Megasthenes and Indian Chronology (Part 2)

Megasthenes and problem of Indian Chronology as based on the Puranas - II [megasthanijah bharatiyah pauranika vamsakalakramasca - 2 ] / By Sri K. D. Sethna; Pondicheri / 276-294

[ nibandho'yam purvato'nuvrttah | asya prathamo bhagah 'purana ' patrikayah purvasmin ( janavari, 1966 ) anke prakasitah | atrapi yavanarajaduta megasthanijavacanani puranavacanani canusrtya yavanaraja- dutoktasya dayonisas ( Dionysus ) namno rajnah adiraja- prthuna saindrokottas (Sar drocottus ) namno rajnasca guptavamsoyena candragupta prathamena abhinnatam sthapanaya anyanyapi pramanani pradattani | guptavamsiya candragupta prathamasya ca rajyarohanakalah 325-324 i0 pu0 nirdharitah | imam kalamasrityaiva ca bharatiyetihasasya punarnirmana- mavasyakamiti lekhaka mahodayena vijnapitam | ] II "Dionysus was the first who invaded India, and was the first of all who triumphed over the vanquished Indians. From the days of Dionysus to Alexander the Great, 6451 years years are reckoned with 3 months additional. From the time of Dionysus to Sandracottus the Indians reckoned 6452 years, the calculation being made by counting the kings who reigned in the intermediate period, to the number of 153 or 154. But among these a republic was thrice established, one extending to ... years, another to 300 and another to 120. The Indians also tell us that Dionysus was earlier than Heracles by fifteen generations, and that except him no one made a hostile invasion of India but that Alexander indeed came and overthrew in war all whom he attacked. This is how we may read the chronological clue from Megasthenes as a result of our scrutiny of the three versions that have come down to us in the reports of Pliny, Solinus and Arrian. The 6452 years, which we have conjectured from Arrian's 6042 in the light of the fact that Sandrocottus was not yet king even when he met Alexander some months after the latter's first entry

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 277 into India, are naturally uncertain to some extent, since we still do not know the exact date of his coronation. But we may very well be correct; for, scholars have suggested 325 B. C. almost as often as 324 B. C. that is to say, about a year later than Alexander's invasion. As Megasthenes could have received his chronological matter from none else than the Puranic pundits of his age, we have compared some of his statements with the information contained in our present Puranas or with whatever those pundits may be expected to have gathered from other traditional sources. We have found: 1. Dionysus is the Greek equivalent of the semi-legendary Prithu Vainya, the first consecrated monarch, a recipient of divine honours who may have been called Raja daivyena sahasa ("King with God-force"), signifying to the Greek ear "King Dionysus", and who initiated a new epoch in India by his achievements and combined himself certain associations of the deities Soma and Shiva, associations corresponding to those of Dionysus in religious myth or ritual practice or popular cult. 2. Sandrocottus, at whose court in Palibothra (Pataliputra) Megasthenes lived from c. 302 B. C. for a few years, is Chandragupta I, founder of the Imperial Guptas, between whom and Prithu there are in the Puranas 153 or 154 kings. 3. Heracles is the Greek equivalent of Krishna, here strictly speaking Hari-Krishna, who belonged to the Yadava, family and who, misunderstood in the primary meaning of "Yadava" as "son of Yadu", would stand in the 15th generation. after Prithu because Yadu, one of the sons of Yayati, marks the 14th generation in the Puranic series. WHAT REMAINS TO BE DONE What now remains to be done is to find the precise startingpoint of the Indian chronology conveyed to Megasthenes, as well as ascertain the missing number of years for the first "republic". Also, the two discoveries, along with the very fact of three "republics" have to be brought into line with the chronological

278 puranam - PURANA [Vol. VIII, No. 2 materials in the Puranas and with the traditional Indian chronology applied to them. This will enable us to see whether more evidence is available from Megasthenes to resolve the rivalry between Chandragupta Maurya and Chandragupta I for identification with Sandrocottus. Before we proceed, we may remind ourselves of three Puranic traditional dates: 1. 3102 B. C., the advent of the Kaliyuga with Krishna's death. 2. 3138 B. C., the year of the Bharata War and Parikshit's birth. 3. 3177 B. C., the year in which the Sapta Rishi, the Seven Rishis, the stars of the constellation Great Bear, are said to have entered the Nakshatra (lunar asterism) Magha in the course of a cycle of 27 centuries supposed to be running through the 27 lunar asterisms of the ecliptic by a stay of 100 years in each of them FROM THE DAYS OF DIONYSUS TO ALEXANDER We must begain our task of reaching the starting-point of Indian chronology in the age of Megasthenes by deciding the date from which to count backward to Dionysus the 6451 years and 3 month, a date connected with Alexander. Obviously, we are concerned here with the question, "When exactly did Alexander invade India and stand as Victor on Indian soil ?" As India Proper is east of the Indus we have to know the year, month and day of Alexander's crossing of this river. The year is 326 B. C. And the consensus of historians is that the crossing occurred in the beginning of spring. But what were the month and the day? For a satisfactory answer we should take note of all the information provided by the classical accounts of Alexander's campaign. THE DATE OF ALEXANDER'S INVASION The opinion that the invasion took place at the beginning of spring in 326 B. C. is derived from a passage in Strabo (XV.17) founded on Aristobulus, a companion of Alexander's. Strabo says about Alexander and his men: "They remained in the mountainous

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 279 country belonging to the Aspasioi and to the Assakenos during the winter. In the Beginning of spring they descended to the plains and the great city of Taxila, whence they went on to the Hydaspes and the land of Poros."1 But evidently the notion of the beginning of spring is general rather than precise: it does not connote the very first day of the season, for that day cannot equally apply to the Indus-crossing and the arrival at Taxila. Vincent Smith tells us that the arrival must have been 3 days later. So Aristobulus must mean a span of several days constituting the initial portion of spring. This is confirmed by another passage in Strabo (XV. i. 61, 62) based again on Aristobulus. Here he speaks of this historian's meeting with two Indian ascetics at Taxila and, referring to climatic condition, observes that "spring had just begun".3 Is there a way to know how small or large we should make Aristobulus span of several days, and how exactly in it place the Indus-crossing? Aristobulus himself supplies no clue. As with the Indus-crossing and the arrival at Taxila, his whole first statement crams together, in quick and uninterrupted sequence, many occurrences which actually stand fairly apart. Smith, following Arrian (V. 8), Diodorus (XVII. 87) and Curtius (VIII. 12, 13), speaks of Alexander's "stay in his comfortable quarters at Taxila for a sufficient time to rest his army". Then the march to the Hydaspes took, by Smith's calculation, probably a fortnight. On the western bank of the Hydaspes there was waiting and watching and foraging, while Porus deployed his army on the eastern bank. Smith supposes 6 or 7 weeks of preliminaries and preparations such as described by Arrian (V. 9, 10): at least a month may be supposed. Aristobulus slurs over all these time-gaps. He slurs similarly over intervals prior to the Indus-crossing. Quoting Curtius (VII.2), Smith' writes that, having left the mountainous country, Alexander "arrived at the Indus after the sixteenth encampment"-that is, at the end of 16 days of marching "through the forests down to the bridgehead at Ohind". On the authority of Diodorus (XVIII. 86) and Arrian (V. 3), Smith 1- Vincent Smith, The Early History af India (London, 1924), p. 64, fn. 2. 2, Ibid., p. 63. 3. J. Mc Crindle, The Invasion of India by Alexander the Great (1896), pp. 389-90. 4. op. cit., p. 66. 5. Ibid., p. 67. 6. Ibid., pp. 68, 90. 7. Ibid., pp. 62-63, 8. Ibid., p. 63.

280 puranam - PURANA [Vol. VIII, No. 2 mentions 20 days' rest and recreation for the army at the bridgehead. This means that the Indus was crossed (16+30) = 46 days after the men had started leaving the mountainous country. In what season should we put these days? Aristobulus has said that Alexander's men "remained" in that country "during the winter". So the 16 days of downward march before touching the plains and the 30 by the Indus before its crossing cannot be "durring the winter". We should thus be led to take Aristobulus "beginning of spring" in a very broad sense: the Indus was crossed 46 days after winter had ended and 49 days of spring had elapsed before Taxila was reached. But to say even after the 49th day that "spring had just begun" is to make nonsense of that expression. Even to designate the 46th as "the beginning of spring" is nonsensical. If the expression is to stand, we must take in rather a broad sense Aristobulus mention of winter also. Some of the 46 days before the Indus-crossing should be put into that season. But we cannot push there much more than half the number. So, approximately, the Induscrossing will take place after the first 23 days of spring and the arrival of Taxila after the first 26 days. Since Arrian (V. 4) informs us that the river was crossed early one morning, we may roughly put the passage in the dawn of the 24th spring-day. When exactly in the year would this day fall? It is surprising that Smith' should write: "The passage of the Indus must be dated in February or at the latest in March." Apparently he is going by that particular Indian Calendar which divides the year into 6 seasons, each of 2 months. Originally, by this Calendar, Vasanta or spring commenced in late February, 2 months after Sisira or dew-time which commenced in late December, strictly speaking at the winter solstice of December 21. But if Smith goes by this Calendar, what becomes of Aristobulus phrase? About 27 days out of a season of 2 months will carry us pretty close to the middle of it and clean beyond the beginning in even the broadest sense. The conclusion is inevitable that Aristobulus, 1. Ibid., p. 64, fn. 2.

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 281 in Strabo's report, was not writing in terms of the Calendar of 6 seasons. And, indeed, woud it not be odd that he should? When we know that the Greeks were writing for Greek readers, then, unless they give warning about a change of meaning in the terms intelligible to such readers, we have to assume for "spring" or for any other season the meaning commonly attached to it in the Greek Calendar. The proof is to hand in Strabo himself. He (XV.1.20) says: "India is watered by the summer rains, and the plains are overflowed." Arrian (V. IX) is clear-cut on the point when he speaks of "the time of the year when the sun is wont to turn towards the summer solstice" and adds: "At this season incessant and heavy rain falls in India." The summer solstice comes on June 21. But the scheme of 6 seasons has Grishma (heat) from late April to late June and then Varsha (rain) from June 21 for 2 months. Its summer is Grishma: it has no regular rains during that season. The Greek historians leave little doubt that, when they do not provide us with a clear sign of a different sense, their "summer" covers with its start the Indian monsoon and that this start is on the solstice of June 21. Here is an index to the usual Greek division of the year into 4 seasons, each of 3 months, in which spring begins on the vernal equinox, March 21, and runs up to the summer solstice. Hence, from all points of view, the first 27 days or so of spring which we have shown to be Aristobulu's "beginning" must extend from March 21 (inclusive) to nearly April 16 (inclusive). Then c. April 16 will mark the arrival at Taxila and c. April 13 the Indus-crossing. C. April 13 is the date of Alexander's invasion of India and the end-point from which we have to count backward by 6451 years and 3 months to reach the starting-point of Indian chronology. THE STARTING POINT OF INDIAN CHRONOLOGY Adding 6451 years and 3 months to c. April 13, 326 B. C., we go backward to c. January 13, 6777 B.C. But if the 3 months

282 puranam - PURANA [Vol. VIII, No. 2 are not meant to be taken quite literally, the starting-point of Indian chronology in the age of Megasthenes was a date somewhere in January 6777 B.C. what is striking about this year is the two end-digits. Immediately we are reminded of the Sapta Rishi cycle. The Seven Rishis enter each asterism in the year 77 of century, just as in the Puranic-traditional chronology they entered Magha in 3177 B. C. It would seem that Megasthenes 6777 B. C. was related to this cycle and that its being the starting-point of Indian chronology implied for this cycle in his day a starting-point_in January 6777 B. C., coinciding with the first year of the intermediate period between Dionysus and Sandrocottus the year in which the reign of the former came to an end and that of his successor, the first king out of the 153, commenced. If we attend to some of the Vedic associations of Sapta, the very use by the Greeks of the name "Dionysus" facilitates our bringing in the Sapta Rishi cycle. "The number seven," writes Sri Aurobindo¹, "plays an exceedingly important part in the Vedic system, as in most very ancient schools of thought." It is also applied to those beings, at once human and divine, called Angirasas, whose parable or legend is "on the whole the most important of all the Vedic myths."" The Angirasas are called sapta rishayah, the Seven Rishis or Seers." "The Angirasa Rishis are ordinarily described as seven in number: they are sapta viprah, the seven sages who have come down to us in the Puranic tradition and are enthroned by Indian astronomy in the constellation of the Great Bear". They are, as described in Hymn VI. 75-9, "the Fathers who dwell in the sweetness (the world of bliss), who establish the wide birth..." Expressive of this world of bliss is the Soma-wine, the heavenly effluence of the god Soma. "The drinking of the Soma-wine as the means of strength, victory and 1. On the Veda (Pondicherry, 1956), p. 111. 2. Ibid., p. 158. 3. Ibid., p. 207. 4. "Not that the names given them by the Puranas need be those which the Vedic tradition would have given them." (Sri Aurobindo's footnote) 5. Ibid., p. 198. 6. Ibid., p. 190.

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 283 attainment is one of the pervading figures in the Veda... The Angirasas also conquer in the strength of the Some."1 "They are brahmanaso pitarah somyasah...ritavridhah (VI. 75-10), the fathers who are full of the Soma and have the word and are therefore increasers of the Truth." The relation of Dionysus, the Greek god of wine, to the Vedic deity Soma, lord of the wine of delight (ananda) and immortality (amrita), pouring himself into gods and men, the deity who is also deep-hidden in the growths of the earth, waiting to be released as a rapture-flow for men and gods-the relation of Dionysus to Soma is obvious. Through Soma, Dionysus can be more easily linked with the seven Rishis and with the astronomical time-calculation known as their cycle. We may even suggest that the same Vedic association of the Seven Angirasas with Soma is related to the name which Arrian (Indica, I, VIII) gives of Dionysus successor who was "the most conversant with Bacchic matters": Spatembas. This name can be thought of as a Greek hearing of the possible Sanskrit compound "Saptambhas", meaning "Seven-watered". Now, the Rigveda (VII. 42.1) speaks of the Angirasas as being not only with "the divine Word, the cry of Heaven..., and of its lightnings thundering out from the Word", but also with "the divine waters...that are set flowing by that heavenly lightning..., and with the cutflowing of the divine waters the outpressing of the immortalising Soma..." These divine waters "are usually designated in the figured Vedic language as the seven Mothers or the seven fostering Cows, sapta dhenavah." And "this Soma-wine is the sweetness that comes flowing from the streams of the upper hidden world, it is that which flows in the seven waters..." Thus the Seven Rishis, Soma and the seven waters or rivers all go together and Spatembas as Saptambhas fits naturally and perfectly into the picture. Through the idea of the Soma-bearing seven waters that are associated with the Angirasas, the successor of Dionysus can also be linked with the cycle of the Sapta Rishi. 1. Ibid., p. 209-10. 2. Ibid., p. 210. 3. Ibid., pp. 215-16. 4. Ibid., p. 146 5. Ibid., p. 210.

284. puranam - PURANA [Vol. VIII, No. 2 Against a Vedic background we may even see a subtle identification of the stars of the Great Bear with the winecarrying waters: for the expressions sapta matarah and saptadhenavah ("the seven mothers" and "the seven fostering cows") are applied in the Rigveda indifferently to Rays and to Rivers.' Spatembas (=Saptambhas) would appear to have a rapport still more close than Dionysus with the Sapta Rishi. Perhaps the Puranic pundits in the age of Megasthenes held that it was he who, seeing the link of the Sapta Rishi with his predecessor, established their cycle as starting with the end of his predecessor's reign and the beginning of his own. In any case, we may well hazard to put the start of the cycle in January 6777 B. C. But the moment we do so we suggest a contact between Megasthenes starting-point of Indian chronology and the chronological statements in Indian tradition, And the question arises: "Initiating the cycle in 6777 B. C. with an appropriate asterism, would we reach in the course of the cycle's repetitions the Magha-century 3177-3077 B. C. within which Indian Puranic tradition places the Kaliyuga's commencement (3102 B. C.) and the Bharata War (3138 B. C.) with Parikshit's birth during its career ?" If that century could be reached, there would be convincing proof that Megasthenes 6777 B. C. was in direct relation to those two dates and that those dates were parts of the traditional Indian chronology as far bac kas c. 302 B.C. The crucial point to settle is: "What asterism out of the 27 should be considered the first one in 6777 B. C. ?" But before we can settle it we must decide whether the Seven Rishis were understood as traversing the asterisms in a forward or in a retrograde motion. Modern scholars have reported two schools of reckoning. Colonel Wilford remarked in 1805 that the direction was supposed to be retrograde. But A. Cunningham³ in 1883 took it to be forward. What was it in ancient times? 1. Ibid., p. 111. 2. "The Kings of Magadha" in the Asiatic Researches, Vol. 9. 3. The Book of Indian Eras.

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY THE DIRECTION OF THE SEVEN RISHIS' CYCLE 285 If we look at the cycle in the light of the Vedic Angirasas, we would take a hint from the verse: "Forward let the Angirasas travel, priests of the Word, forward go the cry of heaven, forward move the fostering Cows that diffuse their waters.." (VIII. 42-1).1 Apropos of the Puranas we have to answer by studying a verse which is found in both the Vishnu and the Bhagavata Puranas and which is the sole one naming another asterism in relation to Magha. It runs in F. E. Pargiter's translation: "When the Great Bear will pass from Maghas to Purva Asadha, then, starting from Nanda, this Kali Age will attain its magnitude." Who exactly is this Nanda and how long after Parikshit does he come and at what remove from Magha is Purva Ashadha ? Let us glance at the sequence of the 27 asterisms, commencing with Asvini as at present: (1) Asvini (2) Bharani (3) Krittika (4) Rohini (5) Mrgasira (6) Ardra (7) Punarvasu (8) Pushya (9) Aslesha (10) Magha (11) Purva Phalguni (12) Uttara Phalguni (13) Hasta (14) Chitra (15) Svati (16) Visakha (17) Anuradha (18) Jyeshtha (19) Mula (20) Purva Ashadha (21) Uttara Ashadha (22) Sravana (23) Dhanishtha (formerly Sravishtha) (24) Satabhisha (25) Purva Bhadrapada (26) Uttara Bhadrapada (27) Revati. If we go forward from Magha to Purva Ashadha we pass from the 10th to the 20th asterism, a space of 1000 years from the beginning of the one to the beginning of the other. By a retrograde motion we do the same after 1700 years. Now, where does Nanda occur in Puranic history? In the Vishnu and the Bhagavata themselves, the name Nanda is used for Mahapadma, who is called Mahananda in the Brahmanda, and ranked as the first of nine Nandas in all the Puranic lists. Also, the period from Parikshit's birth to the coronation of this Nanda, which is given in some Puranas as either 1. Sri Aurobindo's translation, op. cit., p. 215. 2. The Puranic Texts of the Dynasties of the Kali Age (London, 1913), p, 75. 3. lbid., p. 69, fn. 15. 4. Ibid., p. 58, fns. 14, 15.

286 puranam - PURANA [Vol. VIII, No. 2 1500 or 1050 years' or else 11152 is 1015 in certain versions of the Vishnu and the Bhagavata. So the Purva Ashadha century meant by these versions must be the one running from 1000 to 1100 years after the commencement of the Magha century. But that is exactly how that century would run in a forward sequence of the asterisms. Clearly, then, the motion of the Seven Rishis in these Puranas from Magha to Purva Ashadha is in a forward and not in a retrograde direction. THE FIRST ASTERISM FOR 6777 B. C. Now the ground is cleared for us to inquire what asterism should be the first in 6777 B. C. for the Sapta Rishi cycle in a forward motion. As we saw, the list of asterisms at present opens with Asvini. But Whitney informs us that the opening with Asvini was introduced in about 490 A. D. when the vernal equinox took place in the first point of this asterism. And G R. Kaye' rightly tells us: "The early lists all began with Krittika." Shall we make Krittika our initial asterism? But did the early lists put Krittika first because of a linking of it, as of Asvini, with an astronomical phenomenon serving to begin the New Year? And did Krittika always stand first before Asvini took the lead? What Kaye himself has to tell us in full is: "The early lists all began with Krittika, but the Mahabharata puts Sravana first, the Jyotisha Vedanga begins with Sravishtha, the Surya Prajnapti with Abhijit, the Surya Siddhanta with Asvini. But here Asvini is definitely equated with the vernal equinox, while Abhijit, Sravana and Sravishtha, which are continuous, are equated with the winter solstice." As Abhijit stands between Uttara Ashadha and Sravana in a system of 28 instead of 27 asterisms, Uttara Ashadha would replace it in the system with which we are dealing apropos of the Seven Rishis. So we 1. Ibid., pp. 58, 74. 2. Anand Swarup Gupta, "The Problem of Interpretation of the Puranas," Purana, Vol. VI, No. 1, January 1964, p. 67. 3. Pargiter, Op. cit., p. 74, fn. 10. 4. Surya Siddhanta, VIII, 9, p. 211. 5. The Indian Antiquary, Vol. 50, p. 47.

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 287 learn that when the winter solstice occurred successively in Uttara Ashadha, Sravana and Sravishtha, the list successively opened with these asterisms. But when we look further into the the Jyotisha Vedanga, which is admittedly the earliest astronomical treatise in our possession, we find light thrown on the initial position of Krittika in the early lists. For, although Sravishtha (under its old name Dhanishtha) leads the asterism-list, the list of the deities presiding over the various asterisms gives the prime place to Agni, the presiding deity of Krittika. A distinction between the ritualistic or religious primacy and the astronomical is brought out here. Krittika as a list-header is suggested to have a religious and not an astronomical import. And this suggestion is confirmed in the famous statement of Garga quoted by B. G. Tilak and, from Tilak, by Kaye. "Krittika is first for purposes of ritual, Sravishtha for the purpose of the calendar." It is easy to understand the religious primacy accorded to Krittika. As the centre of all ancient ritual was the sacrificial fire, the physical manifestation of the god Agni, and as Agni was the presiding deity of Krittika, this asterism stood the most prominently in the mind of the Brahmanas. But it can have no astronomical significance except when it could be associated either with the winter solstice or with vernal equinox,, the two points at which the New Year used to be started in different ages. Thus, to accept Krittika for starting the Sapta Rishi cycle in 6777 B. C. merely because it heads all the early lists would be a mistake. The asterism we want is one in which the winter solstice or the vernal equinox occured in that year and which on account of that occurrence would open the list. In view of the extreme antiquity of the year concerned we may simplify our search by attending to expert opinion. According to J. B. Fleet, originally the year started at the winter solstice, with Sisira as the first season beginning then. P. C. 1. B. G. Tilak, Orion (Bombay, 1893), p. 41. 2. Ibid., p. 30. 3. The Encyclopaedia Britannica (13th Ed.), Vol. XIII, p. 493.

288 puranam - FURANA [Vol. VIII, No. 2 Sengupta' assures us that the Vedic year began with the winter solstice and the spring came to be reckoned as the first season in a new system of computation. We may add from Sengupta a few other indications. The oldest astronomical reference he discovers is to the winter solstice happening on the full-moon day of the month Phalguna in the year c. 4550 B.C. He has also said: "The Vedic year-long sacrifices were begun in the earliest times on the day following the winter solstice... Winter was thus first season of the year... The Indian years, before the time of Aryabhata I, were generally begun from the winter solstice day, but after his time gradually the years came to be reckoned from the vernal equinoctial day." To find, however, our asterism we must understand the peculiarity of "the precession of the equinoxes". The equinoxvernal or autumnal-moves through the asterisms in the reverse order and the last point of an asterism is reached first and the first point last conversely, the asterism, in which the equinox takes place immediately before it occurs in another, is the one which in the normal order comes after it. This seeming anomaly is caused, as Newton explained, by the action or attraction of the planets, the sun and the moon on the earth's protuberant equatorial ring, so that daily the equinoctial points reach the meridian a little sooner than they otherwise would. The movement of the points is called "precession". The point of the winter solstice lies exactly halfway between those of the autumnal and the vernal equinoxes, that of the summer solstice vice versa. So the seeming anomaly applies to the solstitial points as well. The rate at which the equinoctial and solstitial points shift from asterism to asterism can be known by dividing by 27, which is the number of the asterisms, the time required by these points to 1. "Hindu Astronomy" in The Cultural Heritage of India, (Calcutta, 1937) Old Series, Vol. III, p. 345. 2. Ancient Indian Chronology (Calcutta, 1947), pp. xviii, 169; p. 156; p. 166; p. xx. 3. The New American Encyclapaedia (New York, 1945), p. 1116, "precession of the Equinoxes". 4. Ibid., p. 1265, "Solstice".

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 289 perform one complete circuit of the heavens. The points perform this circuit, called a period of precession, in 25,868 years. Consequently, the passage from asterisms to asterism, in connection with either the equinoxes or the solstices, occurs at the average 2 rate of (25,868/27=) 958 years. 27 To calculate where the winter solstice was in 6767 B. C., we need to start from a definite datum about its occurrence at some time or other. J.C. Ray, working from the accepted identification of Dhanishtha (or Sravishtha) with the star Beta Delphini, has calculated that the winter solstice occurred in the first point of Dhanishtha in 1372 B. C. and in the first point of Sravana in 405 B. C., though the earliest year in which the new moon happened on the day of the winter solstice so as to make Sravana observable as the star of this solstice was 401 B. C. From this it is easy to calculate that the winter solstice began to be in Dhanishtha-that is, at the last point of the asterism in (1372 +958=)2330 B. C. A table based on the average rate of precession can show us at a glance the asterisms of the winter solstice in the ages before 2330 B. C.: From 3288 to 2330 B. C.: Satabhisha 4246, 3288 B. C.: Purva Bhadrapada " " 520,, 4246 B. C.: Uttara Bhadrapada ww 6162,, 5204 B. C.: Revati 7220,, 6162 B. C.: Asvini Now, 6777 B. C. fell between the last two dates. Hence in that year, as throughout the period from 7220 to 6162 B. C., Asvini would head the asterism list. If the Sapta Rishi cycle was thought to have commenced in 6777 B. C., Asvini could be considered its first asterism. 1. lbid., p. 1116. 2. Paper entitled "The First Point of Asvini" (1934) quoted by V. S. Agrawala in India as Known to Panini (Lucknow, 1953), pp. 461-62, but misinterpreted by him owing to neglect of the reverse order of precession.

290 2000 puranam - PURANA [Vol. VIII, No. MEGASTHENES'S STARTING POINT AND THE PURANIC MAGHA Starting with Asvini in 6777 B. C., let us see where the Magha century would come according to a forward movement of the Seven Rishis through the asterisms at the rate of an asterim per century. Magha is the 10th asterism when Asvini is the 1st. So from the beginning of its century to the beginning of Asvini we have 900 years. The Seven Rishis, in the period before Alexander and Sandrocottus, would enter Magha once in (6777-900=)5877 B. C. and a second time, after 2700 years more, in (5877-2700=) 3177 B. C. and a third time in (3177-2700=) 477 B.C. The middle date is a most remarkable result. For, the century from 3177 to 3077 B. C. which it gives as the one during which the Seven Rishis stayed in Magha is precisely the century holding within it, according to the traditional Indian chronology, those two events the birth of Parikshit during the Bharata War and the advent of the Kaliyuga-which the Puranas declare to have occurred in the 100 years of the Seven Rishis' stay in Magha. The conclusion appears inevitable that the chronology communicated to Megasthenes as starting from 6777 B. C. not only employs the Sapta Rishi cycle known to the Puranas but is also related, through this cycle, both to the Puranic associations of Magha and to the dates traditionally going with those associations: 3102 and 3138 B. C. The implication of such a conclusion is that the Puranic pundits who were the informants of Megasthenes had already these dates, together with 3177 B. C., as important points of reference. In other words, the Puranas are linked to these traditional dates through a common background which is a chronology starting from 6777 B. C. and employing, like them, the cycle of the Seven Rishis and having, like tradition, for important points of reference 3138 and 3102 B. C. A strong hint that this chronology was cognisant of these dates is contained in some words of Megasthenes himself. Does he not mention Heracles no less than Dionysus and does he not mention his very epoch and has not scholarship identified his

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 291 Heracles with Krishna who played a central part in the Bharata War of 3138 B. C. and died in 3102 B. C. ? The liaison between Megasthenes starting point and the Puranic Magha, which from 3177 B. C. starts the century holding those traditional dates, has a most critical bearing on Indian history. But we shall touch on this bearing at the end. At the moment let us add a few remarks to render more definite to our minds the place of the starting-point in the January of 6777 B. C., and then pass on to consider the three "republics"? We have two facts to remember in arriving at greater definiteness: it is the Sapta ishi cycle that is concerned and it is the winter solstice of December 21 that begins the ancient. year. The years of the cycle are lunar¹ and get adjusted to the solar by the general arrangement of intercalated and suppressed. months. The lunar year begins after either the first new-moon conjunction or the first full-moon conjunction subsequent to the entrance of the sun into the Zodiacal sign with which the year commences. At present the former conjunction is used in Southern India, the latter in Northern.³ About the India known to Megasthenes, Curtius (VIII. 9) has preserved the information that the Indians "mark the divisions of time by the course of the moon not like most nations when the planet shows a full face but when she begins to appear horned". About still more ancient India Sengupta' says: "The months were begun either from a full moon or a new moon." The Sapta Rishi cycle as observed in Kashmir and thereabouts (from c. 800 A. D.) has its lunar months ending with the full-moon." But, as Jean Filliozat reminds us, they must originally have ended with the new-moon, for their reckoning was from Sudi 1 which is the start of the bright fortnight. What the still older Sapta ishi cycle did we have no knowledge about. According to our inference that it commenced in January 6777 B. C., it would seem to mark the beginning of the first month of the lunar year connected with the 1. The Encylopaedia Britannica (13th Ed.), Vol. XIII, p. 493. 2. Ibid. 3. Ibio. 4. Op. cit., p. 343, 5. The Encylopaedia Brittannica, Vol. XIII, p. 499. 6. L'Inde Classique (Paris, 1953), Vol. II, Appendix, p. 736. 9

292 puranam - PURANA [Vol. VIII, No. 2 winter solstice of December 21, 6778 B. C. And, since there is a gap of about 15 days between the new-moon and the full-moon, one of the two in relation to that solstice must fall in the opening half of January. If an astronomer could calculate which of the two did so, we should know the very day, the first day of the lunar year, from which ancient Indian chronology as formulated in c. 302 B. C. was taken to start. THE THREE "REPUBLICS" AND THE PURANIC CHRONOLOGY Now for the "republics". Do the Puranas contain any suggestions that there were 3 gaps in their dynastic series and do they indicate the time-lengths of these gaps. We may begin with some penetrating comments by D. R. Mankad.' He says: "It is, of course, evident that what Arrian calls a republic may mean a kingless period; and a kingless period would mean a period without a king, but, in the case of an imperial seat like Magadha, an absence of Imperial Dynasty." In other words, a "republic" can cover rule by either freebooters or foreigners. Mankad goes on to refer to what has become famous as the Yuga-Purana. It is the historical chapter of the Gargisamhita. "Scholars," says A. K. Narain," "are almost unanimous in regarding the Yuga-Purana as the earliest among the extant works of Purana type, and as exhibiting an independent tradition." It would be no wonder if Megasthenes received information from it. And the Yuga-Purana speaks of breaks in the dynastic series of Magadha. Mankad³ writes: "Usually, our Puranas say that the Sungas came immediately after the Mauryas... The Yuga-Purana...is unequivocal in saying that there was a period of foreign rule between the Mauryas and the Sungas." Mankad next cites the analysis he has made of the edition published by himself, perhaps the best edition so far, of this old document. From this analysis 1. Puranic Chronology (Anand, 1951), p. 85. 2. The Indo-Greeks. 3. Op. cit., p. 89.

July, 1966] THE PROBLEM OF INDIAN CHRONOLOGY 293 we see that the Kanvas who in the other Puranas immediately succeed the Sungas came also after an intervening period in which foreigners overran the country. Hence we have actually two "republics". But their timelengths are not mentioned. The converse is the case with a gap we may detect in the period earlier than the Sungas and Kanvas. There is the glaring discrepancy among the time-lengths the various Puranas give from the birth of Parikshit to the coronation of Mahapadma Nanda. We have already noted the different numbers of years 1500, 1115, 1050, 1015. The discrepancies between the largest number and the three smaller ones may be taken to point to a gap somewhere in the king-series. Then we shall have a "republic,' of 385, 450 or 485 years, whose unwelcome existence the Puranas have covered up by an increase in the individual reignspans and the dynastic durations, although a memory of it lingers in the shorter versions of the interval. Understood as a break in the king-series, the 385, 450 or 485 years provide us with Megasthenes missing number and complete his statement, while his two other numbers supply a chronological substance to the Yuga-Purana's history. In the matter of the 3 republics, he and the Puranas correspond in a complementary manner. MEGASTHENES'S CHRONOLOGY AND THE IDENTITY OF SANDROCOTTUS Our job of tallying chronologically the Greek evidence and the Indian is complete. But the practically perfect comparison which we have found possible calls for a revolution in our historical ideas. Not only have we to carry to c. 302 B. C. the cognisance of the dates for the Kaliyuga's advent and for thd Bharata War- 3102 and 3138 B. C. respectively-and thus give the lie to the conception dear to modern historians that they were astronomically fabricated after 400 A. D. We have also to see that in the light of this cognisance we come face with the Puranic time-indications about the various dynasties by reference to the birth of Parikshit

294 puranam - PURANA [Vol. VIII, No. 2 during the Bharata War in 3138 B. C. As we have already noted, these time-indications at even their longest stretch bring Chandragupta Maurya not later than the 16th century B. C. And all chronological clues from the Puranas, including the references to the 24th and 27th centuries ot the Seven Rishis after Magha in indicating the length of the Andhra dynasty, combine to put Chandragupta I, founder of the Imperial Guptas, in the age of Alexander the Great. Again, if the last 2 republics out of the 3 enumerated by Megasthenes answer to a couple of breaks of the Puranic dynasties after the Mauryas, then Megasthenes Sandrocottus cannot be Chandragupta Maurya but another Chandragupta much later. This Chandragupta can only be the founder of the Imperial Guptas. So the results, to which we come by commencing Indian chronology in the first half of January 6777 B. C. and by taking into account 3 republics, bear out the result to which we came by identifying Dionysus with Prithu Vainya and counting 153 or 154 kings after him down to the coronation of Sandrocottus. In a triple fashion Megasthenes, contemporary of Sandrocottus, supports the Puranic equation for this King of the Prasii whose coronation took place in c. 325 or 324 B. C. This, of course, does not automatically mean that all the Puranic dates are correct for the several dynasties preceding the Imperial Guptas. All would depend on whether the Bharata War , years before the Kaliyuga's advent, was fought or not in 3238 B. C. The Puranic pundits, accepting this date, have built up their chronology so as to lead from this date down to Chandragupta I in the time of Megasthenes. But the fact remains that they took their stand on the founder of the Guptas as the contemporary terminus of their chronology. The coronation of that king in c. 325 or 324 B. C. is an event we cannot question on the evidence of Megasthenes. Consequently, the whole of Indian history has to be reorientated on the basis of this new date established by Megasthenes for the rise of the Imperial Guptas.