FOREWORD

Jyotisha is the science of celestial luminaries. It deals with the twin fields of mathematics and astronomy. It is classified as one of the essential limbs (constituents) of the Veda (Vedangas).

Vedic literature enjoins that [Yajnas] are to be performed ; these sacrifices are based upon a knowledge of appropriate time for their performance. The science of astronomy gives us a knowledge of time. Hence, indeed, it has been regarded as one of the six [Vedangas]. This is stated in the ninth verse of the [Siddhanta-Siromani](=SS) as under:

´ÉänùɺiÉÉ´ÉiÉ ªÉYÉEò¨ÉÇ|É´ÉÞkÉÉ&

ªÉYÉÉ& |ÉÉäHòÉ& iÉä iÉÖ EòÉ™ôɸɪÉähÉ *

¶ÉÉÛÉÉnùº¨ÉÉiÉ EòÉ™ô¤ÉÉävÉÉä ªÉiÉ& ºªÉÉiÉÂ

´ÉänùɃói´ÉÆ VªÉÉäÊiɹɺªÉÉäHò¨Éº¨ÉÉiÉ **

[Sayana] in his Preface to the [Rgvedabhasya] strikes a note on the use of Jyotisha by citing the [Paniniya-siksa](41-42):

Uôxnù& {ÉÉnùÉè iÉÖ ´ÉänùºªÉ ½þ®úiÉÉè Eò±{ÉÉä%ªÉ {É`ö¬iÉä *

VªÉÉäÊiɹÉɨɪÉxÉÆ SÉIÉÖÌxɯûHÆò ¸ÉÉäjɨÉÖSªÉiÉä **

Metre represents the feet of the [Veda-purusha], Kalpa the hands, the science of Jyotisha its eyes, and Etymology its ears.

The use of the morphemic sequence ªÉYÉEòÉ™ôÉlÉÇʺÉrùªÉä occurring in the [Vedangajyotisha] (3) justifies the view that astronomy arose to establish the times and seasons for conduting sacrifices. [Sayana] further refers to several passages of the [Taittirya] School in support of the above view. The text, [samivatsaram]... indicates the periods of years. The text, [vasante]... refers to the seasons. the text [masi masi]... refers to months. the text [Yam Kama-yeta]... refers to half months . the text [ekastakayam]... days. The text [Krttikasu]refers to naksatras. Hence is the need for [Jyotisha] for determing the times and seasons proper for performing sacrifices.

Occasional referces are made in the Vedas to new moon and the full moon, the number of days in a year, the two halves of the year indicated by the terms, [Devayana] and [Pitryana], the additional or intercalary month(adhikamasa) and the deletory month (Ksayamasa). The text-"These

Krttikas (constellation of pleiades) do not deflect from the East"-suggests that the pleiades were observed to rise always at the east point. This is possible only when the first point of Aries was in the constellation of pleiades, since these are situated in the ecliptic. The point has, of course, shifted backwards now. From this observation, it emerges out that the stars of the zodiac were enumerated, commencing from the Krttikas (ArkaSomayaji, A CriticalStudy of Ancient Hindu Astronomy, p.1,1971).

The [Taittiriya Sam. (==TS) at 4,4, 10 lists the [Naksatras] starting from krttikas to

[Bharani]. This section in TS contains mantras for the naksatra bricks which are arranged in a circle round the naturally perforated brick, beginning on the south-east with krttikas and ending with [Visakha], then continuing on the north-west with [Anuradha] and ending with [Bharani]. The full moon brick is placed at the east point and the new moon at the west point (keith's notes to TS 4,4,10). Vedic [Astronomy cum mathematics shows the determination to count large numerical notations. The TS at 4,4, 11, ii-iii (==Vajasaneyi)[Madhyandina Samhita] XVII -ii) enumerates such numerical terms as [eka], [dasa], [sata], [sahasra], [ayuta]-(10,000), [niyuta]-(100,000), [Prayuta]-(1,000,000), [arbuda]-(10,000,000), [nyarbuda]-(100,000,000), [samudra]-(1,000,000,000), and [Parardha]-(1,000,000,000,000). [Mahidhara's] commentary at [Vaj].[Madh]. [Sam]. XVII -ii is worth quoting here:

+jÉèEòÉÊnù {É®úÉPÉÇ{ɪÉÇxiÉè& ¶É¤nèù°ükÉ®úÉäkÉ®Æú

nù¶ÉMÉÖÊhÉiÉÉ ºÉÆJªÉÉäSªÉiÉä **

Apart from division of the sphere into 27 or 28 [naksatras], Vedic astronomy has contributed to out understanding of the conception of great yugas-of course, carrying out a radical change of the heavenly bodies -- and that of the [Tithi].

In the area of geometry, the [Sulba]- [sutras] (c.200 B.C.) which are concerned with the measurements of sacrificial altars, discuss the construction of squares and rectangles, the relation of the diagonal to the

sides, the equivalence of rectangles and squares and the constrtuction of equivalent squares and circles.

The [Vedanga]-[Jyotisha] is a post-vedic development. The fourth verse of this texts treats Mathematics as standing at the head of all [Vedangas]; and it reads:

ªÉlÉÉ Ê¶ÉJÉÉ ¨ÉªÉÚ®úÉhÉÉÆ xÉÉMÉÉxÉÉÆ ¨ÉhɪÉÉä ªÉlÉÉ *

iÉuùuäùnùiÉÉRÂóMɶÉɺjÉÉhÉÉÆ MÉÊhÉiÉÆ ¨ÉÚvÉÇÊxÉ ÎºlÉiɨÉ **

Thus Mathematics and astronomy are twin disciplinces, the one complementing the other. [Ganitam] includes astronomy, and geometry (Ksetraganitam) belongs to the science of [Kalpasutras]. Geometry included the scope of ganitam. Ganitam also includes fundamental opertaions (Parikrama-), determinations (Vyavahara) and so on. Subsequently arose the [Siddhanta]-literature.

A [Siddhanta] text is an astronomical treatise, dealing with various measures of time, ranging from a [Trti] upto the duration of a [Kalpa] (which culminates in a deluge), planetrary theory, arithmetical computations as well as algebraical processes, problems relating to intricate ideas and their solutions, location of the earth, the stars and the planets and description and usage of instruments (SS vs.6). Of the eighteen [Siddhanta] works that are moticed, mention could down to us: [Suryasiddhanta],[Paitamaha], [siddhanta], [Romaka]-[Siddhanta], [Paulisa-siddhanta], [Vasisthasiddhanta], [Brahmasiddhanta] and

[Vrddhavasisthasiddhanta]. [Varahamihira] wrote the [Pancasiddhantika] and [Paitamahasiddhanta].

Of these the [Suryasiddhanta] and the [Brahmasiddhanta] deserve special mention here, since both these have received correction from time to time. At the same time the former work has shown " the process of adaptation of the new science to Indian ideas in its most pronounced state" (Keith, Hist. of Skt. Lit., 518). It reveals in the theory of kalpas, restores the pre-eminence of mount [Meru] at the north pole and deals with such astronomical concepts as [Naksakras] and others in the Indian context.

The astronomer who wrote the [Aryabhatiyam] (499 A.D.) is [Aryabhata] (born in 478 A.D.) who introduced new ideas into Indian

astronomy. He is the Sanskiritist to write a distinct chapter on mathematics in relation to astronomy. It may not be an exaggeration to say that he was the only Hindu astronomer to propound the doctrine of diurnal rotation of the earth, as stated by [Arka] [Somayaji]in ACritical Study of the Ancient Hindu Astronomy (p.2). The astronomers who followed him were Lalla (500 A.D.), [Varahamihira] (505 A.D.), [Brahmagupta], [Mahavira] (628 A.D.), [Sridhara] (750 A.D.), [Munjala] (932 A.D.),[Sripati] (1039 A.D.), [Bhaskara] II, the author of SS (1150 A.D.), [Makaranda] (1478 A.D.) and [Ganesa] (1520 A.D.). The names of others such as [Garga], [Vrddhagarga] and [Narada], who existed before [Varahamihira], may be added to this list.

The authorities on Ancient Sanskrit astronomy and mathematics are of the opinion that the last scientific work in [Jyotisha] is SS. A temple inscription quoted by O.E.SMITH, (History of Mathematics) runs as under:

[Triumphant] is the illustrious [Bhaskaracharya]

whose feet are revered by the wise, eminently learned.

[Bhaskara] II worked at the [Astronomical observatory in [Ujjain] where [Brahmagupta] is said to have conducted certain experiment several centuries ago.

Actually there were two [Bhaskaracharyas]. the first was a contemporary of the well-known astronomer, [Brahmagupta]. He wrote the [Mahabhaskariya], [Laghu bhaskariya], and the [Aryabhatiya] which are commentaries on the famous work of [Aryabhata]. The second [Bhaskara](1114-1185 A.D.), as has been stated in SS (vs 3.) composed the crest-jewel of astronomical treatises, i,e, SS, after having mastered the science under his talented father, [Mahesvara], a pioneer in astronomy, who championed the cause of [Jyotisha] in the eleventh century A.D.

The chief contribution made by [Bhaskara] II to mathematics [cum] astronomy consisted in realishing the true nature of division by Zero, anticipating the modern theory on convention of signs, representing unknown quantities by phonemes, presenting solutions for quadriatic equations reduced to a single type taking into consideration only positive roots as genuine, solving a few cubic and bi-quadriatic equations and indeterminate

BIBLIOGRAPHY OF

Prof. T.S.K.SASTRY'S WRITINGS

BOOKS

1. [Mahabhaskariya] of Bhaskara I, ed. with the Commentary of [Govindasvamin] and the Super-commentary of [Paramesvara], with detailed Introduction and Indices.Govt. Oriental Mss. Library, Madras, 1957.

2. [Vakyakarana], with the commentary of Sundararaja, with detailed Introduction, Notes and Appendices. K.S. Research Institute, Madras, 1962.

3. [Vedanga Jyotisha], Critically edited with [Translation] and [Notes]. Indian National Science Academy, New Delhi, 1985.

4. Collected Papers on Astronomy and Mathematics. Kendriya Sanskrit Vidyapeetha, Tirupati (1989).

5. [Panchasiddhantika] of [Varahamihira], Critically edited and Translated with Notes. (To be issued by the INSA, New Delhi)

6. [Vakyakarana] with Translation and Notes. ( In Ms. form)

7. Manual on Eclipses (In Ms.form)

8. Moon Tables (In Ms.form)

PAPERS

1. 'The Vasistha Sun and Moon in [Varahamihira,s] [Panchasiddhantika'] Journal of Oriental Research 25 (1955-56) 19-41; Collected papers, pp. 1-28.

2. 'Calendar in Hindu Tradition', Bulletin of the Inst. of Traditional Cultures, Univ. of Madras, 1958, pt.i,Rep. of Seminars, 41-114.

3. 'The [Bijopanaya]: Is it a work of [Bhaskaracharya]?' Jl. of the Oriental Institute (Baroda), 8 (1958-59), 399-409; Collectede Papers, pp. 29-45.

4. 'The Saka era of [Varahamihira] (Salivahana Saka)', Jl. of Indian History 36 (1958) 343-67; Collected Papers, pp. 255-67.

5. 'The untenability of the postulated Saka of 580 B.C.', Jl. of Indian History, 37 (1959) 201-24; Collected Papers, pp. 288-317.

6. 'A historical development of certain Hindu astronomical processes' Indian Jl. of Hist. of Science, 4 (1969) 107-25; Collected Papers pp. 46-75.

7. 'The system of [Vatesvara Siddhanta]', Indian Jl. of Hist. of science, 4(1949) 135-43; Collected Papers, pp. 76-88.

8. 'The school of [Aryabhata] and the peculiarities thereof', Indian Jl. of Hist. of Science, 4 (1969) 126-34; Collected Papers, pp. 89-101.

9. 'Some mis-interpretations and omissions in [Thibaut] and [Sudhakara Dvivedi] in the [Panchasiddhantika] of [Varahamihira]', Vishveshvaranand Indological Journal, 11(1973) 107-18; Collected Papers,pp. 102-17.

10. 'Determination of the date of the [Mahabharata]: The possibility therof ', Vishv. Ind.Jl.14(1976) 48-56; Collected Papers, pp. 318 -28.

11. 'The eposh of the Romaka [Siddhanta] in the [Panchasiddhantika] and the epoch longitudes of the Sun and Moon in the [Vasishtha-Paulisa]' Goverment Arts College, Kumbhakonam and in the Madras Presidency College from where he retired in 1955. Then he taught at the Madras Sanskrit College for about five years as Professor of comparative Philology and History of Sanskrit Literature. Even after retirement he served the college as Honorary Professor of Sanskirit.

Professor SASTRY critically edited six astronomical texts. He brought out a critical edition of the [Mahabhaskariya] with the commentaries of [Govindasvamin] and [parameshvara] with annotation and indices in 1957. Again he edited the [Vakyakarana], the basis of the [Vakya] almanacs of South India, with the commentary of [Sundararaja in 1962. He also critically edited the [Vedangajyotisha]with translation and notes. Subsequently he critically edited the [Pancasiddhantika] with translation and notes.

Dr. K.V. Sarma (now Professor at the Adyar Library Research Centre) who informally collaborated with Professor SASTRY in editing the first two works mentioned above, writes (in the Bio-data of Professor T.S.Kuppanna Sastry) as under:

His )Prof.Sastry's) deep understanding o Indian astronomy...helped him in preparing a rational edition with detailed exposition in English of the [Vedanga Jyotisha] and the [Pancasiddhantika], both of which are master -pieces illustrative of forensic skill in presenting distended facts to prove his point. He prepared also a book on the computation of eclispses incorporating modern corrections, but couched in such a form that it could be used by Indian almanao makers.

His collected papers issued by the Vidyapeetha, is a collection of valuable and original papers-publised in several learned Journals-numbering about twenty. The author has made a systematic, thorough-going and comparative study of the Hindu and Western systems of astronomy. The book deals with such interesting and illuminating topics as the [Vasistha] Sun and Moon, calender in Hindu Tradition, [Varahamihira's Saka Era], Hindu astronomical processes, [Vatesvara] [Siddhanta], [Aryabhata] School of Astronomy, Hindu Astronomy in the age corresponding to pre-copernican European Astronomy, Tamil Astronomy, determining the date of [AdiSankaracharya] (on astronomical grounds), the law of gravitation, the structure of atom and the theory of Relativity and others. Needless to say, among the astromers who have attempted a methodological and cirtical study of Jyotisha, Professor T.S.Kuppannna Sastry, the eminent scholar of ancient and modern astronomy, stands out as preeminent. I state in all hunility that the development of astronomy, marshalled in its historical perspective in the collected papers, will furnish some definite criteria governing the relecvancy and applicability of ancient Indian observations as enshrined in Jyotisha to modern astronomy.

It is now left for us to thank DR.K.V.Sarma sincerely for his hearty cooperation and assistance in printing this book. He read through the proofs, compiled the Bibliography of Prof.SASTRY's writings and sent us the author's Bio-data. Special acknowledgement should

be made to the Rathanam Press, Madras for setting the appropriate types for the book.

Lastly I pay homage to my guru, Professor T.S.Kuppanna Sastry for his excellent contribution to mathematics cum astronomy.

Kendriya Sanskrit

Vidyapeetha M.D. BALASUBRAHMANYAM

TIRUPATI. Principal (1975-85) 20-3-1989.

equations of the first and second degree, computing elaborate tables of sines, studing regularpolygons upto 384 sides, giving the value of as 754/240 and anticipating kepler's method of determining the surface and volume of a sphere (N.N.Sachitanand's] article in the Hindu, Madras dated

1-7-1979 and M.D.Balasubrahmanyam's Foreward to the Annotation of SS by Arka Somayaji, Kendriya Sanskrit Vidyapeetha, Tirupati series No.29,1980). Furthermore, [Bhaskara] II gave a scientific exposition of the sidereal revolution of planets, circumfrence of the earth, lunar eclipses, measurement technique of celestial bodies, longitude of the stars and other astromical facts. Needless to say, the third and fourth parts of SS, --under the heading, [Ganitadhyaya] (or Grahaganita) and (Goladhyaya)--are exclusively devoted to astronomy.

After [Bhaskara] II, very little original work appeared in India in this field. Later scholars were content with the writing of some commentaries on the earlier standard treatises of stalwarts, simply to whet their appetite. But for the scholarly compositions of eminent Sanskritists like [Nilakanta] and the rest, belonging to the productive Kerala School of astronomy, Jyotisha Pandits concentrated their attention more on astrology than on studies and research in mathematical astronomy.

However, in recent times scholars have been attempting to exdamine astronomical theories in the light of western through.*

Realising that specialisation in mathematical astronomy and other sciences has witnessed a decline, the Tirupati Vidyapeetha started a project entitled,'Coordination of Sanskrit and Ancient India', so that unpublised and rare works on Sanskrit mathematics, astronomy, and other disciplines might be critically edited with translation and annotation, besides monographs on historical and descriptive studies on Jyotisha might be brought out. Under this scheme, the Vidyapeetha has already brought out Dr.Arka Somayaji's Exposition in English and Annotation of [Bhaskaracarya's]SS.-(1980). Under the same project, the Vidyapeetha has now come forward to issue Professor T.S.KUPPANNA SASTRY'S Collected papers on Hindu astronomy, mathematics and other related discilines. I record here that it is rather unfortunate for SASTRY and us that he could not live to see his outstanding publication-the last challenging magnum opus of SASTRY. That Professor SASTRY, an eminent scholar in almost all the branches of Sanskrit literature (including mathematics and Astronomy) was specially qualified to write the collected papers, will

become evident, if we look at his curriculum vitae and publications .

Professor SASTRY (1900-1982) alias Srinivasan, was born in

Tirumanilayoor (near Karur, Tiruchirapalli district of Tamilnadu) to Subrahmanya Iyer and Bhagirathi Ammal. A scion of [Nilakantha Diksita, the celebrated Sanskrit polymath of the sixteenth century, Professor SASTRY devoted all his time to a critical study and appreciation of almost all the Sanskrit [Shastras] including [Ganita], [Jyotisha], and modern astronomy. In boyhood he underwent training in the traditional recitation of [Samavedic hymnology]. Having completed his schooling in Karur, he paased the B.A. examination as a student of the famous St.Joseph's College, Tiruchirapalli. He worked as Headmaster of the High School at Tirumayam (erstwhile Pudukkottah State), and then joined the Maharaja's High School, Pudukkottah (later known as Brihadambal High School). Subsequently he worked as lecturer, Assistant Professor in Sanskrit at Maharaja's College, Pudukkottah, Indian Jl. of Hist. of Science 13 (1978) 151-58; Collected papers,pp. 188-200.

12. The Vasishtha -Paulisa Venus in the [Panchasiddhantika] of [Varahamihira]', Indian Jl. of Hist. of Science 14 (1979); Collected Papers, 141-48.

13. 'The main characteristics of Hindu astronomy in the period corresponding to pre-Copernious European astronomy', Collected Papers, pp.118-40.

14. 'Vasishtha-Paulisa Jupiter and Saturn in the [Panchasiddhantika]', Collected Papers,pp.148-68.

15. 'The [Vasishtha -Paulisa] Mars in the [Panchasiddhantika] of Varahamihira]', Collected Papers, pp.169-87.

16. 'The epoch constants of the Vasishtha-Paulisa Star Planets ', Collected Papers, pp.201-05.

17. ['Saurasiddhanta] of [Panchasiddhantika]: Planetary constants and computation ', Collected Papers, pp, 206-40.

18. '[Panchasiddhantika] XVIII. 64-81: An Interpolation', Collected Papers, pp. 241-54.

19. ' A Brief History of Tamil Astronomy', Madras Univ. Jl., Section C, 41.ii; 120-33; Collected Papers, pp. 329-44.

20. 'The Age of Sankara: I', Collected Papers, pp.344-61.

21. 'The Age of Sankara: II', Collected Papers, pp.362 ff.

Complied by

K.V.SARMA

Library Research Centre,

Adyar (Madras).

TABLE OF CONTENTS

Foreword Pages

Bibliography of Prof. SASTRY'S writings xiv

1. The [Vasistha Sun and Moon in

[Varahamihira's [Pancasiddhantika] ... 1- 28

2. The [Bijopanaya: Is it a work of

[Bhaskaracarya]? ... 29-45

3. A Historical Development of certain

Hindu Astronomical Processes ... 46-75

4. The System of the Vatesvara [Siddhanta] ... 76-88

5. The School of [Aryabhata] and the

peculiarities thereof ... 89-101

6. Some Mis-Interpretations and Omissions

by Thibaut and Sudhakara Dvivedi in the

[Pancasiddhantika] of [Varahamihira] ... 102-117

7. The main characteristics of Hindu

Astronomy in the period corresponding

to Pre-Copernican European Astronomy ... 118-140

8. The [Vasistha-Paulisa] Venus in the

[Pancasiddhantika] of [Varahamihira] ... 141-147

9. [Vaistha-Paulisa] Jupiter and Saturn in

the [Pancasiddhantika] ... 148-168

10. The [Vasistha-Paulisa] Mars in the

[Pancasiddhantika] of [Varahamihira] ... 169-187

11. The Epoch of the Romaka Siddhanta in

the [Pancasiddhantika], and the Epoch

Longitudes of the Sun and Moon in the

[Vasistha-Paulisa] ... 188-200

12. The Epoch Constants of the [Vasistha-

Paulisa] Star Planets ... 201-205

13. [Saurasiddhanta] of [Pancasiddhantika]:

Planetary Constants and computation

(PS XVI,XVII, XVIII) ... 206-240

14. [Pancasiddhantika] XVIII 64-81: AN

Interpolation ... 241-254

15. The [Saka Era] of [Varahamihira]

(Salivahana Saka) ... 255-287

16. The Untenability of the Postulated Saka

of 550 B.C. ... 288-317

17. Determination of the Date of the

[Mahabharata]: The Possibility Thereof ... 318-328

18. A Brief History of Tamil Astronomy ... 329-344

19. The Age of Sankara: I ... 345-361

20. The Age of Sankara: II ... 362-370

21. Astronomy ... 371-378

22. The Stars ... 379-384

23. The Structure of the Atom ... 385-390

24. Newton and the Law of Gravitation ... 391-398

25. The Evolution of the Universe

According to Sir James Jeans ... 399-403

26. The Duration of Eclipses ... 404-408

27. The Lunar Eclipse in Hindu Astronomy ... 409-414

28. The Theory of Relativity ... 415-436

29. Computation of the Solar Eclipse in

Hindu Astronomy ... 437-450

30. Hindu Astronomy Through the Ages --

A Short Sketch ... 451-459

THE [VASISTHA]SUN AND MOON IN

[VARAHAMIHIRA'S] [PANCASIDDHATIKA]

(Reprinted from J.O.R, K.S.R.I., Madras, 1955-56)

The [Vasistha] Sun and Moon are contained in Ch.II and in Ch. III. 4 of the [Pancasiddhatika]. II. 1 gives the true sun, II. 2-6 give the cimputation of the true moon, III.4 gives a rule for the daily motion of the moon, II.7 gives rules for the sun or moon's Naksatra and the [Tithi], and II.8-1 deal with topics related with the sun, like the duration of day-light, the length of the noon-shadow when the sun is known and [vice versa], and lastly finding the Orient Ecliptic Point (Lagna) when the shadow is known and [vice versa].

Obviously, the most important parts are II. 1, II.2-6, and III.4, which form the basis for the rest of the work. Also these parts are very interesting from a historical from the more ancient astronomy represented by the [Paitamaha Siddhanta] condensed by [Varaha], giving only mean sun and moon, to the later epicyclic astrronomy represented by the Saura [Siddhanta] condensed by him. Of these, about II.1,Dr.Thibaut(T) makes the remark, "a stanza of obscure import," and leaves it at that, without attempting to translate it, and Sudhakara Dvidevi (5) remarks: "+xÉäxÉ ¶±ÉÉäEäòxÉ ËEò ºÉÉvÉ.ªÉiÉÒÊiÉ xÉ YÉɪÉiÉä, +iªÉ¶ÉÖrùi´É iÉÖ". So much for the sun. About II.2-6, T says (and S echoes him): "Of the above stanzas we have succeded in making out the sense in part only. They manifestly teach how to find the mean and perhaps also the true postitions of the moon by means of a process more compendious than the one usually employed in Indian Astronomy.

" What Preliminary operation is presented in stanza 1, we are altogether unable to say....... It is not apparent why stanza 5 directs us to add for that half-gati six Signs plus four minutes to the moon's mean place, for the moon's meanmotion in one half-gati amounts to considerably more, Viz. six Signs plus about 92 minutes. Nor are we at present able to throw light upon the meaning of the processes prescribed in stanza 6. They possible refer to the operation of finding the moon's true place, although we are more inclined to think that this latter part is treated in stanzas 4-9 of the next (i.e., III) chapter. And S says: "MÉiªÉPÉæ Eò±ÉÉSÉiÉÖ¹]õªÉºÉƺEòÉ®úºªÉ iÉlÉÉ ¶±ÉÉäEòºªÉ <nù xÉÓ{ɪÉÇxiÉÆ xÉ ¤ÉÖrùÉ ={É{ÉÊkÉ." Thus T and S are both unaware that stanza 6 indeed gives the operation of finding the true moon. Further, their

interpretation of II.3-5 is wrong in several places, and by this they have shut out important necessary data. About III. 4. (including th next five stanzas as well), T remarks without being able to give any translation, "Six stanzas referring to the moon. The details, however, are obscure." About the same S says, "+xªÉä¹ÉÆ (i.e. 4-9) ´ªÉÉ®ú´ªÉÉÆ +OÉä ´ÉIªÉä," and he has left it there without coming back to it. In spite of this self-confessed ignorance, T remarks on p.li of his Introduction,"......the methods are so crude and so completely omit to distinguish between mean and true astronomical quantities that the [Vasistha Siddhanta] can hardly be included within scientific HIndu Astronomy".

The [Prancasiddhantika] was first printed in 1889 and during these nearly 70 years nobody, to my knowledge, has thrown light on the true nature of the [Vasistha Siddhanta]. In dealing with the topic in question, it is my desire also to discuss the readings of the text. We have only two manuscripts to go by, one badly vitiated and the other worse, both printed in the edition, one against the emended text and the other as footnotes. Frequent quotations, from the [Pancasiddhantika] are found in [Bhattotpala's] commentary of the [Brhatsamhita], but the range of our topic is limited, the whole thing taking only 14 [aryas], and these are not found quoted anywhere, as far as I know. But the subject matter being scientific, it is possibleto fix the correct text fairly well in most places of doubt, taking for our guidance the [arya] metre and the relics of the slaughtered words, provided we are certain of the intention of the author. I now proceed to the elucidation of the text. The under standing of the text every where comes first, as that is the basis for the correction of the text.

The Sun

II. 1.

EÞòiÉMÉÖhɪɹɨÉÞiÉÖªÉÖiɨÉèEòiÉÖǨÉ- EÞòiÉMÉÖhɨÉÞiÉÖªÉÖiɨÉäEò-

xÉÖ¾þiÉÆ ¹Éb÷¬¨ÉäxnÖùʦÉ̴ɦÉVÉäiÉÂ* iÉÖǨÉxÉÖ¾þiÉÆ ¹Éb÷¬¨ÉäxnÖùʦÉ̴ɦÉVÉäiÉ *

¶ÉʶÉJÉJÉJɪɨɺ´É®úGòiÉ- ¶ÉʶÉJÉJÉJɪɨÉEÞòiɺ´É®ú-

xÉ´ÉxɴɴɺÉÖ¹É]ÂõEòʴɹɪÉÉäxÉè& ** xÉ´ÉxɴɴɺÉÖ¹É]ÂõEòʴɹɪÉÉäxÉè& **

II.1. The days from epoch (Ahargana) are to be multiplied by 4, and 6 added to the product. This is to be divided by 1461 (and the remainder taken). From the reminder should be deducted successively 126 minus 1,0,0,0,2,4,7,9,9,8,6,5,(i.e., the twelve quantities 125, 126, 126,126,

124,122,119,117,117,118,120,121, are to be

deducted successively ). (The sun's [Rasis], Mesa etc., are successively got).

It is to be noted that Ahargana is not mentioned, that we should take the remainder for the operation is not mentioned, and what we get by this rule is not given. We have to presume all these. The word Ahargana can be easily understoood, because that is the starting point of of computation. The rule give, as also the numebrs, point to the necessity of taking the remainder, and to the true sun in [Rasis], as the obect of the operation.

The rule is explained as follows: The days from the epoch by being multiplied by 4 are converted into quater-days. 6 quater days are added to this because the beginning of the year, (in this case the true year), is 11/2 days before the epoch, and by the addition of the 6 quarter -days we get the total quarter-days from the beginning of the true year. According to this [Siddhanta] the year contains 365 1/4 days or 1461 and taking the remiander we get the quater days by 1461 and taking the remainder we get the quater days from the beginning of the current year. During the first 125 quater days (i.e., 31 1/4 days) of the year the sun traverses the first [Rasi], i.e., Mesa. During the next 126 quarter days (i.e.,31 1/2 days) he crosses the second, i.e., [Rsabha Rasi] and so on. It is to be noted that there are 12 quantities for the 12 [Rasis], and that these add up to 1461. The sun's position within a [Rasi] is expected to be found by proportion.

Thus if the days from epoch is 942, says, the quarter days from the beginning of the year just before the epoch is, 942X4+6==3774. Dividing out by 1461, we get 2 full solar years elapsed (which are not wnated), and 852 quarter days are left over in the thrid year. We can take from this 125, 126,126,126,124 and 122,and 103 quarter days are left over, i.e. the sun has passed six full [Rasis], and in the seventh he has gone 103/119 parts or 26o.

At epoch, the sun is 6/125X30o==1o26/ in Mesa, at sunrise at Ujjain on Sunday, (near the end of `saka 427). What is the epoch, and how we

are to find the days from epoch, these two things are not mentioned here. But ch.I.8-13 gives to find the days from epoch and we can adopt it, though given for Romaka and Pauli`sa, for the interval between two points of time is invariant. Only we must take into account the timeof day of epoch. [Vasistha] epoch mist probably is sunrise at Ujjain, Sunday, at the end of `saka 427. This matter will be discussed subsequently.

Now we are in a position to discuss the adopted readings, T and S, because they have not understood the stanza, have simply corected the obvious scriptory errors, and so far as they go, they are correct. The text -reading ªÉ¹É¨ÉÞiÉÖ has been

corrected into ¹É½Âþ@ñiÉÖÇ, following the variant reading¹ÉnÚùGòiÉÖ: ªÉÖiɨÉèEòiÉè is corrected into ªÉÖxɨÉäEòiÉÖÇ. º¤É®úHò is corrected as º´É®úCEò. I have made the following further corrections: (1) ¹ÉbÂ÷ has been deleted and only @ñiÉÖ retained, because the first foot clearly ends with ¨ÉäEò®Âú and there are two [matras] extra. It is better ¹ÉbÂ÷ is omitted because one Ms. has it, the other haivng a corruption, ªÉ¹É. But both Mss. have@ñiÉÖ. (2) In the third foot, I have interchanged º´É®úEÞòiÉ and made it EÞòiɺ´É®ú, because order requires it. There must be a gradual and continuous fall from 126 to 117. The valused indeed must have been obtained empirically, but it is too much to assume that the [Siddhanta] was aware of the gradual nature of the fall and rise, and gave what it saw as º´É®úEÞòiÉ. the error of observation also would be too great if it is º´É®úEÞòiÉ.

The Moon

II.2-6

®úºÉMÉÖhÉxÉ´ÉäxnÖùªÉÖHò 2 ®úºÉMÉÖhÉxÉ´ÉäxnÖùªÉÖHäò

¶ÉʶÉMÉÖhÉJÉMÉÖhÉÉänÂùvÉÞiÉätiÉÉtÖMÉhÉä * ¶ÉʶÉMÉÖhÉJÉMÉÖhÉÉänÂùvÉÞiÉä PÉxÉÉ tÖMÉhÉä *

¶Éä¹Éä xÉ´ÉʦÉMÉÖÇÊhÉiÉä ¶Éä¹Éä xÉ´ÉʦÉMÉÖÇÊhÉiÉä

MÉiɪÉÉä%¹]õÊVÉxÉè& {ÉnÆù ¶Éä¹É ** MÉiɪÉÉä%¹]õÊVÉxÉè& {ÉnÆù ¶Éä¹É¨É **

txÉ (¹ÉÉäb÷¶É) ¾þiÉÆ ¶Éä¹ÉÆ 3 PÉxÉ (¹ÉÉäb÷¶É) ¾þiɶÉä¹ÉÆ

|ÉÉäVVªÉÉtκjÉMÉÖÊhÉiÉÆ SÉiÉÖ¦ÉÇHÆò * |ÉÉäVYªÉÉvÉκjÉMÉÖÊhÉiÉÆ SÉiÉÖ¦ÉÇHò¨É *

¦ÉÉÊnùEò±ÉÉÊuùMÉÖhÉtÉxÉÉ ¦ÉÉÊnù Eò±ÉÉÊuùMÉÖhÉPÉxÉÉ

¶ÉʶɨÉÖÊxÉ xɴɪɨÉÉ·É®úɶÉÉtÉ ** ¶ÉʶɨÉÖÊxÉ xɴɪɨÉÉ·É ®úɶªÉÉt& *

ʴɹªÉvÉÞiɪÉÉä MÉÊiÉPxÉÉ 4. ʴɹɪÉvÉÞiɪÉÉä MÉÊiÉPxÉÉ

MÉiÉÊiɹɹ]õÉƶÉÉäÊxÉiÉÉ Eò±ÉÉ& |ÉÉäHòÉ& * MÉÊiÉEòÉ,#Â]õÉƶÉÉäÊxÉiÉÉ& Eò±ÉÉ& |ÉÉäHòÉ&

´ÉänùÉEòÉÇ& {ÉÉnùºÉÆJªÉÉ ´ÉänùÉEòÉÇ& {ÉnùºÉÆJªÉÉ

MÉiªÉÆPÉÇ vÉxɨÉÞhÉÆ {É®úiÉ& ** MÉiªÉvÉæ vÉxɨÉÞhÉÆ {É®úiÉ& **

MÉiªÉvÉæ ¦ÉMÉhÉÉÆvÉÇ 5. MÉiªÉvÉæ ¦ÉMÉhÉÉÆvÉÇ

näùªÉÆ Ê±É{iÉɶSÉiÉÖ¹EòºÉƪÉÖHÆò * näùªÉÆ Ê±É{iÉÉ SÉiÉÖ¹EòºÉƪÉÖHò¨É *

¶Éä¹É{ÉnùºÉ¨ÉÉ·ÉÉƶÉÉ- ¶Éä¹É{ÉnùºÉ¨ÉɶSÉƶÉÉ-

ºiÉ·É vÉxÉhÉÉÇiÉ ¡ò±ÉÆ nùxiªÉÆ ** ºiÉè·É vÉxÉhÉÉÇi¡ò±ÉÆ näùªÉ¨É **

¤ªÉäEò{ÉnùʨÉÎxpùªÉPxÉÆ 6. ´ªÉäEò{ÉnùʨÉÎxpùªÉPxÉÆ

EÞòiÉxÉ´Énù¶ÉºÉƪÉÖiÉÆ Ê´ÉªÉÖHÆò SÉ * EÞòiÉxÉ´Énù¶ÉºÉƪÉÖiÉÆ Ê´ÉªÉÖHÆò SÉ *

¨ÉxÉÖ´ÉänùªÉ¨Éä¦{É& {Énù- ¨ÉxÉÖ´ÉänùªÉ¨Éä¦{É& {Énù-

MÉÖhÉä Êjɹɹ]õ¬Éä vÉÞiÉä ʱÉ{iÉÉ ** MÉÖhÉä Êjɤɹ]õ¬ÉänÂùvÉÞiÉä ʱÉ{iÉ& **

III. 4

xÉMÉÉi{ÉnùÉqù¶ÉPxÉÉiÉ (Ê´É) xÉ´ÉÉi{ÉnùÉqù¶ÉPxÉÉiÉÂ

ºÉ{iÉÉƶÉ& ºÉÊ·É ºÉÉÆ´É®úÉä¦ÉÖÊHò& * ºÉ{iÉÉƶɺºÉÉÊ·ÉJɺ´É®úÉä ¦ÉÖÊHò& *

MÉiªÉvÉÉÇxiÉÉ UôÉävªÉÉä MÉiªÉvÉÉÇxiÉÉSUôÉävªÉÉä

ʱÉ{iÉɦªÉÉä ´ÉºÉÖ¨ÉÖÊxÉxɴɦªÉ& ** ʱÉ{iÉɦªÉÉä xɴɨÉÖÊxɴɺÉÖ¦ªÉ& **

II.2. 1936 is added to the days from epoch, and the total divided by 3031. The quotient are called Ghanas. The remainder is to be multiplied by 9 and divided by 248. The quotient are called Gatis. The remainder are called Padas.

II.3. The Ghanas are to be divided by 16 and the remainder taken. This should be multiplied by 3 and divided by 4. This is [Rasi] etc. It should be deducated from 12 [Rasis] (and the remainder written down). Minutes of are equal to twice total Ghanas (are to be added). One [Rasi], 7 degrees and 29 minutes are (also to be added).

II. 4. 185 multiplied by the Gatis minus a tenth of the Gatis are minutes (which are also to be added). The (first) 124 Padas are designated a half-gati. (If the Padas are less than 124) they are called plus-padas. If more then 124, 124 is taken from it to form a half-gati, and the remainder are called minus-padas.

II. 5. If there is a half-gati, (for the sake of that half-gati) 6 [Rasis] and

4 minutes are to be added. Also degrees equal in number to the plus-padas or minuspadas (are to be added). With these plus or minus padas respectively, the minutes got by the plus operation or minus operation respectively (in II.6) are to be added.

II. 6. Deduct one from the plus-pada or minuspada, and multiply this by 5. (If plus-pada) add the product to 1094, multiplied the sum by the plus pada and divide by 63. The resulting minutes (are to be added). If minus-pada, the product is to be substracted from 2414, the remainder multiplied by the minus-pada, and divided by 63. The resulting minutes (are to be added). (Thus the true moon is got).

III. 4. Deduct 9 from the number of plus or minus padas, multiply this by 10 and divided by 7. Add this to 702 if plus-padas. Deduct this from 879 if minuspadas. The resulting minutes are the daily motion (for the day ending with the padas).

The following is the explanation of the process. The true moon at a given time t. is, (i) the mean moon at t plus (ii) the equation of the centre for t. (i) is given here in five parts. We shall call them a, b,c,d,e which are to be added up to get the total mean moon. a (usually called the [Mula-dhruva) is the position of the mean moon at a point of time 1936 days before the epoch, when the moon's apogee and the mean moon exactly coincided, according to this [Siddhanta]. This is given as ¶ÉʶɨÉÖÊxÉxɴɪɨÉɸ´É ®úɶªÉÉtÉ:, i.e. 1 [Rasi], 7 degrees, 29 minutes. b is the mean motion during a whole numbers of cycles of 3031 days, each cycle equal to 110 anomalistic revolutions of the moon, from that point of time. This b is found by multiplying the mean motion per cycle,(110 revolutions; 11 [Rasis], 7 degrees, 32 minutes), by the number of cycles, called Ghanas, obtained as quotient by dividing the Ahargana plus 1936 days, by 3031. As full revolutions may be neglected, it is enough if we multiply the Ghanas X2' is given by ÊuùMÉÖhÉPÉxÉÉ: Eò±ÉÉ: (ªÉÉäVªÉÉ:). Because 16 Ghanas X 11 7o30/equal 15 full revolutions, it is enough if we divide out the Ghanas by 16 and take the remainder alone for multiplication, which we are asked to do by PɴɹÉÉb÷¶É½þiɶÉä¹É¨ÉÂ. As 11' 7o30/ is 3/4 [Rasi less than a full revolution, we can

multiply the remaining Ghanas by 3/4 [Rasi] and take this as substractive, which we are instructed to do by |ÉÉäVZÉÚªÉÉ PɺjÉÖËhÉiÉ SÉiÉÖ¦ÉèCEÆò ¦ÉÉÊnù. Thus b is disposed of, c is the mean motion during the subsequent full anomalistic

revolutions called Gatis, which form the quotient got by dividing the remaining days by the anomalistic period, 248/9days; (multiplying the days by 9 and dividing by 248 is only dividing by 248/9). For each Gati the mean motion is 1 revolution and 184 9/10 minutes. Hence the rule to multiply the Gatis by 185'and deduct minutes equal to 1/10 of the Gatis. This is given by ʴɹɪÉvÉÞiɪÉÉä MÉÊiÉ´PÉÉ MÉÊiÉEòÉ´¹ÉÉƶÉÉäÊxÉiÉÉ:Eò±ÉÉ: (ªÉÉäVªÉÉ:). What are now left over are 9ths of days called Padas (and these obviously would be less than 248). The mean motion perpada to complete the mean motion till t. Of this the [Siddhanta] asks us to add 1oper Pada first which is given by ¶Éä¹É{ÉnùºÉ¨Éɸ´ÉÉƶÉÉ: (ªÉÉäVªÉÉ:) This forms d. The residue 27'.843 per Pada forming e is combined with the equation of the centre (ii) and given by the two equations of II.6. If the Pada contain a half-gati, the value of d+e+(ii) for the half-gati part are combined together and given as 180o4'. This180o4'is got as follows: The half -gati is equal to 124 padas. So d==124o. e+(ii) given by the equation is {(1094+5X(124-1)}X124/63==3364'==56o4'; 124o+56o4'==180o4'. Thus it is that we get 180o4' for a half-gati, given by MÉiªÉPÉè ¦ÉMÉhÉÉPÉè näùªÉÆ Ê±Éi{ÉÉSÉiÉÖ¹EòºÉƪÉÖ¨ÉÚ, which instruction has so much puzzled T and S. But of course this is incorrect, and the defect lies in the equation of the centre part of the formulae in II, 6, which give the zero value for the equation of the centre not at 124 Padas, but at 133 Padas,as I shall show presently.

I shall first explain II, 6, and show how the formulae here combine the residual mean motion, viz. Padas X27'. 843(==e), with what is identifiable with the equation of the centre (==ii). If p==plus or minus Padas, the formula for plus-padas can be written down as {1094+5(P--1)}P/63, and the formula for minus-padas, as {2414-5(P-1)}P/63. Let us first take the plus take the plus-pada formula. As we have said e+(ii)=={1094+5(P-1}p/63, and e==27'.843 P.

Therefore (ii)=={1094+5(P-1)} P/63-27'.843P

==(1089+5P)P/63X27'.843P/63.

==(1089-1754+5P) P/63.

==(5 P-665)P/63.

Taking the minus -pada formula,

(ii)=={2414-5(P-1)}P/63-27'.843P

==(2419-5P)P/63-63X27'.843P/63

==(2419-1754-5P)P/63

==(665-5)P/63

Now, the plus -pada representing the original Padas lying within 0 to 124 Padas, correspond to teh anomaly lying between 0 and 180o and 360o. So for the plus-pada the equation of the centre must be negative. We see, (5P-665) P/63 is

indeed negative for all values of plus-pada, and (665-5p) p/63 is positive for all values of minus-pada. It is to be noted that the one is the negative of the other. Again, the former starts from the value 0 for P==0, gradually goes to a minimum for p=661/2 and again gradually, increases reaching 0 for p==133, (but P stops with 124), while the latter starting from 0 value for p=0, goes to a maximum for p=661/2 and falls gradually to 0 in a similar manner. Thus the equations roughly behave like the term of the equation of the centre of modern astronomy, -k sin0, and therefore identiable with the equation of the centre. This is noteworthy, as forming a transition from a stage of no equation of the centre to the stage of epicyclic astronomy giving the equation of the centre in the form, -ksin0. Substituting 661/2 for p we get the maximum or minimum equation of the centre =351.

I now proceed to the explanation of III.4, giving the formulae for the daily motion. They can be written down thus:

For plus -padas, 702'+10'(p-9)/7. For minus -padas, 879'-10'(P-9)/7. We see, the being linear equations, that the increase or decrease in the daily motion is uniform, from 702' to 879' and and back again from 879'to 702'. 702' is the minimum and 879' is the maximum, the former being 88'.5less and the latter 88'.5more than the mean motion790'.6. Though mathematically we can get values less than 702' and more than 879' for P less than 9, this is not envisaged by the [Siddhanta]. I shall now derive these rules from those of II.6 and thus show that these belong to the [Vasistha], to whatever other [Siddhanta] also they belong.

The daily motion for the day ending P padas is clearly, the ture moon for P minus thr true moon for (p-9). So for plus-padas the daily motion is,

[(a+b+c+Po+{1094=5(p-1)p}p/63)]

--[(a+b+c+(P-9)o+ {1094+5(P-1-9)}

(P-9)/63)]in minutes

==540+1089p/63+5P2/63-1089 (p-9)/63-5(p-9)2/63

==540+1089 (p-9+9)/63+5(P-9+9)2/63-1089 (p-9)/63

-5(p-9)2/63

==540+1089X9/63+5X92/63+90 (p-9)/63

==540+162+10(P-9)/7

== 702+10(p-9)/7.

In the same way for the minus -padas,

[(a+b+c+po+{2414-5(P-1)}P/63)]

--[(a+b+c+(p-9)o+{(2414-5(p-1-9))}(P-9)/63]

( in minutes)

==540+2419P/63-5P2/63-2419 (p-9)/63+5(P-9)2/63

==540+2419 (p-9+9)/63-5(p-9+9)2/63-2419

(p-9)/63+5(P-9)2/63.

==540+2419X9/63-5X92/63-90 (p-9)/63

==540+339-10 (p-9)/7

==879-10(P-9)/7.

It is also possible to establish the connection, by summing the two expressions of III.4, and arriving at the formulae of II.6. For the matter of that, there are reasons to surmise that II.6. was got from III. 4. by summation. The difference from the mean position must have been first noticed. This must have been accounted for by a variation in the motion from a minimum to a maximum and vice versa, postulating the variation to be constant as in III.4. The factor (p-9) must have been introduced because the minimum 702' must be obtained for the first day ending 9 Padas, and so on for the others, (though P-4 1/2 would have done better). Then is why the 0 value of the equation of the centre results for 133 padas instead of 124, and for a half-gati we get 180o 4' instead of the correct 181o32'. Otherwise, it is easy to have given the 0value for 124 padas, and the maximum or minimum for 62 padas, by making the equation of the cntre formulae equal to +(5p-620)Xp/63, and combining these with 27'.843 P. The factor (P-1) in the formulae of II. 6, seems to form a relic of a prior summation, which [Varaha] seems to have retained out of respect for the

original [Siddhanta], for the same result will be got by the more simplified forms, (1089+5P) P/63 and (2419-5P) P/63.

I shall now point out some of the errors committed by T and S. They have seriously gone wrong in their interpretation of the second half of II. 3, and this after making two uncalled for emendations of the Ms. text, (See discussion of the text). In this part, as we have laready seen, we are asked to add minutes equal to twice the Ghanas, and also add the [Mula-Dhruva], (or Ksepa), 1 [Rasi] 7o29'. T and S interpret it as, "Multiply the Ghanas by 2 and divide by 2971. The resulting [Rasis] etc. are to be added". This means, instead of 2' per Ghana we add 1'13'', and we do not add the Ksepa 1[Rasi] 7o29' at all. As for the 1'13'', this is unwarranted when both the Mss. say 2' per Ghana. An emendation is called for only when a quantity given is so far removed from the actual that it is not likely to be the quantity given by the [Siddhanta]. Now the value 2' agrees better with the mean motion (siderial)per Ghana, viz. 110 revolutions 11 [Rasis] 7o32' 15'' for the time of [Varaha]. It agrees almost prefectly with the value of [Siddhanta] [Siromani] of [Bhaskara] II. While it is a matter for wonder and admiration how the ancient [Vasistha] achieved a thing not achieved by most later [Siddhantas], T and S come in and spoil the whole thing by their emendations.

As for the Ksepa which T and S have obliterated, it is essential, as it supplies the [Muladhruva]. This can be seen from the mean moon I give according to different systems for sun-set at [Yavanapura], (Alexandria), i.e., for 37 [nadis], 20[vinadis] after mean sun-rise at [Ujjain], Sunday, close to the end of 'saka 427.

Modern astronomy (this from the Vernal equinox of epoch) 354o48'

[Vasistha] (with the Ksepa), taking Ujjain sunrise for the

[Vasistha epoch) 355o6'

[Vasistha](without the Ksepa, taking Ujjain sunrise

for the [Vasistha] epoch) 317o37'

[Siddhanta Siromani] 355o49'

Romaka 356o12'

Saura (with Ujjain noon for epoch, given) 355o6'

[Vasistha] (with the Ksepa, taking Ujjain

37-20 [nadikas] for epoch for [Vasistha] also) 346o54'

[Vasistha](without the Ksepa, taking Ujjain

37-20 [nadikas] for [Vasistha] epoch also) 309o25'

We see that with the Ksepa in tact [Vasistha] mean moon agrees with the other [Siddhantas] and modern astronomy. With the Ksepa gone, there is a defect of 37o.

Incidentally we must discuss another point here, viz. what time of the day is the [Vasistha] epoch. [Varaha] gives different times of day for different [Siddhantas], and sometimes times even for the same [Siddhanta]. For the Saura sun, moon, moon's apogee and [Rahu], midday at Ujjain (Sunday) is given as the time of epoch, and for the planets, the midnight following. For the [Paitamaha] the time of day of epoch is morning (though the year is different). For the Romaka the epoch time of day is sunset at Yavanapura (Alexandria) which is equal to 37 [nadikas] 20 [vinadikas] from sunrise at Ujjain (Sunday), excepting for the moon's apogee, for which sunset at Ujjain is given as the time. In the case of the Paulisa the time is not mentioned, unless we strain xÉÉ%ÊiÉÊSÉ®äú {ÉÉèʱɶÉÉä%{ªÉä´ÉÆ (I.10.) a bit, and make it mean that the Pauli'sa time of epoch also is sunset at Yavanapura like the Romaka. But we can infer that it is indeed so, from III. 13-15. For the [Vasistha] also the time is not mentioned. Is it sunset at Yavanapura, because it has been given as a general instruction? Or is it sunrise at Ujjain, as in the [Paitamaha]? If it is sunrise at Ujjain, the [Vasistha] moon agrees closely with those of other [Siddhantas] as also modern astronomy. If is sunset at Yavanapura it is defective by about 8o. (If the Ksepa is not taken into account the defect will be about 45o). Now 8o difference is too much, especially when we consider the accuracy of the [Vasistha] moon's mean motion. So the time of day of the epoch for [Vasistha] must be mean sunrise at Ujjain (Sunday). If we emend II.3.into,".....¶ÉʶɨÉxÉÖ xɴɪɨÉMɸ´É®úɶªÉÉtÉ:," then the Ksepa is lr-14o29'. In this case the [Vasistha] moon will be 353o54' at Ujjain 37-20 [Nadikas], which agrees approximately with other [Siddhantas]. So we can take the Ksepa as 1-14o29', and keep sun-set at Yavanapura. But here there is the weight of the emendation.

We shall now continue our main discussion. In their interpretation of II. 4-5 also. T and S have fallen into material errors. In II.4, two terms,

plus-padas and minus-padas are defined, to be used in II.6, which have been missed by them.

In II.5, T's translation wants us to add degrees equal to the number of Padas, only in the case of the Padas left over after deducting 124, i.e., in the case of what amount to the minus-padas only as well. But S's commentary gives the correct interpretation. In the second part of this stanza, T's translation is non-committal. S's interpretation is positively erroneous. He says: "+lÉÇiÉÚ ´ÉänùÉEòɱ{É{Énäù¹ÉÖ @ñhÉÆ, +ÊvÉEäò¹ÉÖ vÉxÉʨÉÊiÉ ¤ÉÖÊrù¨ÉÎnÂù¦É:º´ÉªÉ¨Éä´ÉÉäÁ¨ÉÂ". i.e., if the original padas are less then 124, additive. Evidently he thinks that II.6. gives the equation of the centre, pure and simple, and it must be subtractive for padas less than 124, and additive otherwise. But as I have already explained, there are two rules in II.6, the first requiring the addition of two terms, and the second requiring the addition of two terms, and the second requiring the substraction of one term from another. In the first, Padas less than 124 are to be used. But the result of either rule is to be added to the mean moon so far obtained. There is no question of subtraction here at all.

As for II. 6.T and S have said in so many words they do not understand it, but nevertheless given an interpretation, which naturally is wrong. III. 4, they have not touched.

Now we are in aposition to discuss the readings of II.2-6. and III.4.

In II.2. there are the scriptory errors, ªÉÖHò¶ÉʶÉMÉÖhÉ for ªÉÖHäò¶ÉʶÉMÉÖhÉ, and txÉÉ which T and S have corrected. I have adopted these corrections.

In II.3, the scribal errors txÉ forPÉxÉ, ¿iÉÆ ¶Éä¹ÉÆ for ¿iɶÉä¹ÉÆ, tκjÉMÉÖÊhÉiÉÆ for PÉxÉÉ and ¸´É®úɶÉtÉ for ¸´É ®úɶªÉMÉtÉ:, which T and S have corrected, I have adopted. T and S have corrected ¹ÉÉä½þ¶É, and printed within brackets, and I have adopted this.

Apart from these small errors the Ms. is alright. But T and S have carried out two uncalled for emendations which seriously affect the subject matter, The first is the changing ofEò±ÉÉ into ¡ò±ÉÆ, (when both Mss. give Eò±ÉÉ),simply because if the second emendation is made, Eò±ÉÉ would be troublesome. The second is the conversion of ¶ÉʶɨÉÖÊxÉxɴɪɨÉɸ´É into ¶ÉʶɨÉÖÊxÉxɴɪɨɿiÉɸ´É, when neither Ms. has ¼±ÉiÉ, and the [arya] metre is sinned against by the addition of the two [matras], thrus giving 18 [matras] to teh fourth foot and making it a [Giti], which [Varaha] nowhere uses in his text.

Further they have done this thinking they are improving the text. On the other hand they have spoiled the already correct and better text, and also shut out a necessary data (viz., the [Muladhruva]), as we have shown before.

In II.4, T and S have corrected ÊxÉiÉÉ Eò±ÉÉ: into ÊxÉiÉÉ:Eò±ÉÉ:, which I have also done. Metre requires that {ÉÉnùºÉÆJªÉÉ, must be changed into {ÉnùºÉÆJªÉÉ and {Énù will be a better word also. T and S have failed to notice this, and retained {ÉÉnùºÉÆJªÉÉ. ¹É¹ÉÞƶÉ, (a likely corruption for ¹É¹`öÉƶÉ. T and S have corrected {É®úiÉ:, not understanding the text properly, and {É®úiÉ:may stand.

In III.4, ʱÉ{iÉɸ´ÉiÉÖ¹Eò for ʱÉ{iÉÉSÉiÉÖ¹Eò, ºÉ¨Éɸ´ÉÉƶÉÉ: for ºÉ¨Éɸ´ÉÉƶÉÉ. iɸ´É for iÉ踴É, and

nùxiªÉÆ for näùªÉÆ, are all scribal errors corrected by T and S and adopted by me. In II 6,

¹É¹]õªÉÉä vÉÞiÉä for ¹É¹ÉÞªÉÉänÚùvÉÞiÉä is the only error, and that to scribal. T and S have made the correction and I have taken it.

In III. 4, MÉiªÉxiÉÉvÉÉÇUôvªÉÉä for MÉiªÉxiÉÉPÉÉÇSvÉÉäPªÉÉä, is a scribal error corrected by T and S

and adopted by me. As for ºÉÉθ´ÉºÉÉÆ´É®úÉä, corrected by T and S into ºÉÉθ´ÉJÉÉSɱÉÉ, ºÉÉθ´ÉJÉÉSɱÉÉä is a better correction, as retaining the +Éä at the end. The masculine gender that results is not ungrammatical, for making the word qualify ºÉi{ÉÉƶÉ: is better than making it qualify ¦ÉÖÊHò:. An alternative correction would be ºÉÉθ´É´ÉJ´É®úÉä, which seems to be better still as it retains ´É®úÉä (ºÉÉä is given by both Mss.) instead of making it SɱÉÉä, and I have adopted this. All three readings give 702 which is required in the formula. xÉMÉÉiÉÚ and ´ÉºÉÖ¨ÉÖÊxÉxɴɦ´ÉÆ: being apparently correct, T and S have not touched it, which is what all they could have done, because they have refrained from interpreting the stanza, saying the meaning is obscure. But both need correction. First we shall take up xÉMÉiÉ. A [matra] is wanting in the foot. I have shown the rules here are derivable from II.6. The rules require "Padas minus nine" to be used. So I have made xÉMÉÉiÉÚ {ÉnùÉiÉÚ into[Ê´É]xÉ´ÉÉiÉÚ {ÉnùÉiÉÚ, which gives thr required meaning, supplying the wanting [matra]. As for ´ÉºÉÖ¨ÉÖÊxÉxɴɦªÉ: the second rule requires 879, and the numbers are given in the wrong order by the manuscript text. So I have emended it into xɴɨÉÖÊxɴɺÉÖ¦ªÉ:

Tithi and Naksatra

II.7

¶É¶ÉÉnù±ÉÆ ÊjÉEÞòÊiÉ´PɨÉÞ- ¶É¶ÉªÉvÉänù±ÉÆ ÊjÉEÞòÊiÉ-

IɨÉƶÉκUôiÉÉ ¨ÉÖ½þkÉÉÇ ºªÉÖ :* PxɨÉÞIɨÉƶÉκªÉiÉÉ ¨ÉÖ½ÚþiÉÉÇ ºªÉÖ :*

´ªÉ EèòxnÖùnù±ÉÆ Ê´É¹ÉªÉÉä- ´ªÉEèòxnÖùnù±ÉÆ Ê´É¹ÉªÉÉ-

½þiÉÆ ÊiÉÊlɺiÉnÚùnäù´ÉÉäHò:** ½þiÉÆ ÊiÉÊlɺiÉnÚùnäù´ÉÉäHò:**

III. 7. The quarter part of the true moon multiplied by 9 gives the [Naksatra] (in the [Rasi] column), and what is in the degree column are the [Muhurtas]. Half the sun minus moon, multiplied by 5 gives the Tithis, in the same manner, (i.e., in the [Rasi] column, and 30ths of Tithis in the degree column).

The rules are simple and based on the fact that there are 2 1/4 [Naksatras] for a [Rasi] and 2 1/4 Tithis. Two things are noteworthy here. The word [Muhurta] now only a measure of time, (through the time of staying of the moon in a [Naksatra] being used as a measure of time), is used for the 30th of a [Naksatra] segment. Even as applied to teh 30th part of the duration of a [Naksatra], it is different from the period of 2 [Nadikas], in which sense it is usually used.

Now for the text; The scriptory errors ¶ÉκUôiÉÉ for ¶ÉκUôiÉÉ and ¨ÉÖ½ÚþkÉÇ ºªÉÖ : for ¨ÉÖ½ÚþiÉÉǺºªÉÖ :, are corrected by T and S and I have adopted their correction. ¶É¶ÉÉnù±ÉÆ in the manuscipt is corrupt and has been emended into ¶É¶ªÉvÉänù±ÉÆ by T and S because the

rule requires it, and two [matras] are wanting. Also, the foot should end with ÊjÉEÞòÊiÉ. I have adopted their emendation. ®úɶªÉvÉänù±ÉÆ would be an alternative emendation, which would extend the scope of the [Naksatra] rule to the sun as well.

Daylight

II.8

¨ÉEò®úÉnùÉè MÉÖhɪÉÖHòÉä ¨ÉEò®úÉnùÉè MÉÖhɪÉÖHòÉä

¨ÉäJÉÉnùÉè ÊiÉÊlɪÉÖiÉÉä ®ú´ÉäÌnùºÉ:* ¨Éä¹ÉÉnùÉè ÊiÉÊlɪÉÖiÉÉä ®úÊ´ÉÌnù´ÉºÉ:*

EòEÇò]õEòÉÊnù¹ÉÖ ºÉiºÉÖ EòEÇò]õEòÉÊnù¹ÉÖ ¹É]ÚõºÉÖ

jɪÉκjÉEòÉ :¶É´ÉǺÉÒ¨ÉÉxɨÉÚ** jɪÉκjÉEòÉ:¶É´ÉǺÉÒ¨ÉÉxɨÉÚ**

II. 8. When the sun is in Makara etc, (upto Mesa), the sun (in Rasis) plus 3 is the duration of daytime (in Muhurtas). When the sun is from Mesa etc. (upto Karkata), the sun plus 15is the duration of daytime, When the sun is in the 6[Rasis]from Karkata, the sun plus 9, is the measure of the night-time.

The rule is easy to understand. According to this [Siddhanta] the shortest daytime is 12 [Muhurtas] (i.e., 24 Nadikas), which occurs when the sun is at the beginning of Makara. The longest day is 18 [Muhurtas] with sun at the ebeginning of Karkata. Between Makara and Karkata, there is increase in daylight in [Muhurtas] equal to the motion of the sun in [Rasis], and from Karkata to Makara, there is increase of the night-time in the same manner.

It is to be noted that the rule gives the same result as the [Paitamaha] [siddhanta] and the [Vedanga] Jyautisa, with this difference that here the variation of daylight is with the true sun. It is also to be noted that really the daylight increses or decrease not with the sun as given here, but with the tangent of the declination of the sun. The maximum-minimum daylight gives points to the extreme north of India, lat.34o45'.

Now, the meaning easily follows with two small changes made in the text, ®úÊ´É:for ®ú´Éä: and ¹ÉnÚùºÉÖ for ºÉiºÉÖ (which latter T and S have also made). But not understanding the Stanza (or perhaps finding it disagreeing with II.9-10) T and S have made a drastic change in the text, viz. ¦ÉÚº´ÉMÉÇÊiÉÊlÉʨÉiÉÉä for ¨Éä¹ÉÉnùÉè ÊiÉÊlɪÉÖiÉÉä. The letters of ¦ÉÚº´ÉMÉÇÊiÉ are quite unrelated with the letters in ¨Éä¹ÉÉnùÉè. I cannot understand how ¦ÉÚºjÉMÉÇÊiÉÊiÉ欃 means 1591, i.e., how º´ÉMÉÇÊiÉ means 9. ( It is not a misprint, for the commentary also uses the expression.) Again jɪÉκjÉEòÉ: is taken to mean three by three, while it can only mean three threes, i.e.,9. Even if we allow these things, the meaning givenby them is self-condemnatory. At the rate of 3 palas a day the measure of the day is 1865 palas when the sun is at the beginning of mesa, and 2139 when the sun is at the beginning of Karkata, which means the night is 1461 palas; and this disagrees with 1591, because the duration of the day at [Makaradi] is the duration of the night at

[Karkatadi]. The equinoxes according to this interpretation would occur 22

days before the sun reaches Mesa and 22 days after he leaves [Kanya] i.e., from Vernal equinox to Autumnal equinox, there would be 226 days and from Autumnal equinox to Vernal equinox, there would be 139 days!

I have incidentally discussed the readings here. One remains,¨ÉäJÉÉnùÉè, which I have changed into ¨Éä¹ÉÉnùÉä,as being written for ¹É is common in certain parts of India.

The noon-day shadow and the sun

II.9-10

9.EòEÇò]õÉÊnù¹ÉÖ ¦ÉÖHÆò 9.EòEÇò]õEòÉÊnù¹ÉÖ¦ÉHÆò

ÊnùMÉÖhÉÆ ¨ÉvªÉÆÊnùxÉҦɴÉäUôɪÉÉ * ÊnùMÉÖhÉÆ ¨ÉÉvªÉÎxnùxÉÒ ¦É´ÉäSUôɪÉÉ *

¨ÉEò®úÉÊnù¹ÉÖSÉÉ{ªÉä´É ¨ÉEò®úÉÊnù¹ÉÖSÉÉ{ªÉä´ÉÆ

ËEòSÉκ¨É¨É{ɽþ±ÉÉ UäôvªÉÆ ** ÊEòxi´Éκ¨Éx¨É{½þ±ÉÉSUôÉävªÉ¨ÉÚ **

10. ¨ÉvªÉɽþUôɪÉÉvÉÇ 10.¨ÉvªÉɽþxUôɪÉÉvÉÇ

ºÉÊjɨɨÉEòÉè%ªÉxÉä ¦É´Éä tɨªÉä* ºÉÊjɦɨÉEòÉæ%ªÉxÉä ¦É´Éät ¨ªÉä *

=nùMÉ ªÉxÉä ºÉƶÉÉävªÉÆ =nùMɪÉxÉä ºÉƶÉÉävªÉÆ

{ÉS\Énù¶É¦ÉªÉÉä ®úʴɦÉÇ´ÉÊiÉ ** {ÉS\Énù¶É¦ÉªÉÉä ®úʴɦÉÇ´ÉÊiÉ *

II. 9. when (the Sun is ) in the (6) [Rasis] beginning with Karkata, the number of [Rasis] passed beginning with Karkata, multiplied by 2, is the noon-shadow (in Angulis). The same (rule should be followed) when (the Sun is ) in the (6) [Rasis] from Makara. But in this case the result is to be deducted from 12[Rasis] (to get the shadow).

II. 10. when the Sun is in the south-ward course, (i.e. inth e6 [Rasis] from Karkata), half the noon-shadow plus three [Rasis] is the Sun (in [Rasis] etc.) When he is in the north-ward course, (i.e. in the 6 [Rasis] from Makara) half the noon-shadow is to be subtracted from 15 and the Sun is got.

It is supposed that the shadow increases proportionately with the Sun from a starting value zero at the beginning of Karkata. At the beginning of Makara the maximum,12 [angulis], is reached. Then it decreases as it increases, and comes back to zero at the beginning of Karkata. The length of the gnomon is not mentioned, but most probably it is 12 [angulis], commonly given in Hindu astronomy. Zero at the beginning of Karkata would give about 24oLat. to the region where the rule holds, and 12

angulis maximum would give 21o. Thus there is contradiction within the stanza itself. This contradiction may be resolved by taking the gnomon as 11 angulis, but this is not likely. Neither of these agrees with Lat. 34o45'got from the maximum and minimum daylight. The uniform variation is wrong, which means Therefore, the rules must have been taken from different sources and pieced together, and the remark of Brahmagupta, "BiÉÉxªÉä´É MÉÞ½þÒi´ÉÉ ´ÉÉʺɹ`öÉä ʴɹhÉÖ´ÉxnäùhÉ (EÞòiÉ:EòxlÉÉ) seems to hold here, II.10 is

the reverse of II.9.

As for the readings, EòEÇò]õÉÊnù¹ÉÖ for EòEÇò]õEòÉÊnù¹ÉÖ, ¨ÉvªÉÆÊnùxÉÒ for ¨ÉÉvªÉÎxnùxÉÒ, ¦É´ÉäUôɪÉÉ for ¦É´ÉäSUôªÉÉ, ´ÉÉ{ªÉä´É, for SÉ.{ªÉä´ÉÆ, ¨É{½þ±ÉÉUôÉävªÉÆ for¨É{½þ±ÉÉvUôÉäªÉÆ, are scribal errors corrected by T and S, and adopted by me. T and S have corrected ÊEòSÉκ¨ÉxÉÚ intoËEò´ÉÉʺɨÉxÉÚ, while I have made it ÊEòxi´Éκ¨ÉxÉÚ, because iÉÖ is better for the meaining than SÉ, and the likelihood is greater. All this in II.9. There is no error in II.10, worth the name.

Lagna from shadow and shadow from Lagna

II. 11-13

11.uùɶÉʦÉ:ºÉUô ªÉè. 11.uùÉnù¶ÉʦÉ:ºÉSUôɪÉÉ-

¨ÉÇvªÉɽþÉä xÉè¦ÉÇVÉäuùºÉVÉiÉɶÉÆ * ¦ÉÇvªÉɽþÉäxÉè¦ÉÇVÉäºÉ½þiÉɶɨÉ *

+{É®úɽäþ ´ÉxuùÉuùÉÇ- +{É®úɽþÉä SÉEòÉPÉÉÇ-

Êuù¶ÉÉävªÉ ºÉ,EÆò ¦É´ÉÊiÉ ±ÉÉiÉÆ** Êuù¶ÉÉävªÉ ªÉÉEÇò ¦É´ÉÊiÉ ±ÉÉiɨÉ **

12.´ªÉEèò ±ÉMxÉä ʱÉ{iÉ : 12.´ªÉEèò ±ÉMxÉä ʱÉ{iÉÉ :

|ÉÉCªÉ¸´ÉÉ UôÉäÊvÉiÉɺiÉÖ SÉEòÉuùÉÇiÉÂ* |ÉEÚò{ɸÉÉSUôÉäÊPÉiÉɺiÉÖ SÉHòÉvÉÉÇiÉÂ*

EòɪÉUäônù:¶ÉÚxªÉÉÆ EòɪÉǶUäônù:¶ÉÚxªÉÉ-

´É®úɹÉÞ ±É´ÉhÉÉänù¹É½þEòÉxÉÉÆ ** ¨¤É®úɹ]õ±É´ÉhÉÉänù¹É]Úõ EòÉxÉɨÉÚ **

13. ±Év´ÉÆ uùÉnù¶É½þÒxÉÆ 13. ±É¤PÉÆ uùÉnù¶É½þÒ´ÉÆ

¨vɪÉɽþUôɪɪÉÉ ºÉ¨ÉɪÉÖHÆò * ¨ÉvªÉɽþSUôɪɪÉÉ ºÉ¨ÉɪÉÖHò¨É *

ºÉÉ Ê´ÉYÉäªÉÉ UôɪÉÉ ºÉÉ Ê´ÉYÉäªÉÉ UôɪÉÉ

´ÉÉʺɹ]õºÉ¨ÉɺÉʺÉuùÉxiÉä ** ´ÉÉʺɹ`öºÉ¨ÉɺÉʺÉuùÉxiÉä **

11. 11. The shadow at any moment is to be added to 12 and the noon shadow (of date) is to be deducted from that. By this, 36 is to be divided. This result is to be added to the Sun in [Rasis] to get the Lagna, (i.e. Orient Ecliptic Point) (if it is forenoon). If afternoon, the result is to be substracted from 6 [Rasis] adn this added to the Sun to get the Lagna.

11.12-13. The sun is to be deducted from the Lagna and the reminder converted into minutes of arc, if it is forenoon. If afternoon, the remainder should be deducted from 6 [Rasis] and then converted into minutes. 64800 is to be divided by the minutes. The result is to be added to the noon-shadow (of date) and 12 deducted from this. This is the shadow (at the time of the given Lagna).

This according to the brief [Vasistha Siddhanta].

Rule II.11. can be explained as follows. (There is no question of proving it because it is rough and has been constructed empirically, its justification being its serviceability). As in the case of the rough rules current to find the time of the day from the shadow, and because Lagna follows the time roughly, the equation for Lagna, elapsed after sunrise, or to elapse till sun-set, can be expressed in the form, a/(s+b)==L, where a and b are constants, and s is a function of the shadow. If as a first approximation we take the latitude of the place as zero, and the ecliptic as coinciding with the celestial

equator, it follows that there is a rise of 3 lagnas from sun-rise to noon, and 3 more from noon to sunset, that at noon, when L=3, the shadow is zero, and thatmidway between sunrise and noon L is 1 1/2, and the shadow then is 12Xtan 45o,i.e. 12. If really there is no such uniform rise of Lagna, and noon does not correspond to rise of 3 lagnas, but a little more or less, it is because the latitude may not be zero, and the ecliptic does not coincide with the celestial equator. These would also cause shadow at noon where otherwise there would be no shadow, and make the shadow at other times differ correspondingly. Now it is supposed that by deducting the noon-shadow from the shadow the first mentioned ideal condition is established, (because then s would become zero for noon, and the difference in the shadow at other times would be roughly righted)> So it is thought that by taking the shadow minus noon-shadow instead of the shadow itself, everything would be all right (though strictly they would not be). So we may write "shadow minus noon-shadow" for s. As its value, as we have already shown, is 0 for L==3, and 12 for L==1 1/2, we have the two equations, a/(0+b)=3, a/(12+b)=1 1/2, to determine a and b; from which we obtain a=36, and b=12. Thus we get the expression 36/(shadow minus noon-shadow+12) as the value of the Lagna elapsed from sun-rise in

the forenoon, and the Lagna to elapsed till sunset in the afternoon. As at sun-rise, the longitude of the sun is the Lagna, this is to be added to the sun. As at sun-set the longitude of the sun plus 6 [Rasis] is the Lagna, the result is deducted from the sun plus 6 [Rasis], or which is the same, deducted from 6 [Rasis] and the sun added. Thus the rule is explained. T's statement in his explanation, "In order to establish a workable proportion... 12 is added to the first and to the third members..." makes the rule appear more arbitrary than necessary.

The rule II. 12-13. is the inverse operation of the rule II.11. and that explains it. But because the divisor which is in [Risis] in II.11, is converted into minutes here, the dividend, 36, is multiplied by the number of minutes in a [Rasi] and given as 64800.

One thing is to be noted. Except for finding when an auspicious lagna begins or ends by taking the shadow, II.12-13 can only serve as a mathematical exercise as it is, for no rule has been given by the [Siddhanta] for the Lagna other then by using the shadow, and if the Lagna is got from the shadow, what is the meaning of getting the shadow back from the Lagna? At night the Lagna can be got by observation, but then there will be no shdow caused by the sun, to which alone the rule applies.

The concluding word,´ÉÉʺɹ`öºÉ¨ÉɺÉʺÉuùÉxiÉä, though forming a part of the last sentence giving the rule II.12-13 may be taken to mean the whole of chapter II, and even III.4, which we have shown to be [Vasistha's].

Now for the readings: the scribal errors uùɶÉʦÉ: for uùÉnù¶ÉʦÉ:, ºÉUôɪÉè:, ®úºÉVÉiÉɶÉÆ SÉxrùÉiÉÚ for

SÉGòÉvÉÉÇiÉÚ in II.11, |ÉÉCªÉ¸´É UôÉäÊPÉiÉɺjÉ for |ÉÉEÚò{ɸ´ÉÉSUôÉäÊPÉiÉɺiÉÖ, EòɨÉUäônù: for EòɪÉǶUäônù: ¶ÉÚxªÉÉÆ´É®úɹÉÞ for ¶ÉÚxªÉɨ¤É®úɹÉÞ in 11.12, ±Év´ÉÆ for ±É¤vÉÆ, ´ÉÉʺɹÉÞ for ´ÉÉʺɹ`ö in II.13, have been corrected by T andS, which I have also adopted .

The Place of the [Vasistha] in the hostory in the history of Hindu Astronomy

The [Vasistha] marks a stage of development in Hindu Astronomy which is intermediate between those represented by the [Paitamaha] (p) and the Saura siddhantas condensed by [Varaha]. The system we find in the P

is the same as that we find in the [Vedanga] Jyotisa the P is the same as that we find in the [Vedanga] Jyotisa (V.J). The five-year-Yuga, the beginning of the yuga from [Magha] [Sukla], the treatment of the Sun and the moon alone, the absence of the equation of the centre, so that all reckoning is done using only the mean sun and the mean moon, the maximum and minimum duration of daylight and its uniform increase and decrease, are all found in both. Only the methods are different, the V.J.giving the ending moments of Tithis, lunar and solar, Naksatras etc., by an ingenious method which avoids computation using large numbers, while the P gives the [Ahargana] first,

and uses this for computation, like the other [Siddhantas]. The Saura, on the other hand, represents the fully developed stage of Hindu Astronomy, extending the field of operations to the planets as well, using epicycles for the equation of the cnetre and the equation of conjunction, and using spherical trigonometry to solve various problems. In between comes the [Vasistha]. Methods for the true sun and moon are given.

Though the maximum and minimum daylight and its uniform increse and decrease is the same as in p, the [Vasistha] deals with other topics also like computing the noon-shadow from the sun and vice versa, and the Lagna from the shadow and vice versa.

The [Vasistha] gives a solar year of 3651/4 days, which is fairly correct. As for the true sun, the method is crude, and is valuable only as indicating a distinction made between mean and true astronomical quantities, and the ability to express the same. periods of 31 1/4, 31 1/2, 31 1/2, 3 1/2. 31, 30 1/2, 29 3/4, 29 1/4, 29 1/4, 29 1/2, 30, 301/4 days are given for the sun to traverse the [Rasis] Mesa etc., from which we can roughtly say that the slowest motion is in the middle of Mithuna, and the quickest in the middle of Dhanus. All these are fairly accurate for the time of the [Siddhanta]. Only, the days for Mesa should be 31 and for Mina 30 1/2. Even here the error is less than 1/8 day, though it appears to be 1/4 day.

The [Vasistha] must have obtained these values empirically, most probably by an analysis of eclipses. It cannot be that it is [varaha] that gives the values in this empirical form, computing them from the equation of the centre given by the original [Vasistha], for [Varaha] faithfully gives not merely

the values, but also the method of the original. Or else why should he give different methods for the different [Siddhantas]? In the case of the [Vasistha] he gives the sun in periods of days,

correct to quarter days. In the case of the [Paulisa] he gives a correction in minutes to the mean sun, the correction depending upon [Rasis]of "Anomaly", the term anomaly here being used in a peculiar sense. In the case of the Romaka he gives the correction to the Sun in the proper form of the equation of the centre, "K sine anomaly". In the case of the Saura he gives the regular epicyclic theory. It is also reasonable to supoose that before the epicyclic theory was formed, the previous three stages, just mentioned were gone through in the development of Hindu Astronomy.

As regards the moon, we have seen that the mean motion of the moon according to [Vasistha] is more accurate than that of other [Siddhantas], as also its maximum equation of the centre, and its maximum-minimum daily true motion. This accuracy is possible only if observations of the moon with reference to the stars are made, instead of merely depending on anylysis of eclipses, whcih is also necessary. Analysis of eclipses can give the mean motion correctly, but the maximum or minimum equation of the centre got would be only about=301', considerable less than the true values +_ 377', because at syzygies where alone eclipses occur, the second inequality called the [evection] is reduced to the same form as the equation of the centre with the opposite sign, and a maximum numerical value 76', and thus reduces 377' to 301'. It should not be argued that such accuracy is against [Varaha's] statement in I.4. There he compares only the Tithis of different [Siddhantas] (because that is the most essential thing required, and one of the few things that are comparable), and [Vasistha]-Tithi is spoiled not by the moon but by the sun got through using the incorrect mean solar year of 365 1/4 days' and also by the roughness of the rules in II.6. giving the moon's equation of the centre. As already explained (vide above) these rules must have been based on observation of the daily motions, and the theory that they increase or decrease uniformly. Thus the equation of the centre was expressed first in an algebraic form as in the [Vasistha]. Then the trigonometrical expression, "K sine anomaly," was discovered and used as agreeing better with observation. After this the epicycle theory was

propounded, for, everywhere in science theory comes later to explain observed facts. Thus the [Vasistha] is historically important as marking as stage of the development of Hindu Astronomy, where we see the first dawn of the true notion of heavenly bodies.

THE BIJOPANAYA : IS IT A WORK OF BHASKARACARYA?

(REprinted from J.O.I., M.S.U., Baroda, June 1959)

Introduction

The [Bijopanaya] is a short work on Indian astronomy which enunciates two corrections, to be applied to the value of the longitude of the Moon got by applying the usual Equation of the Centre, to make it more accurate. This work with a commentary called [Vasanabhasya] was first publised in 1876 by two pioneers in the field of Drk-almanacs is South India, the late [Chintamani] [Raghunathacharya] of [Nungambakkam] and

[Tadhakamalla] [Venkatakrishna Raya] of [Triplicane]. A short work called [Tithinirnayakarika] by [Srinivasa] [Yajvan] is also found added as an Appendix to this edition. Fifty years later, in 1926, another edition with an Introdution was brought out by Dr.Ekendranath Ghosh. Neither edition speaks anything about the manuscripts used by the editors and it is not definitely known whether the 1926 edition was based on independent manuscript material or only on the older edition, which is not improbable especially in view of the fact that here too the [Tithinirnayakarika] appears as an Appendix. In both editions the text and the commentary are ascribed to the famous Indian astronomer [Bhaskaracarya] II of the 12th cent., author of [Siddhanta-siromani], as stated in the colophons of the editions. But the work seems to be a much later production and an analytical examination tends to show that it cannot be a work of [Bhaskaracarya], and that it is fathered upon him by its later author. It is intended to study the question from this light.

In passing, it may be ntoed that the present work, [Bijopanaya], seems to be little-known, and no Indian astronomer or commentator who followed [Bhaskara] mentions this work or quotes from it, the only exception being the [Tithinirnayakarika] fo [Srinivasa] Yajvan, which is given as an Appendix to both the editions of the [Bijopanaya]. Also modern historians of Indian astronomy like MM. Sudhakara Dvivedi and S.B. Dikshit

have not noticed this work in their histories.

The two corections enunciated in the [Bijopanaya] are asked to be applied after the usual Equation of the Centre has been applied. Of these, the first makes up for the deficiency in the Equation of the Centre when compared with the correct one, and the second is the equivalent of the Inequality called Variation which, when applied to the Moon will take it nearer to its true position in longitude.

A brief recapitulation of the history of the discovery of the three principal Lunar Inequalities will help in the proper evaluation of the [Bijopanaya]. In the West, the first Inequality, viz. the Equation of the Centre, was discovered by the Greek astronomer Hipparchus in c. 140 B.C. Ptolomy of Egypt (c. 140 A.D.) discovered the second Inequality known as Evection. Fourteen centuries later, in c.1580 Tycho Brahe, a Danish astronomer, discovered the thrid Inequality viz. Variation. In India, the ancient [Vasistha Siddhanta] is the first known work that takes into account the Equation of the Centre in the motion of the Sun, the Moon and the Star planets (Taragrahas), and gives its own peculiar formulae thereof, consisting of the summation of an Arithmetical series. The next ancient [Paulisa Siddhanta] does not improve much on this. But the Romaka and the Saura [Siddhantas] that followed, as also the later [Siddhantas] use the usual form, a sin0. [Munjala] (932 A.D.) is the first Indian astronomer--I have read since that Vastesvara (c.904 A.D.) earlier than [Munjala] has given this --who gives the Evection, in his [Laghumanasa],[Prakirnadhyaya], 1-2. What he gives is equivalent to -65.3' sin l +65.3' sin (l 2D), the first term compensating to a great extent for the deficiency in the Equation of the Centre of Hindu astronomy, and the second term forming the Evection proper. (To be exact, the coefficients should be 76' each) Next to [Munjala], [Sripati] (c.1000) in his [Siddhantasekhara], ch.Xl. [Siddhantaraja] gives the

Variation, calling it by the significant name [Paksika], i.e., one having a period equal to a [paksa] or 15 tithis. If the [Bijopanaya] is [Bhaskara's] work, asit purports to be, then he should get the palm for being the first, (and so early as 1151 A.D.), to give the Variation; excepting perhaps an Arabian astronomer by name Muhammad Abul Wefa, who Prof. Godfrey in the Introduction to his Lunar Theory says, "observed at Baghded in 975

A.D., and discovered a thrid inequality called Muhazal". But Iam afraid [Bhaskara] is not the author of the [Bijopanaya] and the Hindus cannot have the credit for this. before proceeding to show this, I shall explain the passage in the [Bijopanaya] giving the two corrections.

The Two Corrections in the [Bijopanaya]

Of the 58 s'lokas in the [Bijopanaya] only 13, viz.20-32, deal with the two corrections, the rest being devoted to various other matters like the praiseworthiness of the Tithis got after applying the corrections. The following are these verses:

20. The greatest subtractive difference between the calculated and observed positions of the Moon, occurs when the Moon is 90o forward from its Apogee and the Sun is 45o forward from the Moon.

21. The greatest additive difference occurs when the Sun is 45o behind the Moon which is situated 270o from the Apogee.

22. There is no difference when the Sun and the Moon or either is at the Moon's Apogee

or Perigee.

23. When the sun and the Moon are situated at 90o or 270o from the Apogee the difference

is--34'.

24. When the Moon is at Apogee and the Sun is at 45o from it, either behind or before,

there is a difference of 34'.

25.a When the Moon is at its Perigee with the Sun at 45o, before or behind it, the difference

is 34'.

25b-28 Thus by means of repeated observations after varying the situations, I have determined the following Periodic corrections; may learned people examine these: For 3o 45', 7o 30', 11o15' etc. of Anomaly there are the following quantities of correction: 6,13.21,27,33,39,45,51,56,61,65,68,70,72,74,75,75,76,76,77,77,78,78,78, all in minutes of arc. These are additive or subtractive; and must be combined with it (and applied to the Mean Moon to get the True Moon).

29-32. This True Moon is then to be subtracted from the True Sun. I fthe remainder is in an odd quadrant the following corrections are subtractive, otherwise additive. For 3o15', 7o30, 11o15' etc. of the remainder, the following corrections are given : 5,9,13,17,22,24,27,30,32,33,34,34,34,33,31,29,26,24,20,16,11,8,3,0, in minutes. These are to be applied to the already corrected Moon and we get the correct True Moon. This is to be used for getting Naksatras Tithis, etc. for ceremonial purposes.

It may be noted that only six and a half s'lokas, 26b to 32, give the actual corrections

with the instructions for applying them.

It can be readily seen that the second set of quantities for correction is approcimately the Eqation of Variation given by modern astronomy, viz.+39.5' sin 2D, where D is the Mean Moon minus the Mean Sun.

As for the first set of quantities, each of them can be resolved into two parts thus: (5.1+.9), (10.1+2.9), (15.2+5.8), (20.2+6.8), (25.1+7.9), (29.8+9.2), (34.4+10.6), (39+12), (43.3+12.7), (47.5+13.5), (51.4+13.6), (55.2+12.8), (58.7+11.3), (61.8+10.2), (64.9+9.1), (67.6+7.4), (70+5), (72.1+3.9), (73.9+2.1), (75.4+1.6). (76.5+.5),(77.3+.7), (77.8+.2), (78+0).

The first parts are equivalent to 78'sin l, and as these have the same sign as the Equation of the Centre given in the [Siddhantas], viz. about-302' sin l, both combined becomes about -380' sin l, which is very nearly the correct Eqation of the Centre according to modern astronomy. The second parts are roughly given by (30'-30 sin l) sin l+5' sin 21. Of these, (30'-30' sin l) sinl,it can be seen, is only the part of the Equation which is in excess due to the difference in the radii of the epicycle at the ends of the even and odd quadrants, given by many Hindu astronomical works. As these corrections purport to be applied after the [Equation of the Centre of the Siddhanta Siromani] has been used, the addition of (30'-30 sin l) sin l (together with -78' sin l) would mean that the [Siddhanta Siromani] wants us to take the Epicycle at even quadrants to be 42 54' and at odd quadrants to be 39o36' as given by it. As for the worth of this part of the correction, if it improves the values a little in two quadrants, it will make them worse in the other two quadrants. Taking the part, (5'sin 2 l), it may seem to be an

improvement because it appears to be a part of the term (+13' sin 2 l) fo the Equation of the cEntre of modern astronomy. But this correction is not an improvement, because this forms part of the first correction which has been mentioned to have the same sign as the Equation of the Centre (corresponding to the principal term thereof of modern astronomy); and as such this is an improvement in the second and thrid quadrants, but will make matters worse in the first and fourth quadrants. What is to be specially noted is that the first correction does not give the Evection at all, and Dr.Ghosh is wrong when he says this gives the Evection. It only corrects the Hindu Equation of the Centre which is defective by about (-76' sin l), and makes it nearly equal to the correct Equation of the centre. As for the (30'-30 sin l) sin 1+5' sin 21, this is worse than useless in certain quadrants as I have already mentioned. Bearing all this in mind we shall now discuss the authorship of the work.

[Bhaskaracarya not the Author of the Bijopanaya]

The [Bijopanaya] itself mentions [Bhaskara] as its author. Sloka 6 says: "I was born in 1036 Saka. I discovered the two [Bijas] when I was 37 years of age." The first part of the sloka giving the year of his birth is the same as is found in the Sid.Siromani, verbatim; the second part leads us to infer that he discovered the [Bijas] the next year after he wrote his Sid. Siromani, for he says in the latter work that he composed it in his thirtysixth year (Sid. Sir., Gola, Prasnadhyaya.58). In slokas 2 and 3 of the [Bijopanaya], the author says: "I wrote the [Siddhanta Siromani] following the ancient texts alone. But this is not sufficient to give

the correct positions of the planets in order to find the auspicious moments for the different rites enjoined by the [sastras]. So I am writing the [Bijopanaya]". The idea of the first half of verse 3 is a repetition from the Sid. Siromani, Ganita, Spasta 1. From the above statements and teh colophon one would think that [Bhaskara] is the author of the work. But it can be shown that [Bhaskara] cannot be the author, and consequently all the above statements are falsehoods and the author is an impostor.

The following are the reasons whu [Bhaskara] cannot be the author. (1) Without the first [Bija] (the second [Bija] may be left out of consideration here, because it vanishes under the circumstances we are discussing here, viz., the syzygies i.e, conjuctions and oppositions) the Moon calculated by

Hindu astronomy (and, of course, by the Sid. Siromani also ) willo give the syzygies almost correctly. The application of the [Bija] will spoil this correctness, with the result that an error upto more than +_ 6 [nadikas] will be introduced into the times of the syzygies. This error at syzygies is proportionate to the Hindu Equation of the Centre,-302'sin l, and therefore has the periodicity of the Moon's apogee. As a result of this error the middle of the eclipses, as calculated, will be later or earlier thatn what they actually are, by an amount of time equal to this error; and the beginning and end of the eclipses will also be affected accordingly. Now, are we to think that [Bhaskara] gives this [Bija] only to introduce this enormous error into his otherwise correct calculation of eclipses? And [Bhjaskara] of all persons? It is by observations made at the times of eclipses that the ancient astronomers have discovered or corrected the mean periods of the Sun and the Moon and their Equation of the Centre, and got such accurate values. Even after a lapse of centuries now, the Hindu Mean Sun and Moon are fairly accurate. As for the Hindu Equation of the Centre, it is what the three major inequalities of modern astronomy reduce to, at syzygies. D being nearly equal to Oo or 180o, the Variation vanishes as already mentioned. For the same reason, the Evection, + 76' sin (l-2D) becomes + 76' sin l, and combined with the correct Equation of the Centre, (-377' sin l) becomes (-301' sin l). This is the reason why the Hindus did not discover the Evection for a long time, i.e. until [Munjala](Vatesvara earlier) discovered it, and after him [Sripati]. Thus the foisting of this [Bija] upon [Bhaskara] means that prince among Hindu astronomers did not make even the usual observations made by Hindu astronomers, that he did not notice the error of upto 6 [nadikas] in the circumstances of eclipses, and that too introduced by himself, an error which any ordinary person, not to speak of astronomers, will be forced to observe, in India eclipses being occasions of important religious rites. So [Bhaskara] cannot have given this [Bija], especially when [Munjala] and [Sripati] had shown him the correct form of the [Bija] which would lead to no such error as above.

(2) There is discrepancy between the observation values given in slokas 20-25 and the [Bijas] given in slokas 26-32 of the work. We shall see whether [Bhaskara] would have allowed this to get into his work, if it is his. In 20-25 the author takes six configuration of the Sun and the Moon

with respect to the Moon's Apogee, and gives the difference between the observed longitude of the Moon and the longitude got by the [Sastra] without the [Bijas].

They are given below in a tabular form together with certain other data that will be useful.

To prove our point, we shall take configuration iv. Here the first correction is= 78' according as the Moon is 90o forward or behind from the Apogee. The second correction is practically 0. So according to the [Bijas] given, the difference should be = 78'. But -34' has been given as observed. If the author had observed first and based the [Bijas] on the observation, how could he construct the [Bijas] in such a way that they would disagree by 44' ( if the Moon is 90o forward from Apogee) or even 112' (if the Moon is 90o behind the Apogee). This means that the author is an extremely ignorant person. Or if the author has taken the [Bijas] from elsewhere (which is very probable) and has 'cooked' the values for the different configurations, this means he cannot even 'cook' properly. One can never expect this of [Bhaskara], that master among astronomers. And it is asserted that he wrote this work after writing the [Siddhanta Siromani], his masterpiece! Configuration vi too shows the above.

(3) I shall now justify the suggestion that the author had probable cooked the values at the configurations and had not made the observations, in spite of his assertaion to have done so in s'loka 25. In configuration iv above, he should have observed zero difference, but he says he had observed-34 difference. There is an error of observation of 34'. In v and vi he gives an observed difference of 34': he does not say whether it is plus or minus. Let us examine v. If the Sun is at 45o, the observed difference should be +39', which we may take as agreeing with the +34' given. But in this case the [Bijas] give -34'. Thus there is discrepancy amounting to 73' between his [Bijas] and observation. If the Sun is at 315o, the observed difference should be -39' which we can take to have been observed by him as -34'. But in this case the [Bijas] give +34' difference, and again there is the same amount of discrepancy between his observation value and [Bijas]. Thus we are driven to one of two alternatives : either the author cannot see the disagreement between his observation and his [Bijas], or there is a discrepancy of 73' between his observed values and the actual values he

should have observed , i.e., there is an error of observation of 73'. In vi, the observed difference is =34' while the actual difference should be =113', which means an error of observation of 79'. With such large errors of observation, very nearly equal to the maximum value of the major [Bija] and more than double the maximum value of the minor [Bija], how at all could it be possible for the author to fix the [Bijas]? Again, with such a large capacity for error how did he get exactly 34' in three configurations, all equal, and 112' in two ? It is extremely unlikely that these observed values would be so closely correct, especially with such a large capacity for error. Then again, with such errors there is little likelihood of his discovering the constant of [Variation], viz., 34' so nearly equal to

the correct value 39', and the other constant, 78', so nearly equal to the correct 76' whether it be the constant of Evection or the defect in the [Equation of the Centre]. These things show that the author must have 'cooked' his observational values from [Bijas] taken from elsewhere, the 'cooking' being done wrongly in certain cases. Now can this person be [Bhaskara]?

Lest it should be thought that the interpretation of [Padardha] to mean 180o, as done by Ghosh, might save the same table the corresponding to his interpretation. It will be seen that except in iii, no observation agrees with the [Bijas]; that there is discrepancy betweenobservation and the value that shouldhave been observed, to the extent of 112' in i and ii, and 34' in iv, v and vi, and that the correct difference observed should have been 0 in all the configurations, because according to Ghos's interpretation all of them are syzygies, the Hindu values being correct at the syzygies, as I have already stated.

(4) Another reason why the [Bijas] cannot be [Bhaskara's] is that it is very unlikely he would have given them in the form of tabular values instead of in the form of an equation, as is generally done by astronomers, and that too without combining the first [Bija] with the [Equation of the Centre] as he could have done. he would only be following the usual practive if he had given in the Sid. ['Siromani]: "The Moon's spicycle is 42o 54' at the end of even quadrants and 3o8' less at the end of odd quadrants:" and then given 5' sin 2l separately as also 34'sin 2D. Or if he had wished to give the Equation of the Centre with the traditional value first and then the [Bijas] in a

separate chapter, as done by [Munjala] and [Sripati], he might have given the epicycles as 34o 44'-31o 36' in the [Spastadhikara] and in a later chapter given-(78' sin l+5' sin 2l) and -34' sin 2D, where D is True Sun minus True Moon. Even if he had discovered the [Bijas] later, he could have no difficulty in incorporating it in the Sid. 'Siromani written by him only one year earlier.

The author of the [Bijopanaya] seems to have anticipated this argument, for he says that [Bhaskara] wanted to keep the [Bijas] secret and therefore did not give them in his regular work; in 'sloka 58 he says: "This should not be given to one who has not served for one year." But this answer is unconvincing in the light of the following: Throughout the [Bijopanaya] the author dins into the readers the importance of the Tithis corrected by the [Bijas] for ceremonial purpose. Then what is the point in making a secret of this knowledge, so as not to reach the hands of astronomers and almanacmakers (for, as already mentioned, no astronomer or

commentator refers to this work, excepting the author of the [Tithinirnayakarika] with which this work seems to be associated; nobody seems to have "been in tutelage for one year" and qualified himself!); for it cannot be kept a secret and at the same time made serviceable. A medicine can be kept a secret and at the same time used by a family of physicaians to cure a disease and make money. Even in predictive astrology knowledge can be kept secret and at the same time profitable. But knowledge of the type we are talking about must be public if it is serviceable; for Tithis computed otherwise, and people would want authority before following them. As for astronomical predictions like eclipses, where agreement with prediction like eclipses, where agreeement with observation is all-important, and not authority, I have already pointed out that the [Bijas] would only serve to spoil even the existing agreement. Further, astronomers usually refer to their science as a secret one, which means that it should not be given to the layman who has no respect or fitness for it; but secrecy of an astronomical work from astronomers themselves is unthinkable. [Munjala] or [Sripati] have not kept their [Bija] secret, and [Bhaskara] must have known it. Then what is the point in suppressing a part of the Equation of the Centre and making a secret of it,

(the first Bija is only this), for anybody can do this.

(5) It has also to be noted that the style of the work is not [Bhaskara's] and the spirit, not that of the Sid. 'Siromani. The Sid. 'Siromani reflects [Bhaskara's] great reverence for [Brahmagupta], whose authority is quoted wherever necessary; and it is also known that [Bhaskara] belongs to the school of [Brahmagupta]. But the [Bijopanaya] is surcharged with the spirit of the [Suryasiddhanta] to which frequent and exhaustive references are made as authority. For instance, at the very beginning of the work there is a long quotation from the [Bijopanayadhyaya] of the [Suryasiddhanta], a passage which [Ranganatha], commentator on the [Suryasiddhanta], characterises as an interpolation. While the Sid. 'Siromani reads like a scientific work, the [Bijopanaya] reads like a [Dharmasastra]. The matter of fact style of [Bhaskara], stands in contrast with the racy style of a controversialist which characterises the [Bijopanaya]. See for instance the following passage from the [Vasanabhasya] appended to the [Bijopanaya] and puporting to be by its author: BEäò SÉ ¤ÉÒVÉÉä{ÉxɪÉ|ɪÉɺɦÉÒiÉÉ:iÉxuùɱɴÉ:¤ÉÒVÉÉä{ÉxɪÉÉMɨɺªÉ |ÉÊIɺÉiÉɨÉÉ®úÉä{ªÉ +|ɨÉÉ{ªÉ-¨ÉÖ®úºÉÒSÉEÖò: iÉx¦ÉªÉÆ ÊxÉʤÉVɨÉä´ÉäÊiÉ xÉÉÊuùªÉɨÉä½äþ, ´ÉIɪɨÉÉhÉ®úÒiªÉÉ iÉnÖùnÚùÊ´ÉiÉ|ÉIÉä{ÉÉxÉÖ¨ÉÉxÉÉ´ÉÉÆ iÉEòʦÉ#ɺÉi´ÉÉiÉÚ * etc.

Incidentally there seems to be a reference here to [Ranganatha's] allegation mentioned

above. This passage also seems to reflect the controversy that raged during the seventies and eighties of the last century in South India between Karungulam Krishna Josyar as the protagonist of the [Vakya] system of almanacs and Sundaresa Srautigal, Nungambakkam Raghunathacharya etc. as the sponsers of the Drk system. On e important point of controversy was whether the centre of the celestial sphere is the same as the centre of the Earth or different, and for what celestial phenomena the centre of the Earth (or of the celestial sphere if it is different) is to be taken as the point of reference, and for what othes the position of the observer was to be taken as that point. [Sundaresvara] [Srautigal] and others argued that the two points are one, while Krishna Josyar argued that they are different, and that planets are given in the [Sastras] with reference to the centre of the celestial sphere. When the former (it may be noted that they were the people who first got the [Bijopanaya] printed at Madras) quoted the following passage from it, in support of their stand: +lÉänùÉxÉÒ JɨÉPªÉä

P¶ÉªÉ¨ÉÉxɺªÉ ¦ÉÚ¨ÉPªÉPHÖò±ÉªÉOɽþºªÉè´É {ÉÉ®ú¨ÉÉÌlÉEòi´Éä ½äþiÉÖ¨Éɽþ etc., no wonder Krishna Josyar accused them of forgery: he also questioned the authority of the [Tithinirnayakarika] which mentions the [Bijopanaya]; cf. " ¤ÉÒVÉÉä{ÉxɪɦÉɹªÉÉxiÉä ¦ÉɺEò®úɪÉè:|É{ÉÉÎ\SÉiÉÉ: *" "¤ÉÒVÉÉä{ÉxɪɦÉɹªÉÉxiÉä ¦ÉɺEò®äúhÉ Ê´É{ÉʸÉiÉÉ * "

To conclude, in the above discussion we have certain pieces of conclusive evidence which have been placed first. Those that follow are valuable as cumulative evidence, the combined effect of which is to show beyong doubt that [Bhaskaracarya] of 1150 A.D. cannot be the author of the [Bijopanaya]. This being so, the real author of the work remains to be identified by further investigation.

A HISTORICAL DEVELOPMENT OF CERTAIN HINDU ASTRONOMICAL

PROCESSES

Reprinted from the Indian Journal of History of Science,

Vol.4, Nos. 1&2,1969

It is interesting to study how in course of time Hindu astronomers developed more refined methods from cruder ones in the matter of (i) finding the true longitude of the sun, the moon, and the star-planets, (ii) applying the equation of time,(iii) computing the eclipses, and (iv) fixing the [Mahapatas]. The earlier [Siddhantas] that could help us in this have been lost, so far as we knoe at present. But [Varahamihira (V.M.) has chosen what were typical of the ancient ones, and given them in a condensed form in his [Panca Siddhantika] (P.S.c.A.D 505), which we can use for our purpose. The extensive quotations of [Bhattotpala] from a [Paulisa Siddhanta] and various other ancient [samhitas], in his commentary on the [Brhat Samhita] are also helpful. We shall consider these one by one.

I The True Longitudes

In the [Vedanga-jyautisa (c.1180 B.C.), the earliest Hindu astronomical work extant, and the less so ancient [Samhitas], and the [Suryaprajnapti] and [Kalalokaprakasa], as also the [Paitamaha Siddhanta] condensed by V.M. in the P.S. only the mean sun and moon are given. The true [naksatras], [tithis], wtc., required for religious rites and observances must

have been obtained by the mean sun and moon as guides.

The [Vasistha Siddhanta] of the P.S.(not the ones extant and available

as separate works) is the first [Siddhanta] giving methods fo rthe true sun, moon and star-planets. P.S. II, i. gives the true sun, and the substance of what it says is as follows: The year begins, i.e. the sun is at the first point of the sign Mesa, one and a half days before the epoch taken by the P.S., and it takes 31 1/4, 31 1/2, 31 1/2, 31 1/2, 31,30 1/2, 29 3/4, 29 1/4, 29 1/4, 29 1/2, 30 and 30 1/4 days successively to traverse the twelve signs and reach the first point of Mesa again. Giving the true sun thus in an empirical form, is facilitated by the absence of the knowledge of motion of the apses, as it is very slow.

In the case of the moon, this [Siddhanta] recognizes the equation of the centre (eq.cent.) superimposed on the uniform mean motion of the moon, recognizing the true motion to be zigzag, and a 'step-liner function' of time; i.e. the rate of motion is supposed to increase uniformly @10'1/7 per pada (==a ninth of a day), from a minimum at apogee of 702', to a maximum at perigee of 879' and then falling at the same rate to the minimum again at the next apogee. Thus the time between one apogee and the next (called a gati) divided into 248 padas (i.e. 27-5/9 days for an anomalistic revolution) forms the period of the zigzag. The mean longitude up to the last apogee is first given, by II, 2-4. We have now to add the true anomaly to get the true logitude. For that prupose the 248 padas are divided into two parts, the first 124 forming the first half gati being called plus padas, and the second set called minus padas. The true anomaly consists of two parts, the accumulated mean motion @ 1o27'.843 per pada, and the accumulated defect or excess in the motion forming the eq. cent. These two parts have to be combined and given. But the [Siddhanta] takes the 1oper pada separately and adds it to the mean longitude already found, for the sake of convenience. The remaining 27'.843 per pada is combined with the eq. cent., which here is a summation of the difference of the linear function of motion from the mean, and thus an algebraic function of the pada of the second degree, and therefore different from the correct eq. cent. of the form =a sin m. The result of the summation is =(665-5p) p/63 in minutes (the upper sign for the plus padas, and the lower for the minus padas). Combining this with the residual mean motion, 27'843 P separately for the plus or minus padas, we get the two formulae given in Ii, 6, {1094+ 5(p-1)} Xp/63, and {2414-5

(p--1)}XP/63, respectively. When the padas are in the second set, the true motion for the first set of 124 padas, equal to 6 [Rasis] and 4 minutes, is added.

The [Vasistha] applies the eq.cent. fo the above form in the computation of Jupiter and Saturn also (Chap. XVIII, 6-20). In the cas eof Venus this is dispensed with, being small. As a first step for every star-planet, the number of days after its mean heliacal rishing is found. The days from the true heliacal rising is foudn by applying the days corresponding to the eq. cent. at the point of heliacal rising. The synodic period of Jupiter, Venus and Saturn is divided into sections, for each of which the degrees of motion is given, according as the motion is 'quick', 'mean', 'slow', 'retrograde', etc., depending on the equation of conjunction (eq. conj.). The days in heliacal sttings, with the corresponding degrees, is also given. From

this set, the degrees corresponding to the true number of days gone is taken and added to the degrees at heliacal rising, with the eq. cent. applied. Thus the geocentric true planet is got.

In the case of Mars and Mercury with large eq. cent. varying with planet's situation in the twelve [rasis], since the planet may be in any of, or, all the 12 [rasis] during its synodic revolution, the above method cannot work. So the varying sectional motion for each of the [rasis] is given (Chap. XVIII, 21-56). Thus we see that the eq. cent. and the eq. conj.are mixed in the case of these two planets; while in the case of Jupiter and Saturn they are separated, which is possible since their motion is comparatively small.

It is said that as early as the third century B.C. the astronomers of Babylonia used the Zigzag or step-linear function, and also gave the star-planets accoring to the sections of motions in the respective synodic periods, starting from their heliacal rising, and this was used in predictive astrology. Thsu there seems to be a connection between the [Vasistha] and [Babylonian] astronomy, and there might have been give and take. Which borrowed which portion from which has to be determined after learning more about these. Or we may consider both as one unit belonging to a part of greater India.

At the end of Chap. XVIII of the P.S. the motion of the star-planets is again given int he last sixteen verses and attributed to the [Paulisa]. But

there is a lot of evidence to prove that this portion is an interpolation. In the case of every planets here the eq. cent. is neglected, so that the motion is simplified and given for sections of the synodic period, on the same model as the [VAsistha] Venus. It is useless, being very rough on account of neglecting the eq. cent., and also fo no historical value, coming as it does long after the age of the eccentric and epicyclic theories.

The next step forward is seen in teh [Paulisa] method for the true sun (P.S.,III, 1-3). The instruction is as follows: Multiply the days from epoch by 120, subtract 33 and divide by 43,831. The revolutions etc. of the 'mean' sun is got. The 'mean' sun plus 20o si called Kendram (Anomaly). For successive 30o of Kendram first subtract 11, 48, 69, 70,54,25 minutes, to get the true sun. Here again we see the zigzag, the variation from the 'mean' being given for the degrees of Kendram instead of the time passed, and the period of variation beginning at 20o before the commencement of the so-called 'mean' sun has practically the maximum eq. cent. of about 139' combined with it, and so the beginning of the 'mean' year is practically that of the ture. The improvement consists in realting the amplitude of the zigzag with the degree of the so-called Kendram paving the way for the appearance of the real Kendram next, in the [Romaka Siddhanta] condensed in the P.S.

As regards the moon, the [Paulisa] has taken over the method of the [Vasistha] in toto, and in III, 5-8 gives a correction to it, in order to yield it own values, conforming more to the '=a sin ,' formula. But owing to the corruption of the text it is not possible to interpret the verses with any certainty.

It is in the Romaka that we see for the first time the mean sun and moon and their respective eq. cent. recognized as separate entities, going to make up the true sun and moon. The eq. cent. is also recognized to depends on teh angular distance called Kendra (now in the

true sense of the word) between the mean body and a point in its orbit (the ucca or apogee, but not named here). Its increase or decrease is given for every 15o of Kendra, with instruction when to add and when to subtract (P.S.VIII, 1-6). By adding the increments the sun's maximum eq.cent. can be found to be 143'23'' and the moon's 4o55'. These values that will be got from the formulae, 143'23'' sin (0o, 15o, 30o, etc.) or 4o55' sin (0o,

15o, 30o, etc). Therefore, we have to conclude that the [Romaka] values were got empirically by observation, and not by using the trigonometrical form. But from better observations and better values, the next step would be the discovery of the trigonometric form, which would lead to the eccentric and epicyclic theories as an explanation of the observed result.

How the early Hindu astronomers explained the zigzag motion, now slower, now faster than the mean, noe direct and noe retrograde, is contained in the [Surya Siddhanta] (Chap.II, 1-5). Deities stationing themselves at certain points int eh orbit ( later to be identified with the apogee and superior conjuction) pull the bodies backwards and forwards with invisible air-reins, which motion combining with the uniform natural motions of the bodies result in the zigzag, it says. But this theory is at best only qualitative, though stasifactory to the minds of the ancients, which could connect anything extra-normal with invisible deities.

As siad before, the discovery of the trigonometrical form for the eq. cent. led to the theory of eccentrics and epicycles, as found in the [Ardharatrika paksa] of [Aryabhata] (reflected in the [Khanda Khadyaka] of [Brahmagupta], and mentioned in the [Mahabhaskariya], VII,20 etc.), and in the Saura of the [P.S.] According to the eccentric theory the planet (either mean or sighra) moves uniformly in a circle whose centre is away from the earth's centre by a distance equal to the maximum equation of the centre (as measured on the arc of the circle) in the direction of the respective [Ucca]. According to the epicyclic theory the mean planet moves in a circle round the earth as centre,while the true planet moves on an epicycle of radius equal to the sine of the maximum eq.cent. or eq. conj., as the case may be, moving round the mean planet as centre. If the maximum eq. cent. is 2e (=sine max.eq. cent.approximately), we can get from the theory the first term of the modern eqn.cent.2e sin m (where m i sthe mean anomaly reckoned from the apogee), taking the distance between the true planet and the earth (called the true hypotenuse) as approximately equal to the radius of the circle, as many [siddhantas] do. If the true hypotenuse aslo is taken into account, as logically following from the theory, which the school of [Aryabhata] does, we get the second term also in the form, -2e. sin 2 m, only slightly different from the modern -5/4 e sin 2 m (cf. the article, 'Some peculiarities of the school of '[Aryabhata]' by the author).

But these theories cannot completely take the place of the actual, for the moon or the planets really move on ellipses, with the parent body at one focus, and the motion is such that equal areas are swept by the radius vector in equal times. The necessary difference between the results of the defective theories and observation was sought to be removed in some works (like the [Aryabhatiyam] and the later [Surya Siddhanta]) by supposing the epicycles to vary; but this has not met with much success. Only in recent times, the adoption of

modern methods of computation based ont he correct theory could give results to any desired degree of accuracy.

In the case of the moon, the distrubance caused by the sun's pull on it is another cause of great deviation from the mean motion. This is analysed into various equations called the Second, Third, etc., inequalities, the eq.cent.itself being the First. Of these the Second inequality, now expressed in the form -76'4. sin (m-2d), (where 2 D is the mean elongation of the moon, and m is the mean anomaly reckoned from the perigee according to modern custim), was first discovered by Ptolemy Ii of Egypt (second century A.D.) and called by him the Evection. O fthe Hindu astronomers, [Munjala] was the first to give this in hi s[Laghumanasa](A.D.932) [prakirnadhyaya], 1-2. It will be interesting to know how he got this. What he gives reduces to --65'.3 sin m+65'.3 sin (m-2D), 9m here being reckoned from apogee according to ancient practice). Of these the first term approximately compensates the deficiency in the eq. cent. of Hindu astronomy(=about 300'maximum, while actually it is 377'.3). The second term forms the Evection proper. Next to [Munjala], Sripati (A.D.999) gives it in his [Siddhanta Sekhara], XI, 2-4.

Then [Nityananda] for the first time in India, in his [Siddhantaraja] (A.D. 1639)gives the Third inequality called the Variation (=39'.5 sin 2D), calling it by the significant name, [Paksika] (i.e. the one having the [paksa] as its period). Some time before, at the end of the sixteenth century, Tycho Brahe had made this discovery in Europe. It is said that an Arabian astronomer by name Muhammad Abul Wefa discovered this inqualtiy first at Baghdad in (A.D.975). In a work called the [Bijopanaya] (alleged A.D.1151) attributed to [Bhaskaracarya] (II) the Variation, as also what may appear to the unwary as the Evection, is given. But the work is by a recent imposter

and not by [Bhaskaracarya], which the author has shown in a study of the work.

[Nityananda] has also given the Reduction to the Ecliptic, which is 7' max. in the case of the moon. The earlier [Siddhantas] neglected this as being small, though it must have been known at least to [Bhaskaracarya] who had understood the need of the [Udayantara[ or [Reduction] to the Equator. The Kerala astronomers like [Madhava] were using this earlier than [Nityananda]. Later, cantact with moden Western astronomy led to the adoption of the fourth major inequaltiy, the Annual Equation (=+11' sin sun's anomaly, having a maximum value of even 14' at syzygies), as also a host of lesser inwqualities.

It is intersting to consider the question why the early Hindu astronomers failed to detect event he major inequalities, excepting the eq.cent. The mean motions of the sun and the moon, as also their eq.cent., were obtained by them by the analsis of the times of eclipses, which occur only at syzygies, and so they are very accurate for the syzygies, a tribute to their powers of analsis. But since 2 D is practically zero at these times, the variation becomes zero, and the Evection is reduced to the form+76'.4 sin m, which combined with the eq. cent.-376'.4 sin m, becomes -300' sin m, very nearly the same as the Hindu eq.cent. Similarly +11'sin (sun's anomaly) merges in the sun's eq.cent., having the same form with its sign reversed, since the sun is subtracted from the moon to get the times of syzygies, so that

the sun's eq.cent., about -119' sin sun's anomaly (this was its value about A.D. 500), becomes about -133' sin anomaly, as given by the [siddhantas]. If instead of analysing the times of eclipses accurate observation of the moon's longitude had been used, at least the major inqualities could have been detected early, one after another, as the accuracy of observation improved.

II The Equation of Time

The Equation of time (eq.time ) is the difference between the mean and true moons or mean and true midnights. This is required in Hindu astronomy to take the sun, moon or star-planet, computed by using the mean days, and therefore got for the end of the mean day, to the true day, which is the practical unit of time. (This is enough for systems that begin

the day at midnight like the [Ardharatrika] of the [Saura Siddhanta] of the [P.S.] or the [Surya Siddhanta] school. But if the day begins at sunrise as in most [Siddhantas], the variation in daytime also will have to be corrected for. If the place is east or west of the prime meridian like Ujjain, a correction for this also will have to be made.)

The eq. time consists of two parts: (a) the part caused by the true sun generating the true day being behind the mean sun or in advance, owing to the eq. cent., and called [Bhujantara] (=-22 sin sun's anomoly in vinadis); (b) the part caused by the true sun moving along the ecliptic, so that its projection on the celestial equator is behind or forward of the mean sun assumed to move along the celestial equator that forms the [nadi mandala] and called [Udayantara] (i.e. reduction to the equator = - 25 sin 2 o in vinadis,o being the [sayana] true sun).

Even at the stage of the [Romaka Siddhanta] of the P.S. the need for correcting the sun etc. for the eq. time was not recognized by Hindu astronomers. In the Saura of the P.S. the [Bhujantara] correction appears for the first time, followed by all later astronomers. The [Aryabhatiyam] does not mention this, but [Aryabhata's] followers have argued that this is intended in [gitika] 2 by the expression, [ajarkodayacca lankayam], as done by Govindaswamy under Mahabhaskariyam, IV, 7. Its appearance in the [Khandakhadyaka] shows this probable. At any rate, [Bhaskara] I gives it in his work (IV, 7, 24, 29-30).

Vatesvara sees the need to refine the minutes of [Bhujantara] by projecting it on the celestial equator, by instructing that the minutes of the eq. cent. should be multiplied by the [Pranas] of ascensional difference of the sign occupied by the sun and divided by 1800. Sripati gives the [Bhujantara] without this refinement since it will be taken care of by the [Udayantara], which he gives for the first time in Hindu astronomy (Siddhantasekhara II], 46; XI, 1.). After this [Bhaskaracarya] and others continue to give this.

III The Computation of Eclipses

Eclipses are mentioned even in the Vedas, the earliest works of the Hindus. Even in the ealry Fifth Mandala of the Rgveda, in the 40th [sukta], the solar eclipse is mentioned as caused by [Svarbhanu] (otherwise called Rahu), and well understood by the Atri family of priests. In th e[Chandogya

Upanisad] the lunar eclipse is mentioned as caused by [Rahu]. Therefore the Saros of something like it might have been known then. But computation as

such could not have been possible before the stage of the [Vasistha] of the P.S., which is the first to give the true sun and moon forming the basis.

P.S. Ch. VI gives the lunar eclipse according to the [Vasistha]. (It does not attempt the solar eclipse ). The instruction is as follows:

(a) Find the time fo true opposition, and the true moon then. (b) Subtract 1o36' from [Rahu] (Head or Tail), whichever node is near the moon), and find the distance between this and the moon in degrees. If this is within 13o, there is an eclipse. If within 15o, there is only a slight darkening. (Rahu is not given by this siddhanta). So we are expected to use the Paulisa's. Subtraction of 1o36' may be an empirical correction.) (c) The minutes of arc of half-duration =[552-L2], Where L is the latitude of the moon in minutes. (This latitude is Paulisa's.) (d) 60Xminutes got in (c) ./. D(i.e. difference of the true daily motions of sun and moon)==time of half duration in [nadikas]. Subtract this from and add to the time of opposition to get the first and last contacts. (A correction is given next for having used the Paulis'a L., Whose significance is not understood by T and S.) (e) 21[25-d2]=minutes of arc of half total phase, where d id moon~[Rahu] in degrees. (f) 60X minutes got in (e)./. D= time of half total phase, from which the beginning an dthe end of total phase is got.

From (e), the critical latitude for total obscurity is 21', from (c), th esame for the eclipse is 55'. Therefore the angular semidiameters of shadow + moon=55' and shadow -moon=21', from which we get the radius of shadow=38' and radius of moon=17', both considered constant. Add to the error caused by these, the roughness of taking the opposition as the middle of the eclipse, and not taking the variation of the latitude during the eclipse into consideration.

In P.S. VII, 6, the [Paulis's] gives the duration of the lunar eclipse directly in [nadikas], equal to 3 [169-d2/4. Since D is not used here, the computation is rougher still. It does not give the total phase.

The instruction, commonm to both [siddhantas], to find the points of firtst and last contacts, is as follows: Divide the moon's half orb on the side opposite to the direction of its latitude into 13 stripts of equal width, by lines

paralled to its east-west (this with reference to the ecliptic) diameter. The eastern and western points of the (d plus 1) paralled line beginning with the diameter are the points of first and last contacts. These have to be reduced to the observer's east-west by the two [valanas], aksa and ayana. The latter is omitted here. The former is given roughly by multiplying a quarter of the moon's rim by the degrees of the hour-angle and the degrees of latitude of the observer, and dividing by 8100. Thus, instead of their sines, the degrees themselves are used for the proportion (VI, 7-s).

The [Paulisa] is the first [siddhanta] to attempt the difficult solar eclipse (P.S. VII). Here the parallax in longitude is transferred to the time of conjunction, as done by all Hindu [siddhantas]. The correction to the conjunction is taken roughly as 4X sin h (=Hour. angle) in [nadis]. The manner of giving the parallax in latitude (1) is peculiar, in that it is done as three corrections to [Rahu], so that the (moon~corrected-Rahu) will give the [parallax-corrected-] direct. (T. and S. have failed to understand the instruction : see

commentary under VII, 2-4).

The first correction to [Rahu] is: multiply the degrees of latitude of the place (o/)by 5 and divide by 27. Add to [Rahu] if Head (i.e. ascending node), subtract if [Tail](descending node). If we resolve the total parallax in latitude (=r sin zenith distance of the nonagesimal) into three parts involving the latitude of the place, the declination of th emoon or sun, and the hour-angle, this approximately does duty for -r cosw. sin o/, where r is the moon (about 49' according to Hindu astronomy), and w is the obliquity of the ecliptic (taken as 24o in do ). Latitudes in India being low, sino/ is taken as o ./. 57 approximately. Since the [Paulisa] gives 55' of moon's latitude (1) for 13o of moon~[Rahu] (d), the correction to [Rahu]=-49X.92X13oXo/./. 27. This being negative decreases north l and increases south. So this is to be added to the head and subtracted from the tail as given.

The second correction is : Add 3 [rasis] to the sun. Take the declination (f) of this point in degrees. Multiply these degrees by the number of [nadikas] of correction to conjuntion, and divide by 22. If the time of eclipse is [Uttarayana] 'forenoon' and 'Head' is chaged, or if all three. This rule does duty for -rsin w cos o/. sec.cos. sin h, . being the sun, and s being the sun,s declination. This [siddhanta] has taken as unity.

sin h=[nadis] of correction ./. 4. Sin wcos o= sinw sin (o plus 3 [rasis])=sin f=ff, approx. Thus we get 49X.9X13XfX[nadis] of parallax got in (1)./. (55X57X4) =fX[nadis] of (1) ./. 22. This has the same sign as -cos o.sin h. Since cos o is positive for [Uttarayana] and sin h for morning, the degrees resulting from the rule are additive to Head. It follows that chage of any one or all three must make it subtractive while the change of any two must keep it additive.

The thrid correction is: find the [nadis] elapsed from sunrise to corrected conjunction if forenoon, or the [nadis] to elapse till sunset if afternoon (i.e. find the unnata nadis of corrected conj.). Multiply this by the degrees of declination of the moon, and divide by 80. If the moon's longitude is within 6 [rasis] the resulting degrees are to be subtracted from the Head or added to the Tail. If more than 6 [rasis], add to Head and subtract from Tail. This serves for the one remaining correction, +r.cosw cos. tan. cos h. Here, cos o/ is roughly taken as unity, and tan equal to approx. As for cosh, this is very roughly taken as [Unnata] [nadis] divided by 15. So the correction takes the form, +49X13X0.92Xunnata [nadis]./. (55X57X15)=xunnata [nadis] ./. 80. Since h is roughly taken to be with in =90o, and therefore cos h always positive, the sign of the correction depends only on the moon's declination, which is positive fir the moon in the first 6 [rasis] and negative otherwise. So the correction is subtractive and additive to the Head respectively, and vice versa to the Tail. These three corrections must have been obtained empirically.

An empirical correction of -1o36' is made on this corrected [Rahu], as done for lunar eclipse. The duration of the eclipse is given in [nadis] by 3[64-d2/4. The corrected conjunction =half, this, gives the first and last contacts. The total [hase is not attempted. As in the lunar eclipse, here too the variation in the moon's latitude is not taken into account, as aslo the true D. From the formula we get, d=8oas the limits, i.e. when the corrected latitude is 55'X8./.13 =33'8. From this, since the angular radious of the moon has already been found

to be 17', that of the sun is 16'8 according to the [Paulisa], both constant.

The [Romaka Siddhanta] gives a very much improved method for the solar eclipse. (Its lunar eclipse is not given in the P.S.) It is the first to use

the [Tribhonalagnam](i.e. the nonagesimal) in the computation (exhibiting thereby knowledge of the trigonometrical problem involved), though the method given is rough. The conjunction corrected for parallax is to be found first, in the same way as in the [Paulisa]. For this time the [lagna] (i.e. Orient Ecliptic point) is to be found. This minus 3 [rasis] is the nonagesimal. The sine of the zenith distance of the nonagesimal (zdn) has to be multiplied by the relative horizontal [parallax r to get the parallax in latitude. But the [Romaka] takes the zdn as equal to the declination of the nonagesimal=o/, the upper sign being used for south declination and the lower for north. (This is given as an approximate method by Brahmagupta.) r is taken as roughly proportionate to the moon's true daily motion, as generally done in Hindu astronomy. Thus the parallax in latitude is given by the formula, tabular sine zdn (=120 sin zdn)Xmoon's true daily motion ./. 1800. (This will give 53'. But as the sun's is really 0'.15, 53' of the [Romaka] is far near the correct 57').

Since the moon is not on the ecliptic when its latitude is other zero, a correction has to be made for this in the parallax in latitude. But this is too small (maximum half a minute) and neglected by some [Siddhantas]. The [Romaka] and some others intend to give it, but by the following method which is wrong (and gives rise to an error of 4' and more): Treat the nonagesimal as the moon and find its latitude. Add this to the declination of the nonagesimal, and use this corrected declination to find the zdn above (P.S.VIII,II). [Bhaskaracarya](II) sees the error in this and criticizes it in his [Bhasya].

The moon's latitude is to be got by multiplying the tabular sine of moon~[Rahu], by 21, and dividing by 9. (This will make the obliquity of the moon's orbit 280', better than the 270' given by other [siddhantas].

The mean angular diameters of the sun and the moon are given as 30' and 34' respectively, and they are made true by multiplying by their true daily motions and dividing by the mean, as done by most [siddhantas]. It must be noted that the sun's diameter here given is too small by 2' and the moon's too large by 3'. The half duration is given in [nadis] by 60 X[^2 -l2]./.D, where ^ is the sum if the semidiameters, l is the parallax corrected latitude at corrected conjunction, and D is, as already said, the difference of the true daily motions. This is subtracted from or added to the corrected

conjunction (as in the Paulisa) to find the first and last contacts. This is defective in so far as the l of the first and last contacts are not found by successive approximation, as done by the later [siddhantas], and used to find the respective time. The total phase is not given by this [siddhanta].

The method for the solar eclipse given by the early [Saura Siddhanta] condensed by V.M. in the P.S. (Chap. IX) is more refined and typical of the later [siddhantas], and uses far better constants. Like the Romaka the Saura also uses the true motions of the moon and the sun to get their respective true parallaxes and angular diameters. But in the case of the moon the true motion at the time of the eclipse is used (which is proper) instread of the true daily

motion (IX, 14). Also these ate not got directly from their mean values as in the Romaka, but from their orbits and diameters in [yojana] as in the regular [siddhantas], through these are not the actual numbers of [yojanas] but those reduced bya factor 270 in the case of the orbits, and other appropriate factors in the case of the diameters and parallaxes. By examination we can find that the mean diameters of the sun and the moon are 32'.1 and 32'.2 respectively, and their mean horizontal parallaxes 3'.8 and 51'.4 (IX.15, 16, 22).

To find the sine of the zdn also a better method than the [Romaka's] is used: The sine of the zenith distance of the meridian ecliptic point is first found. Using th eamplitude of the orient ecliptic point, which is equal to the azimuth of the nonagesimal, the sine of the angular distance between the meridian ecliptic point and the nonagesimal is found next. With these two sines treated as hypotenuse and one side, the sine o f the zdn forming the other side of the right-angled plane triangle is found, approximately, instead of getting it by solving the spherical triangle. After this the parallaxes in latitude and longitude are got by accurate steps.

The parallax corrected conjunction is got from the parallax corrected longitudes, and made accurate by successive approximation. From the parallax corrected latitude of this time, the approximate half duration and first and last contacts are found. Using the parallax corrected latitude of the first or last contact a more accurate time of the respective contact is got, and thus by successive approximation the correct times are got. In this too the [Saura] is an improvement on the [Romaka]. It also gives the

total phase, together with that of the lunar eclipse, in X, 7.

For the shadow also in computing the lunar eclipse, the reduced orbits and diameters are used and the result given in a more reduced form. If we reduce it still more and express the angular diameter in raidans, we get 36 ./. moon's orbit -572./. (5Xsun's orbit), the first term representing double the moon's horizontal parallax of the modern formula, and the second representing double the difference between the sun's angular radius and horizontal parallax. The mean diameter of the shadow can be found to be 78'.4. As already said the total phase also is computed. Successive approximation is used. The method to find the fraction of the moon eclipsed at any given moments is given, as in later works.

A separate chapter (XI) is devoted by the [Saura] to the graphical representation of eclipses. In order to make the figure appear as big as the real, a scale of one (angula) or digit for 2' is instructed to be used near the horizon, and for 3' at the zenith, and proportionately in intermediate positions (XI,6). But the difference in size according to its position in the sky is an optical illusion depending on the atmospheric conditions, and various sacles are given by various [siddhantas], the [Saura] being the first to give it.

In order to mark the points of first and last contacts at the exact positions as seen by the observer, the lay of the segment of the ecliptic where the moon is situated has to be fixed with references to the east-west of the observer. Two corrections, one due to the latitude of the place called [aksavalana] and the other due to the moon's [ayana] called [ayanavalana], have to be applied to the east-west points. (We saw the [aksavalana] given by the [Vasistha]

itself, but very roughly using degrees instead of their sines.) But the versine of the hour-angle is used instead of its sine, a mistake of some ancient astronomers, pointed out by [Bhaskaracarya] (II). While the [Vasistha) dispenses with the [ayanavalana], the [Saura] gives it fairly correctly as sin- (cos moon's longitude X sin w), instead of tan- (cos moon's longitude X tan w(XI, 2-3). It is note worthy that the [Saura] does not use the versine here, a mistaken practice of the ancients condemned by [Bhaskaracarya] (II). (The Romaka is silent about the directions).

The [Aryabhatiyam], and the [Mahabhaskariyam] following it, give the same method as the [Saura] for parallaxes, angular diameters and [valana]. But they find the first separately for the sun and the moon, using in the moon's case its own orbit instead of the ecliptic, to find the parallaxes. We have already mentioned that this trouble is not worth taking in view of other errors. The mean angular diameters and the horizontal parallaxes of the sun and the moon according to these are 33', 31', .5, 3'.9 and 52'.5. It is remarkable that the [Mahabhaskariyam] uses parallax in the lunar eclipse, of course without purpose (Chap. V, 68-70). The mean angular diameter of the shadow can be found to be 79'.8. The later [Surya] [Siddhanta], representing a school to which the [Soma], [Brahma] and [Vrddha] [Vasistha] [Siddhantas] belong, also uses the same method as the [Saura] fo the P.S. with one difference. It gets the [nadis] of parallax in conjunction direct bya formula that reduces to 4xcoszdnX difference of mean daily motions./. 15 (V, 7-8). We see that here the mean horizontal parallax is taken to be the mean daily motion divided by15, which gives 3'.9 for the sun and 52'.7 for the moon. Using the true daily motions as done by the [Romaka] and the [Saura] or using the true hypotenuse like the school of [Aryabhata] would have made the parallax true, which is desirable. But in the case of the [nadis] of parallax in conjunction the constant 4 is justified, since the multiplication by the true motions to get the true parallaxes is cancelled out by the division by the true motions to get the [nadis]. The mean diameters of the sun and the moon according to this school are 32'.4 and 32'.0 and teh shadow 82'.6. This school uses the correct formula for the two [valanas] (Chap. IV, 24-25).

[Bhaskaracarya], following Brahma Gupta, gives the same formula for the parallax as the [Surya Siddhanta], except that he gets sin zdn and cos zdn direct. So the remarks about the [Surya Siddhanta] hold in his case also. The mean diameters according to him are 32'.5, 32'.0 and 80'.8. About the correct formulae for [valana] used by him, we have already written, as also about his criticism of Brahamagupta (and certain others ) whi have given wrong correction in parallax for the moon's latitude.

IV. The [Mahapatas] and Yogas

The [mahapatas] are [Vyatipata] and [vaidhrti], from which the 27 [yogas], Viskambha] etc. were later derived. [Vyatipata] is the situation when the sun and the moon having different [ayanas] have the same north declination or south. [Vaidhrti] is the situation when they have the same [ayana] adn equal and opposite declinations. The 27 [yogas] are computed like the [naksatra], using the sum of the lonfitudes of the sun and the moon for the purpose. The seventeenth in this series called [Sarpa Mastaka](and also Vyatipata) together

with the two [mahapatas] are said to be extremely inauspicious for auspicious rites like marriage, but very efficacious for [dana], [homa], [japa], [tapas], and offering to the manes. The seventeenth [yoga] [Vyatipata], and the twenty-seventh also called [Vaidhrti], are included among the 96 [sraddha] days.

Neither in the Vedas, nor in the [Vedanga] [Jyautisa], are these mentioned. The [Vyatipata] (Mahapata) first occurs in the [Paitamaha][Siddhanta], as seen from its condensation in the P.S. (Chap.XII). The rule is to multiply the days from the beginning of any five-year [Yuga] by 12 and divide by 305 and the mid-[vyatipata] falls when there is no remainder. This is equal to saying that when there is no remainder. Thisis equal to 12[rasis] it is [Vyatipata], which is the same as saying, if the sum is equal to 6 [rasis] reckoned from the Spring Equinox, as defined in later times, it is [Vyatipata]. It must be noted that this can give only the approximate time, the exact time being got in earlier times from observation, and in later times (beginning with the Paulisa Siddhanta) by computing the declination (cf.P.S.,Chap. III.22).

The [Vyatipata] and [Vaidhrti] forming the seventeenth and twenty-seventh of the [yogas], first appear in the [Paulisa] (P.S.III 20). The [Vyatipata] here can be shown to be a relic of the one of the [Paitamaha Siddhanta]. This [Siddhanta] placed the winter solstice at [Sravistha], from where longitudes were reckoned by it (as by the V.J.). The sum of the longitudes thus reckoned, when equal to 12 [rasis], would give the [Vyatipata], satisfying (though approximately) the condition of equal declination. But in course of time Winter Solstice would be occurring earlier and earlier than [Sravistha], so that the criterion of equal declination would not be satisfied, and still people would be observing [Vyatipata] at the time calculated from the sum of longitudes, much as we are now observing the [ayanas] at [Makara] beginning and Karkataka beginning though these have precessed more than 20 degrees. At the period whent he Spring Equinox was observed to be near [Asvini], the astronomers shifted the first point from [Sravistha] to [Asvini], as done in the [Paulisa] by V.M. If longitudes are reckoned from [Asvini], which is 5 [naksatra] segments in advance of ['Sravistha], the traditional [vyatipata] would happen when the sum is 12 [rasis] minus 10 [naksatra] segments (i.e. 17 segments on the whole) and it is this that is given by (P.S.III, 20). Thus was formed this seventeenth [yoga], called also [Sarpamastaka]. Its sanctity, guaranteed by tradition, would still be there. At least from the time when the new position of

the W.S. could be observed, the [Mahapata] [Vyatipata] also should have been computed and used as satisfying the condition [par] excellence of equal declination. But reckoning longitudes from the Srping Equinox, which is 3 [rasis] forward from W.S., the sum should be 6 [rasis] to satisfy the condition, as we have already said, instead of the 12 [rasis] reckoned form W.S. Meanwhile the other [Mahapata](Vaidhrti) arose, whos echaracteristic (already defined) would be approximately given by the sum reckoned form the Sping Equinox being equal to 12 [rasis]. Viewed as formed by the sum of longitudes alone, without the reckoning from the Spring Equinox, this became the twenty-seventh [yoga] [vaidhrti]. It is noteworthy that V.M. does not give the [Mahapata], [Vaidhrti]. He also does not mention any of the

other 25 [yogas] (not even in his Brhatsamhita), which means that the series of 27 [yogas] had not been formed in hi stime.

The Saura of the P.S. also does not give the series, from which we can surmise its absence its from the [ardharatrika] system also. Its figuring in the [Khandakhadyaka] (which follows the ardharatrika) is likely to be an interpolation, since the verse giving it is the same as in the [Brahmasphuta Siddhanta] (Spasta 63), where it is considered by critics to be an interpolation, since it is not commented upon by [Prthudakasvami] and does not figure in the question and answer (Chap. XIV-6,31). The [Aryabhatiyam] gives the two [Mahapatas], [Vyatipata] and [Vaidhrti], under the one name [Vyatipata]. [Bhaskara] I mentions these two separately, and also the seventeenth [yoga], [Vyatipata], calling it [Sarpamastaka]. (It is thesethree that are classified sometimes under the one name [Vyatipata] and considered very inauspicious etc). [Aryabhata's] direct pupil, [Prabhakara] (sixth century A.D.), has mentioned seven inauspicious [yogas] (six of which can be identified in the series of 27 yogas) as seen from a quotation by [Sankaranarayana] in his commentary on the [Laghubhaskariyam], under II, 29. It can also be inferred from the quotation that he does not know the other [yogas], and the practice of giving the [yogas] Viskambha] etc. as an item of the [Pancanga] had not yet come into vogue. But in the [Siddhantas] later than Brahmagupta (seventh century), like those of the [Surya Siddhanta] school, [Vatesvara Siddhanta] etc., the series of 27 [yogas] is seen to be systematically computed.

As for the two [Mahapatas], the early works even up to the time of [Bhaskara] I give only the rules for their approximate time of occurrence as mentioned already, leaving the computation of the exact time and duration to astronomers, from the criterion of equality of declinations. The [Brahmasphuta Siddhanta] is the first, as it itself claims, to deal with them in detail (Chap. XIV, 33-34). First comes the test for possibility and next is found whether the mid-event has gone or is to come. Then taking a point of time arbitrarily before or after, the increase or decrease in the difference between the declinations during the interval is found. Using this, the time when the difference is zero, and the duration for a change of difference equal to the sum of the angular semidiameters, are found. The time of zero difference gives the mid-event, which minus or plus the half duration gives the beginning and end. The (later) [Surya Siddhanta] devotes chap. XI to this topic. This school finds the zero-difference-point by successive approximation, using the longitudes corresponding to the declaintions. It gives the half duration in [nadikas] by multiplying th esum of the semidiameters into 60 and dividing by the difference of the daily motions, which is very rough indeed. Lalla, who immediately followed [Brahmagupta], attacked the problem with zest in his [Sisyadhivrddhida], as also [Sripati] and others, with more or less correctness (cf. note 4.p.73). [Bhaskaracarya] in a special section devoted to this subject (Sid-Siro-Ganita], Patadhikara] gives the most systematic and correct treatment. Using the obliquities of the ecliptic and the moon's orbit, and the [sayana] longitude of [Rahu], the obliquity of the moon's orbit to the celestial equator, with the distance of [Rahu] from the point of intersection, is got. This is fairly fixed, since [Rahu] from the point of intersection, is got.

This is fairly fixed, since [Rahu] moves very slowly (about 3' per day). The correct declination of the moon and its variation is now as easy to compute as those of the sun. Thus the times of zero difference and difference equal to the sum of the semidiameters are got easily. The possibility or impossibility can be visualzed clearly and correctly when the matter is thus simplified. It must be said in conclusion that the [Mahapatas], made so much of by the astronomers, are not even heard of nowadays in connection with religous rites or fixing auspicious moments.

This article is intended to help historinas of Hindu astronomy, but is not exhaustive. Fear of making it too long has prevented considering the development of other concepts like the heliacal setting and rising of planets, etc. The nonavailability of most Hindu astronomical works in the libraries in Madras had limited the matter of even the topics chosen. I expect others to follow up and supply the need.

THE SYSTEM OF THE VATESVARA SIDDHANTA

Reprinted from the Indian Journal of History of Science,

Vol.4, Nos.1&2, 1969

The Vatesvara Siddhanta (A.D.904) is one of the most famous of Hindu Astronomical works, and cited frequently by writers on Dharmasastra. Albiruni mentions the author with another work of his. But later on the study of the work became so rare that only recently it appeared in print from an only manuscript in the Lahore University Library. The new commentary with which it is printed has masked it speculiarities as a work belonging to the school of [Aryabhata], as also many of the fundamental constants like the different numbers of cycles, etc., and given wrong ideas and numbers instead. This has led to further mistakes in interpretation. Hence, a good deal of research has to be done to salvage these and make them available to scholars for futher study.

The findings are as follows: The [yuge] (i.e mahayuga) is divided into four equal quarters. 72 [yugas] make a [manvantara]; fourteen [manvantaras], without any [sandhi] (1008 yugas), form the [kalpa] or half-day of Brahma; 720 [kalpas form his year, and his life-span is 100 such years. At present, 8 1/2 years and 15 days of Brahma's lifetime has gone, up to the beginning of this [Kali]. The constants denoting the number of civil days, revolutions of the Sun, Moon, Jupiter, Saturn, Mercury, Venus as well as of Moon's apogee, nodes, etc., are presented according to teh [Vatesvara Siddhanta] and discussed in relation to wrong numbers given by the commentary. Several other wrong interpretations presented by the commentary are also dealt with and the correct interpretations given. It is pointed out that in the [Spastadhikara] of the commentary also there are a number of mistake to be rectified.

The [Vatesvara Siddhanta] is one of the most famous of Hindu Astronomical works, and cited by writers on [Dharmasastra] like [Kamalakara Bhatta]. The author is [Vatesvara], son of [Mahadatta], of Anandapura], and the work was comsposed in Saka 826 (A.D. 904), when the author was 24 years old (Madhyamadhikara I, 21). With his other work [Karana

Sara], he is mentione dby Albiruni. But now manuscripts of the work have become rare, so much so that the late M.M.Sudhakara Dvivedi regrets, in his history of Hindu Astronomy, that he is not able to procure a transcript. The Indian Institute of Astronomical and Sanskrit Research, Delhi, got the work printed and publised for th efirst time in 1962, from a manuscript in the Lahore Univesity Library, adding a new commentary in Sanskrit with a Hindi translation fo the same. Now, this commentary has masked the system fo this [siddhanta] by superimposing on it its own ideas and constants, unwarranted by the text, so as not to be available to readers without doing research. This has led to further mistakes in interpretation. Hence, this attempt of mine here. For the present I shall confine myself to the [Madhyamadhikara] mainly.

I give the system first, for the sake of scholars who may not be interested in the research. In the [Mahayuga] (usually called yuga), there are 43,20,000 solar years. The yuga is divided into four equal [quarters], [Krta], [Treta], [Dvapara] and [Kali], of 10,80,000 years each, a peculiarity which it shares with the [Aryabhatiyam]. Seventy-two [yugas] make a [manvantara], and 14 [manvantaras] or 1,008 [yugas] make a [kalpa], which ideas also are peculiar to this system and shared by the [Aryabhatiyam].

(Other [siddhantas] assert that the duration of [Krta], [Treta], [Dvapara] and [Kali] are in the ration, 4:3:2:1, that 71 [yugas] make a [manvantara] and that 14 [manvantaras] with 15 [sandhis] make up the [kalpa] consisting of [1,000 yugas].

To continue, two [kalpas] make a day of Brahma, 360 such days from his year, and his life -span is 100 such years. Now, the commentary has missed every one of the peculiarities of the sytem mentioned above, though, strangley enoguh, the introduction seems to accept some of them, at the same time stating that there are 10,000 [yugas] in a [kalpa].

As for the constants, according to this [siddhanta] there are 157,79,17,560 civil days int eh [yuga] as derived from the statement that there are 158,22,37,560 sidereal days in it. The revolutions of the sun are 43,20,000 and of the moon 5,77,53,336. The revolutions (mean) of Mars are 22,96,828, of Jupiter 3,64,220, of Saturn 1,46,568, and thos e(Sighra) of Mercury are 1,79,37,056 and of Venus 70,22,376, and those of the moon's apogee (ucca), 4,88,203, and of its nodes (Pata) 2,32,238. The revolutions of the Seven Sages are 1,692 in the [yuga].

In the life-span of Brahma the revolutions of the apogee of the sun are 1,65,801, of Mars 81, 165, of Mercury 4,77,291, of Jupiter 13,948, of Venus 1,52,842 and of Saturn 72,974, and those of the nodes of Mars 20,684, of Mercury 988271456418719, of Jupiter 39, 202, of Venus 19,6127,48,06,36,835 and of Satrun 1,542. The great circle on the celestial sphere is in [yojanas], 12474720576000 (Madh. VII,3). Of these constants, the commentary gives wrong numbers for the italicized digits, and obliterates the revolution of the Sages.

This [siddhanta] says that time with its indicators, the sun, moon and star-planets were created together with Brahma and coexistent with him, the moment of their sreation being mean sunrise at [Lanka], and the say Saturday. (The other siddhantas say the sun, moon, etc., are created anew in every day of Brahma, getting into dissolution at night-fall; while the

[Aryabhatiyam is silent on this point.) At the beginning of the present [kalpa] eight and a half years together with fifteen days have elapsed sine Brahma's creation. The period of 15 days has to be established by reasoning, since the words giving it in I. 10, have been spoiled by the scribes. (Aryabhata) leaves out the 15 days, which may be a variation or an approxiamte statement. According to the other [siddhantas] 50 years of Brahma's life have passed). Six [manvantaras] and 27 [yugas] in the seventh have gone in the current Kalpa, and three quarters (not nine-tenths) of the twenty-eighth) [yuga], up to the beginning of the present [Kaliyuga], which falls on a Friday accoding to all. To link this with the present time it may be stated inpassing, that this, as also all other [siddhantas] (except the Surya Siddhanta] school, not mentioning any era), say that the ['Saka era] began 3,179 years after [Kali] set in. (The Aryabhatiyam gives the Kali year direct).

Now for the [Ksepas] for the beginning of [Kali]. Alo but two of the [yuga] revolution being divisible by four, the [Ksepa] is naturally zero for them. Of the two exceptions, for the moon's node it is six [rasis] and for its apogee it is three [rasis], as mentioned in IV,55. But the commentary here makes the absurd statemtn that these [Ksepas] are for the beginning of the [Kalpa]. The [Ksepas] of the longer period revolutions can be found by multiplying the numer of revolutions by 24798639, viz. the quarter [yugas] since creation, and dividing the product by the quarter [yugas] in Brahma's life-span, viz. 100X360X2X1008X4. They are:

Sun apogee 2r 18o 51' 37''

Mars " 4 8o 50' 50''

Mercury " 7 16o 42' 54''

Jupiter " 5 22o 48' 31''

Venus " 2 20o 3' 26''

Saturn " 7 26o 55' 4''

Mars node 10 20o 10' 12''

Mercury " 11 10o 19' 54''

Jupiter " 9 0o 54' 2*''

Venus " 5 24o 1' 56*"

Saturn " 8 20o 1' 0*"

The number of civil days from creation to the beginning of [Kali] is 9782551985550210. ( Here too the comentary gives wrong numbers for the italicized digits, or even omits or adds digits. The verse giving the [ksepas] of the nodes of Jupiter and Venus is missing in the text.)

Now, proceeding from [what is undisputed], I shall build up the whole structure given by me as the correct one. Let me begin with showing that the number of civil days in teh [yuga] is 1577917560, and not what it will be according to the commentary, four days more. Using III 26, the civil days in a yuga=4320000X13149313./. 36000=1577917560, as I have given above. Again, using III 23-25, the solar months in the [yuga]=4320000 x12. The

[adhimasas]=66389X4320000X12./.2160000=1593336. The synodic months = solar months+[adhimasas=5343336. The tithis=synodic months X 30=1603000080. The [avamas]=tithisX209021./. (total tithis in the yuga ./. 120)=25082520. Therefore the civil days =tithis -avamas=1577917560. Thus we see it is not 1577917564, as given by the commentary. III 12, 13, 14, 15, each one can also independently give the same number of civil days. But everywhere the commentary assumes a year of days 365-15-31-15, (this is Aryabhata's), and proceeds with the proof, unaware that this will make the yuga days equal to [Aryabhata's]1577917500,64 days short of what the number willbe even according to it. Adding the sun's revolutions in the [yuga], the sidereal days= 1582237560, and not 1582237564 as given by the commentary under II, 1. The mistake lies in its interpreting [jaladhara] as 4, while it actually means zero (Jaladhara=abhram=0).

That [Krta], etc., are equal quater [yugas] is got by the statement [Sadrsanghrayastryah] in I,10. Therefore they are not in the usually givne proportion, 4:3:2:1, as mistaken by the commentary. This is also confirmed by X,7, where the author defends [Aryabhata] against Brahmagupta, on this point. Therefore, the civil days in each quarter [yuga] are a fourth of the total, i.e. 394479390.

I shall now show that according to this [siddhanta]8 1/2 years plus 15 days ,etc., have gone since creation, and there are 72 [yugas] in a [manvantara], etc. III, 18 states that the number of days since creation up to [Kali] is the product of 24798639 and days in a quarter [yuga] (found above to be 394479390) and has given it, viz., 978255198 5550210. ( It must be noted that the commentary omits the two digits 5, 0, and thus makes the number a hundredth of the actual. Strangely enough, it asks us to use for getting the product, the days in a quarter [kalpa] instead of [yuga], which will make the number more than one thousands times the actual! The disputed 15 days or the extra duration of the [kalpa] and [manavantara] in this system are too small to help removing the discrepancy; they will only mar the complete agreement in the digits). Therefore, it is obvious that 24798639./.4=6199659 3/4. I shall show this can be so only with the 15 days more. In 81/2 years there are 8 1/2 X360X2 [Kalpas]. In the next 15 days there are 30 [Kalpas]. Thus there are 6150 in all, in which there are 6150X1008=6199200 [yugas]. The [yugas] gone in the present [kalpa]=6x72+27+3/4=459 3/4. Therefore, the total [yugas] gone=6199200+459 3/4=6199659 3/4, as required. Omitting the 15 days or making the [kalpa] equal to 1000 [yugas], etc., as stated by the commentary, will give 8 1/2 X 360 X 2 X1000+6+71X7.4+27.9=61204567 [yugas] only. Note the difference. Our conclusion is also reinforced by the agreement in the [Ksepas] of the apogees and nodes, as we shall see later. It may be mentioned in passing that the releveant portions of III, 18, and 1,9,10, must have been wrongly read by the commentary or tampered with by somebody; as for instance, [ccandradrayo] in 1,9, must be [ddasradrayo]; [tatha] in 1, 10 must be [tithi] (=15), [bhujabhra] in II, 7 (peculiarly interpreted by the commentary as zero zero), must be [gajabhra] and [s'ara s'ara] in III, 18 must be [khasarasara s'ara]. ( This last one will also make the metre correct.)

We shall now show that the first day of creation according to this [siddhanta] (as also of every kalpa) is Saturday. All [siddhantas] agree that the first day of Kali is Friday. Dividing out the daus from creation given above by 7, the remainder is 6, which means that the week-day of creation is 6 days previous to Friday, i.e. the day immediately follwing Friday, viz., Saturday. This is also mentioned in III, 19, by [mandasitadyo vyastagananaya]. Also logical reasoning, on the pattern of the position taken by the author in X,9-11, points to Sturday as the first day. From this it follows that the first day of every [kalpa] is also a Satruday, because the days in the [kalpa], since they have a factor, 14, must be dividble by deven. In the light of out finding, [vyomacaradhipa] in III, 1, should be taken to mean [grahnanayaka] or Saturn, or it must be read as [vyomacaradhika] and interpreted as the planet having the longest period or the greatest orbit, i.e. Satrun. In II, 9, [raverdine] must be a misreading for [sanerdine]. Let it not be though that the acceptance of the 71 year [manvantara], etc. attributed to the system by the commentary, will give Sunday either for the kalpa or creation. It will not, as can be tested ealily. (UnderX, 11, by the way, in trying to show that the first day of the current kalpa is Thursday according to [Aryabhata], the commentary makes several mistakes, a patent one being the omission from reckoning of the 27 [yugas] in teh current manvantara.)

We shall now pass on to the revolution constants. We have stated that at the commencement of Kali the [Ksepa] of the moon's node is six [rasis] and of its apogee, three [rasis]. The revoluitons of these being whole numbers and given for a [yuga], and three-fourths of the [yuga] having gone at the beginning of [Kali], three-fourths of th enumbers of the respective revoultions celarly give half and a fourth of a revolution , i.e. six and three [rasis]. At the commencement of the [yugas] the [ksepas] must be zero, and therefore at the commencement of the [kalpas] too. But the commentary makes them six [rasis] and three [rasis] at the beginning of the kalpa, stating [kalpadau=srstyadau], exhibiting also ignorance of the fact that according to this [siddhanta] there is no creation of the planets at the beginning of every day [kalpa]. Evidently [kalyadau] has been misread as [kalpadau].

I have given the [yuga] revolutions of Mercury as 17937056, while the commentary gives 80 for 56. Let us verify : by V, in a pyuga],i.e. for 43,20, 000 solar years gone, the cycles of Mercury=4x4320000 + 4320000x20533./.135000=17280000+657056=17937056, agreeing with what I have given. We can reinforce this confirmation by VII, 9, giving Mercury's orbit as 695472+ 11424/5605333 [yojanas]. For, using the correct number of cycles got above in VII, 5, we get Mercury's orbit=4320000x57753336./. (20x17937056)=695472+11424/ 560533, the same as given by VII,9. This will also show that the correct reading of VII,9 is , xÉäjÉÉ%MÉ´ÉänùºÉɪÉEò xÉiÉÖÇʦÉÊxÉxɺɨÉÖuù ¶É榃 SÉxuèù:* ºÉֶɮú®ú´ÉÉ%CMÉ IɱɴÉè:.....' Therefore, in 1,13 giving Mercury's cycles [Kharasairhi] should be read as [rasagni]. This will also correct the error in prosody.

The cycles of the Seven Sages are given in 1, 15, as 1692 (bhujagostayah) per [yuga]. But the commentary misses it, interpreting the verse with the next one, and making the remark, the number 1692 seems to be useless: ¦ÉÖVÉMÉÉä%]õªÉ:½þÊiÉ ÊxÉilÉäEÆò |ÉÊiɦÉÉÊiÉ' (There is

another comment to match, under II, 5, '+lÉÇÊxÉ{ÉÉiÉ:, +lÉÇÊxÉ{ÉÉiɺÉÆYÉEòÉ:......+lÉÉÇiÉ ®úÊ´ÉSÉxuùªÉÉä:ÊuùMÉÖÊhÉiɨÉhɪÉÉäMɺªÉ xÉÉ¨É +lÉÇÊxÉ{ÉÉiÉ ½þÊiÉ *' Apart from the unheard of nature of the name, the commentary is unaware that the [sandhi] in the text cannot give the word. The word is '´ªÉÊiÉ{ÉÉiÉÉ :' and the 'r' is due to [sandhi]. [Vyatipata] here means the two types, [vyatipata] and [vaidhrti], hence [dvigunita], vide the same instruction given by the [Aryabhatiyam], [Kala 3.]

We now pass on to the revolutions given for Brahma's life-span and the [Ksepas] relating to them. Let us start with the cycles fo the apogee of Mercury, 477291, and of the nodes of Mars, 20684, about which there is no dispute. The [Ksepas] for [kali] can be found [conveniently] in two, steps, first up to the beginning of the current [kalpa], and from then up to the beginning of the current [Kali]. (Those who do not want to use this easy method may multiply the cycles by the [yugas] in Brahma's life-span to get the required Ksepas). To get the first part, the fraction to be used for multiplication is the Kalpas gone since creation, by the total kalpas in Brahma's life-span, i.e. (8 1/2x360x2+30)./.(100x360x2)=6150./.72000=41./.480=41c ./.40 =c+c/40. ( This is given by the first half of IV, 53, [Khaka]=40, being misread as [svakha], and [grhadayo] as [gradhadayo] and the meaning given as, 'º´É¶ÉÚxªÉ ¦ÉHò±É¤vɪÉÖiÉ ¦ÉMÉhÉÉ:' with the comment,'+jÉ º´É ®ú´É½þiɱɤPɪÉÖiɦÉMÉhÉÉ : ½þiªÉªÉÞHÆò |ÉÊiɦÉÉÊiÉ !!).

The multiplier to get the second part is the [yugas] gone in the present [kalpa] amounting to 459 3/4 as already found, and the divisor is the same as before. If this is given in minutes of arc, again c being the cycles, the secondpart=cx12x30x60x459 3/4 ./. (72000x1008)=cx613 ./. 4480. (The fraction 613 ./. 4480, or its equivalent, must have been given by the last word of the first half, and the second half of IV, 53, spoiled beyond recognition.) Now, using the two fractions, the [ksepa] of the apogee of Mercury is the [rasis], etc., 7-8-15-0+0-8-27-54=7-16-42-54, exactly as given by the text in IV, 57. Applying to the nodes of Mars, the Ksepa is, 9-3-0-0+1-17-10-12=10-20-10-12, again exactly as in IV, 59. (Incidentally, this confirms again that up to this Kalpa 8 1/2 years and 15 days have gone, and in this kalpa six [manvantras] of 72 [yugas] each have gone, and 27 3/4 [yugas] in the seventh [manvantara], up to Kali. Otherwise we cannot get the agreement in the numbers.)

The above procedure gives us the means of verifying the cycle numbers on the one hand, and the Ksepas on the other, when in doubt about one of them. We shall use it. The revolutions of the apogee of Mars are 81165. From this the Ksepa is 10-3-45-0+6-5-5-50=4-8-50-50. The commentary gives 5 instead of 8 here, interpreting [dhiyah](wrong'y read dhayah) as 5. We see it is 8 according to [Vatesvara]. This is confirmed by [spasta] IV, 11, giving the synodic period of Venus as 584 ({ɪÉÉäÊvÉ ÊvÉ {É´ÉxÉÉ:, where vÉÒ must be taken as 8, but the commentary even here taking is as 5). Thereofre [drhikrtanka dahanendavo], giving Jupiter's apogee-cycles must be 13948, not 13945, as given by the commentary. Further, only the former will give the Ksepa, 5-22-48-31, and not the latter, confirming our conclusion. (The latter will lessen the

Ksepa by morethan 3 [rasis] In the case of Mercury's nodes, to agree with the Ksepa, 11-10-19-54, [sara] (in I, 18a) must be read as [rasa], of course [dhi]as also [mati] (meaning the same), each being interpreted as 8; and we get 988271456418719, vannot give the Ksepa. In the case of Venus's nodes in 1, 18b, 3 for [sikhi] (meaning fire), is most probably intended (thought it may also mean arrow or 5), for in I, 14, it is used for 3, and no author will be confusing (though the word may have two meanings). [dhi] is of course 8. We cannot verify the cycle number here because the verse giving the Ksepa is missing, topgether with that of Jupiter's nodes.

From Venus's number of cycles of apogee, 152842, the Kespa must be 2-20-3-26, and not the commentary's 2-20-13-2; and therefore the reading should be 'ªÉ¨É±ÉÉè xÉ®úµÉɺjɪÉÉä ®úºÉªÉ¨ÉÉ:' From Saturn's number of cycles of apogee, 72974, its Ksepa must be 7-26-55-4, and the reading should be corrected as '¨ÉÖxɪÉÉä %¼MÉP¶ÉÉä%IÉ®úÉ ´ÉänùÉ]õ': Using Saturn's cycles of nodes 1542, the Ksepa got is 8-20-1-0, and the reading is to be corrected as '+¹]õÉä xÉ®ú´ÉɶSÉ ¦ÉÚ:®ú´ÉÉä'. From the Ksepa for the sun's apogee, the number of cycles should be 165801, and not 16511 as given by the commentary. The latter does not yiled the given Ksepa but 5-0-54-13. Therefore in 1, 16, '{ÉÆEòVÉ' is a misreading for '®ú´ÉÆ MÉVÉ'.

The first six significant figures of the node cycles of Mercury and Venus can also be verified by using I, 20: The [sighra] cycles of Mercury and Venus, for Brahma's lifetime, calculated from their [yuga]values, are respectives 988271746560000 and 196127640576000, and we find that their node cycles given by us agree to six significant figures. The difference in the other places is of course expected adn necessary.

The [yojana] measure of the great circle on the celestial sphere, given in VII, can be verified by using VII,5, and thereform it is 4320000X57753336./.=12474720576000 as given by me, there being no doubt about the cycles used and the method. Therefore, the commentary's 1222514920576000 making the number about a hundred times what it is, is wrong. So we see that in VII, 3, [graha] should be interpreted as 7, and not 9, as the commentary does (in II, 7, the commentary itself interprets [graha] as 7), and that '¦ÉÚiÉi´Éº´É'(=2251) is a misreading for '¦ÉÚ¦ÉÖ EÞòiÉ' (=47). The commentary's reading also in against prosody uses 'º´É' for 2, not generally used.

There are scores of other places also where the commentary gives wrong constants, as for example in V,1. These are fundamental, and are used in a number of other places to derive and give other multipliers, divisors and Ksepas for shortening the work. It is funny to see how the commentary ministerprets many of them and adduces proofs, of course wrong, for them too! I have mainly taken the [Madhymadhikara] for study here In the [Spastadhikara] too there is a lot to be corrected.

THE SCHOOL OF [ARYABHATA] AND THE PECULIARITIES

THEREOF

Reprinted from the Indian Journal of History of Science, Vol.4,

1&2, 1969

It is possible to classify early Hindu astronomical works into specific schools, on the

strength of certain peculiarities of each. On esuch school is that of [Aryabhata], revealed in his [Aryabhatiyam]. [Prabhakara], [Bhaskara] I and the Kerala astronomers, [Govindasvamin], [Haridatta], etc.. belong to this shool, as also [Vatesvara] in the north.

[Aryabhata] and [Vatesvara] give peculiar lengths for the different cons. According to them the [yuga] though consisting of th usual 4,320,000 years, is divided into four equal quarters,[Krta], etc., instead of in the ration 4:3:2:1. Seventy-two [yugas] make a [manvantara] and 14 [manvantaras] or 1,008 [yugas] constitute the [kalpa], with no [manvantaras]-[sandhi]. Time and its indicators-the sun, the moon and the star-planets-were created together with Brahma and considered to last as long as Brahma lasts, instead of being dissolved and recreated in each [Kalpa] forming the day-time of Brahma. The moment of this creation was mean sunrise at Lanka and the day, Saturday. At the beginning of the current [kalpa], eight and a half years and 15 days of Brahma's life had elapsed. At the beginning of the present [kali yuga], six [manvantaras] and 27 3/4 [yugas] had gone in the [kalpa]. [Vatesvara] gives the revolutions of the apses and nodes (excepting those of the moon) as so many cycles in the lifetime of Brhama, and therefore all this is not of mere academic interest.

We also find variations from other schools in the number of cycles of the mean or [sighra] motions, as also in the degrees of epicycles. An important peculiarity is the use of the true hypotenuse in computing the equation of the centre (condemned by Bhaskaracarya II). Another peculiarity is dispensing with the first operation (i.e. the application of half-equation of conjunction of the mean) in the case of Mercury and Venus.

It is possible to classify early Hindu astronomers and astronomical works into specific schools on the strength of certain peculiarities of each. One such school is that of [Aryabhata], as revealed in his [Aryabhtiyam] (A.D.499). [Aryabhatta] has written another work known as his [Ardharatrika] system which has been adopted by Brahmagupta in his [Khandakhadyaka] (A.D. 665), Which he claims to give the same result as the [Ardharatrika], whose constants and other peculiarities are given by [Bhaskara] I (c. A.D.628) in his [karmanibandha] (better known as Mahabhaskariyam), chap. VII, and which, the followers of [Aryabhata] think, has been intended only to be examined and refuted, as [Govindasvamin] calims in his [Bhaskariya] [Bhasya] under VII, 35. This [Ardharatrika] system is representative of another school to which belong the [Saura Siddhanta] condensed by [Varahamihira] (V.M.) in his [Panca Siddhantika](P.S.,c.A.D.505) and the [Paulis'a] often quoted by [Bhattotpala] in his commentary on teh [Brhatsamhita]. The school of [Aryabhata] is represented by his direct pupils like [Prabhakara], [Bhaskara] I who has written the [Aryabhatiya Bhasya] the [Aryabhatiyam], Govindasvamin] its commentator, [Haridatta], the author of the [Grahacaranibandhana], and a long line of illustrious Kerala astronomers, as also the famous [Vatesvara] of [Anandapura]. Besides the peculiarities we are going to talk about, the fact that these astronomers belongs to the west coast from [Valabhi] in the north to the Cape in the south, and that they were conscious of belonging to this schools, as gathered from statements made in their works, point to the existence of the

school.

One important peculiarity of the school is that its system of eons differs from that of other schools and the [Puranas], as can be gathered from the [Aryabhatiyam] and the commentary thereon, and from the [Siddhanta] of [Vatesvara], by [gunopasamharanyaya]. According to this, in the [mahayuga] (also called the yuga), there are 4,320,000 years; but this is divided into four equal quarters (Krta, treta, dvapara and kali) of 1,080,000 years each, as seen from Vat. Sid., I,10:'four equal quarters', [Aryabhatiya Bhasya] under [gitika]3, 'for us the four quarters of the [yuga] are equal ', and as inferred from [Vat.Sid., madhya]: X,7, and [gitika]3. Seventy-two [yugas] make a [manvantara], and 14 [manvantaras] make a kalpa (half-day) of Brahma, consisting of 1,008 [yugas]. These are not only mentioned in various places, but also can be inferred from various Ksepas involving these periods. (On the other hand, it is well known that other schools, and the [Puranas], say that the duration of Krta, etc., are in the ration 4:3:2:1, that 71 [yugas] make a [manvantara], and that 14 such [manvantaras], plus 15 [sandhis], each equal to 2/5 [yuga], make up the kalpa fo 1,000 [yugas].

To continue, two kalpas make a day of Brahma, 360 such days form his year, and his life span is 100 such years, as commonly held. But this school asserts that time with its indicators-the sun, the moon and the other heavenly bodies -was created with brahma (the moment of their creation being mean sunrise at Ujjain) and co-existent with him, unlike the statement of others that these bodies are created afresh in every day-time kalpa and get into dissoultion during the night-time. At the beginning of the present kalpa8 1/2 years with 15 days have elapsed since Brahma's creation according to [Vatesvara]. (According to others 50 years have just passed and we are in the next day.) [Aryabhata] is silent on this point, since he has no use for it as he does not give the slow-miving cycles of apogees and nodes. But [Bhaskaracarya] II, in his [Siddhanta][Siromani](Ganita:madh.:[kalamana]: 26), gives it as the opinion of some that Brahma's age is 8 1/2 years. We do not know whether this is made as an approximate statement, or as the opinion of a variant of the [Aryabhata] school, adn held by [Aryabhata] himself. But we can be sure that in Vat. Sid. madh.: I, 10, this extra period of 15 days has been given (but masked by scriptory errors) as a study of madh: III, 17-18, and IV, 56-60, will show. (Before proceeding I wish to state that the commentary of the only printed edition available of the [Vatesvara Siddhanta] exhibits spoiled constants in most places, so that they contradict one another, and the correct constants have to be determined by investigation and research, which I have done in a separate paper with the title, 'The System of the [Vatesvara Siddhanta']. What I am using here is part of that, and for fuller knowledge or in doubt that should be consulted.)

In III, 17-18, the number of quarter [yugas] from creation up to the beginning of [kali] is given as 24,798,639, and this multiplied by the days in one quarter [yuga](=394,479,390) gives the days from creation up to Kali as 9,782,551,985,550,210. Dividing the quarter [yugas] by 4, the [yugas] from creation is 6,199,659 3/4, and this must be got if we compute the number. With the 15 extra days, the (Brahma) days from creation are 8 1/2

x360+15=3,075. The number of kalpas are 3,075x2=6,150. In this period there are 6,150x1,008=6,199,200 [yugas]. The [yugas] gone in the present kalpa are 6x72+27+3/4= 459 3/4. Adding, we get the total number of [yugas] to be 6,199,659 3/4, the same as we want. Thus we see it is 15 days more than 8 1/2 years. Our conclusion is reinforced by the fact that only with the 15 days more can the Ksepas, given in IV, 56-60, be got. The cycles fo the apses and nodes of the sun and the star-planets have been given as so many in Brahma's life span of 72,000 kalpas (or 72,576,000 yugas) in Madh., I, 16-19. Using these it can beseen that for the 6, 199,659 3/4 [yugas] gone up to the present Kali, with the 15 days included, we get these Ksepas to the second.

But though [Aryabhata] has not given the number of the above cycles of the apses and nodes, [Bhaskara] I gives evidence of his knowledge of their motion when he says in his commentary under [gitika] 9 that [Aryabhata] has given these in loose verses. But as the only twoanuscripts texts of the [Bhasya] available are much vitaited, it is not possible to connect these with the cycles given by [Vatesvara].

Another point of distinction is that this school asserts that a great circle on the celestial sphere, measured in [yojanas], is 12,474,720,576,000. This is the distance moved by the sun, the moon or the star-planets in a [yuga], Hindu astronomy asserting that all have the same rate of motion. In the moon's orbit it is 10 [yojanas] per minute of are according to thsi school, as given by [Aryabhata] in [gitika]4, by [Bhaskara] I in [Mahabhaskariyam], VII, 20, and by [Vatesvara] in his [Siddhanta], [Madh]., VII, 3-7. They all give the same manner of getting it, 10 x the minutes in a circle (=21,600) x the moon's cycles per yuga (=57,753,336). As all schools hold that the absolute motion is the same, the ratios of the measures of the celestial sphere, according to them, will give the inverse ratios of the absolute values of the [yojanas] used by them. For example, the celestial sphere is 10 x32,400x57,753,336 according to the [Ardharatrika] system, and the [Surya][Siddhanta] school. Therefore the [yojana] measure they use is 2/3 that of the [Aryabhata] school.

Another important peculiarity of this school is the use of the true hypotenuse in the computation of the equation of the centre. The use of the hypotenuse in the equation of conjunction is common and accepted by all schools, as justified by the eccentric or epicyclic theory of the motion of the planets, which can be readily seen from a geometrical representation of the motion. By the same logic, the hypotenuse should be used for the equation of the centre also, the theory being essentially the same. That is why this school uses it, as a geometrical consequence of this theory set forth by [Aryabhata] in [Kala-Kriya]: 17-21, combined with the theory of uniform motion given in [Kala]:12-14. Thus int eh [mahabhask]., IV, 8-12, the manner of getting the true hypotenuse as based on the theory of epicycles is given, and in 19-20 the same as based on the eccentric theory. In 21, the approximate sine equation of the centre is asked to be multiplied by the radius and divided by the true hypotenuse to get the correct sine equation fo the cntre. [vat.Sid. Spastadhikara], II, 3-4, gives the method of getting the true hypotenuse, and III, 11 instructs its use to divide the approximate equation of the centre to get the correct one.

The use of the hypotenuse is not only a logical result of the theory, but it will also give a better result. It supplies part of the second term of the modern correct equation of the cnetre. Negelcting powers of e (eccentricity) higher than the square, the first two terms are 2e sin m-5/4e sin 2m, where m is the mean anomaly reckoned from the higher apsis, as in Hindu astronomy. The distance between the centres of the original and eccentric circles is equal to 2e sin m./. h (=hypotenuse). But h=sinm./.sin (m-eq,cent.), if the radius of the eccentric circle is taken as unity. Therefore sin eq. cent.=2e sin m xsin (m-eq.cent.)./. sin m=2e sin (m-eq.cent.)=2e sin (m-2esin m) since the eq.cent. is small)==2e sin m-4e as the coefficient of the second term, instead of the correct 5/4 e, it will not make much difference, being the second power of e. Also, the point is that we get the term instead of neglecting it. Using the moon's epicycle of 31 1/2 degrees, which gives 7/80 as the value of 2e, we get for the second term-13' sin 2m, the same as the modern correct one. (The apparent complete agreement is due to the Hindu coefficient of the first term being defective by about a fifth.)

[Bhaskaracharya] Ii discusses the point, why other schools do not use the hypotenuse for the equation of centre. He says that some do not use it thinking that the difference is small. This depends upon what we consider argument he gives, quoting his master Brahmagupta, that the theory itself is that the epicycle instead of being uniform, is proportionate to the true hypotenuse and has to be multiplied by it and divided by the radius, and therefore the division by the true hypotenuse and has to be multiplied by it and divided by the radius, and therefore the division by the true hypotenuse is cancelled out, is untenable, for this kind of argument helps only to shut out a tolerable good theory already existing and nothing more, and is just a way of esacape, as pointed out by [Caturvedacarya] in his commentary on the [Brahmasphuta] [Siddhanta] (cf. Sid. Siromani: Gola: Chedyaka; and commentary thereon).

Another peculiarity is dispensing with the first of the four operations (according to this school) of applying half the equation of the centre to the mean, in the case of the Mercury and the Venus. This is according to [Aryabhata], [Bhaskara] I and [Haridatta]. But [Vatesvara] seems to vary even from this (if the printed text is correct), dispensing with the first two opertations, and chaging the order of the third and fourth, though he himself as also other astronomers all give the application of the equation of the centre as the third, and the application of the equation of conjunction as the fourth. The question naturally arises whether this is an improvement in the method, so as to give a better result. In the first place, even the original method, with its four steps, is far from satisfactory. The third and fourth steps alone are capable of yielding correct results (modern astronomy uses only these), provided (i) the true sun is used for the [sighra] of [Mars], Jupiter and Saturn and the [madhya] of Mercury and Venus, (ii) the true hypotenuse of the sun is introduced in the proper place and (iii) the equation of the cntre is applied to the [Sighra] of Mercury and Venus instead of their [madhya]. It is noteworthy that in the case of Venus, Hindu astronomy actually does (i) though purporting to apply the equation of centre to Venus. Venus's being less, is masked by the sun's. That is why its apogee given is really that fo the

sun. But otherwise dispensing with the [half-manda] application is not going either to make or mar the correctness of the result. As for the very peculiar instruction of the [Vatesvara Siddhanta], to apply the equation fo conjunction and then apply the equation of centre is quite wrong. But [Vatesvara] is too astute to make this mistake, and the text in the printed edition (got from an only manuscript) must be wrong.

Another peculiarity is that the [yuga] is divided into two parts, [utsarpini] and [avasarpini] on the one hand, and [susama] and [dussama] on the other, as given by the [Aryabhatiyam](Kala 9) and the Vat. Sid. (madh., II, 6). Even from a time as early as [Bhaskara] I, who is later than [Aryabhata] only by a little more than a century and who comes in a direct line of pupils, commentators are not sure of the import of these divisions (vide commentary under Kala 9 in the Aryabhatiyabhasya of Bhaskara I and Bhatadipika of Paramesvara). A general interpretation is that during the first half of the [yuga], the longevity, strength, ability, etc., of creatures increase and, in the second half, these decrease, an idea current among the Jains.

Only three sections of the [Vatesvara Siddhanta] are available to us in print now. When the rest is also publised, we can come across more items peculiar to this school.

SOME MIS-INTERPRETATIONS AND OMISSIONS BY THIBAYT AND

SUDHAKARA DVIVEDI IN THE PANCASIDDHANTIKA OF

VARAHAMIHIRA*

The [Pancasiddhantika] (PS) of [Varahamihira] (VM) is very valuable as a source of information on the state of Hindu astronomy before the 6th century A.D. Because in this work VM has condensed the basic texts of five astronomical schools in vogue before him. VM has depicted herein, very faithfully, the constants and methods of the said five schools. The work was edited in 1889 by Thibaut (T) and Sudhakara Dvivedi (S) on the basis of two badly vitiated manuscripts available to them. Since no ancient commentary was to be had, S added to the edition a short Sanskrit commentary, and T supplied it with an English translation and explanatory notes. Unfortunately, certain important verses of the text have not been touched by T-S, as being obscure, and in amny places their interpretation is wrong, some being seriously so. These mistaken ideas have been repeated and made the basis of further research by later scholars, who perhaps, did not have the time or ability to go into the relevant verses and draw consclusions for themselves. During the more than eight decades now, since the work was issued, nobody has re-edited the work, supplying the omissions or rectifying the mistakes. There have just been two reprints (1930, 1968), perpetuating the errors in the earlier edition. An attempt is made in the following pages to correct some of the more important errors,

1. PS III. 1-3 gives the sun according to the Paulisa (P). The meaning is : "Multiplying the days from epoch by 120, subtracting 33, and dividing by 43831, the revolutions etc. of the 'mean' sun is got. 20o added to this is called Kendram, ('anomaly').

For each [Rasi] of Kendram subtract one for one, 11',48', 69',70',54',25', and then add 10', 48', 70', 71', 54', and 25'. The 'mean' sun becomes true.

This straight interpretation supplies all the information required to get the true sun. For e.g., let the days from epoch be 620 (120x620-33)./.43831=1 Rev., 8 [Rasis], 10o52'=8 [Rasis, 10o 52'=mean' sun. This plus 20o, i.e., [Rasis], 0o52', is the Kendram. So, for the 9 full [Rasis] have to be applied,-11',-48',-69',-70',-54', -25',+10', +48',+70', and +2' for the 52' left over, i.e., in all, -147' (= -2o 27). Therefore the true sun=8R 10o52'-2o 27'=8 R 8o 25'.

It can be seen that by [madhyama-Surya] the [Siddhanta] actually means the mean sun plus a large part of the equation of the centre at epoch, and that by [Kendram] it means not the anomaly as we understand it now, but only the argument to get the correction to the so called mean to become true. (It must be noted that the [Vakyakarana] gives a similar method for the sun, in the shape of what are called [Bhupadivakyas], to be applied to the sun from the beginning of the true year).

But the translation of T-S says, "Add 20(o) to the [Kendram," and raises to questions: (1) How are we to get the Kendram without the [ucca] being given, and (2) why should 20[o] be added to the [Kendram] (which is not even used), while to use the equation of the centre for the [Rasis] would be easier and better? If the real anomaly plus 20[o] be called Kendram, why not call the so called mean plus 20[o] so ? Then, against all canons of interpretation, they add together to quantities instructed to be subtracted, given in one sentence, and the quantities instructed to be added, given in the next sentence, and say that these are the equation of the centre values, and thereby face the further difficulty that, according to their interpretation, the instruction where to add and where to subtract has not been given. Why at all should each value be split into two almost equal parts and the two series given in two different places? Further, they do not see that their made up values are so erneous as to represent anomalies 20[o] off even.

2. In PS III 15, the true sun or moon, computed for the mean sunset at Yavanapura (Alexandria), is reduced to the true sunset of a desired place. The meaning is:"From the [nadikas] of [desantara] in longitude (given by 14) if the sun is in the six signs, Mesa to [Kanya], and add the same if the sun is in the next sixsigns. Subtract the motion (of the moon) calculated for this period (from the sun or moon) already found. We get them for the sunset for the sunset of the place.

T-S have missed the importa of the verse and have translated the last part as : "Reject the remaining ascensional difference." If it means this, need it be said? From the instruction it can be inferred that the work is for places in India.

3. In PS III. 10, the computation fo two [yogas], viz., [Vaidhrta] and [Vyatipata], (also called Sarpamastaka), that are days of [sraddha], are given, (Note that the subsequent verses also give the times for the propitiation of the manes.) The meaning is : "When the sum of the true sun and moon is equal to 12 signs, then is the yoga, [Vaidhrta]. When the sum plus 10 [naksatra] segments (ie.,4R 13[o] 20') is equal to 12 signs, then is the yoga, [Vyatipata].

The exact time of day (of their occurrence ) should be sought by using the degrees remaining over (the 12 signs)" (The technical Vyatipata is given in 22). Not realising the purpose of the verse, T-S have unnecessarily chaged [cakre] into [satke], and given a meaning not intended here.

4. PSIII.28-29 give the [Paulisa Rahu] (moon's nodes). To know the position of the node at any time it is necessary that (i) its longitude at epoch should be known, and (ii) its motion during the time from epoch, which should be subtracted from (i) to get the present position.

Verse 28 gives (ii) : "Multiplu the days from epoch by8, and divide by 151, and add minutes equal to the revolutions, to get the degrees of motion during the period."

Then 29 gives (i) and how to use (ii), thus: 7R 25[o] 59' are [Rahu's] (longitude at epoch). The first, i.e. the motion given in 28, being subtrated from this is [Rahu's] Head, (i.e., ascending node). This plus six signs is called [Rahu's] Tail, (i.e., descending node)."

Note that from this we can see the ascending node at epoch, (Saka elapsed 427, Sunday, 37-20 [nadika] from sunrise at Ujjain, which is 509432 days, 22-40 [nadikas] before mean sunrise at Ujjain on Jan. 1,1900) to be 7R 25[o] 59",i.e., 235[o] 59" Compare: The ascending node computed for the time by modern astroomy is 236[o] 11', [Siddhantasiromani 237[o] 6', Romaka 235[o] 49' and [Saura 236[o] 6'. See the close agreement, which will show the need for the Ksepa at the epoch.

But T-S have mistranslated (29) and obliterated the Ksepa as they have done in the case of the [Vasistha] moon. They have missed to note that if left with (28), and the Ksepa was assumed zero at epoch, there should at least be the instruction to subtract the result of (28) from 12 signs to get [Rahu], the motion being retrograde. The relevant section of S's commentary is nonsensical and T confesses ignorance of the expression [vrs'cika-bhaga rahoh]. S says: "There are already 25', being the length of the scorpion-like limb of [Rahu]. Subtracting this 25' from the [Rahu] found in (28), [Rahu's] Head is obtained. For the fact that [Rahu] has this scorpion-like limb, we have only to depend on the statement of the ancients, no other reason can be given." The fact is that T-S do not realise that [vrcika-bhaga]... means 7R 25[o] 59'. This method of stating position is common, as for instance, [sardhah pancalino bhogah (XVIII.1), [Kanyamsan][sadvim'satim] (XVIII.2), [navasardhah kanyamsah](XVIII.11), sodasa vrsabhayamsah (XVIII.18), [mithunadala's] ['sodhyate] (VIII.2); and these expressions have mostly been interpreted by T-S in the way intended by VM.

5. III.31. gives the [Pauli'sa] moon's latitude. Here the maximum given is 280', by both manuscripts. But T-S, for no apparent reason, make it 270', by introducing an unlikely emendation. (The Romaka too gives 280', as can be seen from VIII.11 and 14, though, here too, T-S introduce unnecessary, emendations making them however, very strong reasons to believe that the maximum latitude given is actually 380', That this [Siddhanta] got this by extrapolating the latitude 55' for the clipticv limit 13[o] by the proportaion mentioned hereis, as elsehere in this [Siddhanta], according to the degrees adn not their sines, and that VM's "liptasatatraye..." has been mis-corrected by some scribe into "liptasatadvaye..."

6. T-S have not interpreted III.32-37, as being too obscure. Here VM criticises certain astronomers of his time like Bhadravisnu whose works gave wrong tithis and [naksatras], and the people who follow tham blindly. He criticises the Romaka stating that the tropical year given by it will, in course of time, overthrow the luni-solar year of the Hindus based on the fixed [naksatra] system and play havoc with the [Dharmas'astra rites and national festivals like [Rama-Navami]. By giving this criticism his intention that the [Paulisa] [Siddhanta],VM exhibits his intention that the [Pauli'sa] should be followed by the people, alternatively with the [Saura], for which purpose he gives it so very fully, indicating the days of [Sraddha] etc. It may be remembered that [Pauli'sa] includes the [Vasistha mutatis mutandis], as in teh case of its moon and star-planets.

7. In the second half of IV, 1, VM says, "Here, assuming that (meaning by 'that' the diameter of the circle mentioned immediately before,) to be 4[o], the R sines of the eighth parts of signs (i.e., 3 3/4[o], 7 1/2[o], etc.) are given, (with the method of computing them)". From this it is clear that VM has taken R to be 120', and the sine of 3 signs. But T-S have missed this meaning, and have taken the word'that' to indicate the circumference. They say: "Dividing the circumference into 4 parts, the R Sines of the eighths of the signs are given." So, according to them, the radius is not given as being equal to 120, exactly, but what it would make up by adding the last sines given, to the sine of 2 signs got, and by an error arrive at 120'1" for the radius. This seemingly innocent 1" has perplexed scholars, and some have asked me what VM intends by this tail of 1". In comparative studies they quote this extra 1" with meticulous care. Why, T-S themselves have taken it as sacrosanct, and in filling up the gaps in 13-14, with sine-interv als, according to the context, choose the numbers in such a way that they all add up to exactly 120'1".

The fact is that all this is the result of an apparently wilful mistake T-S have committed. The text gives the sine of 60[o] as 43/55// over the 60/ of the first sign, to make up the correct 103/ 55//. By some freak of thought, T-S have emended [hina manubhir visayaih] in IV.9. into [hina manusa garaih], and,given 43/56//, instead of 43/55//. As a result, all the R sines in the third sign are in excess by 1//, and the last is got as 120/ 1//. It passes one's comprehension why T-S emended the correct [visayaih] into [sagaraih], and created all this trouble for themselves and for others.

8. T-S have left IV. 16-18 untouched, with the remark that these probably give the moon's latitude. Actually they give a rule to find the sun's decliantion, and then the worked out declination intervals for every 7 1/2[o] of the sun's longitude. The meaning is : "The R sine declination being, 120/X61./.150=48/48//, giving 24[o], as the maximum declination). The declination intervals for quarter signs are 183/, 180/, 175/, 166/,156/,144/,128/,104/,90/,63/,40/,11/,(total 1440/)."

9. In IV.41-44, the sun's shadow at a given time is computed. Here, in (42) the sentence, [avisodhanena jivasadghninam eva Kartavya] means, "If the degrees of ascensional difference cannot be subtracted from the degrees of the given time, being larger, take, simplythe sine of the degrees of the taken time." Thsi instruction is necessary, in the case when the sun is in the six signs Mesa etc., happens to be larger. But T-S do not see the need for this gives as a general rule: "Of the [nadikas] multiplied by 6, take the sine without any correction, (to the said nadikas)." In fact the question of correction does not at all arise here. Then what can this means?

10. In IV. 50, the moon's shadow is sought to be calculated, as that of the sun done earlier. here the instruction is : "Take the time after sunset forwhich the moon's shadow is wanted. Add to it the time elapsed from moonrise to sunset, if the moon has risen in the daytime. If the moon rises in the night time, subtract the time of moonrise after sunset. Take this as the given time, and work out as for the sun. The moon's shadow is got." T-S interpret the verse in such a way that the moon's shadow would be got in the day -time, when the sun is shining!

11. In IV.52-54, and then in 55-56, VM uses the word [arkagra] or [suryagra] in the sense of [sarikvagram] of [Bhaskara] I or [sankutalam] of [Bhaskara] II. T-S have understood this to mean what is usually called [agra] or 'sine amplitude of the rising or setting sun'. This has resulted in the wrong interpretation of the verses, though with great difficulty they struggle out, with the correct ideas intact.

12. In V.1-3, giving the time for the first visibility of the moon, the correct expression [ayananukulaviksipte], given by the manuscript text, meaning, `if the latitude has the same direction as the northward or southward course of the moon', has been emended by T-S into `apamanukulaviksipte and the meaning given as, 'if the moon's latitude has the same direction as the difference of the declinations', which is nonsensical.

13. In VI.2, VM instructs that 1[o] 36/ should be subtracted from the calculated [Rahu] (nodes) and then the computation of eclipse proceeded with. (This must be an empirical correction by VM himself.) But T-S interpet [sasatkrtikalam hitvam'sam ] to mean 'subtracting 26', neglecting the word [am'sam], and think that this is the same as given by them in III.29, as correction for "RAhu's scorpion-like limb", (which we have discussed in 4 above). But there it was given as 25/. How has the limb grown into 26/ here? If this is meant here, it has been given according to T-S as an item in getting [Rahu] already, and that [Rahu] is taken in this chapter as ready computed. Why again this instruction?

14. The meaning of VI.4 is: "Multiply (moon~ node), which must be less than 13[o] for eclipse to occur, by 5. These are the [vindadis] that should be added to the total duration if the node is greater than the moon, and subtracted if less." The purpose of this correction is to take into account the empirical correction of 1[o]36/ mentioned before. In the computation of the duration in verse(3), 55/ latitude is taken as the limit. This latitude has been computed according to III. 30-31 from the un-corrected [Rahu], and is not what it would be if the empirical correction is made. Hence this correction, thugh rough.

T-S give a wrong interpretation; T says that he does not understands the rotionale, while S says that it is a correction for the changing velocity of the moon, not understanding how. In truth, the small change in the moon's rate of motion during an eclipse will have practically no effect on the duration, for it has to be noted that it is the lunar eclipse that is being dealt with here, and that there is no question of change due to parallax.

15. In VI.5.T-S propound two extraordinary things in their explanation: (i) The maximum latitude of the moon is 240/, which is against what they themselves have said in III.30, as 270/. (ii) If moon~ node i s10[o] there is total eclipse, while actually it is 5[o], giving 21/ limit in latitude.

16. In VI,7, the points of first and last contacts on the moon's rim, with respect to the ecliptic east-west is given. If (moon~node) is zero, these are due East and West points. If (moon~node is 13[o] it is due North or South, as the latitude is South or North. If (moon~ node) is from 0[o] to 13[o], the quarter-rim from East or West points to the North or South point is divided into 13 parts, one for a degree, and the number of parts from East or West points equal to the difference in degrees, is taken. This is a rough rule and considered satisfactory as far as this [Siddhanta] is concerned. But T-S do not understand the instruction, and ring in [Valana] here, which is treated later. But T actually says: "We do not know the reason for direction given in stanza 7 to divide each quarter circumference into 13 parts ." It can be easily guessed from the 13[o] ecliptic limit.

17. VII. 2-4 containg the parallax in latitude required in the computation of the solar eclipse to correct the moon's latitude, and given in the form of a correction to [Rahu]. The obscurity has been cleared by the present writer elsewhere."

18. In VIII.11, the [Romaka] gives a correction to the declination of the nonagesimal, since the corresponding point on the moon's orbit should be taken instead of the nonagesimal, which is on the ecliptic. The instruction is : "Twice the tabular sine of (nonagesimal~ Rahu) plus 1/6 of itself, is to be applied as correction to the declination of the nonagesimal." This will agree with the 280/ max-latitude of the moon given in verse 14, since (2x120/)+2x120/1/6=280/. But T-S make it 240/ unnecessarily, by carrying out unwarranted emendations.

In (14) the latitude of the moon is given by the rule "Multiply the tabular sine of (moon~node) by 21 and divide by 9//. T-S also give this rule. Still they do not see that this would the max-latitude, 120/x21/9=280/. but assume 270/,and arrive at the strange conclusion that the multiplication by 21/9 is approximate! They do not reaise that in verse 11 too it must be 280/, and their 240/, got by their emendation of the text is wrong.

19. In VIII.4 giving the [Romaka] mean moon, the Ksepa must be 10984, deductive, given by [Krtastanava-khikavarjitat], and not the one given by T-S as 1984. My reading will keep the mean moon at epoch in its correct place 356[o] 12/, while T-S make it 359[o]19/. Their value will take the [Tithi] earlier by a quarter of a day, and spoil the agreement with all other [Siddhantas], and modern astronomy also. In addition, this will disagree with the Ksepa for [avama] 514, given in I.10, in getting the days from epoch according to the Romaka.

20. IX. 15, the [Saura] sun and moon's 'orbits' required in eclipse work, are given. The meaning is, "The true hypotenuse of the sun (given in 14) multiplied by 5347 and divided by 40, is the sun's `orbit'. The moon's true hypotenuse (given in 14) multipled by 10, is the moon's `orbit'." T-S give the divisor 120 in th eplace of 40, and the multiplier 3 in the place of 10. According to all Hindu astronomical works, as also the original [Saura], the mean ration of the sun's orbit to the moon's must be the Moon's [yuga] revolutions divided by the sun's, =2406389./.180000 (see I.14)=13.37. Here according to my meaninmg, the ration got is , (120X5347./. 40)./.(120x10)=13.37 exactly. But T-S's corrections will make it, (120x5347./.120)./.(120x3)=14.86 which is absurd.

21. IX.16 gives the true angular diameters of th sun and the moon according to the [Saura]. The meaning is :"Divide 514787 by the true `orbit' of the sun (found in 15) to get the sun's true angular diameter. For the moon, divide 38640 by its true `orbit'." To agree with (15), there should be only six digits in the dividend for the sun, and five digits in the moon's. But T-S make them seven and six, respectively, and use the numbers 5147080 and 333640, by wrong emendations, and get thirty times what the angular diameters would be, and are dismayed by the result, saying, "But 962.6 can represent that quantity, (i.e., the angular diameter) only if it be divided by 30....For some reasong or other, the text, provided it be correct, does not mention the divisor 30. The rule for finding the true diameter of the moon is analogous, and we there also miss the mention of the divisor30."

22. In IX.22, [Saura's] Parallax in longitude is given. The work given here is : "Square the R sine of altitude given in (21) and subtract from 120[o]. From the remainder subtract the square of the sun's [drksepa] (got in 20). take the root, etc." Here T-S say that 120 minus R sine altitude should be deducted from the square of the [drksepa] and the root taken, not realising that the square of the [drksepa] will always be less then 120 minus R sine altitude.

23. In IX.27, a very important correction to the duration of the solar eclipse found in (26), adn thereby to the first and last contacts, is given. T-S have missed it, and give something unrelated to the point. Now, other things being the same, the nearer the solar exlipse is to noon, the longer is the duration. It is this that is taken into account in this stanza. The meaning is: "Find the parallax in longitude for the moment of first contact (found in 26). Deduct from this the bending of the moment of new-moon (i.e., the correction of the new-moon for parallax) already found, or vice versa. The remainder should be added to the duration. In case the moment of new-moon and the first contact are, one in the forenoon and the other afternoon, the above two parallaxes should be added, and the sun added to the duration. The same should be done for the last contact, (i.e. in the above, substitute last contact for first contact, and do the work)."

24. In XVIII.1-2, the preliminaries for Venus according to the [Vasistha] are given. The meaning is: "Subtract 147-Really it is 167 and 147 is due to a scribal error-from the days from epoch. Divide by 584. The quotient are the risings of Venus. The motion of Venus for each rising is 7 [Rasi] 5[o] 30/ 20//, Venus moving on to 26[o] of [Kanya],i./e., 5 [Rasis]26[o] goes to its first rising (after 30 days). To the remainder above, add days equal to 1/11 of the number of risings gone. The motions during each cycle are given,(for the days remaining)."

T-S here change [gunaptaih] into [gunamsaih], and thus introduce into the correct motion an error of 10 1/2 minutes of arc per cycle, when even the difference between one [Siddhanta] and another does not exceed one minute of arc. They have also changed [Kanyam'san] into [Kalamsan] and shut out the Ksepa for the beginning of the first cycle, as they have in the case of the [Pauli'sa][Rahu], and the [Vasistha] moon, already referred to. It is to be noted that [Kalamsa] is meaningless when applied to motion. It is the interval between the rising or setting sun and the planet, (here Venus), in time-degrees and for Venus it is 8[o] to 10[o], and never 26[o].

25. In the computation of the [Vasistha], Jupiter and Saturn following, VM gives a method similar to that given for the [Vasistha] moon, using the techincal terms [ghana, gati, and pada]. As in the case of the moon, here too, T-S have not understood the nature of the work,and have given wrong interpretations.

26. The last eighteen [aryas] of the work, purporting to deal with the [Pauli'sa] star-planets, are spurious, and not VM's, for which there are strong internal evidences: Stanzas 61, 62,63, patently close, not only the chapter, but the whole work, which can be seen from the words, [avantyakah Varahamihirah .......tara grahaka rikatantram cakre], and after criticising certaing astronomers, occurs, "[drstam Varahamihirena sukhaprabodham], in which the chaged metre itself shows that the work has come to a close. In [drstam] etc., the next three feet are missing, and I feel that it is a purposely done blackout, to obliterate sure signs of the closing of the work, and it must have been done by somebody at alater time, with a view to append the next eighteen [aryas] and pass them off VM's. It is to be noted that if the eighteen[aryas[ also belong to the work, then the work would be without the customary completion.

These eighteen [aryas] form a second set of rules, of which the first two stanzas form a new salutation and introduction. It avers that VM considered this to be a superior work, and that with a liberal heart he gives this to the world. And these stanzas are just worthless stuff. The rules completely neglect theequation fo the apses, and therefore what are given, are only the mean geocantric positions of the planets. That explains the boasted simplicity of these rules. But what is it worth? How on earth can it be styled superior, on which VM is said to take pride and pass them on the posterity, a thing which even a novice can do? Fancy VM, who says in 62, "Let people,who have been discouraged by the inaccurate Mars of Pradyumna, Jupiter of Saura and Mercury of Vijayanadi, have recourse to this part of the manual." Fancy VM giving this second set, which gives mean,i.e., incorrect, results instead of true, i.e., correct results. Further, in Mercury and Venus given in this set, there are mistakes which one cannot presume to have been committed by VM.

T-S do not say anything about the possibility of this being an interpolation, because they themselves say,(vide Introduction PP.xivi-viii), that they do not understand it. They are perplexed at the strange and apparently incorrect constants appearing, and T is very sore that this is atributed to the [Pauli'sa], which he cherishes. He cannot emend the constants into his desired values, because they are "checked and found correct" in the work itself. But actually they are no more strange than familiar friends in different dress. Thus: instead of using the days as unit, the author uses the time taken by the mean sun to move one degreeas the unit of time, and asks us to call this 'day'. This is nothing strange because Hindu astronomers define what they call a `solar day' thus, an dthis is distinguished from the [savana] or civil day, equivalent to the mean solar day of our daily parlance. For certain given periods the planets move stated amounts of degrees forwards or backwards realtive to the sun. Therefore, if the sun is known, we can get the planet by adding or subtracting the degrees. Thus, knowing the sun is a desideratum. So, the author gives the synodic periods in his new units of 'solardays' instead of ordinary days, becuse that will be conducive to uniformity and convenience, and , for the same reason, the period of the motions are also given in the same units . We shall now convert the author's synodic periods into civil days : For 360 solar days there are 365-15030, (this is the Pauli'sa year), ordinary days. Soo we have:

Planet Solar days Civil days The Period according to

given [Vaistha]

Mars 768 3/4 779-58-43-26 779-57-19-0

Mercury 114 6/29 115-52-30-12 115-52-45-0

Jupiter 393 1/7 398-53-7-3 398-53-20-0

Venus 575 1/2 583-54-21-43 583-54-32-44

Saturn 372 2/3 378-6-36-4 378-6-0-0

It is seen that the agreement is close, and the difference is not greater than that among the [Siddhantas] themseves.

It may be pointed out here that only the more important misinterpretations have been dealt with in this paper. A more detailed explanation fo the itmes treated above, besides many more points less important but still deserving notice, has been given by the present writer in his exhaustive commentary in Sanskrit and translation, of PS prepared by him fo rthe Institute fo Astronomical and Sanskrit Research, New Delhi.

THE MAIN CHARACTERISTICS OF HINDU ASTRONOMY IN THE PERIOD

CORRESPONDING TO PRE-COPERNICAN EUROPEAN ASTRONOMY

1. The System of Measuring Longitudes

One characteristic of Hindu Astronomy is that from time immemorical it has been following the sidereal system instead of the tropical system. From the earliest Rgvedic period of their history, the Hindus have been following a lunisolar calendar. The lunar month, which is the period between two conjuctions (or oppositions) of the sun and moon, was observed as a natural unit of time, and used to measure time. It was also discovered that another natural unit, the year, in which the seasons days more. Continued observation showed that nearly once in twenty-seven days, the moon came practically to the same point in tis path, marked by certain stars adn star-groups, twenty-seven in number, [Citra,-Svati,-Visakha]etc., each one roughly corresponding to one day. The twelve lunar months were named after certaing of these asterisms, selected for their suitability, at or near which the full moons occured, like [Caitra], [Vaisakha] etc. When the full moons of the lunar months receded too far away from the asterisms to which they were assigned, by the elevem day defect of the lunar year, in order to bring them back to their assigned asterisms. This automatically tied the lunar year to the solar sidereal year, and has ever been keeping the names of th months [Caitra] etc., meaningful. Why sidereal and not tropical, itmay be asked. The asterisms mark fixed points on the zodiac. (We neglect here proper motions), and the oppositions, (and thereby also the conjunctions) of the moon with the sun, being associated with these fixed points, make the periods of the moon or the sun once round the zodiac, sidereal.

It is true that the solar year, conceived thrugh the coming round of th e seasons, is tropical, i.e. equal to the time taken by the sun to move once round the movable zodiac from equinox to equinox, made shorter than the sidereal year by the precession of the equinoxes. But the difference is small that hundrends of years would have to pass before it would be observed. Indeed, during the long Vedic period itself the difference was observed at long intervals, and since the seasons occur earlier and earlier in the sidereal year in the course of time, earliar and earlier in the sidereal year in the course of time, earliar and earlier new moons or full moons were enjoined by the vedic seers for the rituals depending on the seasons, like "caitrasta amavasta", "caitrasya Purnima". "Phalguna amavasya amavasya" "phalguna Purnima" etc. This itself shows that the sidereal year was taken for granted.

This is confirmed by the fact that in the period immediately following, the [Vedanga Jyotisa](V.J.) (c.1200B.C.) was based on the sidereal system. It divides the zodiac into twenty-seven equal asterismal segments, with the twenty-seven asterisms toughly corresponding to each, and says that in one year the sun traverses these twenty-seven asterismal segments, and in five such years, (called a yuga), the moon makes 67 such revolutions, with the result that there are 62 lunar months in it, two of them being intercalary. The V.J.also says that the zero-point of this sidereal system was at the beginning of the [Sravistha] segment, where the winter solotice was declared to be situated at that time. Earlier, during the vedic period, we have evidence that the zero-point was taken to be situated near the Orion, and then near [Rohini] and then again at [Krttika] at successive periods, where the spring-equinox was situated at those periods.

The V.J. was intended to provide a scheme for a religious and civil calendar, which purpose it could serve excellently, provided an intercalation was dropped now and then on actual observation. We are sure that this was done, so that for more than a thousand years afterwards, even during the period of theastronomical [samhitas], when the winter solstice had receded to the first point of ['Sravana] the zero-point was still taken to be [Sravistha], and predictions made. Only the Jain and Buddhistic astronomical works took the reality of the situation into consideration, and shifted the zero-point of their sidereal system to 'Sravana.

About 570 A.D. the vernal equinox was very near the first point of [A'svini[, near [Zeta-piscium], and the [siddhantic] astronomers took this as the zero-point of their sidereal system, in which the position of the sun, moon and star-planets among the 27 asterismal segments was sought to be found, and their periods were intended to be sidereal. Excepting the [Vasisthasiddhanta] of the [Pancasiddhantika] which use da rough year of 365 1/4 days, and the Romaka of the same, which followed the tropical system accepted to be foreign, all other [siddhantas] give a sidereal period for the sun in the neighbourhood of 365-15-31, (While the actual sidereal period is 365-15-23), so far removed from the tropical that it borders on the anomalistic, though intended to be sidereal. The result is that the zero-point has processed at the rate of about a degree per 420 years, and is now forward by more than 3[o], creating all sorts of confusion in fixing the [ayanam'sa] (i.e. the distance between the Hindu zero-point, and the Vernal equinox forming the zero-point of modern astronomy). Consistent with this processing of the first point, all Hindu astronomers have given in general longer sidereal periods for the moon and the planets and the moon's apogee, and a shorter period for the moon's and larger rate of precession. In comparing the Hindu sidereal periods and rates of precession with the modern ones, this should be borne in mind.

Thus it will be seen that throughout in Hindu astronomical history, the sidereal system has been in use, implicitly in the vedic period, and explicitly from the time of the V.J.onwards. I have made this discussion a little elaborate, because for various reasons, (like astrology and modern convenience) some people have wished that the Hindu system be tropical, and so have asserted that it was tropical during the vedic period, (and according to some even in the later classical period) and tried to distort history instead of loooking facts squarely in the face.

Another thing must be mentioned here. Recently some people have expressed the novel idea that the Hindu longitude is polar, i.e. the great circle are joining the celestial pole, (as distinguished from the pole of the ecliptic), and the moon or planet is projected on the ecliptic, and the longitiude of this point measured from the first point of [Asvini] is given as its longitude by the Hindus. But no [siddhanta] or [Karana] says so. They give the polar longitude for a specific purpose: For the sake of astrological predictions the moon, sun and planet's conjunction in polar longitude is given. This is also given to check their correctness in position by comparing them with the polar longitudes of the stars. For this purpose most [siddhantas] give the co-ordinates of the stars in polar longitude readymade, and this has misled these people. Commentators have all made this point clear, not to speak of the texts themselves.

In contrast to the Hindu system, western astronomers, like the Egyptians, Greeks and Romans, very early adopted the tropical or seasonal year, on which essential civil activities like agriculture depend. When would Father Nile bless them with his yearly floods, was the problem of the Egyptians, and seasonal effects would be more marked in regions of higher latitudes like Greece and Rome. On the other hand the asterisms, and days and months occurrign with these at full or new moon, became sacred to the Hindus even in the vedic period, and important religious rites like [darsapurnamasa] on these, and so the sidereal calendar, originating in the way we indicated, has persisted, requiring the sidereal system of longitude reckoning, so much so that even the rituals depending clearly on the tropical year, later came to be fixed according to the sidereal calendar by the [Dharma Sastras].

2. Cycles in Yugas and Kalpas

In order to give the revolutions of the sun, moon etc., and of their apogees and nodes in whole numbers, and at the same time secure sufficient accuracy, the [siddhantas] have used long periods called the [yuga] and even vastly longer ones called the Kalpa. The [yuga] and Kalpas are not equal in duration for all. There are fourmain schools of [siddhantins] in this respect (1) The Old [Surya Siddhanta] condensed by [Varahamihira] used perhaps a [yuga] of 1577917800 days. [Aryabhata's] [ardharatrika] system clearly had this number, and following it, the [Khanda] [Khadyaka]. A [Paulisasiddhanta], quoted by [Bhattotpala in his commentary on teh [Brahatsamhita] belongs to this school. We do not now whether this school had a Kalpa, nor how long it was.

(2) The School of [Aryabhata], represented by his [Aryabhatiyam], has yuga of 1577917500 days. This divided into four equal sub-yugas [Krta], [Treta], [Dvapara] and [Kali]. 1008 yugas make a kalpa. (3) The [Visnudharmottara] School has 1577916450 days for the yuga. The [Brahmasphutasiddhanta] and the [Siddhanta Siromani] belong to this school. The length of the sub-yugas are in the ration4:3:2:1. One thousand [Yugas] make the kalpa. (4) The later or new [Surya][siddhanta] has 1577917828 days for the [yuga]. A number of new [siddhantas] like the [Soma], the [Brahma], the [Vrddha] [Vasistha], the (new) [Romaka] etc. belong to this school. The sub-yugas are in the ration 4:3:2:1, and the kalpa equal to 1000 yugas. But the revolutions begin, not with the beginning of the kalpa, but 3.95 yugas later, when according to these the creation of the planets was completed.

The same number of revolutions per yuga for the sun, is given by all schools. Excepting school no (3) all other three give the same number of yuga revolutions for the moon also viz. 57753336. No. (3) gives 57753300. As for the revolution of the other planets, the four schools agree or disagree with one another in various ways. It must be noted that even if two schools agree in the number of revolutions of a particular body, the body cannot have the same period, because the yuga days are different. This is natuaral, and must be expected, or else what is the distinction between the shools?

In spite of the above mentioned various differences, all schools take almost the same point of time as the beginning of the Kali sub-yuga. Schools (1) and (4) take the mean midnight at Ujjain, February 17/18, 3102 B.S. beginning Friday, as the beginning of Kaliyuga, (the Kali epoch), while the other two schools take the time of mean sun-rise following six hours later, as the point at which [Kaliyuga] begins. (This would mean that the yugas and Kalpas according to each may vary with regard to the times of their beginning by several years, and even yugas, inspite of the popular feeling that they do not differ, but occur at the same point of time.)

Further all schools agree that the mean sun and moon is at the zero point of [A'svini] at the above mentioned Kali epoch. Excepting shcool(3) the others are also agreed that the mean planets too are at that point at the epoch, and that the moon's apogee is 90[o], and the node180[o] from that. What are we to understand from this? Are we to think that at such an ancient date as 17/18, February, 3102 B.C. the Hindu astronmers gave this result as got from their observation? Or was this point of time fixed by some later astronomers as a convenient epoch for starting their calcualtions, and followed by all the later [siddhantins]? The former alternative cannot be accepted, because the mean sun, moon and planets were not the same but differed widely from one another, nor were they at zero [As'vini] as calculated by modern astronomy for that epoch. Scholars like Bentley showed that starting from the epoch and working by each [siddhanta] the error gradually became less and less, until at the time of the [siddhantas] the error became a minimum, as must be expected. Thus he proved that the second alternative was the correct one, and that the [Kali era] starting from this epoch, was an extrapolated era, founded by the astronomical [siddhantins].

Beore this was done by the astronomers, the concept of the four sub-yugas [Krta] etc., was a vague one. Without any definite number of years attached to them, The concept of the yuga itself had arisen in the vedic times, and was perfected in the [Vedanga Jyautisa] period, the idea being that at the beginning of every yuga, consisting of five solar years, the sun and moon came together at ['Sravistha] where the Winter Solsice was situated at that period. At the time fo the [jyautisa samhitas] a larger sixty year yuga was conceived, in which not only the sun and the moon, but also Jupiter met together at a given point int he zodiac. (The Romaka siddhanta] of the [Pancasiddhantika] gives a lunisolar yuga of 2850 years of which the famous Metonic cycle is the 150th part. (P.S.I, 15). The [Suryasiddhanta] condensed in it gives a lunisolar cycle of 180,000 years (P.S.I, 14). We do not know whether these were of the original [siddhantas] themselves, or [Varahamihira] gave them as fractions fo the orivginal, for convenience). Then the idea of a larger yuga was evoled, at the end fo which the other planets also came together with them, and repeated their motions. Along with this concept of a very long yuga arose the idea that the first part of the yuga was a golden age, the second not so very good, the thrid passable, and the fourth bad, an idea naturally occurring in the minds of ancient peoples. The names [Krta], [Treta], [Dvapara] and [kali], used for very good, good, passable, and bad throws of dice, current from vedci times were borrowed and given to them, until the astronomers used the concepts for their own purpose, giving a definite number of years to them, as we have seen, and fixing the beginning of Kali as an important epoch. The vagueness of the original concept can be inferred from various incidents of different times being taken as occurring at the beginning of Kali, by different ancient authorities. For example, the astronomical kali is accepted by most [Puranas] and given as the date when [Krsna] left off his mortal body, and ascended to heaven. But certain other traditions give the first coronation fo [yudhisthira] at [Indraprastha], others the [Mahabharata] war, and yet others the abdication of [Yudhisthira] and starting on the [mahaprashana], as the starting point of Kali. The Jain and Buddhistic writers take a period 468 years after the astronomical Kali as the time whrn [Yudhisthira] lived and establlished his [Yudhisthira] era (corresponding to 2634 B.C) [Varahamihira] and following him [Kalhana], state that the [Yudhisthira] era began in 2449 B.C. While the [Sakalya Samhita] and the [Matsyapurana] say that the [Saptarsis] were at [Krttika] at the beginning of Kali, the [Garga Samhita] and other [Puranas] say that they were at [Magha] at the beginning of Kali. Thus there is a difference of about 700 years in the times given by them for the Kali epoch. Thus we see that the early astronomers fixed the Kali epoch arbitrarly, as a convenient starting point for their calculations, and others accepted and followed it.

3. The Concept of Precession of the Equinoxes

In most Hindu [siddhantas] the precession fo the equinoxes has been given as ocillatary about the fixed first point of [A'svini] and not as a continuously regressing phenomenon, which it is. The later [Suryasiddhanta] says that there are 600 such to and fro full oscillations in the [Yuga], i.e. the period of oscillation is 7200 years. According to this the equinox coincided with the First point of [A'svini] at the Kali epoch. During the first 1800 years of [Kali] it uniformly moved forward by 27[o], i.e. to a point 20' beyond [Bharani] and then in the next 1800 years, i.e. at 3600 [Kali], it uniformly moved forward by 27[o], i.e. to a point 20' beyond [Bharani] and then in the next 1800 years, i.e.at [Kali], (corresponding to 499 A.D.), it regressed to the First point, and from that time on is continuing the regression till in 2299 A.D. it will stop at 27[o] behind [A'svini] and begin moving forward again (Su.S.III.9-10). From this we can also see that the amplitude of oscillation as 24[o] instead of 27[o]. Some like [Munjala] quoted by [Bhaskara] II, give the phenomenon as continuously regreesing and making 199669 revolutions in a [Kalpa], i.e. about 22000 years per revolution, giving a precession of about 1'per annum. The earlier [Siddhantins] like [Aryabhata] I and [Brahmagupta] do not mention any motion of the equinoxes. But [Bhaskara] I, a senoir contemporary of Brahmagupta and follower of [Aryabhata], discussed it in his [Aryabhatiya] [Bhasya], (pp.174-76 of R 14850, Oriental Mss. library, Madras), mentions the [Romakas] who assert there is precession, characterises them as people who do not know the truth, (aviditaparamarthah) and dismisses their view saying it is a temporary and unnatural phenomenon of the nature of a portend. But [Varahamihira] who came before him and just after [Aryabhata], is more positive and says, (P.S.III. 21, 23 Br.S.III 1-2) "Certainly in the ancient times the Winter Solstice was at ['Sravistha] and the Summer Solstice at the mid-point of [Aslesa] as mentioned in the earlier ['Sastras], (like the Vedanga Jyotisa). Now they are at the beginning of [Makara rasi] and [Karkataka rasi] respectively. If they have moved away even from these points at a later period, the amount of this unnatural change can be found out by observation fo the sun's shadow and examination." Not knowing the rate of precession, he does not mention calculation for this purpose. For the matter of that even [Bhaskaracarya] II, advocates finding the amount of precession by observation, since there is difference of opinion among the [siddhantins] about the rate of motion. Earlier then [Varahamihira], excepting the [Romaka] which is clearly of foreign origin, no Hinduy astronomical work mentions precession, though its effect on the seasons was observed even in the vedic times, as I have already mentioned. Even the shifting of the [Vernal] [Equinox] from [Mrga'siras] to [Rohini], and from [Rohini] to [Krttika] had been observed in the Vedic period. But this had not resulted in the idea of the tropical year as such, as distinguished from the sidereal year, so strong was the hold of the sidereal year upon the astronomical and calendric system of the age. This must be the reason why, even after knowing the coninuous motion of the equinoxes, they considered the phenomenon oscillatory, i.e. something temporary which would rectify itself by an equal motion in the opposite direction. Even if a [Munjala] gives it as continuously regressing, it can have no value unless he had adequate reason to know it for certain (and he had no reason, it could only be a guess), for the real cause of the phenomenon, viz. the behaviour of the earth rotating like a spinning top in resisting the pull of the sun, moon and planets on the extra matter on its equatorial bulge, this was not known at that period.

Another fact, also indicated earlier must be mentioned here. We have see that the Hindu sidereal year, being more than 8 [vinadis] longer than the correct sidereal year, the point of [A'svini] itself has a progressive motion of more than 8'' per annum. Since the correct precession is about fifty and a quarter Seconds, the rate of precession with respect to teh Hindu First Point must be more than 58 1/4, since the [siddhantas] advocate getting the precession by observation of the sun's shadow. Accordingly, [Munjala] and the later works give a rate of precession nearly equal to 1' per annum, which is quite proper. It would be a mistake to suppose as far as ancient Hindu works are concerned that the nearer their rate of precession is to 50 1/4 '' the more correct it is, e.g. it would no tbe proper to commend the [Suryadiddhanta] for its rate of precession of 54// per annum, on the ground that it is so near 50 1/4''.

4. The Eccentric and Epicyclic Theories of Planetary Motion

From the time of the Old [Surya Siddhanta], i.e. the one condensed by [Varahamihira] in his [Pancasiddhantika], the eccentric and epicyclic theories have come to be used in Hindu astronomy in computing and geometrically explaining planetary motion. The ancinet [Vedanga Jyotisa] gives only the mean sun and moon,and problems connected therewith, like the ending moments of [tithis] etc. Even these are rough, keeping in view the convenience of a civil calendar. But the religous rites like [dar'sapurnamasa] require the correct computation fo things, like the day when the moon sets or rises heliacally. Certainly by long observations the ancient priests must have arrived at rules to rectify the positions of the sun and the moon got by the rough rules, so that at least for a few synodic revolutions forward, the heliacal setting and rising of the moon etc. were computed tolerably correctly. But even in the later [Jyotisa samhitas] like the [Garga samhita] we do not see rules for the computation of the true sun and moon, not to speak of the planets. It is so in the [Paitamaha siddhanta] also (the one condensed by (Varahamihira). It is in the [Vasistha siddhanta] of the [Pancasiddhnatika] that we see for the first time distinction made between mean and true sun and moon. A rough and empirical rule is given to get the true sun. The variation in the moon's motion is recognised, but this is supposed to increase and decrease uniformly over its mean motion, during its anomalistic revolution, and the true moon is computed by arithmetically summing the varying motions, the excess or defect over the mean answering for the equation of the centre. Of course this cannot be very correct, since the variation is really with the cosine of the anomaly, and the equation of the centre is a function of sine anomaly. This was recognised and used in the later [Pauli'sa] and [Romaka] [siddhantas], as seen from the empirical values given being roughly proportionate to the sine of the anomaly, and was accepted as such.

A natural tendency would be to picture the motion by a graphical representation. Circular motion, the synodic revolution of planets, and the representation of the equation of conjunction of planets by circles, with which the astronomer's mind was already familiar, readily suggested the epicyles, with its centre moving on a deferent, as a means of representing the equation of the centre, and this picture wa sadopted, as first seen in the Old [Surya siddhanta]. Very soon an exactly equivalent method of represention was discovered, viz. making the body move uniformly on a circle, called the eccentric circle, the centre of which is off the centre of the earth in the direction of the point of slowest motion, by a length equal to the radius of the epicycle. The [Aryabhatiyam] gives both representations.

While both can give the same locus of the moving body forming the orbit of the body, the eccentric would seem more real, because in the former an imaginary non-physical mean body, moving on the deferent would not be acceptable to the mind, and would appear only as a devise to represent the effect fo the equation of centre graphically. This would mean that the distance of the actual planet from the earth's centre called the [manda Karna (hypotenuse of Eq.cent) need not be taken seriously. This explains why most [siddhantas] like the [Siddhanta 'Siromani] do not use the [Karna] in finding the equation of the centre. On the otehr hand the School of [Aryabhata], on the plausibility of the motion of the actual body, which is an ellipse. The equation of the centre of a body moving +5/4 sine 2 anomaly -etc. where e is the eccentricity of the ellipse, and the anomaly is reckoned from the apogee (or aphelion). Taking the [Karana] into consideration would give part of the second term also, while neglecting it would give only the first term. (This has been shown by me in the paper, 'A Historical development of certaing Hindu Astronomical processes,' presented at the seminar of the Indian National Science Academy, held in 1968, and publised in the Indian Journal of History of Science Vol.IV 1-2, 1969) pp.46-75 in this volume).

A question may now be asked: as early as c.225 B.D. in Greece, Apollonius of Perga developed the theories of epicycles and eccentrics. The [Old Surya siddhanta], using the epicycle, must be later and the [Aryabhatiyam] mentioning both theories, is certainly later. There is nor boubt that there was contact between India and Greece from earlier times. Did the Hindu astronomers borrow these ideas from the Greeks, or bid thes occur to the Hindus naturally, as it had occurred to the Greeks? We cannot answer this question with certainty. But an independent origin in India seems more probable, when we see that the celarly western and earlier [Pauli'sa] and [Romaka] did not have either of these theories, that the Hidnu constants were different and better in general, that there was already the anlogy of representation of the equation of conjuction, which must occur to the astronomer naturally to explaing the motion of the star-planets as seen from the earth, and that the Hindus, having already a theory of the variant motion in th eform of the pull or repulsion of the apogee on the planets, required the epicyclic re[presentation as a geometrical model, for which there would certainly be an urge.

Before proceeding we must understand in terms of modern astronomy the meaning of certain Hindu term used in the context of planetary motion. [Madhya graha] is the mean planet in its own orbit round the earth, when applied to the moon. When applied to the sun it means the mean heliocentric longitude, and the longitude of the superior conjunction in the case of [Mercury and Venus], the mean sun being that point. [Sghra graha] means the mean planet in it sown orbits own orbit round the sun in the case of Mercury and, Venus, and the superior conjucntion in th ecase of Mars, Jupiter and Saturn, the mean sun beingthat point. According to modern astronomy, and correctly too, the equation of the cnetre is to be applied to the mean body in its own orbit round its central body. This will give the true longitude in the case of the sun and the moon, and the true heliocntric longitude in the case of the five planets Mercury etc. The equation of conjucntion is then applied to these five planets, using the true sun to convert their heliocentric longitudes into geocentric longitudes. Any instruction given in Hindu astronomy must in essence follow this procedur, or else the instrcution would be wrong.

Now, Hindu [siddhantas] instruct what amounts to applying the equation of the centre, in the case Mercury and Venus, to the mean sun isntead of their own mean longitude, and this is wrong. Further instead of their own mean longitude, and this is wrong. Further instead of taking the true sun as the superior conjunction the mean sun is so taken. On account of these and the error in the equation of the centre itself, the computed true planets did not agree with observation, and various attempts were made to correct these errors, with very little success, since they rarely touched the real cause. For example, the [Aryabhata] and [New Surya Siddhanta] schools made the radius of the epicycles vary with the anomaly. Before applying the equation of the centre to be done first, and dispensing with it altogether in the case of Venus and Mercury. The [Siddhanta 'Siromani] instructed full equation of the centre, and full equation of conjunctions is enjoined using half of both equations in the preliminary work. This disagreement among the [siddhantas] itself shows that the preliminary operation is an unessential hotchpotch.

We have said that the application of the equation of conjucntion conversts the helocentric longitude into the geocentric. This can be readily seen in the case of Mercury and Venus (see fig). If the line joining the earth and sun taken as unity, the sun being the planet's mean longitude in these two cases, the epicycle of conjunction round the sun is the planet's orbit and the radius of the epicycle is the distance of the planet from the sun, taking the distance fo the sun from the earth as unity. Since the ratio of the radius to the circumference in any circle is constant, the degrees of epicycle./.360[o] will give the ratio of the distance of the planet to unity. (If the epicycle varies we can take the mean.) In the case of the superior planets, viz. Mars Jupiter and Saturn also, since the application fo the equation of conjunction converts the heliocentric planet into the geocentric, the geometrical representation must be true, which can be shown thus: Interchance P and E in the figure. This is a representation of the earth circling the sun, as seen from the planet, and the earth will be viewed always near the sun, even as Mercury and Venus are seen always near the sun as viewed from the earth. By comparison, we can now see that if we take the distance of the earth from the sun as before, the degrees of epicycle./. 360[o], gives the reciprocal of the ratio of the distance of the superior planet from the sun to the distance of the earth form the sun. In the table below, these ratios are given. They can be compared with the correct ratios of modern astronomy given.

planet Modern Siddhanta Old Surya New Surya Aryabhatiyam

'Siromani Siddhanta Siddhanta

Earth 1.00 1.00 1.00 1.00 1.00

Mercury .39 .37 .37 .37 .38

Venus .72 .72 .72 .73 .73

Mars 1.52 1.48 1.54 1.54 1.54

Jupiter 5.20 5.29 5.00 5.07 5.16

Saturn 9.55 9.00 9.00 9.11 9.41

We see tha tagreement is fairly close, and the [Aryabhatiyam] comes closest. I fth eHindu [siddhantins] had only taken the geometrical representation more seriously, and if they had not been wedded to the geocentric hypothesis so strongly, theu could have noticed the correct ratios of the distances fo the planets from the sun, and even postulated the heliocentric hypothesis, as Aristactus of [Samos] had done befor, in c.280B.C. They would have also seen that their postulate of equal linear motion for all planets as well as the sun and the moon, is wrong. (The Hindus estimated the distance of the moon to be about 65 times the earth's radius, fairly correctly by using the moon's. This enormous real distance was not realised even by Copernicus, who gives it as about 25 times that of the moon. The [Greek Aristactus] of [Samos] had estimated it to be about 20 times, using a method ingenious but useless, on account of the impossibility of making correct measurement. According to the Hindu theory, Jupiter will be 12 times and Ssaturn 29 times distant as the sun. Quite arbitrarily, they gave a distance to the stars, i.e. the stellar sphere also, as 60 times that of the sun, and for the celestial sphere, 4320000000 times, practically infinity. The distance they gave to the stars is one reason why they clung to the geocentric hypothesis, for they thought that if the earth moved round the sun there would be a large stellar parallax that could be observed, but no such parallax was observed. But in fact there are stellar parallaxes. On account that only modern instruments can measure them. It was left for Copernicus to propound the heliocentric hypothesis anew. Whether it is the epicyclic theory of the eccentric theory, it is the geocentric theory of the heliocentric theory, one is as good as the other as far as the result is concerned, and this was known to the ancients, noth [Greeks] and Hindus. As long as the real reason why the sun should be considered the central body, viz. the strong pull it exerts over the other bodies, is not known, the only recommendation for the heliocentric hypothesis is its simplicity. BUt simplicity can masquerade as truth, and the simpler hypothesis can appear to the human mind as more real, for nature's lawa are characteristically simple. It should be noted that [Copernicus], in propounding the heliocentric hypothesis in his 'Revolutionibus Orbium Clestium', publised in 1543 A.D., very carefully explained that he was adpting the theory only for its simplicity, for if he had not done so he would have lost his valuable influence with the Church, and might have had even to face the inquistion like [Galileo] later.

5.The Concept of The Sidereal Day

Hindu astronomy from the time of the [Vedanga Jyotisa] has conceived the sidereal day as being caused by the rotation fo the stellar sphere round the earth from east to west, the time fo one rotation being one sidereal day. Classical Hindu astronomy supplies a reason for the rotation, by assuming a wind called [Pravaha] in the upper regions, blowing the stellar sphere round. But [Aryabhata] seems to have felt that the same result can be obtained by asuming the earth to rotate on its axis, west to east. This can be seen from his [Aryabhatiyam-Gitika] (1), where together with the number of eastward revolutions of the sun, moon etc. in a [Yuga], he mentions the earth and the number of sidereal days in the [yuga] (1582237500), as the number of its eastward revolutions. This is confirmed by his statement in [Golapada](9): " Just as a man moving along in a ship sees the non-moving mountians moving in the opposite direction, so also the non-moving stars seem to move directly westward in the sky above the equator." But he seems to contradict himself when he says in the very next verse, to contradict himself when he says in the very next verse, "Being blown by the [Pravaha] wind for the sake fo rising and setting, the stellar circle moves westward above the equator at the rate of 1/per [Prana] (i.e. 4 secondsof time). In [Gitika](4) too he says, "The star moves 1/per [Prana." How are we to resolve the contradiction? His successors, even from [Bhaskara]I onwards have taken the latter as his own view, and interpreted the former so as to agree with the latter. The stock argument of the generality of astronomers against a rotating earth, (like the impossibility of birds leaving their nests reaching them againg) seem to have made them interpret [Aryabhata's] conviction and the latter si what he gives as a [Purvapaksa[, i.e. the opinion of others, in the exceptional [sutra] style. For Brahmagupta, following only a century after, in the [dusanadhyaya] of his [siddhnata] condemns [Aryabhata] for holding this view of a rotating earth. He even reads [Gitika] (4) as [Pranenaitikalambhuh] meaning, "The earth mopves 1' per [Prana]", and probably this was [Aryabhata's] original reading. Further whom were the generality of astronomers condemining, if it was not [Aryabhata] and people like him? Thus the scientific mind of [Aryabhata], always seeking simpler assumptions, must have arrived at the view that the rotation fo the earth is simpler to assume, like the [Greek], [Heracledes] of [Poutus] before in c.350 B.C. But he could not have borrowed this idea from Heracledes sine the latter's view had died out even in Greece long ago.

6. Certain Specialities in the Use Of Trigonometry in

Astronomical Work

In Hindu astronomy, trigonometrical functions came to be used only only after the time of the [Pauli'sa siddhanta] condensed in the [Pancasiddhantika]. Upto the time of the [samhitas] and [Paitamahasiddhanta] there was no need for them, when only mean planets were given and rough rules to compute the shadow, [lagna], time, etc. were considered sufficient. The later [Vasisthasiddhanta] found the equation fo the centre as mentioned before by an arithmetical summation. To get the point of contact in the lunar eclipse, it used the rlated degrees of angles themselves instead of their sines (P.S.VI.7-8). The [Pauli'sa] also does not seem to have used the sine etc. Here, in teh case of the sun, empirical values for 30[o] intervals are given to make it true. Its moon is common with that of the [Vasistha]. To find the moon's latitude it uses the degrees of argument of latutude, instead of its sine, though this is not explicity stated. In dealing with the solar eclipse in chapter VII.2-4, three rules are given to represent the parallax correction for latitude as a correction to the moon's nodes, and in all, degrees are used instead of their sines as multipliers. Only in one place, (VII.1), is the sine used to correct the time fo conjunction in longitude for parallax. But this is most probably [Varahamihira's] own, to secure tolerable accuracy. The same verse occurrings as VII. 9, in connection with the [Romaka], confirms this. As for chapter IV, where a table of R sines is given, and spherical trigonometrical problems are dealt with, that is the author's own, and, if it must go with any [siddhanta], it must be the [Suryadiddhanta] there. We see evidence of the use fo sines etc. From the time of the [Romaka] and in [Suryasiddhanta] mentioned above and the later classical works beginning with the [Aryabhatiyam].

As in modern trigonometry, the Hindus use the [jyardha] or half-chord, calling it [jya], (a synonym fo the word [sinjini], from whcih the word sine is derived), unlike the Greeks who used the full chord. The Hindu genius which invented the place-value, decimal notation, and for its sake, the symbol zero, saw the convenience in using the half-chord, and used it. But, instead of taking the raduis as unity like the moderns, the Hindus measure it in certain natural or assumed units, so that the half-chords and arcs are measured int eh same units. So, if the modern word `sine' is to be used, the Hindu half-chords are to be called R sines. The [Aryabhatiyam] is the first extant work in which an R sine table is given. [Aryabhata] uses a natural measure, i.e., he takes the circumference to be 21600 units, so that a minute of arc is the unit. Then the radius will be 3438 units, being the number of minutes in a radian. Many [siddhantas] adopt this measure. But some, for the sake of convenience, assume an arbitrary value for the radius. For example, [Varahamihira] assumes it to be 120 units, because in all computations with R sines etc. R occurs as a multiplier or divisor.

Usually the R sines are tabulated for intervals fo 3[o] 45/, thus giving twentyfour of them in all for the quardrant. The [Vate'svarasiddhanta] tabulates 96 R sines (Plus two more near90[o]), to secure greater accuracy in interpolation. With the R sine table, an R versine table is given by most [siddhantas], and R cosines are got by subtracting these from R, or taking the R sines of the complements. [Aryabhatta gives a geometrical construction to deive the R sine of [o/2,] when R sine [o] and its versine is known, byu using the Pythagoras theorem, (Arya Ganita.11). This can be carried to a very small fraction of an angle where R sine [o] is indistinguishable from R [o]. He has used this to find the circumference of a circle, the radius being given. He says: when the diameter is 0000units, the circumference is 62832 units, thus giving=3.1416, which is correct to four decimal places. [Varahamihira gives two formulae to find the R sines, (P.S.IV.2-5) of successive half-angles, and thence all the twentyfour, tabulated. Using the first tabulated R sine 7/ 51//, he could have arrived at =3.14, which is correct to 2 places of decimals. Using the value of R=3438, the circumference being 21600, the New[Suryasiddhanta] could have found=3.1414, and so also Brahmagupta using his radius. But all these three give what comes to [=/ 10=3.1623], so incorrect, it passes my comprehension why. If, for ease fo computation, the fraction 2 2/7=3.1429 is simpler to use, and far closer to the correct value. The [aryabhtiya], [Ganita]. 12,gives an alternative method to compute the R sines, and the [Suryasiddhanta] repeats it. It depends on the fact that the second differeences fo the R sines vary as the R sines. But the constatn divisor of the R sines, the number 225, seems to have been assumed empirically by a supposed identity with the first R sine. If it had been found from R sine (0-a), R sine0, and R sine (0+a), by calculation, they would have arrived at the correct 234. [Bhaskaracarya] has appended a whole section for getting the R sines etc. by various methods reminiscent of modern trigonometry.

These R sines etc. are used in computing the equation fo the centre and the equation of conjunction which involve the solution of plane triangles, and in solving spherical triangles, which occur in problems like finding the right ascension and declination, and the polar latitude and longitude of heavenly bodies when their latitude -longtude co-ordinates are known, as also in problems requiring the soultions of the spherical triangle SPZ, (S representing the position of the body, P being the pole and Z being the zenith), problems like finding the length of day-light, orient or meridian ecliptic point, amplitude, azimuth, zenith distance, and hour angle of bodies, from which shadow, time, etc. are found. But, instead of using the spherical triangle formulae directly, Hindu astronomy uses plane right angles triangles whose sides correspond to the sides of spherical right angles triangles. These plane triangles are formed by perpendiculars dropped on to the horizontal, prime vertical, and meridian planes, as the sides containing the right angle. These are all simlar to the right-angles triangle formed with the equinoctial shadow and the twelve-unit gnomon, as its base and perpendicular, since in all of them the latitude of the place is involved, and therefore in every on the corresponding sides and hypotenuse are proportional. These are eight in number and are called [aksaksetras], the equinoctial shadow triangle being th earchetype. Problems are solved using the proportionality of these triangles. [Siddhanta 'Siromani], Ganita, Tripra'sna], 13-17 define these, and the [Vasanabhasya] thereon explains them exhaustively. From the earliest [siddhantas], viz., the [Aryabhatiyam] and the [Pancasiddhantika], it is this method of solving spherical triangles that is used, and so it must have been developed by the Hindus earlier still, and is a special feature of Hindu astronomy.

THE [VASISTHA-PAULISA] VENUS IN THE [PANCASIDDHANTIKA]

OF [VARAHAMIHIRA]

(Reprinted from Indian Journal of History of Science

Vol.XIV,No.2,Nov.1979)

In 1889 Thibaut and M.M.Sudhakara Dvivedi (TS) edited the [Pancasiddhantika] with translation and notes, In the XVIII chapter dealing with the star-planets according to the [Vasistha] and the [Pauli'sa] [siddhantas], they have committed several serious mistakes, professing also ignorance of the meaning of several verses. In my paper, 'Some mistakes and omissions of Thibaut and Sudhakara Dvivedi in their edition of the [Pancasiddhantika], Presented at the First World Sanskrit Conference, 1972, (VIJ.XI, 1973, Hoshiarpur), I pointed out the errors without going into the details. About the same time, O.Neugebauer and D.Pingree (Np) brought out an edition of the [Pancasiddhantika]. In that they have imporved upon TS's interpretation in some places, but committed worse mistakes in others. TS and NP have also failed to understand how the equation of conjunction of Jupiter and Saturn. Since they must do something with the verses giving this, they have altered the verses in all sorts of ways, to yield what they thought the meaning might be. Even in the case of Venus, whose computation has been simplified by neglecting several errors. I shall begin with the correct interpretation of the verses dealing with Venus, and point out their errors.

Verse 1. [hitva munijala(?jalamuni)candra]

dyu (ndyu) [ganad] [vedas ta-bhuta-hrtalabdhaha]*

['sukrodaya][gunaptaih]

[sardhah pancalino bhogah]**

Verse 2. [Kanyamsah sa (n sa) dvimsati]

[mitva 'sukro'parena ya tphu(tyu) dayam]*

[udayaikadasabhagan]

dinesu [datva] [tatasta](sca)[rah]**

1. Subtracting 174 from the days from epoch the quotient got by dividing the reminder by 584 are the heliacal risings of Venus. Its motion during the periods is 7 [o] 5[o] 30/ 30//, each.

2. Having gone to 26[o] of Virgo, i.e., at 5r 26[o], Venus rises in the west (for the first time after epoch). Adding an eleventh of the quotient (given in verse 1) to the remining days, the motions (are to be taken from the Table given in verses 3-5).

(The translation is according to my emended text. Material emendations, which require explanation, are given with a question-mark 9in Brackets0. Emendations of mere scribal errors are not).

174 days after epoch, Venus rises heliacally in the west. We have emended [munijala] as [jalamuni] because 147 will not agree with the longitude of Venus at rising given as 5r 26[o]. If Venus is 5r 26[o], the sun must be 5r 18[o], to satisfy the 8[o] given for Venus heliacal rising in verse 58. But since at epoch the sun is in the neighbourhood of 358[o], it can be only near 4r 25[o] after 147 days, i.e., 38[o] from Venus. After 174 days from epoch, the sun would be near 169[o] according to the [Vasistha-Pauli'sa]. This gives the elongation as 7[o] instead of the required 8[o]. But this samll discrepancy can be put to an accumulated error in computing from the original. TS have not felt the need for this emendation because they have emended [Kanyamsan] into [Kalam'san] and omitted the necessary [Ksepa],i.e., the constant equal to 5r 26[o], for beginning the motion. They have not understood the meaning of the word [Kalamsan] which they have brought in. It is the same as the 8[o] mentioned above, for Venus. Np have interpreted [kanyamsan] correctly. But since they have kept 147 days to be deducted intact, they find a serious discrepancy expressed by them on apage 124 of partII. However they derive staisfaction from the fact that the September 10th position of the sun would agree with the time of Venus's rising and their longitudes. This at least must have shown them that the rising takes place only more than 20 days later. They have made all sorts of unnecessary emendations, but they have failed to this necessary one.

We can infer from the instruction to add an eleventh of the quotient, that one synodic revolution takes583 1/1 0/1 days. In this period the sun has moved I revolution 7r 5[o] 30/20// and Venus, 2 revolutions 7r 5[o] 30/20//. From this we can infer that the sun takes 365-15-25 days for a sidereal revoultion. From the methods given from the pther planets also we can see that the sidereal period of the sun used is c.365-15-30, which is an evidence for the system given here being connected with the [Pauli'sa] also, as in the case of the moon.

TS have unnecessarily emended [bhogah] into [bhagah] and taking [sardhah] to mean `together with', instead of `with half', have given a motion of 7r 5[o] 20/ per synodic period. They are unware that this would make the sidereal year c.365-22 days,so wrong. Tthe errorof 10/20// in Venus would accumulate by 1[o] every nine years.

NP interpret [sardhah] correctly, but take pgunaptaih] to mean `with 1/3 degree' and give 7r 5[o] 50/, which would make the sun's sidereal year c.365-3-0, so very wrong. By this the error in Venus would accumulate by 1[o] every

Next, the days for segments fo motion in the sybodic cycle are given by 3-5.

Verse 3. sastitrayena [vedagniyamayutam] amsasaptatim bhunkte*

[ardha](?artha) sta [kair dva] vimsati (?trimsatam)

vimsatyai (tya) [vi] stri(tri) bhis [sapadamsam](?darthan)**

Verse 4. vakram atas tithibhir dvau

[pancabhir evam tato'parastamitah]*

[dasabhih praguditas syan nakhis ca jaladhin mita (tan)

gatva]

Verse 5. [anuvakri parigatvs

viparitam astamait yaidyam (ndryam)*

sasiyamsapancasaptatim itva'parato bhrgur drs'yah**

3. In three periods of 60 days Venus moves 74[o], 73[o], 72[o] respectively. In 40 days he moves 32[o] and in 17 days,51/4[o].

4. From here retrograde motion (begins). In 15 days these are 2[o]; in 5 days the same, i.e., 2[o]. Then, setting in the west, it rises in the east after ten days. Venus is in follow-up retrograde for 20 days, moving 4[o].

5. Then continuing the (direct) motion round, in the order of days and motions reversed, Venus sets in the east. Then moving 75[o] in 60 days, it becomes visible in the west.

The argument to be used in the above table of motions are the days left over, together with the eleventh of the quotient, as mentioned. It can be seen that in my emendations of some of these I have done very little violence to the text. I have been guided in these by the actual motion that must have been oberved, putting it to observational or other error, where the numbers are clear, but deviate from the actual. The ration of Venus's distance from the sun to the earth is c. 72 as given by all Hindu and modern astronomy, and this I have used to compute the segments of actual motion for comparison.

VASISTHA -paulisa VENUS

TABLE I

Motion On the synodic circle (computed)

Superior conjunction Retrograde begins Retrograde ends

30d. 371/2[o] 9d. -2 1/4[o] 243d. 258[o]

Rising West 5d. -2 1/2[o] Setting East

60d. 74[o] Setting West 30d. 37 1/2[o]

60d. 73[o] 5d. -2 1/2[o] Superior conjunction

60d. 68 1/4 Inferior conjunction Total 514d. 575 1/2 [o]

27 1/2d.261/4[o] 5d. -3[o]

12 1/2s. 9 1/4[o] Ri ing East

17d. 7[o] 5d. -2 1/2[o]

6d. 1/4[o] 9d. -2 1/4[o]

Another guide is that the days and motions from the superior to the inferior conjunction must add up to half of the whole, i.e., 292 days, and 287 3/4[o], since the equation of the centre has been dispensed with.

72[o] motion for the third 60 days is about 4 [o] in excess. For the next 40 days the motion has to be 36[o], and I could have filled up the lacuna by [Khaissat] instead of [Khairdva] to get this. But the [siddhanta] seems to have compensated the earlier 4 [o] excess by the 4[o] defect here, which of course is an error. From this we can see that the emendation of `dvim'sat' into `trim'sat' is necessary lest the motion be reduced to 26[o], which is too small.

Next, 1 1/2[o] for 23 days is too small to be correct. Further, the correct total of days and degrees clearly given by the numbers will be spoiled by this. So I have given the meaning as 5 1/4[o] in 17 days, which fairly agrees with the actuality, by introducting a (vi) for the defect fo two [matras] and emending [sapadam'sam] into [sapardhan].

Since the motion is only 15/ for the 6 days near the stationary point as seen in the actual, the [siddhanta] is justified in combining this with the -2[o] 13/ for the next 9 days and giving -2[o] for 15 days. But , for the next 5 days, the motion is about-2 1/2[o] and not-2[o], and this is an observational error. For the 5 days forming the half period of invisibility till the inferior conjunction, the actual motion is about-3[o] but we are constrained to make it -2[o] for agreement with the other numbers, especially when it is left to be understood, no motion being given by the text. A glance at the comparative table will make everything clear.

TS have made the serious mistake of thinking that the segments given begin with the superior conjunction instead fo the rising on the west (vide the scheme given in the Sanskrit Commentary, p.123). By the total of 610 days, and 610 1/4[o] given, they have given that the rising takes place 26 days after superior conjunction, passing 22 1/2[o], which is absurd because it should be 37 1/2 in 30 days, i.e. half the time and degrees given for the time from setting to rising. They do not realise that this is the period of the quickest motion. Within the scheme they get the 77[o] for 85 days, by making a drastic change in the wording of the text. Further, for 85 days in that part, the motion would be more than 91[o]. Giving 1 1/4[o] for 3 days near the stationary point is wrong since it should be practically zero. -4 [o] 5days in retrograde is too great. But they have given the same -4[o] for the 10days in the [ativakra] region where the rate should be the greatest.

As for NP they have correctly interpreted that in the three 60 day periods after rising, the motions are 74[o], 73[o] and 72[o], that from stting in the east and rising in the west there are 75[o] for 60 days, half before the superior conjucntion and half after, and that near the inferior conjunction there are 50 days of retrogression and -12[o], half of each falling on each side of the point, as given by the text. Adding these we can account for 250 1/2[o] in 235 days. Since we should get 287 3/4[o] for the 292 days from the superior to the inferior conjunction, we have still to account for 37 1/4[o] in 57 days. This we must seek in the second half of verse 3. By some likely emendations we can secure this, as I have done. But NP have drastically changed the text as

[arthas takavimsatya vim'satyam ['saha] stribhis sa padam'sam],

also sinning against prosody, and given only 28[o] for the 27 1/2 days, after the third sixty-day period, and 1 1/2 [o] for the next 3 days, thus, not accounting for 16[o] and 30 1/2 days.

[ardhastakavim'satya] cannot mean 27 1/2, besides being an un-Sankritic formation. Further, for the 27 1/2 days in that part fo the synodic circle the motion should be more than 26[o] and for the next 3 days, more than 2 1/4[o] as can be seen by examining the actual. This error of 16[o] and 30 1/2 days is doubled for the whole cycle, and the weight of this error of 32[o] and 61 days has been carried by them to the 60-day period of invisibility and drawn the remark on page 121, Part II:"....or 54d and 32[o] more than for a rather implausible conclusion. At any event, the description of the motion of Venus as given in our text seems incomplete". The footnote here is uncalled for.

[VASISTHA-PAULISA JUPITER AND SATURN IN THE

PANCASIDDHANTIKA]

Introduction

In chap. XVIII (of the edition of Dr. Thibaut and M.MSudhakara Dvivedi (TS) of the [Pancasiddhantika] (PS) the computation of Jupiter and Saturn follow next to Venus. This is because their treatment is next simple on account of their small mean motion and equation fo conjunction, owing to their great distance. TS and O.Naugebauer and D.Pingree (NP) in their edition of the PS (chap.XVII) have expressed inability to understand the part of the computation where the equation of the centre is obtained and applied, before the application of the eq. of conjunction. Still, they have attempted to interpret the concerned verses. chagning the wordings drastically, to yield their fancied ideas. In getting the eq. of conj., too, they have made several mistakes.

As in the case of Venus, here too, the true motion is traced from one heliacal rising to the next. The method of getting the true anamoly of the eq. cent. is similar to that of the moon given by the [Vasistha] in chap. II, and based on the same theory of the uniform increase and decrease of the rate of rate of motion, forming a linear zigzag. Even the same technical term, [pada], is used here. All these are reminiscent of the Babylonain astronomy of the Selucid period as I have suggested in my article dealing with Venus. (Ind.J. of Hist. of Sci.Vol.14, no.2, Nov.1979). As between Jupiter and Saturn, their treatment is exactly similar, so that explaining one would suffice for both. Verses 6-13, deal with Jupiter and 14-23 with Saturn. My main aim here is to state and explain the procedure in the computation, a thing not understood by investigatiors. The verificationof the epoch constants depends mainly on comparison with other systems and modern astronomy. So this will be done in a separate paper.

The better reading of the two original manuscripts (printed by TS in the left hand column) has been taken by me. In writing the verses below, the correction of obvious scriptory errors are put in brackets, but the more important ones are indicated also by a question mark.

Jupiter (verses 6-13)

6. Ê´ÉSÉiÉÖκjÉ¶É ÊuùMÉÖ (tÖMÉ)hÉÆ

xÉɽþÒʦɺiÉÉ´ÉiÉÉʦɮúÊ{É SÉ MÉÖ¯û:(®úÉä:)*

½þ(½Úþi´ÉÉ xÉ´ÉxÉ´Énù½þ xÉè-

iÉÖ(¯û)nùªÉÉ ±É¤vÉÉκlÉ (ººÉlÉ) iÉÉ Ênù´ÉºÉÉ:*

7. =nùªÉxÉi´ÉÉÆ¶É (¶ÉÆ) nùS´ÉÉ

ÊnùxÉä¹ÉÖ ¹É½Úþ´ÉMÉǺÉÆMÉÖhÉè¯ûnùªÉ:(ªÉè:)*

BEòxÉ´ÉÉÊOÉUôzÉä (SUôzÉè:)

´É ({É) nùʨÉÊiÉ ºÉɹÉÞnù¶ÉÆ ¶Éä¹É¨ÉÚ**

6. The days of Jupiter from epoch minus 34 d. 34m.divided by 399, give the number of risings. The remianing are days (after rising).

7. Add to these days a ninth of the number fo risings. Multiply the number of risings by 36, add 18, and divide by 391. The remainder here are [pada].

We can conclude the following from these two verses:i. At d.34 N from epoch, the first period from rising to rising begins. ii. The interval between the risings, i.e., the synodic period is 399-1/9=398 8/9 days. iii. 391 [padas] make one full sidereal revolution of Jupiter, i.e., 360[o] of mean motion. In one synodic period, Jupiter moves 36 [padas]. One [pada]=55'15''. 36 [pada]=36[o] 9'. At epocahl days 34-34, Jupiter's longitude is 18 [pada] (=16[o] 35']. But NP have taken the 18 given here as degrees. This is wrong. The difference of 1[o] 25' is too snall to show itself in their verification, Table 32, Part II. But for Saturn this [pada] constant is 89, and for Mars, 85. Taking these as degrees have resulted in big differences and puzzled them. See Part II, page 124. iv. In 391 syn. revolutions there are 391+36=427 solar sidereal revolutions=36 Jupiter's sid. revolutions... one sid.rev. of Jupiter takes 4332-22-48 days, and one sid. rev. fo the sun=365-15-30, we conclude that, there too, as in the moon, it is mixed up with the [Vasistha's].

TS and NP give the same interpretation though making more than necessary emendations. In their verification, TS use the rough syn. period of 399 days instead fo the correct 398 8/9, making the sid. period=4333-35-0.

8. Gò¨É¶ÉÉä ¨ÉvÉªÉ (vɪÉ) º¡Öò]õ¶SÉ

®ú´É{½þÉä [EòɪÉÉê]iÉlÉÉä (ªÉÉä)¶SÉ Ê´É¶ÉÉä (¶±Éä)¹ÉÉiÉÚ*

º¡Öò]õ½þ xÉ tÖ¹ÉÖ nùtÉ-

iÉÚ¦ÉvªÉiÉÚ (x¨ÉvªÉÉiÉÚ)ºÉÉè®äú (ªÉè)%xªÉlÉÉ ½þÉÊxÉ:**

8. One after another, mean and true segments are to be arranged. Taking their difference, if the true is group of Jupiter's days (left over in the synodic cycle as the remining days). Otherwise, (i.e., if the true is more, the dfference is to be substracted.

'True" here means 'true as corrected for the eq. of the cent'. How to get these true positions is given in verses 9-11, and the segments are to be got using these. The mean positions are to be got by using the remaining [padas], extending the work done in verse7.

EòɪÉÉè is introduced to make up for the [matras] wanting. ºÉÉè®äú would mean 'pertaining to either Sun or Saturn'. But we are dealing with Juipter (ºÉÚÊ®ú:) here. ºÉÉèªÉè would alone mean, 'pertaining to Jupiter'. One ms.has no Êuù.

Computing and arranging the mean and true segments against each other is to facilitate interpolation to any required day. It will also be useful to prepare an ephemeride. The example worked will make things clear.

TS have expressed doubts about their translation, since they have not understood verses 9-11. They have retained ʽþ, not supplied the wanting [matras], and not noticed the grammatical error in ºÉÉè®äú. Np have made three drasitc emendations, quite unrealted to the lettering of the text, "ÊxÉʽþiÉ:","¨É{½þ±É:" and "iÉx¨ÉvªÉ®ú´É{½äþ", thought generally following TS.

9. ®úºÉʴɹɪÉEÞòxɶÉÉRÂóEòÉ:

Iɪɮú´É{½äþ Ê´É (®ú´É)vÉ޴ɪÉ:{ÉnÆù ªÉÉ´ÉiÉÚ*

ʴɹɪɮúºÉÉäxÉÉ(? ±ÉºÉä¶É')´ÉÞuùÉè

VÉä´É:ºªÉÉi{É\SÉiÉ´ÉÊiɶÉiÉÉiÉÚ**

10. ¤É½þjɺÉÖ¨ÉxÉ´ÉÉä ½þÉxÉÉè

iÉÞxÉҪɮú´É{½äþ MÉÖ¯ûºiÉÖ ¹ÉÉä½þ¶ÉEäò*

{É\SÉ(?nù)MÉÖÊhÉiÉä jªÉ{ÉÞEò¦ÉÉ-

ÊVÉiÉä Eò±ÉÉ:{ÉÚ´ÉÇiÉÉä%¦ªÉÖÊnùiÉä**

11. xÉ´É ºÉÉvÉÉÇ :EòxªÉÉƶÉÉ:

|ÉlɨÉä ®ú´É{½äþ ÊuùiÉҪɮú´É{½äþ ®ú¡Öò 'ºªÉÖ:)*

SÉGòvÉÆ SÉ MÉÖhÉÉÆ ¶ÉÉ:

nù¶É ¶É (SÉ) Eò±ÉÉ näù´É{ÉÚVɺªÉ**

Ver.9. jupiter being in the diminishing-motion. sector upto 180 [padas], there is the constant 1456 (to work with, in order to get the eq. cent-corrected-Jupiter). Being in the increasing-motion-sector in the next 195 [padas] (i.e.181 to 375), there is the constant 1165.

Ver.10. Jupiter being in the diminishing-motion. sector (again) in the next 16 [padas], there is the constant we want computation from these numbers, in the respective sectors), multiplying them by the [padas] and dividing by the [padas] and dividing by 24, minutes of arc are got, (as the eq.cent.corrected total motion in the respective sector) at the rising in the east (and also thereafter if wanted).

Ver 11. The total of such motion of Jupiter in the first sector is 5r 9[o] 30'. In the second sector, it is 6r 4[o] 10'.

Breifly, expressed as formulae, the eq.cent.corrected Jupiter is given by:

1. If [padas] are from 0 to 180,(1456 -padas)x [padas]'./.24.

2. If [padas] are in the next increasing sector, i.e. from 181 to 375,(1165+padas) [padas']./.24+5r 9[o] 30', where the [padas] used are those gone in that sector.

3. If the [padas] are in the next following sector, i.e. 376 to 391, (1486-padas',.,24+5r 9[o] 30'+6r 4[o] 10', where the [padas] used are those gone inthat sector.

Thought the instructions are laconic, comparison with the moon's computation makes things clear. The increasing-motion sector is obviously the 180[o] from apogee to perigee, where the rate of motion is supposed by this [siddhanta] to increase uniformly from a minimum to a maximum. The apogee is at 180 [padas] (=166[o]) and the perigee at 376 [padas] (=345[o]). The last 16 [padas], contined by the forst 180 [padas] form the diminishing half circle where the rate of motion diminishes uniformly from the perigee to the apogee, Differentiating the formula, (constant+pada) pada'/24, the increase or decrease in the rate of motion is found to be [2'=/2 4 1'/12 per pada]. There may be a small hiatus at the juntion, apogee and perigee, owing to the unequal division of 391 into 196, to avaoid half [pada]. But the average of the rates at apogee and perigee, (1165' and 1486')/24=55'1/4, agrees with the mean motion forming one [pada]. (Incidentally, this justifies our amendment of ʴɹɪɮúºÉÉäxÉÉ into ʴɹɪɮúºÉä¶ÉÉ:. There are other justifications also, as we shall show later). Further, the first [pada] in the first sector must follow next to the rate during the 16th [pada] of the third sector. Since 1486'/24 is taken as the motion of the first [pada],the motion of the 16th is (1486-30)/24=1456'/24, This must be the commencement of the third sector, and this is what is given. We can also see that the fastest rate, (at perigee), is 1486'/24=62', and the slowest, 1165'/24=48'1.2, (at apogee), giving the mean value 55 1/4, of the [pada], ending which there is the apogee, is, (1486-195X2)'/24=1096'/24. But the minimum motion falling at apogee is given as 1165'/24. This hiatus must also be due to the fact that the 2'/24 increase int he rate per [pada] is only approximate, and the actual is a little less than 2'/24. But the formulae are so given that the total of the first sector is ( 1456-180) 180'/24=5r9[o]30', as given. The total of the second sector is, (1165+195) 195'/24=6r 4[o] 10'. The total of the third sector is, (1486-16) 16'/24-16[o] 20'. These add up to 12 [rasis], exactly, as they should. Incidentally, this justifies my emendation of MÉÖhÉÉƶÉÉ: into ªÉÖMÉÉƶÉÉ:, {É\SÉMÉÖÊhÉiÉä into {ÉnùMÉÖÊhÉiÉä, and giving the mening of \ªÉ¹ÉÞEò as 3x8=24. The justicfication for correcting Ê´ÉvÉ޴ɪÉ: to get 180, and ®úºÉÉäxÉÉ to get 1165, are also reinforced by this perfect agreement found here.

TS and NP also give ®ú´ÉvÉÞiɪÉ:, seeing the reason for that. TS emend ®úºÉÉäxÉÉ: into ®úºÉäxÉÉ (=1265), which will give the total 6r 17[o] 43', far from the correct 6r 4[o] 10'. The text itself gives 6r 3[o] 10', one degree off. TS give 6 [rasis] exactly, not knowing the peculiarity of this [Siddhanta]. Using SÉGòÉvÉè thus, they areleft with MÉÖhÉɶÉÉ: nù¶É SÉ Eò±ÉÉ:. This they interpret as 13[o] (wrongly, for it can mean only 30 or 103). Emending nù¶É SÉ Eò±ÉÉ: into {É®ú¶ÉÉEò±Éä, they say that this 13[o] is the total motion of the third sector. They do not realise that the 16[padas] of the third sector is near perigee, and the total motion must be greater than the mean motion, 14[o] 44'. Not knowing the nature of the method-here, they think that the total fo the third sector also should be given. It has no use, and [Varahamihira] has not given it.

About {É\SÉMÉÖÊhÉiÉä jªÉ¹]õEò¦ÉÉÊVÉiÉä, I have emended {É\SÉ into {Énù, to delete the one [matra] in excess, and to give the agreement already seen. jªÉ¹ÉÞEò is 24, as already said. TS retain the {É\SÉ, but emend \ªÉ¹]õEò into +¹]õEò, making it 5/8, leading nowhere.

As for NP, they generally follow TS's emendations. But, for the divisor 8 they suggest the alternative 83 (\ªÉ¹]õEò). Unlike TS, they realise that the three sectors must add upto 12[rasis] and make their own emendation of the last part of verse 11, as ÊuùMÉÖhÉÉƶÉ: nù¶ÉÉ ºÉnù±ÉÉ:, interpreting it as 20[o] 30'. NP have given gist of verse 8 correctly, but making a lot of unnecessary emendations. They have wondered in Part II, why such small units, as [padas], have been taken. This is because, they seem to think, that the three sectors are each taken wholly to get intermediate values by interpolation. An examination of the total of each sector would show how wrong it would be. The true eq.cent. corrected Jupiter is given for the end of any [pada] we want. We are expected to use these to get the true motion through any segmentation of the total [padas], for correct interpolation, and the ends of the segments may fall anywhere, from [pada] 0 to pada 390. Therefore the small [pada] segments are used. I shall work out an example at the end to make everything clear.

I shall explain the rationale of the instruction in verse 8, of adding or subtracting the difference. The eq. cent. corrected Jupiter is subtracted from the sun to get the anomaly of conjunction. So, a positive eq.cent. means, less anomaly of conjunction. The days of the synodic period, corresponding to 360[o] of anomaly. So the day is taken as roughly equal to the degree of anomaly, and the difference in degree subtracted. Vice versa for the eq.cent. corrected Jupiter, it being less tahn the mean. [Varahamihira] is too astute to confuse day and degree, as NP think. (In verses 64-81 too, there is no confusion in the author's mind, as NP seem to think. There he has deliberately chosen the time taken by the sun to move one degree as the unit of time, and call it 'day', for convenience. This is patent ofn the face of the synodic periods given, though TS have not even see it, and are perplexed. We have reason to think that verses 64-81 are by somebody else). (Cf; item 26 of my paper 'Some errors and Omissions etc. [Vishveshvaranand Indological Journal, Hoshiarpur], Vol.XI.1973).

Ver.12:

12. ÊnùxÉ ¹É¹ÉÞlÉÆ¶É (¹É¹]õlÉÉƶÉÉxÉÂ) uùÉnù¶É

®ú´ÉEÞòiÉè´ÉènùÉ: (nùÉxÉÂ) EÞòiÉÉ·ÉʦÉuùÉè SÉ*

ºÉ{iÉɹ]õEäòxÉ ´ÉEòÒ

¹É½Úþ ´ÉMÉÉÇ:(¦ÉÉMÉÉxÉÚ) ¹Éι]õiÉ: ¹É]Âõ SÉ **

Ver.13:

13. +xÉÖ´ÉGòÒ (GòÉä)%¶ÉÒiªÉÉEòÉÇ.

½þÒ (xuùlÉÚ) xÉÉPÉǪÉÖ (¶É) iÉäxÉ xÉ´É iÉiÉÉä %ºiÉʨÉiÉ:*

κlÉi´ÉÉ ºÉèEÆò ¨ÉɺÉÆ

º¡Öò]õÉänùªÉɹ]õÉkÉ®Æú (?ªÉÉä%¹]õÉäkÉ®èú®úRÂóMÉè:)** ´ÉÞ½þº{ÉÊiÉ&**

12: By 60 days, (Jupiter moves) 12[o], by 40 days 4[o] and by 24 days 2[o]. Becoming retrograde, by 56 days he moves 6[o] (i.e.-6[o] and by 60 days, 6[o] (i.e.-6[o]).

13: Following after retrograde, he moves 12[o] in 80 days, and 9[o] in 48 days. Then setting, staying so for a month plus one day, he clearly rises moving 6[o]8'. Ends Jupiter.

The Scheme given

Days Rising in 60 40 24 56 60 80 48 Setting 31 Rising =339

Degrees the east 12[o]4[o] 2[o]-6[o] West East

-6[o] 12[o]9[o] 6[o]8' =33[o]8'

These values agree well with actualities, considering that whole days and whole degrees are given, excepting the last 6[o]8', given to complete the value for the synodic cycle. 6[o]12' would be better at that region and for the whole number, 399 days. 56 days for -6[o], and 60 days for the same-6[o] must be explained by the intention to give whole degrees and segmentation.´ÉMÉÉÇ: is an obvious mistake for ºÉ{iÉɹ]õEäòxÉ, to mean 15, which such an expression never means. It can mean either 56 or 87. They understand another 60 days by the word used. All this, to make up the wrong scheme used by them, based on the mistaken idea that the statement of motions here begins with conjunction and ends with the rising in the east after the next conjunction. The following is their scheme:-

Days Conjunc 60 40 24 15 60 60 80 45 Setting 30 Rising =414

Degrees tion 12[o]4[o] 2[o] 0[o]-6[o] West East

-6[o]12[o]9[o] 15[o] =42[o]

qùÒxÉÉvɶÉiÉäxÉ is emended by TS intovªÉÚxÉÉvÉǶÉiÉäxÉ, but how can this word mean their 45? As for the last part,κlÉi´ÉÉ ¶ÉèEÆò ¨ÉɺÉÆ,they have taken it to mean 30 days instead of the corect 31 days. Let that be. They have not given any motion for it in their interpretation. It cannot be left to be guessed and completed by an ordinary, computer. They, who can be expected to know, have guessed, quite wrongly, 15[o] motion for 30 days, not realising that it can be only 6[o] and a few minutes more. For the 414 days from conj. to the rising after the conjunction, the total can only be about, 33[o] 9'3[o]=36[o] 9', and not the 42[o] given by them.

As for NP, they have emended qùÒxÉÉvÉǶÉiÉäxÉ into ÊnùxÉÉvÉǶÉiÉäxÉ to mean 50 days. Since they take 30 days for the setting, i.e. one day less, they make the total of days, 400. They give 7[o] motion for the 30 days (which they make even 29 days in the last part). They have changed the wording to some ununderstandable form here, t´ÉxiªÉä¨ÉɺɺªÉ. Further, the 7[o] is far too much for 30 days. But there is no 7[o] in the text. They have corrected the text ºÉèEÆò into "<·ÉÆ>", thinking that +·ÉÆ in [bhutasankhya] means 7[o].

Incidentally, one other matter may be considered here, viz., the degrees of heliacal rising, for Jupiter. During the set-period of 31days, the sun moves about 30 1/2 degrees, and Jupiter, about 6[o] 8', and the relative motion is 30 1/2[o]-6[o]8'=about 24[o], from setting to rising. This gives about 12[o], for the helicacal rising of Jupiter, which is fairly accurate, especially for very high latitudes. (Classical Hindu astronomy gives 11[o]). Verse XVIII. 58 gives the [Vasistha-Pauli'sa's] degrees of heliacal rising as 12[o], 14[o], 12[o], 15[o], 8[o], 15[o], from moon onwards, by SÉxpùÉnùÒxÉÉÆ uùÉnù¶É¨ÉxÉÖ®úÊ´ÉÊiÉlªÉ¹]õÊiÉÊlɺÉÆ®ú´ªÉè:. 15[o] for Jupiter given here is too much, and 14[o] for Mars is too low. (Classical Hindu astronomy gives 17[o] for Mars). So, the scribe seems to have made a small change in the order, and the corect order is "SÉxuùÉnùÒxÉÉÆ uùÉnù¶ÉÊiÉÊlÉ-¨ÉxÉÖ®ú´ªÉ¹]õÊiÉÊlɺÉÆ®ú´ªÉè:", 12[o], 15[o],14[o],12[o], 8[o],15[o], with only one change of place.

Example : Find the True Jupiter at 2415 days from epoch.

i. The beginning of the first cycle after rising next to teh epoch is 34-34 days later.

The days after this, required to find the number of cycles gone+2415-34-34+2380-26.

Dividing by 399, cycles gone =2380-26|399=5, with 385-26 remainder.

Adding 5 x1/9 days,(=0-33), we have 385-59 days left over after 5 cycles gone.

ii. The Padas at 5 cycles gone=18+5x36=198.

Mean Jupiter=198 Padas=[198x360[o]\391=6r2[o] 18'.

True Jupiter:-

For the 198 Padas, 180 Padas formingthe first sector has gone and 18 Padas are left over in the second sector.

5r 9[o] 30' +14[o]47'=5r 24[o]17'.

Eq.cent.=True-Mean =5r 4[o] 17'-6r 2[o] 18'=-8[o] 1'.

iii. The Padas at 399 days in the cycle,i.e., the begining of 6 cycles gone=198+36=234=180+54.

Mean Jupiter =234X360./. 391=7r 5[o] 27'.

True Jupiter =5r 9[o] 30'+(1165+54)54'\24=6r 25[o] 13'

True -mean Eq.cent.=10[o] 14'.

Eq.cent. at 0 day of 6th cycle =-8[o]1'

" 399 days of " =-10[o]14'

" at remaining days (385-59)=

=(385-59)X-2[o] 13' ./. 399+-8[o]1'=

-- 10[o] 10'.

iv. True Jup. is less than Mean Jup. by 10[o] 10'... days of Anomaly of Conj.=385-59+10-10=396-9.

v. True an.of conj.=

for 60 days +12[o]

for 40 days + 4[o]

for 24 days + 2[o] 28-9\31X6[o] 8'=5[o] 32'

for 56 days -6[o]

for 60 days -6[o]

for 80 days +12[o]

for 48 days + 9[o]

Total 368 ... +27[o]

for 28 -9 ... 5[o] 32'

396-9 32[o] 32'

vi. True Jup.=Mean Jup. at 0 day of An. of conj.

+ eq. cent. + true ano.of conj.

=6r 2[o] 18'-10[o] 10' +32[o] 32'=6r 24[o] 40'.

Note 1: The need for interpolating the eq. cent. to the remaining days in the cycle be seen by working for 399 days of the 6th cycle and 0 day of the 7the cycle and comparing. They must be the same.

Note 2: The eq. cent. is computed for 0 day of each cycle, i.e., for intervals of 36 Padas=33[o] 9'. Interpolation using these as we have done,can be only rough. To get better interpolations, we can divide the 36 Padas into desired segments, find the eq. cent. of each, and use. We can form an ephemeride, giving the values at teh ends of these smaller segments, each. Or we can form an ephemeride of values at the ends of the day segments given, 60, 40, 24 etc. and use for interpolation. All these logically follow from the instruction, though not specifically stated.

Saturn (Verses 14-20)

As I have already said, the treatment of Saturn is similar to that of Jupiter. So there will be little need for frsh explanations.

14. +vªÉvÉǶÉiÉÆ ¶É (ºÉ)jªÉÆ-

¶É¨É{ÉxɪÉäiºÉÚªÉÇVɺªÉ Ênù´ÉºÉ䦪É:*

´ÉºÉÖ¨ÉÖÊxÉMÉÖhÉÉäPÉÞ (nÚùvÉÞ)iÉ䦪É:

κlÉiÉÆ (iÉÉ) ÊnùxÉÉtɺiÉ(ººÉ)¨É¦ªÉÖnùªÉÉiÉÚ**

15. VÉZÉÉ tÖ(nÖù) nùªÉnù¶ÉÉƶÉÆ

tÖ¦ªÉÉä xɴɺÉÆMÉÖhÉÉnÂù¦É (x¦É) VÉänÖùnùªÉÉiÉÚ(xÉÚ)*

¹É½ÚþʴɹɪɪɨÉè: ¶Éä¹ÉÆ

{ÉnèùªÉÖÇiÉÆ iÉxiÉ (zÉ) ´ÉɶÉÒiªÉÉ**

14. Regarding Saturn, 150-20 days are to be subtracted from the days from epoch. These being divided by 378, the reminder are the days from the rising gone, the quotient being the number of risings gone.

15. One tenth of the rising, (i.e, the quotient), in days, is to be subtracted from the remainder. The number of risings got is to be multiplied by 9, and divided out by 256. The remainder plus 89 Padas form (the padas required for using in the computation). (The idea is that 89 is to be added to (quotient x9), and then divided by 256, to find the padas for use).

In (15), I have emended ºÉÆMÉÖhÉÉnÚù and °ünùªÉÉiÉÚ into ºÉÆMÉÖhÉÉxÉÚ and °ünùªÉÉxÉÚ to agree with ¦ÉVÉäiÉÚ requiring accusatives as als NP. BUt TS have kept them. In NP's emendation ÊnùxÉÉtÉ{iÉÆ, +É{iÉÆ does not agree with the word κlÉiÉÉ and, the meaning also is redundant. Both TS and NP have emended ¡ònèù: into {Énäù, thinking that xÉ´ÉɶÉÒÊiÉ: is degrees. Even this they doubt as seen in the translation, because as mentioned by them in Part II, page, 124, it has led to disagreement.{Énèù:, as it is, clearly says that the 89 is padas. So is the 18 of Jupiter and the 85 of Mars.

We understand from the instruction that the synodic revolution of Saturn takes 378 '/10 days, that in one synodic revolution Saturn moves 9 padas, that 256 padas make nine sidereal revolutions of Saturn, that there are 256 +9=265 sidereal revolutions of the sun in 256 synodic periods of Saturn, and that at 150-20 days from epoch, Saturn's mean longitude is 89 padas. (NP give intheir translation, "89[o]", as mentioned already Therefore, one sidereal revolution of Saturn takes 378 '/10x256./.9=10754.84 days. One sid. revolution of the sun=378'/10x256./. 265=365-15-32.

Again, the Sun's sid. period got is [Pauli'sa's].

One pada =360[o]/256=84'22''.5. The motion in one synodic revolution =9x84'22''.5=12[o] 39'22''.5.

Mean Saturn at 150-20 days after epoch=89x84'

22".5=125[o] 9'.4.

16. ¹É½Úþ°ü{É´Éänù{ÉIÉÉnÚù

´ÉÞÊuùκjɶÉi{ÉnùÉÊxÉ ºÉÉè®úºªÉ*

xÉ´É°ü{ÉʴɹɪɪɨɱÉÉ-

½þÉ (nÚùvOÉÉ) ºÉ: º´É®ú¦ÉɺEò®ú{ÉnùÉ®ú´ªÉ(xiÉ:) *

17. |ÉSɪÉ:º´É®úÉÊOÉ®ú´ÉªÉ¨ÉÉ-

xÉ(zÉ)´ÉxÉ´ÉiÉ (ÊiÉ) κjÉ´ÉPÉxɦÉÉMÉʱÉ{iÉÉxÉɨÉÚ*

IɪɴÉÞÊuù (uùÒ) uùÒ) ÊuùMÉÖhÉ{Énèù-

®äúEòMÉÖhÉ´PÉ: ¶ÉiÉè (xÉä) °ü (°ü) nùªÉ:**

18. ¹ÉÉä½þ¶É ´É޹ɦɺÉäªÉÉƶÉÉ

xÉ´ÉʱÉ{iÉÉ´ÉÌVÉiÉÉ: |Élɨɮú´É{½þÉ:({½äþ)*

ʴɹɪÉÉκjÉMÉxÉ (xÉÉ) ϺjɶÉ-

SÉ (SSÉ) iÉÖªÉÖÇiÉÉ ¨ÉvªÉ¨Éä ®ú´É{½äþ**

Ver. 16 Reegarding Saturn, there is an increase (of the rate of motion) for thirty padas, from 2416. Then, there is a decrease for 127 padas from 2519.

Ver. 17 Next there is an increase for99 padas from 2037. The amount of decrease and increase are by the padas multiplied by 2. The divisor of the total minutes is 27, its multiplier beng one.

Ver 18: The total of the first sector is 1r 15[o] 51' and the total of the middle sector is 5 r 27[o] 34'.

Note: The multiplication by one is unnecessary, but given to clear the doubt that may arise by the instruction to multiply the padas by two for subtraction and additions coming before.

The meaning is clear, and no material change has been needed. I shall give what is given in the form of formulae:

The total motion upto any pada in the first sector

viz.,(1-30)=(2416+2xpadas) padas./.27, in minutes.

-do- -do- second sector.,viz.,

(31-157)=(2519-2xpadas) Padas./.27, in minutes.

-do- -do- third sector, viz.,

(158-256)=(2037 + 2xpadas)padas./. 27 in minutes

The total of the whole of first sector sector=

(2416-2x127) 127./.27=10654'=5r 27[o] 34', given.

Being unnecessary, the total of the third sector is not given. But we can calculate it and use it to see if all those add up to 12 [rasis], as necessary, and this will verify every instruction given.

The total of the third sector=(2037+2x99)x99'./. 27=8195'=4r 16[o] 35'. Now, 11r 15[o] 51'+5r 27[o] 34'+ 4r 16[o] 35'=12r. Examining the constants, we find that the maximum motion per pada is 2519'./.27=93'.3. The minimum rate is 2037'./. 27 =75'.4. The mean rate is =84'.35 as already found, as the mean motion equal to the pada. Differentiating as before, the increase or decrease in the rate is 4'/27. Actually it is slightly less than this, the multiplier being slightly less than 2, given. (2037+4X98)=2416 shows this. The perigee falls at end of 30 padas, i.e., 1r 14[o], and the apogee, 127 padas later, at 7r 13[o].

The instruction howto use the result of these verses has not been given, because it is the same as that given in verse 8 for Jupiter. Indeed, the un-emended reading ºÉÉè®äú there means, "with reference to Saturn".

As in the case of Jupiter, here too TS and NP have not understood what exactly is given in these verses, how it is got by applying the three formulae, how the eq. cent. is got, and why the instruction to apply this to the days remaining is given, in the manner said.

So, their emendations of the readings, done without knowing the subjectmatter, need not be taken seriously. TS have emended the correct ÊuùMÉÖhɽþiÉ, meaning "divided by 32", applied to the risings and not to the number got in the formulae. NP have kept the reading, but given the translation as, "There is a subtraction or addition of 12 degrees and minutes, (i.e., 12[o] 12'). Multiply by 31 and divide (the product) by 32 (or by 32 padas.) (The result is ) Saturn's rising." Where is 12[o] 12' mentioned? They take the 32, not as a number, but as a segment of longitude equal to 32 padas, i.e., 45[o]. Again, how can this give the risings? And the risings have already been given in verse 14. All these show that they do not understand what is said.

19.¹É½Úþ ½þiÉɺjÉÒhÉÉÆ(?EÞòiªÉÉ jÉÒxÉÆ) ¶ÉÉxÉÚ

¨ÉxÉÖ(¨ÉÖÊxÉ)ʦÉ̱É{iÉÉ·ÉiÉÖ®Âú (·Éä¹ÉÖ) MÉÖhÉɺºÉ{iÉ*

¹Éäb÷¶ÉʦɷÉɶÉÒ iÉ (ÊiÉ)

EÞòiÉÉäxɹɹ]õlÉÉ ÊuùMÉÖhÉ(?´ÉänùªÉ¨É) {ÉIÉÉxÉÂ**

20. ´ÉGòÒ Ê´É¦ÉÚiɹÉ{]õlÉÉ

ÊjÉ(jÉÒ)xÉƶÉÉxÉ ¹Éι]õiÉ: EÞòiÉÉxÉ ºÉÉè®ú:*

+xÉÖMÉÉä%EÇò¶ÉiÉèxÉÉÇ (iÉäxÉÉ) ¹]õÉè

¹É]ÚõEÞòiªÉÉ SÉÉ kÉMÉä (MÉÉä) nù½þxɨÉ ** ¶ÉÊxÉ:**

19: Saturn (moves)3[o] in 36 days, 35' in 7 days, 80'in 16 days, and 224' in 56 days.

20: Then becoming retrograde, he moves 3[o] in 55 days, and 4[o] in 60 days. Then following up direct, he moves 8[o] in 112 days, and setting, he moves 3[o] in 36 days in the set period, (i.e. rises in the east after that). Ends Saturn.

This is the scheme given

days 36 7 16 56 55 60 112 36 =378

distance moved Rising East 3o 35' 1o20'3o44' Retrogr -3o -4o Direct 8oSetting West 3oRising East=

I shall now discuss the values given, justifying the three emendations I have made. The corrupt ¹ÉbÚ÷½þ´ÉɺjÉÒhÉÉƶÉÉxÉ has to be emended as 3[o] for 36 days, considering the position, and the fact that it must practically be equal to the rate between setting and rising, 3o for 36 days. The days must add up to 378 days from rising to rising also as from conjunction to conjunction. All the numbers for days are clear. Therefore, the days for the second segment must be 7. So I have emended ¨ÉxÉÖ into ¨ÉÖÊxÉ. The motion given there, 28', gives the rate 2', too absurd for that position, if the original 14 days are accepted, and it cannot be that the [siddhanta] does not know the absurdity. Even for the emended 7 days, it so too low, being only 4' rate, while the rate on both sides is 5', and also consistent with facts. Therefore, ·ÉiÉÖMÉÖÇMÉÉ is emended into ¶SÉäʹÉMÉÖhÉÉ. Now, these three segments can be combined into 4[o] 55' for 59 days, without affecting the result. I do not know why the [Siddhanta] has broken it into such bits.

Next, the total for the 378 days must be the mean motion for the period, i.e., 9 padas, equal to 12[o] 39'.4, roughly taken by the [Siddhanta] as 12[o] 39'. Therefore the motion for the fourth segment, 56 days, must be 224'. so I have emended ÊuùMÉÖhÉ into ´ÉänùªÉ¨É.

In the case of Saturn, too, as in the case of Jupiter, TS and NP have thought that the unnecessary total motion for the third sector has been given. Finding no wording answering to that, they have changed drastically the first half of verse 19, and obliterated the first two segments of days and motion. They have emended the half verse into ®ú´É{bä÷%xiªÉä ʺɽþɶÉɨÉÖxɪÉÉä ʱÉ{iÉÉ·ÉiÉÖMÉÖÇhÉɺºÉ{iÉ as if they are writing their own book. This means int eh last sector the total is 4r 7[o] 28'. But even this does not help to get 12 [rasis], the total coming to only 11r 20[o] 53'.

With the other half and the next verse, they make up the whole scheme as:--

days 16 56 55 68 60 105 36 =396

motion +3o +232' +4o -3o -4o +8o +3o =15o

Chaning +¶ÉÒÊiÉ into +ƶÉÉMxÉÒxÉÚ, not giving any word for the motion of 4o in 55 days, but simply putting the motion there, jÉÒxÉƶÉÉxÉÚ into +¹]õ®úºÉèºjÉÒxÉ newly introducing 68 days, and giving it the retrograde motion -3o, and +EÇò¶ÉiÉäxÉ into +lÉǶÉiÉäxÉ to mean 105 days. As in the case fo Jupiter, they trace the motion from conjunction to the rising after the next conjunction, taking 396 days. But the total motions must then be, 12o 39'+1o 30'=14o 9' and not 15o given. They must know that 3o for 16 days, giving the rate 11 1/4' per day is very much wrong, when the rate is only 5' for the nearer segment got from the motion 3o for 36 days.

As for NP, they emend the first half of verse 19, as {ÉÊ®ú½þÒxÉÉ: ºjÉÒ ®ú´ÉÉƶÉÉ: ¨ÉxÉÖʦÉ̱É{iÉɶSÉä¹ÉÖMÉÖhÉɺºÉ{iÉ meaning "Zero degree of Virgo diminished by 14o, plus 35'", i.e., 4r 16o 35'. They are here better than TS because they have seen that the aim should be to get the total of 12 [rasis] for the three sectors combined. They have also kept closer to the lettering of the text, though the manner in which they have got their total for the third sector is farfetched. After this, they follow the text without chaging it. Only at the end they interpret that the motion of 3 o for 36 days comes before the setting, and leave the period set without any days or motion given. Thus, their scheme is:

days 16 56 55 60 112 36 ? Total 378

motion Rising East 80' retrogrrade 232' -3o -4o direct 8o 3o Setting West ? Rissing East Total 12 o 39'

To make up the totals, a motion of 3 o 27' for 43 days has to be given. But it must be at least 3 o 35'. For the 43 days of the set period, the sun's motion is 42 o 20'. Therefore, the degrees of Saturn for heliacal rising comes to (42o 20'-3o27')./.2=19o 26'. This is far greater than the 15o given in verse 58, and also in all [Siddhantas]. Further, the opposition must occur at the middle of the period from rising to setting and also the middle of the retrograde period. The one falls 169 days after rising, and the other 130 days after, as great as 30 days off. I am sure NP have noted all these discrepancies, but given them as they understood the wording, just to mark time.

I shall now give an example, to make the method clear.

Example:

Find true Saturn at 5000 days gone from epoch.

i Days 5000

To be subtracted 150-20

Dividing by 378)4849-40(12=full cyucles gone)

313-40 (=Remaining days)

Days to be deducted 12/10: 1-12

312-28(Corrected remainder)

ii. Padas at 0 day of the 13th cycle :89+12x9

256 =197

remainder

Mean longitude =197 padas=9r 7o 2'

197=30+127+40(in the third sector)

Eq.cent. corrected mean longitude:-

=1r 15o 51'+ 5r 27o 34' +(2037+2x40)40'./.27

=1r 15o 51'+5r 27o34'+1r 22o16'=9r 5o41'

Eq. cent =9r 5o 41'-9r 7o 2'=-1o21'

iii. Padas at 378 days gone in the cycle =197+9=206

Mean longitude =206 padas =9r 19o 41'.

206 padas=30+127+49 (in the third sector)

Eq.cent.corrected mean longitude=

1r 15o 51'+5r 27o 34' +(2037+2x47)4/2 7/7'=9r 18o 0'

Eq. cent. =9r 18o 0' -9r 19o 41' =- 1o 41'

Interpolated for days 312-28, the eq cent=

-1o 21' -0o 17'=-1o38'.

iv. Correcting the remaining days 312-28 by this, 312-28 +1-38 -6 days, to be used to find anamoly of conjuction.

v. 36 days +3o

7 ... +0o 35' Mean Sat. at 0 day 9r 7o 2'

16 ... +1o20' Eq.cent. --1o 38'

56 ... +3o 42' An.of conj. +7o 40'

55 ... -3o -------

60 ... -4o True Saturn = 9r 13o 4'

Remaining 84-6+6o1 ---------

314-6 +7o40' 8/112 x84-6

As per Ephemeris : 9r 11o.7

THE [VASISTHA-PAULISA] MARS IN THE [PANCASIDDHANTIKA] OF

[VARAHAMIHIRA]

This paper on Mars is contination fo my two earlier papers "[Vasistha-Paulisa][Vanus]" and "[Vasistha-Paulisa] Jupiter and Saturn", contained in verses 21-35 in Cahp.XVIII of Thibaut and Sudhakara Dvivedi's (TS) editiojn, (Chap. XVII in O. Neugebaur and D.Pingree's (NP) edition) of the [Pancasiddhantika](PS) of [Varahamihira] (VM). As I have already said, Mars, and Mercury need elaborate treatment owing to certain peculiarities about them, and so are reserved to the end of PS by VM. The synodic period of Mars on which the equation of conjunction depends is 780 days, during which there are more than two revolutions of the Sun and one revolution of Mars, so that one full anomalistic period of Mars is contained within this period. This, with the large equation of the centre, and the large equation of conjunction causes large variations in its motion from sign to sign and even in the same sign, according to the different types of motion governed by the anomaly of conjunction, like fast, slow, retrograde etc. Hence is the need for detailed treatment.

Further, we have reason to think that the various motions given are all empirical, based on long observation, synodic period after synodic period. The separation into the equation fo the centre, and the equation of conjunction is yet to come, it seems, unlike the cases of Jupiter and Saturn, where it is easy. This would explain certain discrepancies found in the values given.

Regarding the constants given, some can be verified by mutual comparison and corrected where necessary when there is a doubt about the reading itself. But some, like the epoch constants, which are peculiar to the [Siddhanta] itself, cannot be so verified and corrected when necessary. Only in such cases, where we can argue that no [siddhanta] is likely to give such wrong values and those so far from the real, that we can make some plausible corrections. Therefore,not only in the case of Mars, but also Venus, Jupiter and Saturn, I mean writing a separate paper, on their epoch constants.

TS and NP have not understood the nature of the motion of Mars, just as they havve not understood Jupiter and Saturn. While TS have not even attempted translating some verses, wrongly interpreting those attempted, NP have attemped translating all, but many wrongly. I shall point out these after my own translation and discussion of the verses, step by step.

Verse21:

21. tÖMÉhÉä ¹É]ÚõEÆò ´É ({É\SÉ)ªÉ¨ÉÉxÉÚ

ʴɽþÉªÉ {É\SÉɹ]õEÆò SÉ xÉÉb÷ÒxÉɨÉÂ

MÉMÉxÉɹÉÞ¨ÉÖÊxÉʦɰünùªÉÉ

±É¦ªÉxiÉä |ÉÉRÚó ¨É½þÒVɺªÉ**

22. =nùªÉMÉÖÊhÉiÉÉ Ê´ÉxÉÉb÷lÉ:

º´É®úÉÊiÉlɪÉÉä%¤vÉÉ (vªÉ)Îx´ÉiÉÉ ÊnùxÉIÉä{É:*

vÉÞÊiÉMÉÖÊhÉiÉɺªÉ©ÉÒnÖù (iÉÉƺjªÉMxÉÒxnÖù)ʦÉ-

°ünùªÉÉx½þ (xÉ ½þ) i´ÉÉ ÎºlÉEòÉä iÉÉ亨ÉÉ:(iÉκ¨ÉxÉÂ)**

23. {É\SÉɶÉÒËiÉ EÞòi´ÉÉ

|ÉÊiÉ®úɶªÉ ¨ÉvªÉ¨É: @ñ¨É¶É:*

®úÉʶÉ|ɨÉÉhÉiÉ%ºÉ (ºªÉ)

º¡Öò]õiÉÉ SÉÉ (iÉ·É) ®úGò¨ÉÆ EÖòªÉÉÇiÉÂ**

24. º¡Öò]õ¨ÉvªÉ¨ÉÊ´ÉIÉä{ÉÉÆ (¶±Éä¹ÉÉÆ)-

¶ÉÉÎxIÉ{Éäx¨ÉvªÉ¨Éä [%ÊvÉEäò] tÖ¦ªÉ:*

¨ÉvɪɨɽþÉxÉÉè VÉZÉÉiÉÂ

MÉÊEòEòÉä%lÉ SÉÉ®úÉ¨É (xÉ) ʦÉvÉɺªÉä**

21: 'Subtracting 256-40-0 days (-nadis-vinadis) from the days from epoch, and dividing by 780, the synodic risings of Mars in teh east are got.'

22-23: '(157 plus 4) [vindis], multiplied by the risings got, are to be added to the remaining days. Multiply the risings (got by verse 18), and adding 85, divide by 133. The reminder, converted into [rasis] is Mars at rising. According to the whole or portions of [rasis], the true motions are to be taken one after another, and pieced together.'

24: 'The difference between the mean and true degrees should be added (to the reamining days got in verse21), if the mean is greater. I fthe mean is less, the difference should be subtracted from the remaining days. This done, I shall give the true motions according to each type of motion.'

From these verses we learn the following:

(1) 256-40-0 days from epoch, Mars rises int he east, after which the counting of risings begin.

(2) One synodic revolution takes 779-57-19 days (=780 days minus 161 vinadis). The addition of [vinadis] multiplied by revolutions, is for taking the synodic period as approximately 780 days.

(3) For this period of 779-57-19 days, we get 1+18/133 sidereal revolution of Mars, and 2+18/133 sidereal revolutions of the Sun. So, in one synodic period Mars moves 408o 43'.3.

(4) In 133 syn. periods=103734-3-7 days, there are 151 sid.rev. of Mars and 284 sid.rev. of the Sun. From this, the sun's sid. period got is 365-15-38 days and Mar's 686-58-50 days. The sun's period is 38 [vinadis] more than that given for it by the [Vasistha], and near the 365-15-30 of the [Pauli'sa]. Therefore, like the moon, Venus, Jupiter and Saturn, Mars also is common to [Pauli'sa].

(5) At the first rising when calculation commences, mean Mars=85/133 rev.=7r 20o 4'.5.

(6) The addition or subtraction of the difference from the remaining days has been already explained with reeference to Jupiter and Saturn.

No method, however, is given to find the equation of the centre. Now, the true motion is affected by both the equation of the centre and equation of conjunction. The segments of motion given in verses 25-26 are as affected by the equation of conjunction alone. By making the days given for true motion in 27-35, conform to the segments, we can get the rdegrees and through that the days affected by the equation of the centre alone.

This can be of use only for the remainder of days. But the equation of the centre at the beginning of each cycle must be given. It has not been given by any rule. since its period is about 687 days, and it has its own rise and fall ofabout 11o from perigee to apogee and back, it cannot be associated with the synodic period of 780 days. So this is an omission.

I have corrected the corrupt iÉɺjªÉÉ©ÉÒnÖùʦÉ: into iÉÉƺjªÉMxÉÒxnÖùʦÉ:, to mean 133. This is necessary for agreement with the actuals, and the effect of my emendation is seen in my discussion (3) above. TS have made it ¤ÉÉhÉäxnÖùÊ¦É :. How can ¤ÉÉhÉ, with such different lettering, come in here? Further, this will give 18/15 rev. =1 rev. 72o as the mean motion of Mars in one syn. period, 23o wrong per period. They have made {É\SÉɶÉÒËiÉ EÞòi´ÉÉ into {É\SÉÉƶÉÉäxÉÆ and thus shut out the position constant of Mars on the first day where reckoning begins, viz., the point of time 256-40-0 days from epoch. (It will be remembered that in every case, the Moon, Venus, Jupiter and Saturn, they have made this mistake). By this emendation they reduce the motion by 86o 24', and make the mean motion of Mars 345o 36' per syn. period of 780 days!!

As for NP, they have made the correct emendation jªÉMxÉÒxnÖùʦÉ: giving correctly 151 revolutions of mean Mars in 133 syn. periods, and identified it with that given by the Babylonian astronomy of the Seclucid period.

But they have not seen that {É\SÉɶÉÒÊiÉ, meaning 85, is correct as it is and gives the constant 7r 25o 4'.5 at 256-40-0 days from epoch, (see item (5) above). They think it is the constant in degrees, though no word meaning degrees is found here. (This kind of mistake they have made in the case fo Jupiter and Saturn also, as we have shown). But 85o would not do; so they have substituted ºÉÊjÉ®úÉ覃 for |ÉÊiÉ®úɶªÉ and made it 175o. But even this would not do and therefore they have chaged the days from Epoch itself into 216-40-0, by emending ¹É]ÚõEÆò´ÉªÉ¨ÉÉxÉ into ¹É]ÚõEèòEòªÉ¨ÉÉxÉÂ. But this has led to other troubles, leading to theirremark, "For Mars this would mean a longitude of 175o (instead of 194o derived on the basis of a. in table 32). This longitude would correspond to September 27, and a solar position at 186o, hernce to an elongation of 11o" (124, Part II). It is to be noted that 11o for the first visibility of Mars is given by nobody. It is in the range of 14o to 17o. (For details, see separate paper below, "The epoch constants of the [Vasistha-Pauli'sa] star planets").

Verse 25:

25. |ÉÉMÉÖnùªÉä ¹É]Úõ SɹiɺiÉäEò (SÉi´ÉÉ ªÉèEò)-

¨É¹]õÉnù¶É¨É (MÉ) ºiÉiÉÉä ´ÉHò¨ÉÚ*

+iªÉvÉæ SÉ ¶ÉiÉÆ ¶ÉÒ´PÉÉÆ-

nÚùªÉÖxÉÉ ¹Éι]õºiÉ iÉÉä %ºiÉÊ´ÉiÉ:**

(? +vªÉvÉÇ SÉ ¶ÉiÉÆ ¶ÉÒQÉÉÆ-

ºiÉiÉÉä%ºiÉʨÉiÉÉä nÚùªÉÖxÉÉÆ ¹É¹]õ¬É)**

26. ºÉ¨ÉiÉÒPÉÇ SÉ nù¶ÉÊjɪÉÖiÉÉ (iÉÆ)

ÊxÉ®Æú¶ÉÉMÉiÉÉä ʴɶÉËiÉ (? iÉζjɶÉiÉÆ) ´ªÉiÉÒiªÉ EÖòVÉ:*

=nùªÉ¨ÉÖ{ɪÉÉÊiÉ ´ÉIªÉä

MÉÊiÉSÉÉ®ú ÊnùxÉÉ (xÉ) Hò¨Éä SÉÉiÉ:**

25-26: 'After rising in the east, Mars moves 146o (in quick motion) and then 18o each (of slow motion), retrograde, and follow up after retrograde (anuvakra) come, and after that, 150o fo quick (sighra) motion, Then setting, it reaches conjunction (i.e. niram'sagatah) in 60 days, moving 13 plus 30 (=43) degrees. Then it rises, (moving the same degrees in the same days). Beginning from here I shall mention the series of motions, with their days.'

Rises east

+ 176o {146o=I type of motion (sighragati)

{ 18o=II " " (mandagati)

-C.18o {- 7o=III ... ... (vakragati)

{-11o=IV ... ... (ativakragati)

+C.18o V ... ... (anuvakragati)

+ 150o VI ... ... (sighragati)

Sets west

+ 43o VII (in 60 days) (atisighragati)

Conjunction

+ 43o VIII (in 60 days) "

------

Total 412o Rises east

------

The numbers I have given in the scheme are practically what are found in the text, without emendation, excepting three. In verse 25, I have emended SÉ{iɺiÉäEò into SÉi´ÉɪÉæEò to get 146o, the most plausible value. TS have made it ºÉ{iÉiªÉäEò, meaning 176o which is too large. See discussion following 150o is given by +vªÉvÉǶÉiÉÆ, where SÉiÉ is emended into ¶ÉiÉ. This is necessary to make up the total 410o motion in 780 days. Secondly, 43o motion for the days given from stting to superior conjunction is required to agree with the 17o usually given for heliacal rising. This is made up by emending ʴɶÉÊiÉ into ËjɶÉiÉÆ, with the 13o given by nù¶ÉÊjɪÉÖiÉÆ added. I shall now show that the motion of Mars is near 43o in distance nearly 1.53 that of the sun given by modern astronomy and also as computed from Hindu astronomy, the equation of conjunction at this region is 12' per day, (as can be verified) which, plus the daily mean motion of 31'.4, gives 43'.4 perday, making, in 60 days roughly 43o. This also agrees with the angle for heliacal rising of Mars, nearly 17o, given by Hindu astronomy. (In 60 days the sun moves 59o. So the elongation is 59o-43o=16o roughly. If ʴɶÉËiÉ is taken as it is, we get 20o +13o=33o, which is 10o short of the actual 43o and which also gives the angle for heliacal rising as great as 26o, so far from the 14o-17o given by all).

In the mean, the motion from setting to conjunction must be equal to the motion from conj. to rising. That is why it is not given by the text separately. That the motion segments given in the two verses in mean is also clear, since no position of Mars from its apogee is taken into account. So the total motion must be equal to 409o. But the total got by adding the segments is 401o. This must be due to the defective method of the original or the empirical natur e of the motions, and rounding off to whole degrees, as seen from 43o being given for 43o.4. The opposition must fall at the mid-point of the retrograde motion, -18o, and divide it into -9, -9. The total motion from conj.to opposition must be equal to that from opposition to conj. But actually, 43o+146o+18o-9o=198o, and -9o +18o+ 150o+43o=202o, is given. It may be that the angle segments given are empirical, and also there are errors in the apparently correct numbers giving the segments, needing emendation.

The days on the sybodic cycle taken to pass each type of motion must be nearly equal to the average of the days given in 27-35 for that type of motio. This has been used to check the degrees of each type. But the synodic period, as also the mean motion of 1+18/133 revolution during it, are very near accurate, and they must have been got by analysis of the observed motions. So the [Siddhanta] must have known that the motions and times are half and half on both sides of the opposition.

Another point: Beginning from rising, type I is [sighra](quick) motion. II is manda (slow) motion. The distinction seems to be 'faster than the mean' or 'slower than the mean'. so, the dividing point must be where the tangent from the earth touches the synodic circle. Since the distance fo Mars is 1.53 times that of the earth from the sun, this point falls about 189.5 degrees from conjunction. Subtracting 43o.5 from conjunction to rising (given as type VIII) 146o is left for I. This segment extends upto the point where retrograde begins. As the planet is stationary here, a small error of observation can make this lesser or greater than the actual. The text seems to give it as 18o. Types III and IV form the retrograde motion. III is called Vakra (retrograde) and IV, [ativakra] (faster retrograde). This text is defective here, and we cannot fix the exact extent of the retrograde segment. But III and IV seem to be divided as 5:7 of the total. V is [anuvakra], is (i.e., 'follow up after vakra'). In the detailed motions given, this is the sum of segments III and IV, but direct motion. This must be the counterpart of II. Type VI is [sighra],and so the counter part of I. It sextent is given as 150o/ Type VII is the very quick motion (atisighra) from setting to conjunction, and given as 43o in 60 days. Type VIII (atisighra) is the counterpart of VII, from conjunction to rising.

These divisions are mostly based on convention. But as these divisions are given only in the case of Mars, and Classical Astronomy does not give them, we have only to guess regarding the segments. To add to the difficulty, the text is corrupt in the places giving the numbers.

TS have expressed inability to understand verse 25. Still they have made some emendations which do not give any cogent meaning. No translation is given. There is only a question mark. In verse 26, they give 20o motion from conjucntion to rising. This can give only 28 days, as against the 60 days given by the text. By this the elongation for heliacal rising would be 7 1/2, so absurdly low.

As for NP , in both verses, they have needlessly emended correct forms, wrongly emeded the corrupt ones, some in faulty Sanskrit, and given an untenable scheme. The following is their scheme: Rising east /186o motion/18o retrograde motion/180o motion /Setting/30o motion/Conj./30o motion /Rising east. They have made the emedations and substitutions with their eye on the total motion of 409o in the synodic period. They make the total 408o, nearly correct. But they do not identify the vestiges of the different types of motion found in these verses. Further, 30o motion from setting to conj. and then from conj. to rising, is short by 13 1/2o from the actual 43 1/2o. The time required to move 30 o is 42.2 days, and the sun would move 41o.5 for heliacal rising, far short of the actual , especially for such high latitudes as the [Vasistha-Pauli'sa] envisages.

Verse 27:

27. SÉi´ÉÉË®ú¶É¶É (¶ÉUô)ʶÉxÉ¨É (MÉ)

[¨ÉÖ]vªÉ (xªÉ) ¹]õªÉ¨ÉÉÎx´ÉiÉÉ Ê´É{ÉIÉÉ SÉ *

|ÉlɨÉMÉiÉÉè EÖòªÉÉÇÊnù (? Hò¨ÉÊnù)´ÉºÉÉ

¨ÉÒxÉuùÉʶÉuùªÉºÉ¨ÉÉxÉÉ:**

27: In the I type motion, there are 40 +1(=41), 40+7(=47), 40+7(+47), +8 (=48), 40+2(=42), 40-2(+38), days per motion of 30o each, respectively, in each month of the diad of [rasis] beginning from [Mina], (i.e. Pisces).

This means, that for 30o of motion, the time taken is 41 days in the [rasis] is ¨ÉÒxÉ (Pisces) and hÉä¹É (Aries), 47 days in @ñ¹É¨É (Taurus) and ʨÉlÉÖxÉ (Gemini), 47 days in EòEÇò]õEò (Cancer) and ˶ɽþ (Leo), 48 days in EòxªÉÉ (Virgo) and iÉÖ±ÉÉ(Libra), 42 days in ´ÉÞÊ·ÉEò (Scorpio) and vÉxÉÖºÉÚ (Sagittarius), and 38 days in ¨ÉEò®ú (Capricorn) and EÞò¨¦É (Aquarius).

An examination of the rate shows that the perigee is situated at the end of ¨ÉEò®ú, and the apogee at teh end of EòEÇò]õEò, which both fairly agree with the actual.

TS say that they do not understand this vese, and no translation is given, its place being taken by a question mark. NP translate thus: "In the first gati 240 plus 28 minus half (=267 1/2) (days). One should calculate days for every two signs from Pisces." It can be seen that they do not see that this verse gives the detailed rate of motion of the I gati in the diads of [rasis] from Pisces, as affected by the equation of the centre. They think that the first motion given in verse 25, 186o according to them, takes 267 1/2 days, as given by them here. If so, what is the use of the instruction of calculate for "every two signs from Pisces"?

Ê´É{ÉªÉ [®úºÉ] ®ú´´É®ú ºÉ{iÉ ? ®úºÉ )iÉÖÇ-

{É\SÉEòÉnù (xnù) ¶ÉMÉÖhÉÉxÉ Êuù [iÉÒ] ªÉMÉiÉÉè*

ºÉʽþiÉÉxº´É ®èúEò {ÉIÉ-

iÉÖÇ SÉxuù¶ÉÒiÉÉƶÉÖÊ¦É Gò¨É¶É:**

28: In the II type motion, in the same order, (i.e., for each month of the diads, Pisces-Aries, etc.), 18o takes 5x10+7, 6x10+1, 7x10+2, 6x10+6, 6x10+1, and 5x10+1 days.

This gives 57 days each to move in each of the signs Pisces-Aries, 61 days for each of Taurus-Gemini, 72 days for each of Cancer-Leo, 66 days for each of Virgo-Libra, 61 days for each of Scorpio-Sagittarius, and 51 days for each of Capricorn-Aquarius. From the days given it can be seen that there is a slight tilt in the apogee towards Leo, and in the Perigee towards Aquarious. This small difference from the findings in verse 27 shows that the values are empirical.

As for the readings, [®úºÉ] has been inserted because we want six numbers for the six diads, and one is wanting. Sy,,etry requires that it must be ®úºÉ (=6) there. Also, two [matras] are wanting. ºÉ{iÉ is emended into ®úºÉ because, 76 for Virgo-Libra, with 72 on one side, and 61 ont he other, will take the apogee to the end of Virgo, 60o off from its place.

The average in the II type motion is 30o in 61 days, which is less than the mean rate. From this we can conclude that the I type motion is faster thanthe mean, and the II type, slower, as me have surmised.

TS have expressed inability to interpret this verse also, and not translated it. Yet they have made an emendation which need nto be taken seriously, since it has been done without understanding.

NP have interpreted the verse as giving 57, 71, 72 66, 61 and 51, by inserting @ñiÉÖ,as the fourth. But symmetry shows that the second number 71 is wrong, and it must be 61, to avoid the jump from 57 to 72. At any rate, read with their interpretation of verse 27, we can see they do not understand the use of this series of numbers. They do not even say that these are days.

Verse 29:

ZɹɴÉÞÊ·ÉEòÉVÉ´ÉÉ (?SÉÉ){Éä

¹ÉÖ´ÉGäò¹ÉÖ (´ÉGÆò) ¹É]ÚõºÉ{iÉEäòxÉ ´É´É (?xÉ®úÉ) ¦ÉMÉÆ (MÉÉxÉÚ)*

Ê´É (Êuù) EÞòiÉäxÉ ÊnùxÉMÉ (?iÉèxÉÇ´ÉÉ)%ÊiÉ´ÉGòÒ

Ênùxɹɹ]õlÉÉ ¹ÉÉäb÷¶ÉÉxÉÖMÉÊiÉ: **

Verse 30:

MÉäʦÉlÉÖxÉiÉÉèʱÉEòxªÉÉ

xÉÖ´ÉɺÉxÉèºiÉ (? º´ÉÊOɺÉÉMÉ®èú: º´É) ®úÉxÉƶÉÉxÉÂ*

®ú´É (? ÊjÉ) EÞòiÉènÇù¶É Êjɹɹ]õÒ(¹]õªÉÉ)

ºÉ{iÉnù¶É ªÉlÉÉGòɶÉÉ(iÉÂ) **

Verse 31:

EòEÇò]õʺɽþªÉÉä´Éènù-

ºÉÉMÉ®èúººÉ{iÉ ºÉ{iÉ®ú´ÉÉhÉÇ´Éè·É Ênù´ÉºÉÉxÉ (?®ú±ÉÉhÉÇ´Éè: ʶɴÉÉxÉƶÉÉxÉÂ)

¹É]Âõ¹ÉÉ (¹É) ¹]õlÉɹ]õÉnù¶É

Gò¨ÉÉiEÖòVÉÉä ´ÉGò{ÉÚ´ÉÉǺÉÖ **

Verse 32:

PÉ]õ¨ÉÞMɪÉäªÉÇ¨É (?xÉÇMÉ) nù½þxÉè:

{ÉbÚ÷¦ÉÉMÉÉxÉ´É´ÉnÂùMÉ (?zɴɽÖþ) iÉɶÉxÉè®äú (xÉÆ) ´ÉjÉ (?SÉ)*

¨ÉÖÊxÉʴɹɪÉè {É\SÉnù¶ÉÉÆ-

¶ÉEòÉÆ·É iÉnÚùjÉ (iÉÚjÉ)ªÉä%{ªÉÉ®ú: **

Verse 33:

´ÉEäòÊnùxÉÊjɦÉÉMÉè-

xÉÇ´ÉÉƶɪÉÖEòEÖò±ªÉÊVÉxÉè¦ÉÖÇHèò:*

+ÊiÉ´ÉGäò Ê´É{É®úÒiÉÆ

´ÉGò¨ÉxÉÖ´ÉGòMɺjɪÉƶɨÉ **

Ver. 29: In the signs Pisces, Scorpio, Aries and Sagittarius, Mars moves 7o in 42 days when retrograde (vakra), and 9o in 42 days when extra-retrograde (ativakra). In the follow up after retrograde (anuvakra) Mars moves 16o in 60 days.

Ver.30: In the sign Taurus, Gemini, Libra and Virgo, Mars Moves 7o in 43 days retrograde, 10o in 43 days extra retrograde,and 17o in 63 days in the follow-up-after retrograde.

Ver.31: In Cancer and Leo, Mars moves 7o in 44 days, 11o in 46 days, and 18o in 66 days, respectively in the three types retrograde etc.

Verse 32: In Capricorn and Aquarius, Mars moves 6o in 37 days, 9o in 39 days, and 15o in 57 days, respectively int eh three types of motion.

Verse 33: ? ? ? ? ? ?

Types III, IV, and V, called respectively retrograde (vakra), extra-retrograde (ativakra) and follow up-after-retrograde (anuvakra), are given in these verses. The first two are actual retrograde motion, and the third is the slow direct motion following. They are shown hereunder ina tabular form:

Signs Pis-Aries Taur-Gemi Can-Leo Virg-Libr Scor-Sagit Capri-Aquar

------------------------------------------------------------------------

Type

III -7o/42d -7o/43d -7o/44d -7o/43d -7o/42d -6o/37d

------------------------------------------------------------------------

IV -9o/42d -10o/43d -11o/46d -10o/43d -9o/42d -9o/39d

------------------------------------------------------------------------

V +16o/60d +17o/63d +18o/66d +17o/63d +16o/60d +15o/57d

------------------------------------------------------------------------

The division into the three types is arbitrary, based on some convention. By examining the table we can see two things to be noteworthy. The total of I and II is equal to III though III is positive. The days for I and II are the same, except for Cancer-Leo, and Capri-Aquarius. There is symmetry on both sides of these sets. Guided by the above, I have emended certain numbers which glaringly go against these points. In verse 29 xÉ´É for [vakra] is corrected into xÉMÉ since it must be less than 9o given for [ativakra], and both equal to 16o, clearly given for [anuvakra]. In verse 30 the corrupt ºiÉ®ú is changed into º´É®ú to make up the total 17o. The corrupt xÉÖ´ÉɺÉxÉè: is amended into ÊOɺÉÉMÉ®èú:, guided by symmetry. ®ú´ÉEÞòiÉè : is emended into ÊjÉEÞòiÉè: since the number should be greater than 42, by symmetry, In verse 31 the corrupt ºÉ{iÉ®ú´ÉÉhÉÇ´Éè·É Ênù´ÉºÉÉxÉ makes no sense and is corrected into ®úºÉÉhÉÇ´Éè:ʶɴÉÉxÉƶÉÉxÉÂ, because it is reqired to make up the 18o total for [anuvakra]. ºÉ{iÉ is a repetition, and ®ú´ÉÉhÉÇ´É, 40, does not fit since maximum is wanted there, and 46 eminently fits. In verse 32 ªÉ¨É is corrected into xÉMÉ since ªÉ¨Énù½þxÉè: (=32) is too short a period, and far from the 42 days on both sides, and the number should also be a little less than 39. ®ú´ÉäSÉ corrupt is emended into xÉ´É SÉ which will make up the total 15o of [anuvakra].

As for verse 33 the words init are all perfect, without any corruption. But they do not make any sense. It seems that some rules are given here for the division into the three types with days, and the proportion is roughly 5:7:12 of the degrees of all three combined. At any rate, this instruction does not seem to serve any purpose.

[ativakra] represents the faster retrograde motion near opposition plus the slower [Vakra] motion on the other side. That is why it is greater and faster. But why exactly the same number of days? This seems to be a convention. But this is against logic. For, only in Cancer-Leo and Capricorn-Aquarius, there is a small excess of days for [ativakra], but even this is far too small. The sum of [vakra] and [ativakra] is 18o and a mximum at and fairly evenly distributed in between. But actually, at Capricorn-Aquarius, the sum is near 9o, as a comparison with the motion of Mars given in the [Vakyakarana] will show.

TS have translated verses 29-32, omitting verse 33 as obsure. But they think that all three types are retrograde motion, (while only III and IV are retrograde, and V is direct motion). This would make the range of the retrograde motion from 36o in Cancer-Leo to 30o in Capricorn-Aquarius, while the range given is 18o to 15o, and actually the latter is even as small as 9o. Also they do not see that the IV type should be greater than the III. So they give the numbers as they get from the words, instead of emending them appropriately. So, in verse 29, xÉ´É ¦ÉÉMÉÉxÉ as emended by them should be correctly xÉMɦÉÉMÉÉxÉÂ, xÉMÉÉx´ÉGòÒ should be xÉ´ÉÉÊiÉ´ÉGòÒ. In verse 30 their emendation +ΤvɺɨÉÖuèù: should be +ÊOɺÉÉMÉè®ú:, ®ú´ÉEÞòiÉè: should be ÊjÉEÞòiÉè: In verse 31 their ®ú´ÉÉhÉÇ´ÉèÌnù´ÉºÉè:, giving no degree at all for the days, and defective in [matras], should have been emended into ®úºÉÉhÉÇ´Éè : ʶɴÉÉxÉƶÉÉxÉÂ. In 32, symmetry requires our emendation of ªÉ¨ÉnÚù½þxÉè:, while TS have kept it.

As for NP, they have understood that type III gives retrograde and type Iv, extreme retrtograde, though the numbers they give for degrees and days are untenable in many cases. Seeing that the degrees of V is the total of the retrograde motions, while actually V is direct motion. They do not see that if the degrees of V are the total of III and IV, the days too must be the sum of the days, and therefore V is not total retrograde. They have translated verse 33, but this does not give any sense.

BEäòÎxuùªÉ´ÉºÉÖʶɴɨÉxÉÖ-

[¨ÉxÉÖ]¦É´É ÊjÉ´ÉMÉè iÉÖ {ÉIÉ (? MÉÇiÉÖÇ{É\SÉ) (¶ÉÆ) ¶ÉƪÉÖHò¨É *

¶ÉÒQÉMÉiÉÉè {É\SÉɹ]õEò-

¨ÉÚxÉÆ SÉ ¶É¶ÉÉRÅóEòEÞòiÉ´Éänèù:**

Verse 34: In the quick motion following (=sighragati) (type VI), there are the days, 40 +1,40+5, 40+8,40+11, 40+14, 40+14, 40+11, 40+9, 40+5, 40-1, 40-4, and 40-4, for every 30 degrees.

Obviously, these days are given for each one fo the signs beginning from Pisces. The numbers are almost perfectly symmetrical on both sides of the inter-section of Cancer-Leo and Capricorn-Aquarius, reinforcing the conclusion that the former intersection is apogee, and the latter perigee.

[¨ÉxÉÖ] for Leo is a glaring omission and is inserted. Symmetry requires 45 for Scorpio, and so {ÉIÉ is emended into {É\SÉ. The number-wrong for ÊjÉ´ÉMÉÇ and so has been emeded. iÉÖ is removed being an extra [matra] and purposeless. This verse is of the same kind as 27-28 combined, giving types I and II.

TS have not understood what is given here and its purpose, as they have not understood the corresponding verses 27-28. BUt they have translated this verse as the words go, and wrongly too, the numbers given by them forming a mere jumble, without any instruction, 2+1, 2+5, 2+8, 2+14, 2+11. 2+9, 40-1, 40-4, 40-4. No wonder, they append this with a question mark.

NP have emended the already correct {É\SÉɹÉÞEÆò, meaning 40, into {É\SɹÉϹ]õ which they think would mean 5x60=300. {É\SɹÉϹ]õ would mean only 65. To mean 5x60, the form should be {É\Sɹɹ]õªÉ: (nominative) or {É\Sɹɹ]õÒ:(accusative). The troble is that they have not undertsood that the days given are for every 30o, every where, except in types III, IV and V. That is why they make such unauthorised corrections.

¹ÉÏ]ÂõjɶÉiºÉƪÉÖGòÉ

ÊuùEò±ÉɽþÉEòÉÇ (? ZÉxɱÉɽÚþEòÉEÇò) ÊjÉ´ÉMÉÇMÉÖhɶÉÚxªÉè:*

Ênù´ÉºÉɺºÉ{iɨÉMÉiªÉÉÆ

SÉÉ´ÉÉä (®úÉä) ªÉ [] [] nÚùnù¹]õ¨ªÉɨÉÂ**

Verse 35: In the VII type, (i.e., atisighra), for every 30o of the diads Pisces-Aries, Taurus-Gemini etc., there are the days, 36+3, 36+9, 36+9, 36+12, 36+9, 36_3, 36+0. The motion as given for the VII type, is for the VIII type too.

Here, too, the symmetry on both sides of the apogee and perigee is necessary, and the corrupt ÊuùEò±ÉɽÚþEòÉÇ, is corrected into ZÉxɱÉÉRÂóEòÉEÇò corresponding to ÊjÉ´ÉMÉÇMÉÖhÉ. This forms the motion from setting to conjunction. As the VIII type, forming the motion from conjunction to rising is exactly the counterpart of the VII inthe synodic cycle, it follows that type VIII is the same.

TS have emended ÊuùEò±ÉɽþÉ into nÚùlÉxɱÉRÂóEòÉ resulting in 7 quantities of days, while only 6 are wanted for the 6 diads of signs. NP, here too, as in verse 34, have made an unauthorised correction under the same misapprehension. They have substituted ¹Éι]õ< ʦÉjÉ> for the correct {ÉÏnÂùjɶÉiÉÂ. They do not see tha contradiction this leads to. From setting to conj. or from conj. to rising the days they give are not less than 60, 66, on the average. For this, the average motion would be about 46o, in this region of the synodic cycle. But they have interpreted that it is 30o, in verse 26. They have not seen the contradiction.

As I have done for Jupiter and Saturn, here too, I shall work out an example, to make my explanations clear.

Before closing, I wish to say something about Mercury. The case of Mercury is as involved as that of Mars, but in a different way. In one sidereal year Mercury traverses three full synodic cycles, and more. But it moves in the zodiacal circle together with teh sun in the mean, and its equation of the centre depends on its position in the 12 signs. So the motion types vary from synodic clcle to synodic cycle in the same year, needing a large number of day-groups,and these are given by VM. But it is these numbers that are spoiled by scribes most, and the reconstruction is atedious job. (The methods though sometimes peculiar, can be guessed and explained). I have neither the leisure nor the equanimity of mind to undertake this work at present, and hope someone else would do it.

Example: Find true Mars at 800 days from Epoch.

Verses 21-23 Days from Epoch 800-0-0

Subtract days at first rising 256-40-0

----------

543-20-0

----------

No. of revolutions gone 543-20-0

---------=0

780

Remainder 543-20-0

Correction for revolutions gone 0- 0-0

After rising; remaining days --------

543-20-0

---------

Mean Mars at Rising =18x0+85

--------- revolutions =230o

133

Verse 24-35. Motion of mean Mars in 543-20 days=408o 43'X543 1/3 ./. 780=284o 42'. Starting from 230o at rising, true Mars moves in 543-20 days, 20o in ´ÉÞζSÉ 543--20 Type I 146o

To go 10o in ´ÉÞζSÉ takes 14 days. Remaining days 529--20

" 30o in vÉxÉÖ " 42 " 487--20

" 30o in ¨ÉEò " 38 " 449--20

" 30o in EÖò¨¦É " 38 " 411--20

" 30o in ¨ÉÒxÉ " 41 " 370--20

" 16o in ¨Éä¹É " 21--52 " 348--28

II 18o " 14o in ¨Éä¹É " 26--36 " 321--52

" 4o in ´ÉÞ¹É " 8-8 " 313-44

III-9o " -4o in " " 24-36 " 289--8

" -5o in ¨Éä¹É " 30-0 " 259--8

IV-9o " -9o in " " 42-0 " 217--8

V -18o " +16o in " " 60-0 " 157--8

" +2o in ´ÉÞ¹É " 7-25 " 149--43

VI 150o " 28o in ´ÉÞ¹É " 44-48 " 104--55

" 30o in lÉÖ欃 " 51-0 " 53-55

" 29o-57' in Eò]õ " 53-55 " 0

True motion in 543 -20 days (form 20o ´ÉÞζSÉ 29o 57'Eò]õ)=249o 57'.

Mean motion =248o 42'.

Days to be added ==34-45.

True motion for 34-45 days =30o X34 3/4 ./. 54=19o 18'.

True Mars : 29o 57' in Eò]õ +19o 18'=19o 15' in ˺ɽþ.

THE EPOCH OF THE ROMAKA SIDDHANTA IN THE [PANCASIDDHANTIKA], AND THE EPOCH LONGITUDES OF THE SUN AND MOON IN THE

[VASISTHA-PAULI"SA]

I Romaka Epoch

[Varahamihira] (VM) gives in chap, I.8 of his [Pancasiddhantika] (PS) the Epoch of the Romaka [Siddhanta] as mean sunset at Yavanapura (Alexandria in Egypt) ending Sunday and beginning Monday, close to the beginning of the Hindu Caitra 'Sukla of 'saka 427 elapsed, equivalent to 6 p.m. local mean time at Alexandria on the Julian Sunday 20th March, 505 A.D. He says this is the Epoch of the [Pauli'sa Siddhanta] as well. Since in III. 13 he says that the local mean time at Avanti (Ujjain) is 7-20 [nadis] in advance of that of Yavanapura, the moment of the Epoch is 37-20 [nadis] from mean sunrise at Ujjain on Sunday 20the March, 505 A.D. Thibaut and Sudhankara Dvivedi (TS), the first editors of the PS, agree with this. But Neugebaur and Pingree (NP) in their edition of PS (Kobenhavn 1970, 1971) say (Part I, p.8) that it is one day later, i.e., Yavanapura, Monday/Tuesday, equal to 6 p.m. 21 st March. This is wrong, and I shall show in this paper that the Sunday/ Monday, one day earlier, is the Epoch.

The matter can be clinched by comparing the mean new moon of the Romaka with those of the other [Siddhantas] and modern astronomy. They should be reasonably near each other. Since the [tithi] is independent of the origin, the new moon is eminently fit for comparison. VM says that the Romaka [tithi] is tolerably accurate. The epoch constants of PSI. 9-10, can also be used to check the agreement.

The following table gives the mean sun and moon at 37-20[nadis] from mean sunrise at Ujjain on Sunday 20th March, 505A.D. (corresponding to sunset at Yavanapura, Sunday/Monday) according to various [Siddhantas], and modern astronomy. The modern values are tropical, but since the [Ayanamsa] (precession) is near zero and the [tithi] is independent of the origin, this will not affect the result. The [kali] days of the point taken is 1,317,122-37-20 days from the first day of mean [Kali], i.e., Friday mean sunrise at Ujjain, 18th February 3102 B.C.

The Julian days of this point is 1,905, 588-9-43. The titles of the columns I to III are I, Modern Astronomy (Newcombe, Brown), II, [Brahmasphutasiddhanta], and [Siddhanta 'Siromani, III, Modern Suryadiddhanta, Aryabhatiy], and the [Suryasiddhanta] of P.S. Columns IV to VII are for the Romaka at 37-20 [nadis], at Ujjain on Epoch day; IV if the moon's constant is 1984, given for Ujjain sunset, V if the moon's constant is 1984 for Ujjain at 37-20 [nadis], VI, the constant 10984 given for Ujjain at sunset, VII the constant 10984 given for Ujjain at 37-20 [nadis].

I II III IV V VI VII

Mean Sun 359o 37' 0o 42' 359o 49' 359o 42' 359o34' 359o42' 359o 34' Mean Moon 354o 48' 355o 49' 355o6' 0o 56' 359o19' 357o49' 356o 12'

Moon-Sun -4o 49' -4o 53 -4o 43' +1o 14' -0o15' -1o53' -3o 22'

Moment of New

Moon (in nadis)

23-41 24-0 23-11 6-4 1-15 9-11 16-33

later later later earlier later later later

Beginning of Mean

Solar Year 23 43 11 18 26 18 26

(in nadis) later earlier later later later later later

The following must be noted: (1) The positions of the sun and moon are apparently greater by about one degree in the [Brahmasphuta Siddhanta] and [Siddhanta]['Siromani] because thesetake the tropical zero-point about 100years later (c.628 A.D.) as the zero-point of [Mesa Rasi]. But, as we have already said this will not affect our investigation.

(2) The constants for the sun and moon in PS IX. 1-2 are taken for mean noon at Ujjain on Sunday 20th March, 505 A.D. This agrees with the [Khandakhadyaka].

(3) Columns II and III exhaust all main classical [siddhantic] schools. The [Vasistha] and [Pauli'sa] will be treated separately.

We see from these tables that the new moons fo all the classical [Siddhantas] and modern astronomy fall within one [nadi] of one another, showing their accuracy. Their moments are 23 to 24 [nadis] after the time taken. Since the classical group includes the [Suryasiddhanta] fo PS, the point of the time taken, Yavanapura sunset at Sunday/Monday, is the epoch. Now the Romaka, described as nearly correct compared to the Pauli'sa and [Surya] in its [tithi] (cf.PSI.4) must have its new moon also near other new moons, may be not very close. We must rule out its falling, say about 30 [nadis] earlier or later, which can be due to a mistake in the epoch data given. There is aboubt about the time of day for which Romaka constants of the sun and moon are given. Is it 37-20 [nadis] after Ujjain sunrise (sunset at Yavanapura) as normally it should be ? Or is it sunset at Ujjain (7-20 nadis earlier), taking the time for the constant of the moon's anaomaly given in PS VIII. 5 for these also? Columns IV and V of the table give the results of the two cases. Both are unsatisfactory. In the former ease, if Sun/Mon is the Epoch, new moon falls 1-15 [nadis] later than the point of time, about 22 [nadis] earlier than the rest. In the latter, new moon falls 6-4[nadis] earlier than the point of time, 29 [nadis] earlier than the others. If Mon/Tues is taken as eopch, in the former case new moon falls about 38 [nadis] later, and in the latter about 30 [nadis] later. Both are unsatisfactory and so the constants must be wrong.

There is fair agreement between the Romaka mean sun and that of the others. But the mean moon differs byt several degrees if the constant is taken as 1984 [Krtastanavakaika]. By emending the word into [Krtastanavakhaika], we get the constant as 10984. With this constant, given for Ujjain sunset, we get 357o 49' for the mean moon, and 356o 12' if it is given for Ujjain 37-20 [nadis], i.e., at the epoch time of day. The former gives the new moon 9-11 [nadis] later, while the latter gives the new moon 16-33 [nadis] later, as close as 7 [nadis] from the rest. This is the best we can get. If the epoch is taken as Mon/Tues, these two new moon moments will be 46 [nadis] and 53 [nadis] later. Hence the epoch day is Sun/Mon.

This new moon, coming 19-33 [nadis] after zero, Sun/Mon, agrees with the statement [Caitra-'sukladau], the absence of the constant for [adhimasa], and the constant 514 given for [avama]. The sun being at 350o 34', the solar year begins after 26 [nadis], i.e., about 9-30 [nadis] from new moon, for which the [adhimasa] constant will be as small as one, and is therefore neglected in PS 1.9. It is also the beginning of Caitra as stated. If the new moon is taken to the next day, 50 [nadis] in the new moon is taken to the next day, 50 [nadis] in the new solar year would have passed, the new moon would initiate the [adhika-Vai'sakha], and the [adhi-constant] would be as large as 222, disagreeing with the above instruction. These confirm the Sun/Mon epoch. As for the [avama] constant 514, it represents an [avama'sesa] 43-52 [nadis] giving the end of new moon as 16-8 [nadis] from epoch which agrees fairly well with the 16-35 [nadis] got after Sun/Mon.

All these show that the epoch is Sun/Mon and not Mon/Tues. (The Vasistha-Pauli'sa] constants also give this result, as we shall show later). This is confirmed by I. 17-20, which give the Lords of the [savana] year, month and day. At the moment of epoch, (i.e., zero day gone) we have to work with 2227. +0-2227. Dividing by 2520, the remainder is 2227. Dividing this by 7, the remainder is 1. The instruction is that the Lrods are to be counted from the Sun (I.19). So we get, Sunday ending at Zero Epoch, and the next day beginning is Monday. So the Epoch, and the next day beginning is Monday. So the Epoch is Sun/Mon. Now, in part II, p.14, NP add one to 2227 to get the Lord of the first day after Zero Epoch. and dividing out 2228 by 7, get remainder 2 all right, which is correctly Monday, counting from Sunday as one. But they mistake it to be Tuesday, and incorrectly take Mon/Tues to be the Epoch.

Now, we come to the actual day mentioned in PS I.8. The reading of MS. is 'Saumya-divasadye', that of B is 'bhaumya-divasadye'. Bhattotpala's reading is 'soma-divasadye'. It is this reading, meaning 'beginning Monday' that is correct. But NP discredit it as one likely to have been emended by [Sudhakara] Dvivedi in his edition. Whatever it is, the discussion in this paper shows that it is correct. The word [Saumya] also can be taken to mean Monday. It has two meanings, (1) 'the son of Moon', i.e., Budha. If this is taken it would denote wednesday, which is two days later and obviously wrong; (2) 'realated to Moon' o.e., moon's, day, i.e., Monday. The reading, 'bhaumya' is in the worse vitiated manuscript B, and a corrupt form of 'saumya'. Further, [bhaumya] is meaningless, and has to be corrected into [bhauma] to mean Tuesday.

The emendation into [bhauma] is due to Dikshit. In part I. 18 NP day: "Dikshit concludes that the Epoch of VM is Tuesday 22nd March, 505, but that according to the [Suryasiddhanta], the [Ksepakas] (epoch constants) in IX.1-4 are for the noon of Sunday 20th Matrch, and the [Ksepakas] in XVI. 10-11 are for midnight 20/21 March in the same year. The [Ksepakas] in VIII.1,4-5, 8 are computed for sunset of 20th March 505, and this is not the epoch fo the original Romaka [Siddhanta], which he claims was written between the time of Hipparchus and A.D. 150." Now, the days for computing are from the zero day of Epoch, which is precisely the time of the Epoch, and the Ksepakas also are for the time of Epoch. So, according to these statements of Dikshit himself, the Romaka Epoch is sunset, Sunset, Sunday 20th March, near Yavanapura sunset, ending Sunday, beginning Monday, for the Julian date 20th March. Similarly the [Surya siddhanta] Epoch for the sun and moon is Ujjain noon on the same sunday, 22 [nadis] before the Romaka Epoch, and the Epoch for the star planets in PS XVI is 8 [nadis] later than the Romaka Epoch, which is the julian Sun 20/Mon 21, March. Then how could Dikshit say that the Epcoh is Tuesday 22nd, when the purpose of the Epoch is only to give the beginning of the time to be taken in a computation? To compute modern values for comparison, NP go even to 5 P.M. Tuesday 22nd March, one day later than the Epoch they themselves have fixed, and two days later than the correct Romaka Epoch.

Thus Sun/Mon is the Epoch. As already mentioned the constants of the [Vasistha-Pauli'sa], and [Surya siddhanta] (in XVI), also agree with this.

II. The Epcoh Longitute of the Sun And Moon in the [Vasistha-Pauli'sa]

In II.1 the sun's Epoch constant is given by [yasamamrtuyuta], corrected into [mrtuyuta], and thus taking+6. This verse gives the true sun. so the true sun at Epoch is (0X4+6) X30o/127=1o25'. Since on examination we see that the apogee falls near the middle of [Mithuna], we may roughly take the equation of the centre at the beginning of Mesa to be 2o 10', (the maximum being=135') and subtracting this from the true sun, the mean sun is 359o 15' at Epoch.

The mean moon at Epoch can be got from the constant+1936 days, given in PS II. 2. The days to work with is Epoch days+1936=1936. The mean moon=the mean motion in 1936 days+a constant (PS II. 2-4)= 309o 25' +1r-14o-29'=353o 54'. Obviously this constant is the mean moon at the point 1936 days before epoch. This is only 54' less than the modern, given in the table, and need not be ascribed even to the error of the [siddhantas] (its mean motion being remarkably equal to the modern) but to it first point of t (i.e. the vernal equinox) of 505 A.D. as it should be about 60 years earlier. (Its mean sun at Epoch 359o 15, confirms this idea.) Mean sun [minus] mean moon 5o 21, the new moon falls 21-18 [nadis] later than the epcoh time, and if it should fall in line even approximately with the other [siddhantas] and the modern, the Epoch itself should be sun/Moon. Since the Vasistha moon is the same as the [Pauli'sa], the Epoch constants of the [Pauli'sa] got from PSI. 11 and III. 1-3 should condfirm this, and they do, as we shall see.

In part II, p.22, NP go back 1936 days from 5 p.m. 22nd March 505 A.D. arriving at Dee 3, 499, and state that the mean moon of that day at 5 p.m. at Ujjain is 2r 9o 20'. This being=2r 9'7'1" (the constant given in this text by ['sa'smuninavayama's ca ra'syadyah) they say that the 5p.m. 22nd March, and the constant given, are verified at one stroke. But the 5 p.m. 22nd March is nearly one day later than the Epoch they fix, namely Mon/Tues Ujjain 37-20 [nadis] on 21 st March. This new Epoch is given nowhere in the text of [Vasistha] of [Pauli'sa].

The following is their mistake: The correct date to go back is Dec. Ist 499, 6p.m. Ujjain, and at that time the mean moon is nearly 1r 15o 23' by modern astronomy. This is only one degree off the emended text value 1r 14o 29'. Which error can be reasonably attributed to the [Siddhanta]. It should be noted that the apparent agreement brought about by NP is due to two mistaakes made by them, one equal to the other. First the two days error of about 26o mentioned above. Secondly, they have interpreted ['sa'simuninavayama'sca ra'syadyah] as 2r 9o 7' 1'', not paying attention to the word [ra'syadyah] meaning 'in the order beginning from [ra'sis]etc'. The correct value, as given by the emended text is 1r 14o 29'. So the error here is nearly 25o. (cf, the previous footnote).

Next we take the [Pauli'sa]. Since no separate instruction is given, the new moon is the same as the [Vasistha]. To fix the mean sun at Epoch without any doubt, the whole of III.1-3, giving the true sun must be studied, since these verses have not been understood by TS or NP. LIke the [Vasistha], the [Pauli'sa] also beings with giving what is actually the true sun at the begininng of the year, and calling it the mean sun. Therefore, the equation of the centre at the begininng of the year, about 2o 13'.5 minus the 7' got by the instruction to subtract 11'x20/30, equal to 2o 6'.5, is included in the so-called mean sun. I shall translate the verses PS III.1-3.

III I. Multiply the days from Epoch by 120, subtract 33, and divide by 43831. The revolutions etc,, of the mean sun is got. This, plus 20o is the [Kendra] (i.e., anomaly, but here used in the sense of the argument to be used).

III. 2-3. For each [rasi] of Kendra subtract continuously, one for one, 11',48', 69', 70',54', 25', and then add 10', 48' 70', 71', 54' and 25'. The mean sun becomes true.

This stright interpretation gives clearly the method of computation. Example: Let the days from Epoch be 620. Then (120X620-33)/43831=1rev. 8r 10o 52'. This plus 20o gives 9r 0o 52' as the [kendram]. So the true sun=8r 10o 52'-11'-48'-69'-70'-54'-25'+10'+48'+70'+2'(the last term for the 52'left over)=8r 8o 25'. (This is similar to the method given in the Vakyakarana).

I pointed out TS's mistakes in teh paper mentioned in the previous footnote. I subsequently found that in their edition (Part II, P.21) NP have imroved upon TS' Translation, but have made a great mistake in concluding that the sun's maximum eqauion of the centre to be 72'. This is due to their mistaking the 11', 48' etc. as actual values, instead of differences, of the equation of the centre and that the anomalies taken are 290o onwards. But actually the apogee is taken to fall at 70o, i.e., 10o Gemini, not at 80o (i.e.,20o Gemini) as they say. The 70o may appear to be quite wrong, but it may be the relic of a very early period when the vernal equinox was forward by about 5o (the equivalent of 360 years). This shifting of the first point backwards, together with the actual small movement of the apogee, could have been the reason for taking the differences as -11', -48', etc. instead of the correct original differences for 270o, 300o etc., namely-19',-51',-70',-70.-51',-19',+19', +51',+70',+70',+51',+19'.

On this basis, we can fix the mean sun fo the [Pauli'sa] at zero day Epoch. The so-called 'mean sun' is (0x120-33)/4383'=-16'.5. This contains the eqauation of the centre of that point, 133'.5 less the 7' got by computation, i.e., 126'.5. The real mean sun at Epoch is, -16'.5-126'.5=-2o23'=357o37'. The mean sun will reach zero Mesa 2 days 25 [nadis] later than the Epoch. Mean Sun-Mean moon=357o37'-353o54'=3o43'. So the new moon occurs 18-17 [nadis] later than the Epoch. This is [nadis] before the classical and modern new moon, and closer than the Romaka. The days corresponding to the [adhimasasesa] are from mean new moon to mean zero point of Mesa, 2 days, 6 [nadis], 43 [vinadis].

That these findings are correct will be confirmed by the ksepas (constants) for [adhimasa] and [avama], in PS I.11, according to the [Pauli'sa]. The former is 698, giving 698/9761 of a lunar month, nearly equal to the 2 days, 6 [nadis], 43 [vinadis] found above.

The [avama] constant is given by the last foot, both Mss. combined giving the reading, [trikrtadinanya-vamasamksepah]. ['trikrta]' is corrupt, as taken by all. So, whatever is given by '[trikrta]', that is the avama constant given. I emend it into '[vikrtakrtany]' and take it to mean "the [avamaksepa] is less(than the Ksepa given before, i.e. 698) by 44'. So the [avamaksepa] is 654. This gives very nearly 41-43 [nadis], which subtracted from 60 [nadis] gives nearly 18-17 [nadis] as the time of new moon after epoch, giving perfect agreement.

I shall justify my emendation. There must be an [avamaksepa], in the absence of which the [Pauli'sa] new moon will have to be taken as falling at Epoch itself, so far away from others, and so wrong. We cannot take the Romaka [Ksepa] for the [Pauli'sa] since in that case both new moons will fall at the same moment, very unlikely, and also the [Pauli'sas's] will be as wrong as the [Romaka's]. Therefore the Ksepa is given, and that must be 654, if we want the word [Krta] to be kept. TS and NP also have to emend [Krta] to mean 6. TS make it [rta]. (NP's sat is very unlikely, and unnecessary). Their emendation means "63 days are the avamaksepa". What they mean is that the lunar days ar eto be divided by 63 days to get the avamas. But where is the Ksepa? They neglect the word'Ksepah'. We can understand the lunar days to be divided. But these have not become days yet, and there is no meaning in saying 'divide by 63 days'. If 'days' mean lunar days here, then it should be 64. Further, the instrumental case is required to instruct division, not the nominative. My emendation sets everything right.

This detailed discussion, though lengthy, has been required to fix the Epoch and the Sun and Moon at Epoch precisely, removing any doubt. This is necessary for the determination of the Epoch constants of the star-planets of the [Vasistha-Pauli'sa], and to compare them with the constants of the later [Siddhantas], and modern astronomy, which will be done inanother paper.

THE EPOCH CONSTANTS OF THE [VASISTHA-PAULISA] STAR PLANETS

In my paper, 'The Epoch of the Romaa [Siddhanta]' (See above, pp. 188ff.), I have shown that the epoch of the Romaka-Pauli'sa therein is sunset at Yavanapura, ending Sunday and beginning Monday, March 20, 505 A.D., (equal to 37-20 nadis from mean sunrise at Ujjain, as given, 20th March). I have also shown in Section Ii of the paper that the [Vasistha] mean sun at that time is 359o 16', taking the vernal equinox of circa 505 A.D. as the zero point, and the [Pauli'sa] mean sun, 357o 37'. The mean moon according to both is 353o 54'. In this paper the epoch-constants of the star-planets fo the [Vasistha-Pauli'sa] are sougnt to be determined.

Several of the words giving the numbers are corrupt owing to the carelessness of the scribes, and some have been deliberately tampered with inadvertently owing to ingorance. Yet it is possible to determine many of these numbers by mutual comparison, aided by a knowledge of the methods that are being used and the related information given in other verses. But some cannot be checked in these ways. A knowledge of the original [siddhantas] alone can help, but they have not been found yet. In these cases we can only see whether our conclusions agree with other [siddhantas] and modern astronomy tolerably well, explaining small disagreements, and rejecting those that are too far away to be correct. In mythree papers dealing with the [Vasistha-Pauli'sa] Venus, Jupiter-Saturn, and Mars, I was mostly content with correcting the numbers in so far as they mutually agreeed, so that my purpose, viz., to explain the method of computation was not impeded by too much extraneous details. Hence is this paper, where I have attempted to fix the constants as correctly as possible.

1. The Epoch Position of Venus

In my paper on [Vas-Pauli'sa] Venus, I emended the word giving the days of first rising after epoch, [munijala-candran], (=147) into 'jalamunicandran' (=174), as doing the least violence to the text. At 174 days from epoch, the mean sun is =16 1/2o degrees according to [Vasistha] and 16o degrees according to [Pauli'sa]. since Venus is 176o, given, the elongation is 6 1/2o degrees and 7 degrees, respectively, a little less than the 8o given in verse 58 of the chapter. So the emendation into [jalamuni] is plausible, the words having interchanged places as in III. 4. Since the [Vas-Pauli'sa] sun has been fixed near 358o at epoch, the sun would be near 169 1/2o at 174 days, and the elongation fo Venus would be near the 8o required. The small difference can be attributed toa small error accumulated through along period.

This also agrees with the 30 days given from stting to superior conj. and 30 days again from sup. conj. to rising, the motion given being 37 1/2o(verse 5). The sun's longitude would be 30o less, i.e., 139o. Venus would be 37 1/2 less, i.e., 1381/2 mean longitude, both nearly equal. The time of sup.conj. is 142 or 143 dyas, respectively, from epoch. Subtracting the mean motion for the respective days, we get the mean position of Venus at Epoch as 138 1/2o- 227 1/2o=271o, or 138 1/2o-229o6'= 269o24', respectively. Modern astronomy as well as the [Suryasiddhanta] of the PS give 267o. The difference is not great, and can be put to the variation from [siddhanta] to [siddhanta] itself.

2. The Epoch Position Of Jupiter

The numbers giving the Epoch constants are clear. So, we have to put any discrepancy to other causes than wrong numbers. The first rising is 34-34 days from Epoch (verse 6). The [Pauli'sa] mean sun of date is 34o+357o37'=31o37'. The mean Jupiter, as corrected for the equation of conjunction alone is 18 padas=16o 35'. The equation of the centre at this point is that for the 18 padas at rising, minus 16o 35'=(1456-18) 18'/24-16o35'=1o24' (by verses 8,9)... the true Jupiter at this point is 16o 35' +1o24'=17o59', i.e., the same as (1456-18) 18'/24. The elongation for heliacal rising got is 31o 37'-17o 59'=13o38', which is reasonably near that given in verse 58, especially when we see taht we have used only the mean sun here, so that the treatment may be general. The elongation for heliacal rising as computed from the time given from setting to rising, 31days, and the motion for that, 6o 8', (verse 13), viz. (30o 32'-6o8')/2=12o12', is fairly close, when we consider that various values, from 11o in Hindu [siddhantas] to 14o in verse 58, are given.

Continuing , Mean Jupiter 15 1/2 days before 34-34 days=16o 35'-3o4'=13o31'. Subtracting the mean motion for the 19-4 days remaining to reach the epoch, mean Jupiter at Epoch is 13o31'-19-4X5'(perday)= 11o 56'. Modern astronomy gives 9o and the [Surya siddhanta] of the PS 8o.

4. The Epoch Position Of Saturn

Here too there is no doubt about the numbers giving the epoch constants. The method of procedure is similar to that of Jupiter: The first rising is 150-20days (verse 14) after epoch, and the meansun then=the epoch position of the sun+150-20 day's mean motion+ 357o37'+148o 12'=145o49'. The mean longitude of Saturn as corrected for the equation of conjunction =89 padas,=125o9'. The equation of the centre-corrected mean Saturn=1r 15o 51' (for the first 30 padas+(2519-2x59)59'/27, (for the remaining 59 padas)=45o 51+ 87o27'=133o 18', (verse 15,16). The equation of the centre=133o 18'-125o9'=8o9'. (Since 8-9 days will have to be subtracted from the 0 day (verse 8), the actual conjunction will fall 8 days later). The uncorrected elongation=145o49'-125o9'=20o40'. Taken to the date of actual conj., and corrected for its equation of the centre, the sun=145o 49'+8o 2'-2o 10'=151o 41'. True Saturn carried to this point of time =133o 18'+41=133o59'. The elongation is 151o 41'- 133o 59'=17o42', fairly near the 15o for Saturn given by verse 58, when we consider that the values might be empirical.

Continuing, mean Saturn, 18 days before rising,+125o9'-1o30', (=18days motion at the rate 5' near conjunction)=123o39'. Subtracting the mean motion for the 132-20 days remining to reach epoch, at mean motion 2' per day, mean Saturn at epoch=123o 39'-4o15'=119o 24', Modern astronomy gives 122o, and the [Suryadiddhanta] of PS 122 1/2.

4. Epoch Position of Mars

In the case of Mars in verses 21-23, the days of first rising after epoch is given by [satkamvayaman], (which is corrupt) plus 40 [nadis]. TS and I have emended it into [satpancayaman] (=256) while NP have corrected into [satkaikayaman](=216). But 216 is certainly wrong since the true longitude of Mars at that time is given as 85 padas=(=230o), unmistakably. For, 216 days would give the sun in the neighbourhood of 211o, but it must be greater than 230o by about 14o to 17o, for heliacal rising to take place at that time. So we have to take the emendation [satpancayaman], in the absence of any other [plausible]ome. There is doubt only about the mid-digit and no other number can fit, as can be verified.

At 256-40 days from Epoch, the mean sun=[Vas-Pauli'sa] Mean Sun at epoch +256-40 day's motion=357o 37'+253o=250o 37'. The uncorrected elongation is seen to be 20o 37', not far from the 14o to 17o given for Mars by verse 58, and Hindu astronomy. The mean longitude of Mars, 85 padas(=230o) at rising, actually contains the +12o of the equation of conjunction at the point of rising. Since, according to the method of the text, this is constant, and the mean motion for the 780 days from rising to rising is 409o, (got from the 18 padas given plus the one revolution or 360o, understood) which is also constant, the 12o can be combined with this for ease of work. So the actual mean Mars is 218o at rising. Subracting from this the mean motion for 256 days 40 [nadis], equal to 409ox256 2/3 ./. 780=134o 36', we get 218o-134o 36'=83o24', as the mean longitude of Mars at epoch. According to modern astronomy it is about 75o, and according to the [Suryasiddhanta] of the PS 75 1/2o. There is a difference of about 8o. This cannot be reduced by correcting the days alone, since both the 256-40 days, and the 85 padas will have to be corrected. But the 85 is given unmistakably. So it must be put to small errors accumulating over a long period and /or empirical values, and /or some mistake made by somebody somewhere.

Iam not taking Mercury for investigation now,

[SAURASIDDHANTA OF PANCASIDDHANTIKA]: PLANETARY CONSTANTS AND

COMPUTATION (PS XVI, XVII, XVIII)

Chapter XVI of the [Pancasiddhantika] (PS) deals with the compuation of the mean star-planets, Mars etc. according to the [Saura Siddhanta], and chapter XVII of the true motions, with their heliacal risings and latitudes. The mean planets are made true by employing the method of epicycles,a s in the case of the sun and the moon, in chapters XI and X. Of the five [siddhantas] condensed by [Varahamihira] (VM) the [Saura] alone uses epicycles, and there is no evidence of its use in any other. So, in the originals also only the [Saura] must have used epicycles, since V.M. follows the original as far as necessary. Thus the [Saura] is the most mature and may be considered to begin the highest developed stage of Hindu astronomy, represented by the [Aryabhatiya], the [Brahma-sphuta-siddhanta], the Later [Surya Siddhanta]etc.

Though VM's [Saura], being a [Karana], does not use [yuga-cycles] for the planets, the original must have had them, and they can be re-constructed from the epoch-constants given, as we have done in the case of the sun, moon, moon's apogee, and nodes. These can be seen to agree with the corresponding parameters of the [Pauli'sa] quoted by Bhattotpala in his commentary on the [Brhatsamhita], and with the [Ardharatrika-paksa] of [Aryabhata], a work now lost, but reconstructable from its description given in the [Mahabhaskariya], chapter VII.21-35, and from the [Khandakhadyaka] of [Brahmagupta] which latter expressly follows the [Ardharatrika-Paksa]. Not only the [Yuga-cycles], but also the [yuga-days], and epicycles and apogee positions and nodes agree in these. Strangely enough, the Later [Surya Siddhanta] does not agree with the 'Old' in many things. In the matter of computing the latitudes of the star-planets, the [Saura] gives the same method as the [Ardharatrika-paksa] combining two types of latitudes, but the [Khandakhadyaka] follows the [Aryabhatiya] itself exactly as propounded in the [Mahabhaskariya], VI.52-55.

As for agreement of VM's [Saura] with the other [siddhantas] of the period, a perusal of the table appended will show this. But it must be ntoed that the agreement in mere number of cycles is not real agreement, because, the [yuga-days] being different, there will be difference in the calculated mean values. But at the period we are considering, c.500 A.D, the mean positions fairly agree with one another, and also with what would be got by modern astronomy, showing thereby the accuracy of their observations. There is agreement in the degrees for heliacal rising and setting and the method of computi9ng the star-planets between the [Saura] fo the PS and the Later [Surya Siddhanta], though the epicycles differin many ways.

Another important matter should be mentioned. In XVI.10-11, and XVII.10-11a, VM gives corrections, which are his own, to secure agreement with observation to make the [saura] fit for correct almanac-making, which naturally will be demanded by the literate. Thus, in XVI. 10-11, certain [bijas] are given to correct the mean of Mars, Jupiter and Saturn and the [Sighra] of Mercury and Venus. The corrections amount, in terms of [yuga-cycles], to: Mars+57; Mercury +400, Jupiter-33 1/2; Venus-150; and Saturn+25. These corrections ar similar and approximately equal to the famous ´ÉÉM¨ÉÉ´ÉÉäxÉ correction on the [Aryabhatiya], propounded by his successors in his school, to correct his cycles to agree with their observation. I do not suggest that VM was aware of the ´ÉÉM¨ÉÉ¤É correction in that form, but the tendency to correct the earlier results with [bijas] based on observations, is found everywhere, whether north or south, a healthy sign of the growth of the sceince. In XVII.10-11a, VM attempts to correct Mercury and Venus to secure agreement with observation. (See notes under 10-11a).

Another thing is to be noted. In (1) the [Aryabhatiya], in (2) the [Ardharatrika-paksa] (which means [ipse facto] the [Khandakhadyaka], VM's [Saura] and [Bhattopala-quoted Pauli'sa] and in (3) the Later [Surya Siddhanta], the yuga cycles are such that the mean planets are all Zero at the beginning of Kali, the moon's apogee is 90o, and the moon's node 180o. Now the [Aryabhatiya] has equal [yuga-padas], [Krta], [Treta], [Dvapara] and [Kali], i.e., they are equal in llength. The other [Siddhantas] have unequal [yuga] divisions, [Krta] being 4 parts, [Treta] 3 parts, [Dvapara] 2 parts and Kali 1 part. If the other [siddhantas] also postulate, like the [Aryabhatiya], that the planets were created and began to move from the beginning of the [Kalpa] from a zero position, then the cycles should be divisible by 20. But they are not so divisble in all. This necessity is avoided by postulating a time later than the beginning of the [Kalpa] called 'the time of creation of planets' by the Later [Surya Siddhanta], as stated int eh verse,

OɽþIÉnäùiªÉÉÊnù ºÉÞÇVÉiÉÉä %ºªÉ SÉ®úÉSÉ®ú¨ÉÂ*

EÞòiÉÉΤvÉ´ÉänùÉ Ênù´ªÉɤnùÉ-¶ÉiÉ´PÉÉ ´ÉävɺÉÉä MÉiÉÉ& ** (I.24)

and by having both the number of cycles and [yuga-days] divisible by four. In the case of the moon's apogee, the cycles should be odd, and in the case of [Rahu], the cycles should b even, but not divisible by 4. These necessary conditions are indeed found in the Later [Surya Siddhanta] and its kind. Thus, if there is any observed difference in the mean planets, moon's apogee and nodes, they must be due to the 3600 years elapsed after Kali, for the period c.499 A.D. But the observed differences should be only small, and due to error of observation. The cycles must have been, and have been, constructed with an eye to this also. In fact, the number of cylces have been determined by observation, and by using the Diophantine equation (Kuttaka). The difference of just 300 days int eh length of the [yuga, (it does not matter much if it is 328 days, as in the Later [Surya Siddhanta), to secure equality at c.499 A.D., between the [Ardharatrikapaksa] and the [Aryabhatiya], which is called, for distinction, the [Audayika-paksa], meaning the type beginning the day from mean sunrise at Ujjain, provided the number of cycles are the same. (See table). There is a difference of just a quarter of a day accumulated from zero Kali to c. 499 A.D. and the difference is made zero at this point of time. We shall noe proceed to examine the verses, one by one, in chapter XVI and XVII.

B¹É ÊxɶÉÉvÉè%´ÉxiªÉÉÆ

iÉÉ®úÉOɽþÊxÉhÉÉæ (? ÊxÉhÉǪÉÉä%)EÇòʺÉuùÉxiÉä*

iÉjÉäxnÖù{ÉÖjɶÉÖGòÉè iÉÖ±ªÉMÉiÉÉè

¨É½þ (?¨ÉvªÉ)¨ÉÉEæòhÉ** 1**

`The following is the determined position of the star-planets at midnight at Ujjain according to the [Saura Siddhanta]. For their computation, the mean sun should be taken as the mean Mercury and Venus.'

Note : 1 follow TS's emendations.

Example: Find the mean Venus at 1,20,553 days after epoch for the star planets, viz. 427 'Saka elapsed midnight at Ujjain. This is the mean sun at 1,20,553.5 days from midday, of epoch taken for the sun (IX.1). Therefore the mean Venus =the mean sun=(1,20553.5x800-442)./.2,92,207=17o 18'27''.

SÉҴɺªÉ ¶ÉEòɨªÉºiÉÆ

ÊuùÊjɪɨÉÉÊOÉÊSÉ (? ÊjÉ) ºÉÉMÉ®èú´ÉÇ (?Ì´É)¦ÉSÉÉäiÉÂ*

tÖMÉhÉÆ EÖòSɺªÉ SÉxuùÉ-

%%½þiÉÆ iÉÖ ºÉ{iÉɹ]õ[¹ÉbÚ÷]¦ÉHò¨ÉÂ**2**

ºÉÉè®úºªÉ ºÉ½þºjÉMÉÖhÉÉ-

nù (où) iÉÖ®úºÉ (¶ÉÚ) xªÉiÉÖǹÉ]ÂõEò¨ÉÖÊxɶÉèEèò&*

ªÉ±±É¤vÉÆ iÉä ¦ÉMÉhÉÉ

¶Éä¹ÉÉ ¨ÉvªÉÉ (vªÉ) OɽþÉ& Gò ¨ÉähÉè´É **3**

nù¶É nù¶É ¦ÉMÉhÉä ¦ÉMÉhÉä

ºÉƶÉÉävªÉɺiÉi{É®úÉ&¶ÉÖ®äúVªÉºªÉ*

¨ÉxÉ´É& EÖòVɺªÉ näùªÉÉ&

¶ÉxÉÉä¶SÉ ¤ÉÉhÉÉ Ê´É¶ÉÉävªÉɺjÉÖ(ºiÉÖ)**4**

®úÉʶÉSÉiÉÖ¹]õªÉ¨ÉƶÉnÚùªÉÆ-

Eò±ÉÉ Ë´É¶ÉÊiÉ´ÉǺÉֺɨÉäiÉÉ&*

xÉ´É ´ÉänùÉ·É Ê´ÉʱÉ{iÉÉ

¶ÉxÉävÉÇxÉä (xÉÆ) ¨ÉvɪɨÉɺlÉä (¨ÉºlÉä) ´ÉÉ**5**

+¹É]õÉè ¦ÉɪÉÉ (MÉÉ) ʱÉ{iÉkÉÇ({iÉÉ

@ñiÉ)´É&®ú´É¨ÉÖIÉÉä ({ɶIÉÉè) MÉÖ®úÉè Ê´ÉʱÉ}iÉÉ·É *

IÉÉä{É& EÖòSɺªÉ VÉ(ªÉ) ¨ÉÊiÉÊlÉ-

{É\SÉËjɶÉS´É ®úɶªÉÉt&**6**

¶ÉiÉMÉÖÊhÉiÉä ¤ÉÖvɶÉÒQÉÆ

º´É®úxɴɺÉ{iÉɹÉÞ¦ÉÉÊVÉiÉä Gò¨É¶É&*

+jÉÉvÉÇ{É\SɨÉÉ& iÉ-

i{É®úÉ·É ¦ÉMÉhɽþiÉ (?iÉÉ&) IÉä{É&**7**

ʺÉiɶÉÒQÉÆ nù¶ÉMÉÖÊhÉiÉä

ÊuùMÉÖhÉä ¦ÉHäò º´É®úÉhÉÇ´ÉÉÊ·É (·É)ªÉ¨Éè*

+Eêò (vÉê) EòÉnù¶ÉÉ näùªÉÉ

Ê´ÉʱÉ{iÉÉ (Î{iÉEòÉ) ¦ÉMÉhɺÉÆMÉÖÊhÉiÉÉ **8**

˶ɽþºªÉ ´ÉºÉ֪ɨÉÉƺÉÉ&

º´É®úÉänù (®äúxnù)´ÉÉä ʱÉÎ{iÉ EòÉ YɶÉÒPÉvÉxɨÉÂ*

¶ÉÉä[vªÉÉ:]ʺÉiɺªÉ Ê´ÉEò±ÉÉ&

¶ÉʶɮúºÉxÉ´É{ÉIÉÉ(IÉ) MÉÖhÉnù½þxÉ&**9**

To get the mean Jupiter, multiply the days from epoch by 100, and divide by 4,33,232. Revolutions etc. are got. Deduct 10"' per revolution. Add 8o 6'20", the mean at epoch. (This is called Ksepa.) (A bija correction is given by VM, to this, for which see verses 10-11 below.)

To get mean Mars, divide the days by 687. Revolutions etc. are got. Add 14"' per revolution. Add 2r 15o 35'0", the mean at epoch. (See verses 10-11), below for (bija correction).

To get mean Saturn, multiply the days by 1000 and divide by 1,07,66,066. Revolutions etc. are got. Deduct 5"' per revolution. Add 4r 2o 28'49", the mean at epoch. (SEE verses 10-11 below for bija correction).

To get the ['Sighra of Mercury, multiplyt the days by 100 and divide by 8797. Revolutions etc. are got. Add 4 1/2"' per revolution. Add 4r 28o 17'0'', for the [sighra] at epoch. (See verses 10-11, below for [bija] correction.)

To get the ['Sighra] of Venus, multiply the days by 10 and divide by 2247. Revolutions etc are got. Add 101/2" per revolution. Add 8r 27o 30'39'', the [sighra] at epoch. (See verses 10-11 below for [bija] correction).

Note i. I follow TS's emendations, except in vese 6. Where I have read ®úµÉ¨ÉIÉÉè as ®ú´É{ÉIÉÉè instead of their ®ú´É¨ÉIÉÉä, which is meaningless. But their meaning, 20, is all right. In 7, the word IÉä{É can stand, and need not be emended as done by them.

Note ii. The word ¨ÉvªÉ with reference to Mars, Jupiter and Saturn in mean planet in modern parlance, and ¶ÉÒOÉ with reference to Mercury and Venus is mean planet according to modern terminology.

Note III. How to get the days from epoch has already been explained, and it should only to be brought to the midnight following , to be used here.

Example 1. Find the mean Mars at 1,20,553 days from the midnight following the Romaka epoch, which is the epoch given for the star-planets.

1,20,553 ./.687=175 revolutions, 5r 21o 52'40''

The revolution correction=175x14"'= + 41''

Ksepa or mean at epoch = 2r 15o 35' 0''

----------------

Mean Mars at required date = 8r 7o 28'21''

-----------------

Example 2. Find the ['Sighra] Venus at 1,20,553 days for epoch;.

1,20,553X10./.2247=536 revolutions, 6r 2o 19' 23''

Revolution correction =536X10 1/2"= + 1o 33' 48''

['Sighra] at epoch = 8r 27o 30' 39''

-----------------

['Sighra of Venus at 1,20,553 days= 3r 1o 23' 50''

-----------------

Note iv. The rules given to find mean planets etc. depend on the fact that there are approximately 100 revolutions of Jupiter in 4,33,232 days, one revolution of Mars in 687 days, 1000 revolutions of Saturn in 1,07,66,066 days 100['Sighra] (truly mean) revolution of Mercury in 8,797 days and 10 of Venus in 2247 days. The revolution correction make these exact. The epoch constants ar ethe means at epoch.

Note v. From the rules given we can reconstruct the [yuga] cycles of the original [Saura-siddhanta], of which the [Saura] of the PS is a [karana], and from these the epoch constants. These we shall do now. The [yuga] days of the original [Saura] ar e1,57,79,17,800, as computed from the short [Saura-yuga] given in I.14, from which it can be computed that in 1,80,000 years there are 6,57,46,575 days, since the [yuga] is 43,20,000 years, being 24 times the short [yuga].

We might now verify, by calculation, the [yuga] revolutions (Yuga-paryaya) and epoch constants (Ksepa) of the several planets.

Jupiter: [Yuga revolutions]

1,57,79,17,800x100./.4,33,232=

3,64,220 rev., 0r 17o 25' 1"

Revolution correction 3,64,220 x10''' = - 16o 51'43''

---------------

3,64,220 rev., 0r 0o 31'18''

----------------

The error in the [Karana] method is 31'18'' in 43,20,000 years, which is negligible when we consider that the rule is given in a [Karana], which is not intended to be used for such a long period. The [yuga] revolutions 3,64220, is indeed that given in the original, as seen from the [Ardharatrika-paksa] and [Bhattotpala's] [Pauli'sa], and [Khandakhadyaka].

Epoch constant (Ksepa) for Jupiter

The epochis 427 'Saka, i.e. 427+3179=3606 years from zero Kali, i.e. midnight.-3 [nadis], 9 [vindadis]. For 3606 years the motion is 3,64,220X(1/1200+1/1200X600)

==304 rev., 0r 8o 6'36''

Subtracting the motionfor 3

[nadis],9[vinadis] -16''

-----------

Jupiter's epoch constant got= 0r 8o 6' 20''

------------

This is exatly what is given above in verse 6.

Saturn: Yuga revolutions

1,57,79,17,800X1000./.1,07,66,066=

1,46,564 rev., 0r 3o 26' 12''

The cycle correction = 1,46,564X5"' = -- 3o 23' 34''

----------------

----------------

This is indeed the [yuga] cycles given in the [Ardharatrikapaksa] etc. neglecting the small error of 2' 38'' accumulating in 43,20,000 years, due to the [Karana] roughness.

Epoch constant for Saturn

1,46,564 (1/1200 +1/100X1200)=122 rev. 4r 2o 28' 55"

Deducting for 3[nadis], 9[vinadis] - 6"

----------------

The epoch constant got = 4r 2o 28' 49"

----------------

Mars: [Yuga revolutions]

1,57,79,17,800./.== 22,96,823 rev. 6r 29o 4' 59"

Rev. correction== 22,96,823X14"'== + 428o 52' 5"

----------------

----------------

== 22,96,824, in round numbers, being short only by 2o 3', negligible in the long period. We see agreement with the original.

Epoch constant for Mars

The epoch constant is 22,96,824 (1/1200 +1/600X1200)

==1917 rev., 2r 15o 36' 43"

For 3[nadis],9[vindis] -- 1'39.5"

less 1917"' ./.5 -- 6"

---------------

The epoch constant = 2r 15o 34' 58"

---------------

There is agreement.

Mercury : [Yuga revolutions]

1,57,79,17,800X100./.8997=

1,79,36,998 rev. 11r 21o 41' 33"

Rev. correction == 1,79,36,999X 41/2"'= + 1r 13o 41' 15"

---------------

The [Yuga cycles] == 1,79,37,000 rev., 0r 5o 22' 48"

---------------

There is fair agreement with the [Ardharatrika-paksa] etc. with an excess of 5o 22' 48'' in the [Yuga], which need not be considered great in a [Karana] rule.4 7/16 instead of 4 1/2"' would have taken this difference also into account.

Epoch constant for Mercury

1,79,37,000(1/1200+1/600X1200)=

14972 Rev., 4r 28o 30' 0"

Subtracting for the excess 3[nadis], 9[vinadis]- 23'53"

For 1/16 repeat the correction + 16"

--------------

Epoch constant = 4r 28o 17'23"

--------------

Here the constant seems to have been given to the nearest minute.

Venus: [Yuga revolutions

1,57,79,17,800X10./.2247==

70,22,331 Rev., 1r 8o 55' 54"

Revolution correction==

70,22,331X10 1/2=56 Rev., 10r 21o47' 52"

-------------

[Yuga cycles]== 70,22,388 0r 0o 43'46"

-------------

There is a small error of 43'46", negligible in the long period of [yuga], due to the [Karana] rule. 10 85/178 would have been very correct.

Epoch constant for Venus

Epoch constant, 70,22,388(1/1200+1/600X1200)=

5861 rev., 8r27o 35' 38" .4

Less for 3 [nadis], 9[vinadis] - 5' 2"

Extra in the correction + 2' 4"

----------------

Epoch correction (in full agreement)= 8r 27o 30' 39"

----------------

In the sun, moon, [Rahu], and moon's apogee too we see such exact agreement with the [Ardharatrika-paksa], [Khandakhadyaka] and [Bhattotpala-quoted Pauli'sa], from which we can conclude that the source of [VM's[ [Saura] is the [Old Saura Siddhanta].

IÉä{ªÉÉ& º´É®äúxnÖùÊ´ÉEò±ÉÉ&

|ÉÊiɴɹÉÇ (¹Éæ) ¨ÉvɪɨÉÊIÉÊiÉVÉÉä(VÉä)*

nù¶É nù¶É MÉÖ®úÉä̴ɶÉÉävªÉÉ&

¶ÉxÉè·É®äú ºÉÉvÉǺÉ{iɪÉÖiÉÉ&**10**

{É\Sɤnù (¤vÉ) ªÉÉä ʴɶÉÉävªÉÉ&

ʺÉiÉä, ¤ÉÖvÉä ºiÉÉ (®ú´ÉÉ) ʷɺÉ{iɪÉÖiÉÉ&*

®ú´É®ú´É´ÉänäùxnÖùÊ´ÉEòÉ (Eò) ʱÉEòÉ&

¶ÉÉävɪÉÉ& ºªÉÖ& ¶ÉÖ®ú{ÉÚÊVÉÊiɺªÉ ¨ÉvɪÉÉ&ºªÉÖ**11**

'Deduct 10" per year from mean Jupiter. Add 17" per year to mean Mars. Add 7 1/2" per year to mean Saturn. Add 720" per year to teh ['Sighra] ('mean according to modern parlance,) of Mercury. Subtract 45" per year from the 'Sighra] (modern 'mean') Venus. In addition, subtract 1400" or 23'20", constant, from Jupiter's mean.

Note 1. I follow TS's corrections

Note 2. These corrections are obviously Vm's own, to secure agreement with observation, because VM sees the [Saura] used widely for almanac making (besides himself being it follower) and used these [bija] corrections to the [Saura]. Being VM's own, we cannot verify the numbers used, but we can compare these correction with those given by the followers of the [Aryabhatiya] belonging nearly to his time. Note how close they are, and commende the tendency to observe and correct,instead of blindly following the masters

The Kerala school following the [Aryabhatiya] gives the famous [Vagbhava]correction:

¤ÉÉM¦É´ÉäxÉÉSUôEòɤnùÉnÂù PÉxɶÉEò±ÉªÉªÉ½þÉx¨Éxnù´Éè±ÉIɪɮúÉ&*

|ÉÉ{iÉÉʦÉ̱ÉÎ{iÉEòÉʦÉÊ´É®úʽþiÉiɨɴɷÉxuù´ÉkÉÖRÂóMÉ{ÉÉiÉÉ& *

¶ÉÉä¦ÉÉxÉÒ°üføºÉÆÊ´ÉnÚùMÉhÉEòxÉ®ú½þiÉÉx¨ÉÉMÉ®úÉ{iÉÉ& EÖòVÉÉtÉ&

ºÉƪÉÖHòÉ YÉÉ®úºÉÉè®úÉ& ºÉÖ®úMÉÖ°ü¦ÉÞMÉÖVÉÉè´ÉÌVÉiÉÉè ¦ÉÉxÉÖ¤ÉVÉǨÉÂ**

(Katapayadi notation is used here.)According to this, the corrections per annum are for Mars+11.5", for Mercury +105", for Jupiter -12", for Venus-39", and for Saturn +5". See that these compare well with VM's.

Example: Give the [bija] corrections for Mars and Venus at 1,20,553 days from epoch.

This is 330 years.

The correction for Mars=330X17"(positive)=+1o33'30". The correction for Venus=330X45 (negative)=-4o7'30".

Thus [bija]-corrected mean Mars of date is 8r 9o 1' 51", and [bija]-corrected Venus, 2r 27o 16'20".

¶ÉÒQÉÉ®ú´ªÉ%EòÉæ %xªÉÉä¹ÉÉÆ

¦Éè¨ÉÉnùÒxÉÉÆ iÉÖ {ÉÊ®úvɪÉÉä%ÊuùMÉÖhÉÉ&*

{É\SÉËjɶÉÉiºÉ(x¨É)xÉ´ÉÉä-

%¹ÉÞlÉ& ºÉÖ(º´É)®úÉϺjÉ¶É [jSÉ ¦ÉÉMÉ]*:**XVII.1

`For the other planets (i.e. other than Mercury and Venus, viz., for Mars Jupiter and Saturn), the sun is their [sighra]. The epicycles of equation fo the apsis of Mars etc. are twice, 35o, 14o, 16o, 7o, and 30o, (i.e., of Mars 70o, of Mercury 28o, of Jupiter 32o, of Venus 14o and of Saturn60o.)'

Note 1.I follow TS's emendation in {É\SÉËjɶÉx¨ÉxÉ´É& But I read ºÉÖ®úÉ as º´É®úÉ&, like TS, because º´É®úÉ& is nearer the given reading ºÉÖ®úÉ& and also 14o is the epicycle given in the [Ardharatrika-paksa] etc. Five [matras] are wanting in the last foot, and it must be supplied with some such words as ¦ÉÉMÉÉ&, as all numbers are already given. But TS make it ¹ÉbÖ÷ªÉÖ´ÉÉκjɶÉÉ& and, strangely, enought, translate it as 24, confusing the 'addition' mentioned by themselves for 'subtraction'. (However, on page XXIII of the Introduction Thibaut gives the correct 60o, 30o X 2=60o.)

Note2. The first foot is to be read with verses 7 and 8 of chap.XVI wher the ['sighra] of Mercury and Venus have already given. Properly speaking, the matter in this foot should have been given in chap.XVI.

®úºÉ¦É´É´ÉºÉÖnäù´ÉÉEòÉÇ

˴ɶÉÊiÉMÉÖÊhÉiÉÉ& EÖòVɺªÉ nù¶ÉEòÉähÉÉ(xÉÉ)#&*

¨ÉxnùMÉÊiÉxÉɨÉ(?)¦ÉÉMÉÉ&

EÖòVɤÉÖvÉMÉÖ°ü¶ÉÖGòºÉÉä(ºÉÉè)®úÉhÉɨÉÂ**2**

'6, 11,8,4,12, multipied by 20, Mar's being less by 10o, (i.e.120o, 120o,160o,80o, and 240o) are the apogee positions of Mars, Mercury, Jupiter, Venus and Saturn.'

Note 1. I follow TS's emendations ¨ÉxnùMÉÊiÉxÉɨɦÉÉ®úÉ&

does not make any sense. But the meaning is obvious, it must mean apogee positions. Some drastic emendation of the word can be made to give this meaning, but I am against such an emendation.

Note 2. These positions agree with those given in the [Ardharatrika-paksa] etc., as also in the [Aryabhatiya]. The correct positions according to modern astronomy are 128o, 234o, 170o, 290o, and 244o, respectively.

Note 3. The apogee 80o for Venus and the epicycle 14o are the same as given for sun. The apogee position of modern astronomy, so far away. We shall explain this under verses 10-11a, below.

¶ÉÒQÉ{ÉÊ®úvÉÉ´ÉlÉÉƶÉÉ&

EÞòiÉMÉÖhÉ{ÉIÉ(IÉ)Êuù´É μxɶÉÒiÉEò®úÉ&*

{ÉIɺ´É®úÉ ®ú´ÉÆ (®ú´É)¹Ét (bÚ÷ªÉ)-

¨ÉÉ (¨ÉÉ&)®ú´ÉEÞòiÉÉ&EÖòVÉÉnùÒxÉɨÉ ** 3**

'The degrees of epicycles of conjunction of Mars is 234, of Mercury 132, of Jupiter 72, of Venus 260, and of Saturn 40.'

Note 1. I have generally adopted TS's corrections. But the text is corrupt in the third foot, and TS's correction itself wants one[matra]. I would read the third and fourth foot thus:

{ÉIɺ´É®úÉ·É ®ú´ÉÆ ¹Éb÷lɨÉÉ ®ú´ÉEÞòiÉÉ·É EÖòVÉÉnùÒxÉÉ*

This would follow the original work.

Note 2. The values agree with the [Ardharatrika-paksa], [Khandakhadyaka] and [Bhattotpala-quoted-Pauli'sa] group, as to be expected.

Verse 4-9 give the method of computation of true planets.

¶ÉÒOÉÉx¨ÉvªÉ½þÒxÉÉnÂù

´ÉÉ (®úÉ) ʶÉÊjÉiɪÉä MÉiÉè¹{É(¹ªÉ)nÆù¶Éä (¶É)VªÉä*

¦ÉÖ EòÉä]õÒ iÉ{iÉ®úiÉ&

¹ÉbÚ÷¦ÉªÉÉ& (ʦÉ&){ÉiÉ(ÊiÉ)iÉä ºÉ B´É Ê´ÉÊvÉ&**4**

®ú´É{ÉÊ®úÊPÉMÉÖÊhÉiÉä ¦ÉÉVªÉä

®ú´ÉiÉÖÇMÉÖhÉè [ºiÉä]Ê´É{ÉÊ®úMÉ (Ê®úhÉ) iÉä iÉV´É*

EòÉäÊ]õ¡ò±ÉÆ ´ªÉɺÉvÉè

¨ÉÞMÉEòGòªÉÉÇnèù SÉÉ{ÉSɪÉÉ&(ªÉ¨ÉÂ)**5**

iÉnÚù¦ÉÖVÉEÞòÊiɪÉÉäMÉ{Énèù-

¦ÉÉÇVɪÉäjÉxɦÉÚVÉ®ú´ÉÆ (¦ÉÇVªÉÆ ®ú´É)ªÉÇ´PÉ*

iÉV´ÉÉ{ÉÉPÉæ ¨Éxnäù

½þÉÊxÉvÉxÉÆ ¶ÉÒQÉEäòxuù´É¶ÉÉiÉ **6**

º¡Öò]õʪÉi´Éè´ÉÆ ¨ÉxnÆù

¨ÉvªÉÉV´É ʴɶÉävªÉ iɺªÉ ¦ÉÖVɨÉÂ*

{ÉÊ®úhɨªÉ EòɨÉÖÇEòÉvÉæ

iÉx¨ÉxnäùxÉè´É vÉxɽþÉxÉÒ **7**

¨ÉvªÉÉiÉ {ÉÖ®úÉä (xÉ)´Éæ¶ÉÉävªÉ

ºiÉ(iÉ)º¨ÉÉnÂù ¤ÉɽÖÆþxÉ (½ÖþxÉÇ)iɺªÉ ªÉSSÉÉ{ɨÉÂ*

iÉx¨ÉvªÉ¨Éä IɪÉvÉxÉÆ

EòiÉÇ´lÉÆ ¨ÉxnùEäòxnÚù´É¶ÉÉiÉ **8**

BÆ´É º¡Öò]õ¨ÉvɪÉÉ®ú´ÉªÉÉÆ(®ú´ªÉÆ)

¶ÉÒQÉÉiÉ ºÉƶÉÉävªÉ {ÉÚ´ÉÇÊ´ÉÊvÉxÉè´É*

+ÉÊnù´ÉnùÉi{Éä ({iÉÆ) SÉÉ{ÉÆ

º¡Öò]õ¨Évɪɮú´ªÉÉä{É (®ú´ªÉä) SɪÉÉ{ÉSÉ´É&(ªÉ¨ÉÂ)**9**

The first step:

'Deduct the mean from the ['sighra]. If the remainder (called 'sighra-kendra) is within 90o, sin. [sighra-kendra] is called [bhuja], and sin (90o-sighra-Kendra) is called [Koti].

If ['sighra-Kendra] is more than 90o and less than 180o, subtract it from 180o. Taking this as the ['sighra-Kendra'] sin ['sighra-Kendra] is [bhuja] and sin (180o-'sighra-kendra) is Koti. If ['sighra-kendra] is more than 180o and less than 270o, deduct 180o from it and take this as ['sighra-kendra]. sin. ['sighra-kendra] is [bhuja] and sin.90o-['sighra-kendra] is Koti. If ['sighra-kendra] is from 270o to 360, deduct it from 360o and take its sine as the [bhuja] and sin90o-['sighra-kendra] is the Koti. (The bhuja of manda-kendra is to be found in the same way using manda-Kendra in the place of 'sighra-Kendra.)

The [bhuja] and Koti must b emultiplied by the planet's epicycle of conjunction and divided by 360. Thus transformed, they are called [bhuja-result] and [Koti-result] pertaining to the equation of conjunction. If the ['sighra-madhya] is from 270o to 90o, the [Koti-result] is to be added to 120 (the R. of the PS). If ['sighra-madhya] is from 90o to 270o, the Koti-result is to be subtracted from 120. Square this and add it to the square of the [bhuja-result]. Find its square root, and by this divide 120X[bhuja-result]. Find are-sine of this. Subtract half this from the longitude of apsis if the ['sighra-kendra] is from 0o to 180o. Add if from 180o to 360o.

Second step: Halfrectifying the apogee position, thus, deduct it from the mean. The result is to be used as the anomaly of the apsis in the second step. As we find the [bhuja] of the anomaly of conjuynction ['sighrakendra), so find the [bhuja] of the anomaly of apsis. Multiply the [bhuja] by the [manda] epicycle and divide by 360 and get the transformed [bhuja-result] of the apsis. (This is sine equation of the centre.) Find its are-sine. Add half this are to the half rectified longitude of apogee if the anomaly of apsis is from 0o to 180o and subtract if 180o to 360o. Thus the apogee is rectified completely.

Third step: Subtract this rectified apogee from the mean and thus get the anomaly of apsis. Find it s[bhuja] and multiply it by the epicycle of the apsis and divide by 360o. The [bhuja-result], (this is the equation of the centre), is got. Find the are-sine of this, and subtract the whole of this are form the eman of the anomaly fo apsis is from 0o to 180o, and add if from 180o to 360o. The result is rectified mean.

Fourth step: Deduct the rectified mean from the ['sighra]. The anomaly of conjunction is got. Find the [bhuja] and [Koti] of this in the same manner as we did in the first step. Multiply the [bhuja] by the epicycle of conjucntion and divide by 360o. Sine anomaly of conj. is got. Multiply the Koti, i.e., cos. anomaly of conjunction, by the epicycle of conj. and divide by 360o. The realted cosine is got. Add this to 120 if from 90o to 270o. Square this, add the square of the [Bhuja] (i.e.equation of conjunction) and find the square root. Divide the equation of conj.X120 by this square root. The are sine of this is the result. Add this result to the rectified mean if the anomaly of conj. is from 0o to 180o. Subtract otherwise. The geocentric true planet is got.

Note 1. In verse 4, I follow TS's reading, except that I have emended¹ÉbÚ÷¦ÉªÉÉ& into ¹ÉbÚ÷ʦÉ& instead TS's ¹ÉbÚ÷¦ÉªÉUô because mu reading allows us subtraction or addition, as is wanted. In verse 5, I follow TS, except in the second foot, where I give the [Brhatsamhita] reading. Either reading gives the same sense. In verse 6, I follow TS, except in the seconmd foot, where I have given ¦ÉÉVªÉÆ for ʴɦÉVÉäiÉÚ, as being more likely. But the meaning is the same. In verse 7, the text requires no emending, and TS's vÉxɽþÉÊxÉ is unnecessary, In verse 8, like TS, I have corrected {ÉÖxÉUô into {ÉÖxÉ®Âú but I have also corrected ¦ÉɽÖÆþ into ¤ÉɽÖþ®Âú which is required by grammar. In verse 9, I have corrected ¨ÉvɪÉÉ®ú´ªÉÉÆ into ¨ÉvɪÉÉ®ú´ªÉÆ, which is the reading of the other manuscript. Otherwise I follow TS.

Note2. VM here, as elsewhere in the PS, uses his tabular sines, where R is 120', as given in chap IV. So we must use his tabular values to get the R sines and R cosines. Of course, we may use the modern table, or the [Siddhantic] table with R=3438. But there the R, I for the modern tables, and 3438, for the [Siddhantic] tables, is to be used instead of 120', which is instructed here. (VM uses bhuja to mean sine and Koti to mean cosine, instead using the word jya).

Note 3. The method is the same as what si found in the Later [Surya Siddhanta], with some changes for converience. But in the matter of the number or order of the steps, the [Aryabhatiya] and the [Siddhanta-'Siromani] differ. This is because, correctly speaking, the first two steps are useless, and the last two steps alone are necessary. In essence, the third serves to get the true heliocentric position, and the fourth to convert the heliocentric position into geocentric. The earlier steps are in the found hope of getting correct positions agreeing with observation, while the real trouble is in the inexact parameters folloowed by the [Siddhantas].

Note 4. Teh second and thrid steps are merely akin to finding the equation of the centre and applying to the mean. The first and fourth steps are conversion of heliocentric to geocentric positions, neglecting the latitude, which issmall, and does not affect the result much.

Example 1. Find the true, i.e. geocentric, Mars at 1,20,553 days from epoch, given:

Mean Mars, already found with [bija] corr. 8r 9o 2'

'Sighra Mars=mean sun of date=== 0 17 18

Aphelion (apogee) of Mars assumed, 3o 20o

(for 120o given)

Epicycle of apsis==70o, Apsis of conj.=234o.

First step: Anomaly of conj.=

'Sighra-mean =17o 18'-8r 9o2'

=128o 16'.

This is more tahn 90o and less than 180o.

So, subtracting from 180o, [bhujam'sa] is 51o 44', Kotiamsa=38o16'

Bhuja=94'2". Koti =74'17".

Bhuja result=94'2"X234o./.360o=61'14"

Koti result =74'17"X234o./.360o=48'17"

As anomaly oc conj,. is between 90o and 270o, this is subtractive from 120'.

120'Xbhuja-result./.[71'43"+bhuja-result]

=120'X61'14"

----------

71'43"+61'14" =77'55"

Arc for 77'55"=40o 30'

1/2 arc =20o 15'

Subtracting from aphelion (since anomaly of conj. is from 0o to 180o) 110o-20o 15'=89o 45', which is the half corrected aphelion.

Second step:

An. of apsis=mean -half cor.aphelion=17o 18' -65o 43'=311o 35'.

The [bhuja] degrees are 48o 25'.

[Bhuja]=89'44".

[Bhuja]result =89'44"X14o./.360o=29".

Arc. 3'29"=1o40'. Half arc=50', subtractive as an.conj.is from 180o to 360o.

Corrected aphelion=Half corrected aphelion-50'=65o43'-50'+64o53'.

Third step:

An. of apsis=mean corrected aphelion

=17o 18'-64o53'=312o25'.

[Bhuja] degrees =47o 35'. [Bhuja]=88'33".

[Bhuja] result=88'33"X14o./.360o=3'27".

Arc sine 3'27"=1o39', additive as An. of apsis is 180o to 360o.

So, mean Venus +arc=17o18'+ 1o 39'=18o57' is the eq.cent.correctedmean.

Fourth step:

An.conj.=[Sighra-corrected mean

=87o16'-18o57'=68o 19'.

[Bhuja] degrees =68o 19'. Koti degrees =21o 41'

[Bhuja]=111'29". Koti=44'19".

[Bhuja]-result=111'29"X260o./.360o=80'31"

[Koti]-result=44'19"X260o./.360o=32'0".

As An. conj. is from 270o to 90o, additive to 120'.

So, 32'o"+120'=152'o"

[Bhuja] result X 120./.[Bhuja-result+152==62'12".

Arc sine=62'12". Half arc=31o 14', additive as an.conj. is 0o to 180o.

Geocentric true Venus=18o 57' +31o 14', =50o11'.

In verse 11, below VM requires us to subtract 67' or 1o7' constant, as [bija]-correction, after all work is over.

Geocentric True Venus=50o11'-1o7'=49o41',

ºÉ´Éæ º¡Öò]õÉ& ºªÉÖ®äú´ÉÆ

¶ÉºªÉ iÉÖ ¶ÉÒQÉÉÊuù½þÉªÉ ®úʴɨÉxuù¨ÉÂ*

®úÊ´É MÉÊ®úÊvÉxÉiÉÆ ¤ÉɽÖÆþ

¤ÉÖvÉ´ÉiÉ (¤ÉÖvÉä%EÇò´ÉiÉÂ) IɪÉvÉxÉä EÖòªÉÉÇiÉÂ**10**

¶ÉÖGòºªÉ ºÉ{iÉ´ªÉι]õ (¹Éι]õ)ʱÉ{iÉÉ&

¶ÉÉävɪÉÉ& º¡Öò]õÒEÞòiɺ{Éè (ºªÉè) ´É **11a**

10. All star-planets are (geocentrically) made true in the above manner. But in the case of Mercury, this additional work is to be done: subtract its apogee from the [sighra], and using the sun's epicycle, find the [bhuja]-result and apply it to the mean Mercury (which, of course, is the same as teh sun's), with the addition or subtraction done, as the sun's [bhuja]-result is additive or subtractive.

11a. From Venus, subtract 67', constant, after all the earlier [sptuta] work instructed has been done.

Note 1. The reading IɪÉvÉxÉä is better as it is, and T.S. need not have corrected it into IɪÉvÉxɨÉä. The reading ¤ÉÖvÉ´ÉiÉ is deficient by two syllables. Lalla's reading ¤ÉÖvÉ´ÉiÉÂ. Being an arbitrary rule, we cannot decide which gives the original meaning, ¤ÉÖvÉä%EÇò´ÉiÉ or ¤ÉÖvÉ´ÉiÉÂ. But since Lalla's reading is not defective, at least as far as the [matras] are concerned, I have adopted it. Also, it is clear that it is not a substitute for any of the four steps because, if so, the separate epicycle for Mercury will be useless.

Note 2. It is clear that the rules given here are VM's own, to secure, in his opinion, better agreement with observation, because they are not given in the [Ardhratrika-paksa] etc. and the original four steps are all in line with them, as also the Modern [Surya Siddhanta] and [Siddhanta 'Siromani].

Note 3. The whole work of finding the true positions, especially of the star-planets, is defective in Hindu astronomy in that the equation of the centre of Hindu astronomy neglects the second, third, etc. terms, which is considerable in the case of the moon, Mars, Saturn and Mercury, in which last case the second term is as large as 3o. In the case of Mercury and Venus it is applicable to the sun, instead of their [sighra] which is really their mean. In the equation of conjunction, the sun's true distance from the earth, and true longitude should be used, instead of the mean distance and mean longitude, as is done by Hindu astronomy. On account of these defects, computation does not agree with observation, and all sorts of hotch-potch rules are given in different astronomical works. The disagreement among themselves would itself show that they are beside the mark. When these defects are remedied,the thrid step giving the heliocentric true planet, and the fourth step converting the heliocentric position to the geocentric.

Note 4. In the case of Venus, there is another kind of defect. Its maximum eq. of cent. being small, it is confused with the sun's, adn the sun's epicycle and apogee are given to Venus also. While its aphelion position is 290o according to modern astronomy, its apogee is given as 80o, the same as the sun's.

TABLE I

Heliocentric Star-planets at Epoch, (for mutual comparison)

------------------------------------------------------------

Planets Modern Siddh. Later Earlier Vasistha Interpola-

astronomy 'Siromani Surya Surya Pauli'sa tion of

Siddhanta Siddh. of PS PS xviii

pf PS

--------------------------------------------------------------

Mercury 151o 148 1/2o 166o* 148o 161 1/2o

Venus 269o 268 1/2o 264o 267o 269 1/2o 269 1/2o

Mars 75o 76 1/2o 78o 75 1/2o 83 1/2o 83 1/2o

Jupiter 9o 9 1/2o 9o 8o 12o 9o

Saturn 122o 122o 123 1/2o 122 1/2o 120o 118o

---------------------------------------------------------------

For values in column (5) see the papers ['Vasistha-Pauli'sa] epoch constant's (pp.201-5) and 'Mars' (above). All values have been computed by me.

TABLE II

Synodic periods of the Star-planets

-------------------------------------------------------------------------------

1 2 3 4 5 6 7

Planets Mod. Sid,'Sir. New PS Su. PS Vas.- Interpol- Ptolomy

Ast. Su.Siddh Siddh. Paul. ated

section

--------------------------------------------------------------------------------Mer.115. 87747766 8784290 8780110 8785195 8791307 8750556 .879

Venus583.92136655 8968279 9001782 8975750 9092440 9060301 584.000

Mars 779.93610175 9222494 9242712 9211734 9553326 9787326 .943

Jup. 398.88404760 8894794 8891768 8891698 8891358 8852917 .886

Sat. 378.09190150 0859936 0863874 0860183 0997090 1100185 .093

-------------------------------------------------------------------------------- All values except column (7) have been computed by me. In column (6) the solar 'days given have been converted into ordinary days.

´ÉGòÉxÉÖ´ÉGòEòɱÉÉè

¦ÉÖÊHòʴɶÉä¹ÉähÉ Ê´ÉYÉäªÉUô**11b**

11b. The times from the beginning of the retograde motion to is end and the follow up period can be found by the dauly motion (being negative), during this period, and the convention regarding these.

Note: The term [vakra] (retograde) and [anuvakra] (follow-up at the end of retrograde) are techincal. They are eight in number according to the [Surya Siddhanta], (given by the verse):

´ÉGòÉÊiÉ´ÉGòÉ EÖòÊ]õ±ÉÉ ¨ÉxnùÉ ¨ÉxnùiÉ®úÉ ºÉ¨ÉÉ*

iÉlÉÉ ¶ÉÒQÉiÉ®úÉ ¶ÉÒQÉÉ OɽþÉhÉɨɹ]õvÉÉ MÉÊiÉ&**

The generally given reading ´ÉGòÉxÉÖ´ÉGòÉ is wrong in my opinion and I have read it as ´ÉGòÉÊiÉ´ÉGòÉ, and +ÉÊiÉ´ÉEòÉ has taken the place of +xÉÖ´ÉGòÉ in the verse. The expression ªÉÉ ´ÉGòÉ ºÉÉxÉÖ´ÉGòMÉÉ in the next verse makes it clear.

Generally the near are subsumed into one another. But in the case of Mars VM gives all these eight and their degrees and periods, (See paper, ['vasistha-Pauli'sa] Mars in VM's P.S.' above, pp. 169ff., under verses 33-34).

º¡Öò]õÊnùxÉEò®úÉxiÉ®úÉxiÉ®úÉÆ (Eò®úÉxiÉ®úÉÆ)¶ÉÉUô

SÉxuùÉnùÒxÉÉÆ SÉ nù¶ÉÇxÉÒ (xÉä)YÉäªÉÉUô*

Ê´É(Ë´É) ¶ÉÊiÉ°ü (°ü)xÉÉ ´ÉºÉÖ-

¶ÉÉʶÉ(delete)ʶÉÊJɨÉÖÊxÉxÉ´É°üuäùÊnù(Îxuù)ªÉä& Gò¨É¶É&**12**

12. The heliacal rising and setting of the moon, Mars, Mercury, Jupiter, Venus and Saturn are when their elongation (from the true sun) are 12o, 17o, 13o, 11o, 9o and 15o.

Note 1. I generally adopt TS's readings. But ¶É榃 is extra, and evidently a mistake which has crept into TS's reading. To make up for this they have removed °ünÚù, which is necessary, and this emendation has spoiled the correct agreement with other Siddhantas.

Note2. These are time-degrees, i.e. time expression degrees (Kalabhaga)and are arbitrary in essence, and depend on the keenness of the observer's eyesight, as also the atmopheric conditions. The Later [Surya Siddhanta] gives 10o and 8o for Venus at superior and inferior conjunctions, and 14o and 12o for Mercury, respectively, while the [Surya Siddhanta] here and some other give the mean of each. (The Mahabhaskariya gives even 4o or 4 1/2o for Venus at inferior conj. and 8o at superior conj.)

¨ÉxnùOɽþÉxiÉ®úVªÉÉ

º´Éɹ]õÉƶɪÉÖiÉÉ%EæòVÉҴɶÉÖHòÉhÉɨÉÂ*

ºÉ訪ÉÉx{É (xªÉ) ·É {ÉnùÉäxÉÉÆ (xÉÉ)

Ê´ÉIÉä{ÉÉäx{É (xªÉ) ·É ¶ÉÒQÉÊ´ÉvÉÉè **13**

MÉÖ°ü¦ÉÚiÉxɪÉÉ%%º¡ÖòÊVÉiÉÉÆ

{ÉÉnùÉäxÉÉ YɪɨɨɪÉÉä¨ÉÖ¶ÉÉÆ (YɪɨɪÉÉä¶iÉÖ ¶ÉÉ)¹]õÉƶÉÉ&*

ÊjÉVªÉÉ´PÉÒ EòhÉÉÇ{iÉÉ

Ê´É (Ê´É) ªÉÉäMɪÉÉä¶É ºÉ (ªÉÉäMɶÉ)Ê´ÉIÉä{É&**14**

13. Add one eighth of itself to the R (120') sine of (mean planet-apogee,) in the case of Saturn, Jupiter and Venus. For the two others, (i.e., Mercury and Mars) subtract one fourth of itself. (This is one part of latitude ). There is another part of latitude using the Anomaly of conjunction.

14. From the R sine anomaly of conjunction of Jupiter, Mars and Venus subtract one fourth of itself. From that of the rest, (viz., Mercury and Saturn), add an eighth. Add both algebraically and note the direction, north or south. Multiply this by R (i.e.120') and divide by the hypotenuse got in the last step. The latitude is got, its direction being that of the noted direction.

Note 1. This is a peculiar primitive way of finding the latitude of the star-planets. It is not found in the allied [Khandakhadyaka] and the quoted part of the [Bhattotpala]-quoted [Pauli'sa]. It is found in [Aryabhata's] [Ardharatrika-paksa] given in the [Mahabhaskariya] (VII.28-33). But there are some differences between the two, and we cannot decide which follows the original [Saura] here, and which has slightly modified the original. They both mention two kinds of latitudes for each star-planet which are to be added algebraically. But there is a difference in the maximum latitudes and int eh ascending nodes to be subtracted from the mean longitudes or [sighras]. VM's [Saura] implies the max. latitude 90',90', 135', 135', and 135', for Mars, Mercury, Jupiter, Venus and Saturn, respectively, to be multiplied by sine anomaly of conjunction, and 90', 135', 90', 90', and 135' to be multiplied by sine anomaly of conjunction, no separate node being given, which means that the apogee itself is the node for the one kind of latitude , and the mean planet itself for the other. But the [Ardharatrika] gives only one set of maximum latitudes for both, viz., 90', 120',60',120',120o. It gives the nodes, 20o, 40o, 70o,260o,and 150o for the former and 20o, nil, 70o,260o, and 150o, for the latter. Govindasvami's [Bhasya], and the [Mahabhaskariya] being meagre, does not help us.

Note 2. By implication, we had better take the arguments of the eq.cent. used in teh thrid step for the former, and the anomaly of conj. used and hypotenuse obtained in the fopurth step for the latter.

Example:Find the latitude of Mars at 120,583 days from epoch.

In the thrid step of the earlier example, the sine of the argument of eq. conj. is 42'24". As it is Mars, deducting a quarter of itself, the latitude is 31' 48", north, as this argumentis between is between 0o and 180o. In the fourth step, the sine of the argument of conj is 88'20". For Mars one fourth is to be subtracted. So, the latitude due to this is 66'15", again north, since the argument is from0o to 180o. Adding, 31'48"+66'15"=98'3", north.

The hypotenuse obtained there, in the fourth step, is 88'52".

98'3"X120./.88'52"'=132" north, is the true latitude of Mars for the day.

(As it is, this is far from the latitude obtained from using the later Siddhantas.)

Note 3. In XVIII. 57-60 this latitude is used to correct the mean elongation given in verse 12, for the heliacal setting and rising. For this reason, we shall deal with those two verses here, to complete the [Saura]

VªÉÉÊ´ÉÊvÉÊ´ÉIÉä{É´PÉÉ-

SÉ(\´É)®úEòɱÉÉnù¨¤É®úɹÉÞ´ÉänùÉƶɨÉÂ*

VÉZÉÉ [iÉÂ]ÊIÉ´Éä\´É ªÉɨªÉä-

kÉ®Æú (®äú) OɽþÉxiÉ º´ÉÆ ªÉlÉÉEòIÉÆ (EòɹÉÞ¨ÉÂ)**XVIII.57**

B¤ÉÆ EÞòiÉä OɽþÉxiÉ-

®úÉƶÉEèò®úºiÉnù¶ÉÇxÉÆ iÉä¹ÉɨÉÂ** 58a **

57. Find the R since of the latitude of the body. Multiply the (maximum) half-cara (i.e. the half difference between the half day-time or night-time from 15 [naids] in [vinadis], by this R sine, and divide by 480. Add this or subtract this to or from the moon or star-planet if the latitude is north or south, according to the proper direction, i.e. according as the phenomenon (of setting or rising), takes place in the west or east respectively.

58a. When this is done, their setting or rising happens according to the interval in degrees between the sun. and the planet (given in XVII.12, or the moon).

Note 1. There is a lot of lacunae in verse 57. The [half-cara-vinadis] meant is the maximum for the place. ªÉlÉÉEòIÉ is meaningless here and corrected into ªÉlÉÉEòɹÉÞ¨ÉÂ, i.e. according to the direction, but the direction is not mentioned. If north latitude, the addition is for the moon or planet in the west. Also, if north latitude, the subtraction from the moon or star planet is to be done for the east. If south latitude, the subtraction is for the west, and the addition for the east. These things can be got by a little reflection.

Note 2. The amount of degrees to be applied can simply be got by multiplying the degrees of latitude of the body by the tangent of the latitude of the place. (The equinoctial mid-day shadow of the 12" gnomon ./.12, is tan. latitude of a place.) The amount got is very rough.

The degrees wanted =Sin. half-cara (i.e., tan latitude of place x tan. declination of the sun)x sin (90o-angle for heliacal rising), nearly. The last term is neglected here, tan. decliantion is roughly taken as 48', half-cara is converted into degrees by division by 10, and the conversion into sine-function is applied to the latitude of the planet instead of the half-cara, as roughly equal.

Note 3. This application is what is technically called [Aksa-drkkarma]. The [Ayana-drkkarma] is neglected.

Note 4. The heliacal rising of the moon, and of Venus and Mercury when retrograde, takes place in the west. The heliacal setting of the moon and retrograde Venus and retrograde Mercury takes place in teh east. Otherwise, all star-planets set in the west and rise in the east. (This Siddhanta does not envisage the setting or rising of Venus and Mercury when retrograde, no separate degree for that being given.)

SÉxuùnùÒxÉÉÆ uùÉnù¶É

¨ÉxÉÖ®úÊ´ÉÊiÉiªÉ¹]õÊiÉÊlɺÉÆYÉè.**58b**

58b. The setting and rising (mentioned in 58a) is by 12o, 14o, 12o, 15o,8o, and 15o, for the moon etc.

Note1. This has to be taken with verse 58a. The degrees given here separately are according to the [Vasistha Pauli'sa], which do not instruct the correction due to the latitude of the planet or for even the latitude of place (aksa-valana). The result will therefore be very rough.

Note 2. These degrees are necessary, as mentioned already, and the correctness of the numbers cannot be verified in the absence of the original [siddhantas] which are now lost. But we can guess the probable values as we are sure of the relative luminosities of the planets. The numbers seem to have been misplaced. They should be [dvada'sa], [tithi],[manu], [ravi], [asta],[tithi] (12o, 15o, 14o,12o,8o,15o for moon etc.). All [siddhantas] give 17o for Maars instead of 15o. The rest are nearly correctly given, according to one [siddhanta] or other.

ËjÉ (ÊjÉ) ¶ÉiÉÊ´ÉxÉÉb÷ÒMÉÖÊhÉiÉè-

vÉ(°ü) nù¶É (ªÉ)[Ê´É]xÉÉb÷Ò|ɨÉÉhɽþiÉè&*

±É¤vÉÉRÂóEò|ɨÉÉhÉÉ-

nÖùnùªÉÉä%ºiÉÆ ´ÉÉ ¶¡Öò]Æõ ´ÉÉSªÉ¨ÉÂ**59**

59. Multiply the degree by 300 and divide by the [vinadis] of oblique ascensional difference of the sign rising at that moment, (near sunset or sunrise, as the case may be), and get the respective degrees. When the distance between the sun and the planet is that much, the respective setting or rising takes place.

Note 1. This work is what is known as the conversion of time-degrees (Kalabhaga) into degrees of distance on the ecliptic (Ksetrabhaga). Since the tule has to apply commpnly to [Saura] on the one hand and the [Vasistha-Pauli'sa] on the other, it has been placed last.

Note 2. Since the [positions of the sun, moon and planets are given only on the ecliptic, this conversion is necessary to measure distances.

YÉʺÉiÉÉ%%VªÉÉEçò(CªÉÚÇ)xÉÉ&

¶ÉÉʶÉxÉ& |ÉiªÉÖkÉ®Æú JÉ®úÉƶÉÉxÉÉ (¶ÉÖ·É)*

YÉÉi´Éè´ÉÆ Ê´ÉIÉä{ÉÉ-

nùÉnäù¶É¨ÉxÉÉMÉiÉè EÖòªÉÉÇiÉÂ**60**

60. (The rising takes place in the east when) Mercury, Venus, Mars, Jupiter and Saturn areless in longitude than the sun, and the sun is less than the moon in the opposite direction, (i.e., west). Making the computation according to the instruction given above using the latitude etc, the phenomenon should be predicted.

Note. The verse is very corrupt. But knowing what it is about we can give the meanin, making possible corrections. The rising is mentioned here as it is more important for application to [dharmasastra] etc. But rising also envisages settings, with the word 'less' taken for 'more', and 'more' for 'less'.

Example 1. The latitude of the moon is 3o 45'N. The maximum half-cara of the place is 150[vinadis]. The oblique-ascensional difference of the rising sign is 280 [vinadis]. near sunset. Find the ecliptic distance between the sun and moon for the heliacal rising of the moon.

The time-degrees for the moon i s12o. The heliacal rising of the moon takes place in the evening. R sin 3o 45'x150./.480=7 38/60X150./.480=2o 23'.

This is additive since the moon's latitude is north, and the phenomena pertains to the west. Therefore, the corrected time-degrees =14o.

14oX300./.280=15o is the distance on the ecliptic between the sun and the moon, required.

Example 2. For the same place, (i.e., max. half-cara 150 vinadis) find the ecliptic distance required for heliacal rising, given: the latitude of Mars 1o 15'N. and the oblique ascensional difference near sunrise at taht time is 330 [vinadis].

The latitude correction to teh time degree (17o for Mars)=R sin 1o 15'X150./.480=2'33"X150./.480 taken as degrees=48'.

As the latitude is north, and the phenomenon pertains to the east (since it is the rising of Mars that is considered), it is subtractive.

17o-48'=16o 12' is the corrected time-degrees. 16o 12'X300./.330=14o44', is the distance on the ecliptic required.

+É´ÉxiªÉEò& ºÉ¨ÉɺÉÉ-

ÊUÅô(ÎSUô) ¹ªÉʽþiÉÉlÉÇ iɨÉnÂùMÉ (?)º¡Öò]õÉRÂóEòºÉ¨É¨É*

SÉGäò ´É®úɽþʨÉʽþ®ú&

iÉÉ®úÉOɽþEòÉÊ®úEòÉiÉxjɨÉÂ**61**

|ÉtÖ©É(¨xÉ)¦ÉÚʨÉiÉxɪÉä

VÉÒ´Éè (´Éä) ¶ÉÉè®úlÉ´ÉÉ´ÉÒ (?ºÉÉè®äú%lÉ) VɪÉxÉÎxnùEÞòiÉä*

´ÉÖvÉä´É xÉOÉÉ (SÉ ¦ÉOÉÉä)[iºÉɽþ&]

º¡Öò]õʨÉnÆù Eò®úhÉÆ ¦ÉVÉiÉÉÆ (iÉÉiÉÂ) **62>>

où¹ÉÞÆ ´É®úɽþʨÉʽþ®äúhÉ ºÉÖJÉ |É´ÉÉävÉÆ

61. For the good of hi disciples, [Varahamihira], born in the Avanti country (Ujjain region), wrote this section dealing with the star-planets, briefly and with the constants such as to agree with the originals.

62. A learner, discouraged by the computation of Mars by the astronomer Pradyumna, the computation of Jupiter according to the [Saurasiddhanta], and the compuation of Mercury by Vijayanandi, can have recourse to this section of the manual.

63. By [Varahamihira] has been seen, (i.e., written) (this Karana) easy to understand,...........

Note 1. Verses 61 and 62 clearly close the section dealing with the star-planets. Since VM says that he has improved on the earlier authors, he must be referring to chapters XVI and XVII, dealing with the [Saura]. His reference to his improvement on the [saura] itself in the case of Jupiter must refer to the [bija] correction made by him in XVI. Indeed, his dissatifaction with teh Jupiter of the [Saura] is relected in his formula for computing Jupiter to give the years of the sixty-year Jovian cycle, given in his [Brhatsamhita], in the cha[ter dealing with the motion of [Brhaspati] (Jupiter). As for cahp. XVIII he could not have meant the [Vasistha-Pauli'sa] star-planets there as an improvement , theybeing crude.

Note 2. Verse 63 evidently closes the [Pancasiddhantika], as indicated by the [Vasantatilaka] metre instead of the regular [arya] metre. But unfortunately the last three feet are missing. Perhaps it is a purposely done 'black-out' by a later astronomer-scribe, to append his spurious verses 64-81, and unfortunately only his manuscript has survived as the archetype of the few extant manuscripts.

[PANCASIDDHANTIKA XVIII 64-81: AN INTERPOLATION]

That verses 64-81 of [pancasiddhantika] (PS) form only an appendage to amanuscript of the work is evident from its occuring after the work has closed in the customary way, with a concluding colophonic verse, with its metre chaged to [Vasantatilaka] from the [Arya] metre in which all the previous verses of the cahpter had been couched and the author speaking about himself in the said concluding verse. It is also to be noted that this set of verses begin with a new salutation. Had these verses really belonged to the PS, the customary finish must come at its end, but there are no such finishing verses at the end. Further, in verse 65 it is said that VM considers this as a superior set containg a previous method or matter and that he was giving it, with a liberal mind, to the generality of astronomers without hiding it from them. But actually it is inferior stuff, and can give only very rough results since the equation of the centre is dispensed with, only the equation of conjunction being given, which makes it valueless. VM is alleged to boast here that he has made things easy, and takes credit for thsi which only a novice could have done. Fancy VM speaking thus, when in verse 62 he is so intent on accuracy that he says, "Let people whi have been dissatisfied with the inaccuracy of astronomers like [Pradyumna], [Vijayanandi] etc. have recourse to his treatment of the [Saura]". Further, there are mistakes in the computation of Venus and Mercury, unpardonable in any astronmer.

There spurious verses are dealt with below for the sake of completeness of the text as given int eh editions of the work.

|ɺiÉÉ´Éä%Ê{É xÉ nùÉä¹ÉÉ (¹ÉÉxÉÂ)

VÉÉxÉzÉÉ (zÉ) Ê{É xÉ (delete)´ÉÊHò ªÉ& {É®úÉäIɺªÉ*

|ÉlɪÉÊiÉ MÉÖhÉÉ (hÉÉÆ) ·É iɺ¨Éè

ºÉÖVÉxÉÉªÉ xÉ¨É (¨É&){É®úʽþiÉɪÉ**64**

+¹ÉÞnù¹É¶ÉʦɤÉÇuùÉ-

xlÉÉiÉÉ®úÉOɽþiÉxjɨÉäiÉnùɪÉÉÇʦÉ&.

´É®úɽþʨɽþ (´É®úɨÉÊiÉ)´É®úɽþʨÉʽþ®úÉä

nùnùÊiÉ ÊxɨÉÇiºÉ®ú& Eò®úhɨÉÂ**65**

63. Salutation to the good people, ever interested in the welfare of others, who even when knowing the faults of othes, and even when there is an opportunity, do not mention their faults, but proclaim their good qualities.

64. In eighteen [arya] verses, [Varahamihira], without feeling any jealousy, gives this manual to the world, ending with teh treatement of the planets, thinking that it is good.

Note 1. The emendation are TS's also.

Note 2. Verse 64 is a paranomasia and means also, "Salutation to the good sceince of astronomy called techincally [Parahita-ganita] (prevalent in Kerala in South India), which at the beginning deals only with mean motions, though knowing its defective nature as not being true motions, and which furnishes tabular values of equations going by the names, [mandajya] (R sine table of the equation of the centre), [Karkijya] (R sine table of the equation of conjunction for the anomaly 90o to 270o) and [makarajya] (R sine table of the equation of conjunction for the anomaly 270o to 90o).

Since this meaning being not obvious to laymen, I give the phrase by phrase meaning:

|ɺiÉÉ´Éä OÉxlÉÉ®ú¨¨Éä, {É®úÉäIɺªÉ +º¡Öò]õOɽþºªÉ, ]õÉä¹ÉÉxÉ Bº¡Öò]õi´ÉÉÊnù-

´ÉÉä¹ÉÉxÉÂ, VÉÉxÉzÉÊ{É ªÉUô xÉ ´ÉÊHò VÉÉxÉzÉÊ{É xÉ ´ÉnùÊiÉ, (¨ÉvªÉMÉiÉä ®äú´É |ÉEÞòiÉi´ÉÉiÉÂ) MÉÖhÉÉxÉ ¨ÉxnùVªÉÉ-EòÉÇVªÉÉ-¨ÉEò®úVªÉÉ <iªÉÉÊnùMÉÖhɶɤnù´ÉÉSªÉÉ VªÉÉ& |ÉlɪÉÊiÉ |ÉEò]õÒEò®úÉäÊiÉ MɨÉʪÉi´ÉÉ Ê±ÉJÉiÉÒÊiÉ ªÉÉ´ÉiÉÂ, iɺ¨Éè ºÉÖVÉxÉÉªÉ iɺ¨Éè ¶ÉÉä¦ÉxÉVÉx¨ÉxÉä {É®úʽþiÉÉªÉ {É®úʽþiÉMÉÊhÉiÉÉªÉ xɨÉ& xɨÉ%ºiÉÖ**

Note 3. These two verses also form part of the 18 verses mentioned. So, actually there are only 16 verses (66-81), giving the computation.

+ÉEò®úhÉÉnÚù ®úʴɦÉÉMÉÉ

Ênù´Éº±ÉÉ·É(·ÉÉ) ®úÉƶÉEòÉ ®ú´Éè EòɪÉÉÇ&*

+ÉÊvÉEòÉ ªÉÇ (ªÉ) nùÉ ÊnùiÉä (xÉä) ¦ÉªÉ&

¦ÉÉMÉÉ YÉäªÉɺiÉnùÉ SÉGòÉiÉÂ**

66. From the epoch to the time of computation of the planet, find the sun's degrees passed. These are to be techincally called 'days' (and used int he computation). Find the remainder after dividing by the cycle number given for the respective star-planets. Take the 'days' of motion corresponding to the set of motions given to the respective planet. These are degrees of planetary motion. Add this to the sun's longitude. The true planet is got.

Note 1. The 'days' mentioned here is only what is called [sauradina] (sun'sday) as distinguised from the [savana] or civil day, and are actually degrees. (This is like the word light-year, which is used as a unit of distance). This instruction is given with respect to all planets.

Note 2. The cycle given for each planet is only the period of the planet's synodic revolution converted into solar days, i.e., it is the synodic periodX360o ./.365-15-30, nearly. When so converted, we have:

Mercury Venus Mars Jup. Saturn

The regular

synodic days 115-52-45 583-55 779-57 398-53 378-6

Converted into

solar days 114 6/29 575 1/2 768-45 393 1/7 372 2/3

The second set is given for the respective planet, saying that the synodic period is so many 'days'. Not knowing this TS have remarked that they do not understand why there is so much difference in the periods from the regular days generally known. Also, they say they cannot dismiss them as wrong, since the numbers given are checked in teh computation itself. (See pages Ix-iii-Ixiv of Introduction in TS's edition).

Pingree and Neugebaur have understood (as seen from their edition of the (PS) that the cycles are in solae 'days'. But they have remarked that 'VM' has confused the days and degrees, not realising that there is a purposr in giving the cycles in the solar day units. These units have been used because, now, the 'days' and the 'degrees' will have the same meaning, and they can be combined without, at every point, instrcuting it . Thus, ultimately, the combined value is the degrees of the true planet.

Example. Days from epoch 1,20,553. Find the 'days', and assuming the sun at epoch as zero, find the sun at the end of 1,20,553 days.

1,20,553X360./.365-15-30=1,18,187.5'days'.

This plus zero, and divided out by 360o=17o.5, sun's longitude.

xɴɨɪÉMÉÞhÉÉxÉÖǽþÒxÉÉ (MÉÖhÉiÉÖǽþÒxÉä)

EÞòiÉɽþiÉä (½þiÉä) ʴɹɪɺɹiÉJÉÉÊOɽþiÉä*

¦ÉÞ(¦ÉÚ)ªÉÉä ½þiÉÉä (½þiÉä) SÉiÉÖ̦É&

Ê´É®ÆúºÉ (ÊxÉ®Æú¶É)Ênù´ÉºÉÉ ¨É½þÒVɺªÉ * 67 **

¹É]ÚõËjÉiºÉ´Éè (ʴɹɪÉè) κiɹɪÉè) κiÉlªÉÖ (lªÉÚ) xÉ&

où¹]õ´ÉEÖòvÉÞÊiÉ[ʦÉ]®Æú¶ÉEòÉSUô(:¹É) ι]õ&*

+¹]õ¶ÉiÉäxÉ ´É (SÉ) ¹Éι]õ&

ºÉ{iÉiªÉÉ nÚùlÉ (jªÉ) ÊvÉEòªÉÉ xÉ´ÉÊiÉ&**68**

¹É¹]õlÉÉä (¹]õlÉÉ) ]õªÉÖHòªÉÉ ºÉÆ (¶É)iÉ-

nù±ÉÆ SÉ JÉÉʤÉvÉ (Ê·É)ÊuùEèò& JÉ®úÉÊuù&(´PÉÉ&)*

+ºiÉʨÉiÉÉä%iÉ& ºÉ{iÉɹ]õEäòxÉ

ÊiÉlɪÉÉä ÊxÉ®Æú¶ÉÉMɶÉÊxÉ& (MÉÊiÉ&)**69**

67. Subtract 6329 from the 'days'. Muliply the remainder by 4 and divide out by 3075. Take the remainder and divide by 4. These are the 'days' after conjunction (i.e., the anomaly of conj.) for Mars.

68-69. After 56 'daus' he goes behind the sun by 15o, and becomes observable, (i.e., the heliacal setting ends). In 188, 108, 73, 68, 220, 'days' Mars lags behind by 60o, 60o, 90o,50o,70o, respectively. Then it sets heliacally, and in 56'days' lags 15o behind and goes into conjucntion

NOte 1. I generally agree with TS's emendations. But in verse 68, I give ¹ÉÎbÂ÷´É¹ÉªÉè& for ¹ÉÎ]ÂõjɶÉiºÉ´Éè&, which latter is both meaningless and has one [matra] extra. TS's ¹ÉbÂ÷´ÉMÉê& does not agree or at least nearly agree. ºÉ{iÉiªÉÉ nÚùlÉÊvÉEòªÉÉ has been emended by me into º{É´ÉiªÉÉ jªÉÊvÉEòªÉÉ, and JÉÉΤvÉ into JÉÉÎJ´É.These