 9. Yuvan 10. Dhātr 11. Īśvara 12. Bahu-dhānya 13. Pramāthin 14. Vikrama 15. Vṛṣan 16. Citrabhānu 17. Subhānu 18. Tāraṇa 19. Pārthiva 20. Vyaya 21. Sarvajit 22. Sarvadhārin 23. Virodhin 24. Vikṛti 25. Khara 26. Nandana 27. Vijaya 28. Jaya 29. Manmatha 30. Durmukha 31. Hemalam-bin 32. Vilambin 33. Vikārin 34. Śārvarin (or Śar) 35. Plava 36. Subhakṛt 37. Śobhakṛt (Śobhana) 38. Krodhin 39. Viśvāvasu 40. Parābhava 41. Plavaṅga 42. Kilaka 43. Saumya 44. Sādhāraṇa 45. Virodhakṛt 46. Paridhāvin 47. Pramādin 48. Ānanda 49. Rākṣasa 50. Anala 51. Piṅgala 52. Kālayukta 53. Siddhārtha 54. Raudra 55. Durmati 56. Duṅdubhi 57. Rudhirodgā-rin 58. Raktākṣa 59. Krodhana 60. Kṣaya The names of the Bārhaspatya years occur rather early in Inscriptionṣ Among the earliest references are two, one to Siddhārtha-samvatsara in the 5th year of the reign of the Calukya king Mangaleśa (I. A. 19 p. 1 ff at p. 18) and the other is to Saumya-samvatsara in the Alas plate of Rāṣṭrakūṭa king Govinda in Śake 692 (E. I. vol. VI p. 209). How the samvatsara names were supposed to indicate different consequences for the years concerned can be illustrated by citing two verses from the Bṛhatsaṁhitā (8. 28-29) ‘when the samvatsara Prabhava starts there are drought, strong winds and destruction by fire and also other calamities (like excessive rain, locusts &c.), there are diseases due to phlegm, but still people are not distresseḍ’ The next four samvatsaras are auspicious, viz. the king rules over a kingdom which is full of paddy, sugarcane and barley and other crops, which is free from dangers, in which there are no hatreds, in which the people are joyous and free from the evil results of kaliyuga (such as adharma, poverty, accidental deaths &c.).

The subject of months is one of the most complicateḍ Indians followed the luni-solar calendar from very ancient times like the Babylonians, Chaldeans, Jews and the Chinese. All had therefore to resort to the device of intercalary months (adhika-māsa or malamāsa). But Indians had rarely a kṣaya month also which is absent in the systems of the other peopleṣ This difference arose owing to the different ways of the calcula tions of the motions and positions of the sun and moon, in which I do not propose to enter for reasons already stateḍ But a few words to indicate how an additional month arises may be said here. The solar year exceeds the lunar year by a little over eleven tithiṣ This excess amounts to one lunar month in about 32 solar monthṣ The Babylonians had a cycle of 19 solar years in which there were 7 intercalary months (i.e. in all 235 lunar months). The same cycle was called Metonic cycle after Meton, an Athenian, who introduced it in Greece. This cycle is the foundation of the Jewish and Christian calendars, especi ally so far as Easter is concerneḍ From the Vedāṅga-Jyotiṣa ⁴³ it appears that in a cycle of five years there were two intercalary months, one being a 2nd Āṣāḍha at the end of 2½ years and the other a second Pausa at the end of the yuga of five yearṣ Kauṭilya's Arthaśāstra merely echoes thiṣ The Pitāmaha- siddhānta provided that an intercalary month was to be added after thirty monthṣ⁴⁴ In the Virāṭaparva⁴⁵ it is said that the period of 13 years (12 years in exile and one year incognito) was more than completed by the Pāṇḍavas if the cāndra recko- ning were followed, as two months were to be added in five yearṣ The purāṇas and other works give varying periods after which an intercalary month occurṣ A few examples may be citeḍ Viṣṇudharmottara⁴⁶ states that the solar year has eleven days more than the lunar one and that after two years eight months, sixteen days and four ghaṭis a month has to be addeḍ This is not to be taken literally, otherwise it would have to be held that after four ghaṭikās on 2nd tithi of the dark half of a month an intercalary month has to be addeḍ All that the passage of the Purāṇa does is to indicate the number of days in the lunar reckoning after which an additional month is possible. Similarly, a passage of the Kāṭhakagṛhya⁴⁷ says that in the lunar month in which there is no saṁkrānti or there are two saṁkrāntis a month is added after thirty months have ex- pired from a previous intercalary month. This too indicates only a possibility and not a certainty. The Nirṇayāmṛta⁴⁸ declared that a month is added between thirty and thirtysix monthṣ This too is only approximate. A more definite rule is that the lunar month in which no saṁkrānti occurs is called adhika (intercalary) and bears the same name as that of the

māsa occurs there is an adhikamāsa before and after it within three months (Grahagaṇita, Madhyamādhikāra, verse 6, explanation). The curious reader may consult pp. 246-251 of the C. R. O. Report for light on the intercalary monthṣ The Siddhānta-śiromaṇi⁴⁹ mentions some śaka years in which kṣaya-māsa occurred or would occur, viz. in śaka year 974 (expired) occurred a kṣaya month and a kṣaya month will occur in śaka 1115, śaka 1256, śaka 1378 and a kṣaya month generally occurs after 141 years and rarely in 19 yearṣ In No. 1387 in the list of Inscriptions by D. R. Bhandarkar there is a reference to the 2nd tithi of the bright half of the first⁵⁰ Pausa (dated samvat 34, probably of Harṣa era), for a reference to 2nd Āṣādha see Sohawal plate of Śarvanātha in year 191 (probably Gupta) in E. I. vol. 19 at p. 130, for 2nd Pausa Jesar plate of Śīlāditya in 357 Valabhi era in E. I. vol 22 p. 120 may be seeṇ

The rules about what religious acts should be done or not done in an intercalary month will be dealt with a little later. The Śāntiparva refers to the kṣaya of saṁvatsaras, months, pakṣas and divasas (tithis)⁵¹ When there is a kṣayamāsa, the adhikamāsa before it is like all ordinary māṣas i. e there is no prohibition of doing religious acts in it, but the adhikamāsa that follows a kṣayamāsa is the one that is declared unsuitable for religious riteṣ Adhikamāsa and kṣayamāsa may be illustrated by one example for each. Suppose that there was Mesa saṅkrānti on Caitra amāvāsyā, that from the first tithi following the amāvāsyā to the next amāvāsyā (which is Vaiśākha) there is no saṅkrānti and then Vṛṣabha saṅkrānti occurs on the 1st tithi thereafter; the month that had no saṅkrānti would be adhika Vaiśākha and the month on which Vṛṣabha saṅkrānti takes place would be the śuddha ( proper ) Vaiśākha. Now as to kṣaya māsa; suppose that there is Kanyā saṅkrānti on Bhādrapada amāvāsyā, then there is adhika Āśvina followed by Śuddha Āśvina, on the first tithi of which there is Tulā saṅkrānti, then on Kārtika-śukla 1st there is Vṛścika saṅkrānti, then on Mārgaśirṣa-śukla pratipadā there is Dhanus saṅkrānti and in the same month on amāvāsyā there is Makara saṅkrānti. Then that month having two saṅkrāntis ( Dhanus and Makara ) would be called kṣayamāsa and so there would be one month Pausa ( constituted by two months, Mārgaśirṣa and Pausa ). Then on Māgha amāvāsyā there is Kumbha saṅkrānti, then Phālguna would be adhika māsa and there would be Mīna saṅkrānti on śuddha Phālguna śukla pratipadā. Thus in the year in which there is a kṣayamāsa, there are still 13 months and the days of the year amount to some days less than 390.

The four kinds of months, Candra, Saura, Sāvana and Nākṣatra ⁵² have already been described (p. 657) when speaking of the Candra and other yearṣ As observed by the Kṛtyaratnākara (p. 60) the nākṣatra month is not required in Dharmaśāstra, but is known only in Jyotiṣa-śāstra. Pañcāṅgas are generally prepared for each year. They contain separate pages for the two pakṣas (fortnights) of each of the twelve months (or 13 months if there be an intercalary month). The five important parts of the Indian calendar are tithi, week day, nakṣatra, yoga and karaṇa. The Muhūrtadarśana ⁵³ says that it may have six parts if the signs of the zodiac are given and seven if the positions of the planets are mentioneḍ Vara (weekday) is the period of time between one sunrise and the next and not much will have to be said about it except about the origin of week dayṣ A great deal has already been said about tithis and nakṣatraṣ What remains to be said about these, if at all, and the treatment of Yoga and Karaṇa will be dealt with later oṇ We must revert to the topic of nakṣatraṣ

The names of the twelve solar months distributed two each among six seasons are very ancient. They occur in the Taittirīya-saṃhitā IV 3 11. 1, Vāj. S. 13 25. They are Madhu, Śuci, Nabhas, Nabhasya, Īṣa, Ūrja, Sahas, and Tapasya⁵⁵. In the Brāhmaṇas we meet with months (lunar) named form nakṣatraṣ It is therefore that some works speak of ṛtus (seasons) as of two kinds, viz. Saura and Cāndra. The former begin either with Mīnarāśi or Mesa-rāśi, while the latter are called Caitra and the rest ⁵⁶ The name of the nakṣatra is often formed into a derivative and joined to the word Paurṇamāsī, Pūrṇamāsa or Amāvāsyā and sometimes the names of the months such as Phālguna and Caitra occur in the Brāhmaṇas (vide note 1015 above ⁵⁷)

Pāṇini derives the names of the months like Caitra from Paurṇamāsī and expressly names the Āgrahāyaṇī (IV. 2. 23), Phālguni, Śravaṇā, Kārtikī and Caitrī (IV 2 23). 'Paurṇamāsī' itself is derived from Pūrṇamāsa by the Vārtika on Pāṇini IV. 2. 35. The Full Moon tithi which has Puṣya nakṣatra is called Pauṣī by the sūtras (Pāṇ. IV. 2. 3 and IV, 2 21). Thus there were three stages; first the 27 nakṣatras were distinguished and named in the very ancient Vedic saṁhitās, then the Full Moon day was called Caitrī Paurṇamāsī (and so on) because on that tithi the Moon was in Citrā nakṣatra and so on; then The months were called Caitra, Vaiśākha &c, because they had Caitri or Vaiśākhi Paurṇamāsī in it. All this occurred long before Pāṇini. Later the saura names Madhu, Mādhava &c. came to be identified with lunar months Caitra, Vaiśākha &c. and became synonymṣ When this occurred it is difficult to say. But it is much older than the Christian era. It must be long before Kālidāsa, who in his Meghadūta first speaks of Āsādha and then refers to the next month Śrāvana as Nabhaṣ Often on the Full Moon day the Moon might not actually be in or very near Citrā or Śravaṇa nakṣatra but that did not matter; the month would still be called Caitra or Śrāvaṇa as the case may be.⁵⁷

It has already been shown that in ancient Brāhmaṇa times the months ended with the Full Moon (i. e. were pūrṇimānta). When foreigners like Kaniṣka and Huviska ruled over Northern India they struck to the Pūrṇimānta Indian months in their records, though here and there they employ Macedonian month nameṣ For example, the Greek month Gurppiya is used in Huviska's reign (28th year) in the Mathurā Brāhmī Inscription and the Greek month⁵⁹ Artemīsios also occurṣ In the Zeda Inscription the month must be held to be pūrṇimānta as the nakṣatra Uttarāphalgunī is there said to have occurred on the 20th day of Āsādha.⁶⁰ The Tekkāli plates of Devendravarman record a grant on the occasion of a lunar eclipse on the 30th day of Magha.⁶¹ Similarly, in the Lodhia plates of Sivagupta Bālārjuna of Kośala (i. e. Chhattisgarh region) published in E.I. vol. 27 p. 319 the date ‘Sambatsare sapta pañcāśattame Kārtika- porṇamāsyāṁ ankena samvat 57, Kārtika-dina 30’ (p. 325) shows that the month was pūrṇimānta. The Rājatarangiṇī ⁶² says that king Harirāja came to the throne at the beginning of Āṣāḍha in the 4th year (of the Laukika era), who ruled only for twenty-two days and died on 8th of the bright half of Śuci (i e. Āṣāḍha) and thereby shows that the month was pūrṇimānta in Kaśhmir then as now.

Though inscriptions before the Christian era rarely mentioned the names of months (such as Kārtika 14th in the Kharosthī Inscription of the Greek king Menander, ⁶³ 2nd century B. C.), the usual mode is to mention the rtu (season), then the number of the month (and not name) in the season and then divasa, e. g. vide the seven inscriptions from ⁶⁴ Mathurā and the Śārnath Inscription of Mahārāja Kaniṣka. ⁶⁵ In some inscriptions only the rtu (season), the number of the pakṣa and the day occur; vide Nāgarjunikoṇḍa ⁶⁶ Inṣ (3rd century A. D.), the Omgudu grant of Vijayaskandavarman (E. I. 15 p. 252 ‘Hemantapakse trtiye trayodaśyāṁ’), the Nāga image Inṣ of Kaniṣka’s time (E. I. 17. p. 11) in sa. (Samvat) 8, grī. 4, di (divasa) 5. Vide also Lūder's list No. 922 and No. 1202 (the latter of Madhariputa Purisadata, samvacchare 20, Vāsāpakham 8, divasam 10). Sometimes the month is mentioned but no pakṣa and the days are counted continuously from 1 to 30, e. g. Khoh plates of Mahārāja Jayanātha (in year 177 and 22nd day of Caitra) and Khoh plate of Śarvanātha in year 193 month Caitra day 10th without the mention of the fortnight (vide Gupta Inscriptions pp 121 and 125), E. I. vol. 20 p. 59 at p. 63 (sam 159, apparently of the Gupta era, Magha di. 7), E. I. 19 p. 96 (the Mathurā pedestal Inṣ of Kuṣhāṇa year 14 in Pauṣamāsa divasa 10), Gupta Inṣ of Gupta year 129 in Jyestha-māsa di 18 (Gupta Inscriptions p 45 at p. 46). The non-mention of the paksa and the continuous counting of days from 1 to 30 continued even in the 9th century, vide Kota Inscription of Samanta Doradatta (I. A. 14. p 351 ) of samvat 879 (923 A. D) Māgha pūrṇimānta su. 20. It may be noted that here the month is pūrṇimānta. In modern times people use such abbreviations as su di or va di. The first is abridged from śukla dina (or divasa) or tadya. The first is abridged from śukla dina (or divasa) or śuddha dina and the latter from bahula-dina or divasa (va and ba being often interchangeable). For ‘bahula’, vide E. I. vol. 18. p. 236 at p. 238 (inscription of Rudrasena in the year 127 held to be of the sake era), Rudradāman’s Junagadh inscription in the year 72 (150 A. D.) of Māgasirsa, bahula-pratipad (Luder’s list No. 965 and E I. vol. VIII p 36 at p. 40), Udayagiri cave inscription in Gupta era 106 (425-26 A. D ) ‘bahuladine’ (Gupta inscriptions p 258). For śukla, vide Gupta inscriptions pp. 95, 102, 114 and for suddha, vide the Nasik cave inscription in the year 42 of the reign of Usavadāta, son-in-law of Naha-pāna, where we have ‘Kātīkasudhe’ (in E I VIII at p 82) and Pardi plates of Dahrāsena in Kalacuri year 207 (i. e. 455-56 A. D.) in C. I. I. vol IV pp 22, 24 ( Vaiśākha-śuddha-trayo-daśyām), for su di, vide E. I 17 p 110 (of the Valabhi year 106, 525-26 A. D.). For ‘ba di’ vide JBBRAS (New Series) vol. I. p. 66 ‘Bhādrapada ba 13’ in Valabhi year 210 and E. I. vol. I. p. 126 plate of Dhruvasena in Valabhi year 210 ‘Bhādra-pada ba di 9’). It may be noted that ‘sudi’ and ‘tadi’ are treated as indeclinables (avyaya) and included in the ‘svara-digana’ (‘svarādimpītamavyayam’ Pān I. 1 37) by later grammatical writerṣ The Kāsika, however, does not include these two in the ‘svarādi-gana’ which is an ākṛtīgana and therefore liable to be added to. ‘Vadyā’ is inexplicable except as an attempt to have some word with some assonance of sound parallel to śuddha and related to va di. It is not clear why the word ‘paksa’ is not used in many inscriptions before and after the Christian era even though the word paksa occurs in such ancient works as the brāhmaṇas ⁶⁷ and Upanisadṣ The Chāndogya-upaniṣad (IV. 15 5 and V. 10. 1), the Brhad-āraṇyaka-upaniṣad VI. 2 15-16 employ the words ‘āpūrvya-māna-paksa’ (also pūrvapaksa' and 'aparapaksa' in Br Up.

appears to mean 'soiled or dirty clothes'. The word 'malimluca' occurs in Vāj. S. 22. 30, Śāṅkhīyāna-śrautasūtra VI. 12.15. The Malamāsatatva derives the word as follows: being male (soiled) 'it passes on'. The words 'samsarpa' and 'abhasaspati' occur in Vāj. S. 22. 30 and 31 ⁶⁸ respectively, and 'ambasa- spati' in Vāj. S. 7. 31 after the mantras referring to the twelve saura months, Madhu, Mādhava &c. The Tai. S. (I. 4. 14. 1 and I. 5. 3. 4) has the formula 'samsarpsyaham-spataye tvā' ⁶⁹. 'Amhaspati' literally means 'the lord of sins or evil' (i. e. a condemned month). Later medieval writers make a distinction⁷⁰ between 'samsarpa' and amhaspati'. When there are in one year two adhimāsas and a kṣaya māsa, the first of the two adhimāsas is called 'samsarpa' and it is not condemned for all religious matters but only for marriages, such sacrifices as Agniṣṭoma, festivals (that are not obligatory) and auspicious rites of children (such as nāmakaraṇa'). The name amhaspati is confined to kṣayamāsa. The intercalary month is called Purusottama-māsa (Viṣṇu being called Puruṣottama) in some of the Purāṇas as in Padma VI 64, probably with the purpose of reducing the opprobrium that attached to an intercalary month.

In the dharmaśāstra Literature considerable space is devoted to the intercalary month e. g. in the Agnipurāṇa 175 29-30, the Rājamārtaṇḍa (folios 68a-69, verses 1078-1115), Kalaviveka of Vimarṣāvasena pp. 113-168, Hemādri on Kāla pp 26-66, Kālanirṇaya of Mādhava, Madanapārijāta pp. 544-550, Suddhikamudi pp. 265-296, Malamāsatatva (Jīvananda eḍ) pp. 736-856, Samayaprakāśa pp 140-155 (part of Viramitrodaya), Nirṇayasindhu pp. 9-17, Smṛtikaustubha (samvatsarādidīhti) pp. 520-529, Dharmasindhu pp. 4-7, Puruṣārthacintāmaṇi pp. 17-31. A brief treatment is all that can be attempted here. The Agnipurāṇa provides:⁷¹ one should avoid in an intercalary month the performance of the following acts, viz. of consecration of Vedic fires, the establishment of images, sacrifices, religious gifts, vratas, the recitation of the Veda after undertaking it, observance of certain niyamas such as mahānāmnī, the letting loose of a bull (after sapinḍana), tonsure of a child, upanayana, rites such as the saṁskāras of nāmakaraṇa, inauguration of a king. Long lists of what may or may not be done in an intercalary month occur in many works such as in Hemādri on Kāla pp. 36-66, Nirṇayasindhu pp 10–15, Dharma- sindhu pp. 5–7 It would be enough if some general provisions are stated The general⁷² rule is that one should perform obligatory acts and acts prescribed on certain occasions even in an intercalary month. For example, one must perform even in an intercalary month daily sandhyā worship, the five daily Mahāyajñas (such as Brahmayajña, Vaiśvadeva), the offering into fire (as an agnihotrin) and must take a bath if there is an eclipse (in adhimāsa (this also is naimittika); so also one must perform the funerary rites after a relative's death even if the death takes place in an adhimāsa (this also is naimittika). But an obligatory rite should not be performed in adhimāsa if śāstra allows its postponement, e g. Soma sacrifice. The general rule is that no kāmya rite (not obligatory, but to be performed for securing some object or result) is to be performed in adhimāsa. There are exceptions to this also if a certain religious rite is begun before an adhimāsa falls (such as a Prājāpatya penance for twelve days or a cāndrāyaṇa penance for a month) it may be continued even in adhimāsa. If there is drought and one desires to perform the Kārīrīṣṭi for bringing down rain, this may be done even in adhimāsa, since to wait might defeat the desired object These provisions are neatly summarised in the Kālanirṇaya-kārikās quoted below ⁷³

Certain matters were to be done only in malamāsa such as the gift ⁷⁴ of 33 apūpas (cakes) every day to brāhmaṇas or at least once in the intercalary month; certain acts were not to be done in adhimāsa, but only in the śuddha ⁷⁴ month, such as the sinking of wells, construction of tanks, solemn sacrifices, the mahādānas and vratas Certain matters can be done in adhimāsa as well as in śuddha-māsa ⁷⁵ e. g. rites (such as puṁsavana), taking interest on money lent, wages of workmen for fixed periods, monthly śrāddha (on amāvāsyā), daily gifts, bath at a holy place, japa (muttering mantras), homa with barley grains or rice or sesame, rites on the birth of a child, funeral rites, nava-śrāddhas, śrāddha on 13th tithi with Magha-nakṣatra, the sixteen śrāddhas, bath and śrāddha and gifts and japa on lunar or solar eclipse, obligatory and naimittika acts; these may be done in malamāsa also A great deal of discussion is held on śrāddhas in malamāsa by Hemādri on Kāla (pp 56–63), Malamāsatattva (pp. 841–856), Nīrṇayasindhu (pp. 11–15). Samayaprakāśa pp. 151–155 This is passed over for reasons of space For some remarks on the śrāddha of a person dying in an intercalary month, vide H of Dh. vol. IV. pp. 546–547

Just as the 13th month was to be discarded for religious rites, in Western countries number 13 is supposed to be unlucky, particularly in the case of company at a table

One of the five angas in the Indian pañcāṅgas is the week-day. A brief discussion on days and week-days has therefore to be helḍ The interval between two sunrises was considered as the most important unit of time This is solar and is called the civil day But tithi is a lunar division of time (as described above) and begins and ends at various spaces in the solar day. The word ‘day’ has two meanings, viz (1) from sunrise to sunset and (2) from sunrise to sunrise ⁷⁶ In Rg. (i. e ‘rātri’ night) and to the bright part.⁷⁷ The word ‘ahan’ the word ‘dina’ is very frequently used in compounds like ‘sudinatva’, ‘sudina’, ‘madhyandina’; ‘ahorātra’ (day and night) occurs only once in the Rg. X. 190.2 The word ‘pūrvāhna’ (first part of the day or forenoon) occurs in the Ṛgveda X. 34 11. Three parts of the day also occur in the Ṛgveda viz prātah, sangava and madhyandina ⁷⁸ (mid-day). These are three out of the five parts of the day, the other two being aparāhṇa and astamaya, astagamana or sāyāhna, which five are mentioned in the Śatapatha Br. II. 3 2 9 Both ‘prātah’ (morning) and ‘sāyam’ (evening) occur in Rg. V. 77.2 and ‘sāyam’ alone in Rg VIII. 2 20, X. 146. 3 and 40. The day and night were also divided into eight parts by Kauṭilya (I. 19) and by Daksa and Kātyāyana The division of the day and night each into 15 muhūrtas has been already dealt with at some length on pp. 537–541.

Great divergence prevailed about the beginning of the day. The Jews placed the beginning of the civil day in the evening (vide Genesis I, 5 ‘The evening and the morning were the first day’ and also I 13.) The Egyptians divided the day from sunrise to sunset into twelve parts and thus obtained hours whose length depended on the season, while the Babylonians ⁷⁹ began the day with sunrise and divided day and night into twelve parts, each of which corresponded to two of the equinoctial hourṣ In Athens and in Greece generally the day commenced in historic times with the evening for the purposes of the calendar (vide Heath's 'Aristarchus of Samos', 1913, p. 234). In Rome the day began with midnight. With Indian writers the day begins with sun- rise (as in Brāhmasphuṭa-siddhānta XI. 33), but they were not unaware of different beginnings of the day. The Panca- siddhāntikā ⁸⁰ remarks that Āryabhaṭa declared that the day began at midnight in Lankā and again he said that it began with sunrise and that sunrise in Lankā coincides with sunset in Siddhāpura, with mid-day in Yamakoṭi and with midnight in the Romaka country.

In modern times the civil day begins at midnight.

The seven day week is an artificial man-made arrangement. It corresponds to no astronomical or celestial phenomenoṇ The Mexicans had a week of five days ⁸¹ before the Spanish conquest. The seven days week prevailed among Jews, Babylonians and the Incas of South America. The Romans had a week of eight days in Republican times, while the Egyptians and ancient Athenians had a week of ten dayṣ In the Old testament God is said to have gone on with the work of creation for six days, to have rested on the seventh and to have blessed it and sanctifieḍ it (Genesis 2. 1-3). In Exodus 20. 8-11, 23. 12-14 and Deuter-onomy 5. 12-15 God is said to have commanded the Jews to labour for six days, take rest on the seventh and to observe it as the Sabbath of the Lord God and to keep it sacreḍ The Jews gave no special names to the week days except the Sabbath which was the last day of the week and which they observed on Saturday (and not on Sunday).

The Old Testament contains no traces of proper names having been given to week dayṣ It appears that even in the New Testament week days were indicated by numberṣ Vide Matthew 28 1 'In the end of the sabbath as it began to dawn towards the first day of the week came marry Magdalen &c '; Mark 16 9 'Now when Jesus was risen early the first day of the week, he appeared first to Mary Magdalen &c ', Luke 24 1 ' Now upon the first day of the week being early in the morning they came into the sepulchre' Each of the seven days of the week is designated as the sabbath or holy day by various nationalities and religious communities i e Monday is the Greek Sabbath, Tuesday the Persian, Wednesday the Assyrian, Thursday the Egyptian, Friday the Moslem, Saturday the Jewish and Sunday the Christiaṇ

On the origin and development of the seven-day cycle the leading work is F. H. Colson's 'the week' (Cambridge University Press, 1926 ). Some of his arguments may be briefly indicated Dion Cassius (first quarter of 3rd century A.D.) in his 37th book says that Jerusalem was captured⁸² by Pompey in 63 B C. owing to the reverence of the Jews for Saturday (their Sabbath day) and he further makes the statement that planetary week originated in Egypt, that it was of recent growth and that it was in general use in his day; vide Dio's 'Roman History' vol. III pp 129, 131 where two traditional explanations are given as to how the week days were named after planets, the 2nd of which is based on the 24 hours of the day and night What must be emphasized is that Dio is positive that week days did not originate in Greece, but in Egypt and their introduction was recent Dio was a Roman and wrote his History between 200 to 222 A. D Therefore, the introduction of week days in Greece could not have been earlier than the first century of the Gupta year⁸³ 165 (i. e. A. D. 484), which mentions Thursday and the 12th tithi of the bright half of Āṣāḍha. Supposing for argument that weekdays were originally borrowed from some foreign country, at least a few hundred years would be necessary before such a borrowed system could become general and could be utilized even in Indian proclamations by a great Indian imperial dynasty like the Guptaṣ

Let us now turn to the literary evidence. The Aryabhaṭīya (Daśagītikā, verso 3) mentions Thursday.⁸⁴ In Bṛhat-saṃhitā⁸⁵ occurs this 'when Pitamaha declared that the week-day Tuesday does not lead to auspicious results or when (I declare) that the day of the son of the earth is inauspicious, what difference is there between (the śāstra) composed by a divine author and that by a human author (like myself )?' It has been seen above (p 458 ) that the Pañcasiddhāntikā tells us that a five year cycle started according to Pitamaha in the second year (expired) of the Śaka era. Therefore it appears that Pitamaha who probably wrote about 80 A D mentioned Tuesday. The Pañcasiddhāntikā (in I. 8) mentions Monday in summarising the features of the Romaka siddhānta (vide note 1004). The Bṛhatsaṁhitī in chapter 103 (verses 61-63) sets out what actions are proper to be done on the weekdays from Sunday to Saturday and Utpala quotes eighteen Anuṣṭubh verses in all from the ancient astronomer Garga on the same topic.⁸⁶ Kern assigns Garga to the 1st Century B. C. This would carry the knowledge of weekdays in India to the first century before Christ. Philostratus in his life of Apollonius of Tyana (who died in 98 A D ) relates ( vide note 900 above) how when Apollonius travelled in India he received from Iarchus, the chief of the brāhmaṇas, seven rings bearing names of seven planets, each one of which Apollonius was to wear on each of the weekdayṣ As Philostratus wrote in the first quarter of the 3rd century A. D., this would carry the knowledge of the planetary week in India at least to the 1st century of the Christian era. The above references furnish a terminus ad quem for India's knowledge of the planetary week days (viz. first century B. C. to first century A D.). The terminus a quo cannot be stated with certainty.

Now some other literary evidence about the dates of which there are likely to be doubts and disputes would be set out. In the Vaikhānasa-smārta-sūtra (I 4) and Baudhayānadharmasūtra (II. 5. 23) tarpaṇa (satiation with water) is offered to the Sun, the Moon, Mars, Mercury, Jupiter, Venus, Saturn and to Rāhu and Ketu and the former mentions Budhavāra (Wednesday) in II 12. It has to be conceded that other sūtras such as Aśv gr III, 4. 1-5, the Kauśitaki gr (IV. 9-10), the Hir. gr. II. 19-20, Bhāradvāja gr. III. 9-11 contain no tarpaṇa to the weekdays or their deitieṣ The Atharvaṇa-Vedāṅga Jyotiṣa has a section on weekdays (of eight verses) and points out what should be done on each of the seven weekdays from Sunday to Saturday (two of which are quoted below by way of sample⁸⁷). The Gāthāsaptaśatī of Hāla (a prākṛt anthology) mentions Tuesday and Visti (III. 61). The Yājñavalkya-smṛti (I. 296) enumerates the planets in the modern weekday order from Sunday to Saturday and adds Rāhu and Ketu as grahas and provides for a sacrifice to nine grahas (I 295-308). The Nāradapurāṇa (I. 51. 80) also does the same. The Matsya (chap. 93. 7 ff) provides for a more elaborate sacrifice to planetṣ The Viṣṇudharmasūtra (78 1-7) refers to śrāddhas performed on the seven weekdays from Sunday to Saturday and the benefits derived from such performance (vide H of Dh vol IV.p 373-374 for details). The Purāṇas contain numerous references to the weekdayṣ The Nāradīya provides that Monday⁸⁸, Wednes- day, Thursday and Friday are auspicious for all rites and actions, while the other weekdays yield the desired result only in cruel or violent actions The Matsya provides that all week- days except Sunday and Tuesday are auspicious (253 7 ) The Kūrmapurāṇa ( Uttarārdha 20. 16-17) provides that śrāddhas on weekdays from Sunday to Saturday respectively yield the benefits of health, happiness ( or prosperity ), success, all desires, desired learning, wealth and long life The Viṣṇudharmottara (I. 142. 30-32 ) has similar provisionṣ The Brahmapurāṇa (118. 30) recommends touching an aśvattha tree on Saturday. The Vāmana (14. 49-50 ) forbids certain actions on weekdays (such as an oil bath on Sunday or Tuesday ) The Matsya specifically refers to Sunday ( in chap 55 4, 65. 5, 70. 46 ), to Monday ( in 57. 4), to Tuesday ( in 72. 27, 193 8 ) Though the chronology of all the above works is somewhat uncertain, yet it cannot be gainsaid that at least in the first two centuries of the Christian era planets had become objects of worship and the weekdays had been named and their order as it exists today had been settleḍ The Mahābhārata, in spite of its vast extent and copious des- cription of Dharmaśāstra matters, is entirely silent about the names of weekdayṣ The cumulative effect of all the evidence adduced above is to render it extremely probable, if not certain, that the worship of the planets and knowledge of the weekdays presided over by them was known to the Indians at least bet- ween 100 B C. to 100 A D One further circumstance is that the names of weekdays in India are purely Sanskrit and there is hardly any reference to a weekday in Sanskrit works under any Greek or other non-Indian name, while in the case of the signs of the zodiac frequent use is made of such foreign words as kriya or leya Fotheringham ( in explanations in Nautical Almanac for 1935 at p 769 ) states that the mean temporal hour with the length that it had at equinoxes is first found in Hipparchus ( 140 B. C ), that Ptolemy in the Tables in the Almagest reckon- ed the day from mean noon of Alexandria and divided the day into sixty equal parts each of which was divided sexagesimally and that Ptolemy in his manual Tables divides the whole day into equinoctial hours reckoned from noon in Alexandria and subdivided these sexagesimally. This shows that two systems of dividing the day into 60 parts and also into 24 parts were accepted by Ptolemy himself (who wrote only about 150 A D ). Therefore, Indians could very well have used very early the 60 divisions system for finding out the order of the weekdays instead of 24 divisions systeṃ The Indians did not calculate their day from noon or night but from the morning. The Aśvamedhika- parva⁸⁹ expressly says that the day comes first and then night.

It is possible to advance several theories about the origin of the seven day week in India. The five planets (Mercury, Venus, Mars, Jupiter and Saturn) were especially regarded by the ancient Babylonians as the powers controlling the destinies of men and therefore five leading Babylonian divinities were identified with the above five planets In course of transmission Babylonian names of divinities assumed Roman forms i e Ishtar, goddess of love, became Venus, the great god Marduk became Jupiter and so on (vide note 748 above) The five planets together with the Sun and the Moon made a group of seven heavenly bodies In Chaldean temple worship which spread to Syria, it became the custom to sing the praises of each god on a particular day. The name of the god worshipped on a particular day was transferred to the day itself. The days which were sacred to the Sun and the Moon were called Sunday and Monday. Some Norse words like Wednesday (Woden's day) and Thursday (Thor's day) were introduced in England But weekdays in Europe are ultimately derived from the Babylonian godṣ It has already been shown (pp 598–600 above) that India and Babylonia had close commercial and other contacts from very ancient timeṣ Vide the recent work on 'Hittites' by O. R. Gurney (in Pelican Books, 1958), pp 104–105 and 124 for technical Sanskrit words in horse breeding Vedic gods and royal names in Sanskrit in the 14th century B C It is possible to argue that seeing the temples of planetary gods in which worship was performed on certain days some Indians got the suggestion of a planetary week, as it has been suggested above that the ideas of the Zodiacal signs were derived from Babylonian boundary stones and similar monuments There were Sun temples in ancient India⁹⁰ such as that of Mārtanda in Kashmir, that of Modhera in North Gujarat and that of Koṇārka in Orissa. Even now in some towns and villages there are temples of Saturn and Rāhu ( as at Rāhuri in the Ahmednagar District of the Bombay State, which bears that name after Rāhu ). Kauṭilya mentions the several divisions of time from ‘truth’ to yuga and states that two nāḍikās are equal to a muhūrta and an ahorātra ( day and night ) has 30 muhūrtas ( vide note 681 above ). From this it follows that Kauṭilya knew only the division of the day into 60 nāḍis and that a nāḍī was the same as ghaṭī. The Āryabhaṭṭiya gives the following table: 60 vināḍīs are equal to a nāḍī, 60 nāḍīs are equal to a day (Kalakriyāpāda, verse 1) Another way of citing the measures of time is: Time taken for repeating ten long letters is prāṇa, 6 prāṇas are equal to pala, 60 palas are equal to a daṇḍa, ghaṭī or nāḍī (Sūryasiddhānta I. 11, Jyotiṣat-attva p. 562). Pāṇini derives the word 'nāḍindhama'⁹¹ from nāḍī Nāḍī is a very ancient word ⁹² It occurs in the Ṛgveda and appears to mean some cylindrical instrument like a flute ⁹³ It appears that the word nāḍī came to mean a period of time when a conch or a trumpet or similar instrument was blown to announce the lapse of one nāḍī ( 60th part of the day ) as there were no clocks theṇ Hence division into 60 nāḍīs or ghaṭīs (both words being men- tioned by Patañjali who is generally held to have flourished about 150 B. C.) was the ancient mode of dividing the day in India and if any Sanskrit work (like the Sūryasiddhānta, chap 12. 78-79) refers to 24 parts of the day, that work may be regarded as one of later times, but the whole system of time measuring by weekdays in India cannot be said to have been borrowed from a foreign source. Even Patañjali refers to the words nāḍī and ghaṭī as long in use before his time Therefore, the division of the day into nāḍis or ghaṭis must be held to be long anterior to 200 B. C. On the whole it would not be far from the truth to say that weekdays were not borrowed by ancient India en bloc from elsewhere It is arguable that usages in Babylon and Syria might have suggested the arrangement of days into weeks, but there is hardly any cogent evidence beyond similarity and prejudice to show that the origin of the nomenclature and the method of arriving at the order of the weekdays were not indi- genous Vide Cunningham in I A. vol 14 pp. 1 ff. in which he shows by diagrams how the calculation by Indian ghaṭis results in the same order of weekdays as the European divisioṇ Sham Shastri in Annals of Bhandarkar Oriental Research Institute, vol IV pp 1–31 after mentioning some rather hazy and obscure passages from the Vedic literature arrives at the conclusion that India was not indebted to the Greeks or Babylonians for week days or for the discovery of planetṣ It may be mentioned that Alberuni (tr by Sachau, vol I. chap XIX pp 214–215) illustrates how week-days received names from planets, narrates (p. 215) that it was the custom with Hindus to enumerate the planets in the order of the weekdays and that the Hindus persisted in using these in their astronomical books and declined to use any other order or method even if it was more correct.