Connections between Vedanga Jyotisha and other Vedic Texts

Indian Journal of His/ol)) of Science. 44.3 (2009) 357-368

CONNECTIONS BETWEEN THE VEDANGAJYOTISA AND OTHER VEDIC LITERATURE.

R.N. IVENGAR*

(Received 14 October, 2008; revised 25 Ferbruary, 2009)

The VedaligaJyoti~'arepresents a sophisticated calendar system,stated cryptically in terms of symbols and formulas. Notwithstanding thefact that it is approximate, it represents a development closely linked withthe previous Vedic literature. The connections between Vedanga Jyoti~'aand the Vedic Saq1hitaare investigated in this paper. It is argued that theyear length of the system was based on the Vedic long count 3339, firstappearing in the ~gveda, but interpreted by the BrahmalJ1a PuraIJa asthe number of deities drinking Soma on daily basis in the dark fortnights.

Key words: Bralll11al}a,Brahma1J1a PuralJa, Calendar, Five year

Yuga, Sa'11hita, Vedic Astronomy

INTRODUCTION

The VedcuigaJyoti!ja (VJ) has been studied and discussed by severalauthors over the last hundred years and more. Ini.tiallythe effort was to edit,amend and to interpret the text. After Kuppanna Sastry (1984) brought out acritical edition of VJ,interest in the text has been to understand the basis of theVedic calendar through the help of VJ.It is known that)he Vedic people had aluni-solarcalendarwheretheyearwassolarbutthemonthswerelunar.Intercalationwas practicedto bring the solar and lunaryear into harmonyby variousmethods.The central theme of VJ is to provide an algorithm to find in advance the tithi,parvan, nalqatra in the formalizedVedicfiveyearcycliccalendar.Thus its focusis not observational but essentiallycomputational.This does not mean that thebasic elementsofthe calendarwere borrowedfrom an out of Indiasource. It alsodoes not suggest that the users did not observe the sky to correct the dates asand when necessary.However,the observationalelementis less obvious than the

* Raja Ramanna Fellow, Centre for Advanced Research and Development, Jain Group ofInstitutions, Bangalore, Jakkasandra, Bangalorc - 562112; e-mail: [email protected]

358 INDIAN JOURNAL OF HISTORY OF SCIENCE

computational bias that is quite conspicuous all through the text. Fonnulaic algOlithmscombined with the mnemonic devices presented in VJ indicate a level ofsophisticationpossibleonly afterlong observation.This raises the question,whatcould havebeen the observationalfeatureshiddenbehind the developmentof VJ.To findtIllSmaterial,willchwouldhavebeensupersededby thetraditionmentionedin VJoriginatingaround 1400BC, we have to go by the intemal evidencesin VJand possible dependenceof its basic elements on anterior literaturenamely thefigveda (RV), the Yajurveda (YV) and the ancillary texts. It is demonstrated in

. the present paper that there are strongreasons to link the basic parameters of VJwith the characteristic number 3339 of the RV which has been previouslydemonstrated to be the 18-yeareclipse period number (Iyengar 2005).

BACKGROUND

VJhas come down to us in two branches,namelyarca-jyoti~aof RV andyajwia-jyoti~aof yv. The basic elementsare common to both and hence the twoare here consideredtogetheras a singletraditionof ancientVedicastronomy.Thebasis of VJ is the five yearyuga period equated to 62 synodic lunar months of1830 days, taken equal to 67 sidereal months. There are 1768moon rises and1835risings of the eclipticstar Sravi~!ha(Dhani~!ha,Vasava-j3-Delphini), withwhich sun and moon came togetherat the winter solsticec 1400BC. The lengthof a solar year according to VJ is 371 tithi or 366 days.Any three independentelementsamongthe aboveparametersleadto the completeluni-solarcalendarofVJ.There are severalpublicationsdiscussingthe strengthand weakness of VJasa calendar.The glaringinaccuracyis with the lengthof the solaryearwhich is toolong. Hence if the fonnulae are used blindly,the resultswould perceptiblymissrealitywithina few years.Howeverit hasbeenpointedout in the past (Abhyankar2002) corrections were done in the fom1of intercalarymonths and droppingbftithi to keep the calendarin tunewith the sun.Therehavebeen effortsto interpretVJ, claiming that the 19-yearMetonic cycle is implied by the ~gvedic VJ text,willch,in turnwouldhave lead to a nearperfectsynchrOlllzationbetweenthe lunarand solarmovements (Holay 1989).But this does not explainwhy VJ went in fora five year cycle with excess length of the solar year. It may be noted.here thatthe five year cyclic calendar with 366 days per year was adopted by the non-VedicJain astronomicaltext popularlyknownas Si1ryaprajliapti-Candraprajnapti(1989).There are no direct evidencesin the VJtext to link its parametersto longterm observations and/or eclipses. This leads one to look for the basis of VJ inthe SaIl)hita,the Bralm1al).aand the Sl1traliterature.This would be natural since

CONNECTIONSBETWEENViANDOTHERVEDICLITERATURE 359

VJ.was developed as a guide for timing the rituals appropriately.It may not beerroneous to presume that VJ, as available now, was formally enunciated onlyafter the philosophy,principlesand canonsbehind the Vedicritualswere fixed.Itis not an observationaltextbut is skillfullyalgorithmicin its contents.Theproblemof tracingthe backgroundis made difficultdue to absenceof explicitastronomicalstatements in Vedicliterature,as we understandthem today.In addition to this isthedistinctpossibilityofvariousVedicschoolsfollowingdifferentmodesof countingdays/nightswhich showsup alreadyin VJadvocatingmonths ending/startingonNew Moon (amiinta).It is well knownthat months ending/staliingon Full Moon(purlJimiinta)was in vogue and continues to be followed to this day in someparts of India. Notwithstandingthis bifurcationof allolder practice, the conceptof tithi or,what may be looselycalled, lunar day is of prime importanceall overIndia. This amply indicates that once upon a time the month and the year werebased on the moon and not on the sun. Several scholars in the past (Roy 1976)have noted this point citing support from the R V statement samiiniim miisaiikrtilJ,that is, moon is the maker of the years. The transition of the ancientlunar calendarwould have takena fairlylong time to evolveto its finalluni-solarform that VJ represents. Both R V and YV contain sufficient evidences to showthat strictlysolarpositionssuch as equinoxesandsolsticeswereimportantin someof the rituals even as moon's positionwas importalltforother observances.Thusthe cosmicviewof theVedasis neitherexclusivelylunarnor solar,but is admittedlyluni-solar.However,observationof the sky would havebeen possibleonly in thenights givingprominenceto moon and its position,in the backgroundof the fixedstars, as the marker oftime. Hence it would be reasonable to start looking forlunar numbers and how they might have got cOlmectedwith later numbers suchas the lengthofthe solal'year.This,however,is subjectto an importantlinlitation.The available Vedic literature at best reflects astronomical truths in an indirectfashionhavingtrallsformedtheminto ritualsto be observedby the faithful.Hencethe effort here is to seek the source of a number or a rule from the context in

which it arises. This is the typical inverse problem wherein searchingthe causefrom the effect may not always lead to a unique conclusion.

THEVEDICYEAR

Vedic people had recognized several types of years such as the niilqatra,the lunar, the siivana, the solar and the intercalary year. The Nidiina-siJtra (Y.11-12) belonging tg the school of the Samaveda states this as,


fa.ttrbpiono navonasca fac}hono'tha savano 'fflidasabhirjyiiyanahoMi/!siivaniitparo niik~atramiti miisasca tasya caiva trayodasacii n dram iis ass iivan as ca u bh iiva t hii:f!a d asyu tt ana af fa-s apt a tr i '?Isate

pourqol11iisyiil11praslidhayet IThe year that is less (than the siivanayear) by 36, the year that is lessby 9, that which is less by 6, then the siivana year, then the year greaterby 18days.The siderealyear (lessby 9) has thirteenmonths(of 27 each).Thetwokindsof yearsare the lunarand the silvana.Theyeargreaterthan18 days has to be observedon 37-38 Full Moon.

This was understood by taking a Savana year of 360 ti/hi as the reference.The first one deficient by 36 was the nak~atra year of twelve sidereal monthsmaking 324 ti/hi. Then the one less by 9 was a lunar year consisting of 13 monthsof27 tithi each. The savana lunar year of360 iithi was made up of 12 synodicmonths. The solar year of 366 days and intercalary years oflonger duration werealso recognized. This has been discussed in the past by Shamasastry (1938) andothers. For our purpose the points to note are that the month was always reckonedwith the help of moon's position and the VJ solar year with 371 tithi was ~approximate effort at making the savana year match with the position of the suq.The rich variety of years clearly indicates an effort at synchronization of two orthree different observable celestial rhythms. Since it is the moon that was observedwe surmise, the ancients would have first noted the synchronization between thesidereal and the synodic months. It is conjectured that the near equivalence of 12synodic months with 13 sidereal months, counted in terms of sunrises, would havelead to the concept of year as a longer measure of time than the month. Thisharmonizes with the earliest ~gvedic word denoting year as sarna, used in thesense of being same, coincident or equivalent. With the recognition of seasons asdependent on the sun, synchronization of three rhythms seems to have becomeimportant. Nidfu1a-sUtrarefers to this equivalence more accurately in the form ofa verse already well known to the Vedic community it was addressing.

yasmin vai parivatsare sallryo 11Iiiso'tha ciindramaso I

nlik~'atro na vilupyate kasvittam veda kasvit II

a~!lisaptatri1!lSatetasmin samvatsare mite I

sallryo mliso'tha clindramaso nalqatro na vilupyate II

Who knows that year in which the solar, the lunar, and the siderealmonths are not lost, who knows that? In the year measured by 37

CONNECfIONSBETWEENVJANDOTHER ~ICLITERATURE 361

or 38 (full-moons),the solar,lunarand the siderealmonths are notlost. (Shamasastry)

This points to the approximationof37 synodicmonths with 40 siderealmonths, even though the latter number is not mentioned. Similarly,for the solarcycle also to have matched, 37 synodicmonths shouldhave been taken equal tothree (solar) years. The number word a~{il-sapta-tri,!rfategives the meaning ofbeing between 37 and 38, not of 37 or 38, as in the above translation ofShamasastry (1938). Taking 30 tithi per month, one gets 1110 ti/hi in three(solar) years giving 370 tithi per year,which is nearly the value used in VJ.Theabove also hints at the presence of a three year cycle that should have existedbefore the improvedfiveyearcycleof VJcame intovogue.This leadsone to inferthat the basic VJ parameters have had their roots in more ancient beliefs andobservations. The text ofthe Nidilna-si1tracited above is available on the web

(at http://is1.mum.edulvedicreserve/kalpa/shrauta/nidana_shrauta_sutra.pdf).However, it is to be noted that on page 72 of this web edition the words athasavana~ appear wrongly as atha sadhanam.

There is evidencein the Vedicliteratureto note that not all Vedicgroupssynchronized their arumalcalendar with the seasons. In the darsa-paurI}amilsa(DP) rites, which are eminentlysuited for time keeping,some driftedthroughtheseasonsfor thirtyyears.Darsa is linkedwith the visiblenew moon andhence notconcernedwiththepreciseconjunctionofthe moonandthesun.Similarly,pourlJimilneed not have been the precise oppositionofthe two celestialbodies. However,when a ritual was observed for such a long period, surelytherewould have beena few lunar eclipseson Full Moon days.Ritualsordainedduringeclipsesare firstnoticeable in the Atharva VedaPariSi~!a,a very late text that also names planetsand comets. hl the available core Vedicliterature there are no direct references

to occurrenceof eclipsesduringa ritual.But, a closereadingof the hymns showsup several interesting statementspointing to a relationbetween eclipses and theritualistic numbers. 'For example, the Nidilna-si1tramentions a special siderealyear that falls short by nine (navona) in relation to the savana year of360 tithi.This year had 13months of 27 ti/hi making the lengthof the year to be 351 tithi.This corresponds to a year of346-347 (solar) days. What was being achievedby this, unless this had some hiddenconnectionwith the eclipseyear? In modemparlance, eclipse year is the time taken for the lunar nodes to be in line with the


sun and the moon, when an eclipse is possible. The well known eclipse periodof223 lunations is equal to 18.03solar years or 6585.32 days. This consists of19eclipse years of346.6 days.The unknown elementhere is the ancientway ofmeasuring tithi. We can however be reasonably certain that it was associatedwith the moon but the earliestVedicway of fixingthe tithi is not yet deciphered.Nevertheless it was known to be less than the mean solar day with its valuevaryingin the interval0.984-0.986,statedto be equal to (61/62).Desire to avoidfractions in the remote period of Vedic astronomy would have given place toapproximations in tenns of integers with an error of one unit. Thus, the eclipseyear length would have been approximated to 351 ti/hi, while its actual lengthwas nearer to 351.5 tithi. Nineteen such years lead to 6669-6678.5 tithi as theequivalentof223 lunations.This makes a case for postulatingthe knowledge ofa Vedic long count as the basis of the VJ calendar.

VEDIC LONG COUNT

The lJ.gveda presents many small and large numbers, their physical meaning

hiddenbehindan almostforgottenesotericworldview that can be labeledancientHindu scientificnaturalism.A case in point is the ~gvedic number 3339which ishalf of 6678 cited above. This number appears in two places in the R V (III.9.9;x'52.6) and once in the RV-khila.TraditionalVedicinterpretershave taken thisto be a mystic count of a ;groupof deities called Visvedeva~,with no knownphysicalbasis.RV statesthis figureto be the munberof deitieswho worshipAgni,hiding in unknown places. The present author (Iyengar2005), with the help ofevidences internal to R V and the Brahmaf}cJa Pura/.1G,has shown this number to

be associated with the recurrence of similar lunar eclipses. The counting methodgiven in the Purfu}.a,in line with the naturalistic mysticism of the Vedas, may bebriefly described here.

Cipfirayan surum1]ena bhCigam bhfigamahal] kramCit Isu~ul111JCiapyayamanasya litkla vardhc.nti vai kalal] II

(Brah. Pu. Part 1,23.61)

The bright parts (of Moon) increase in the llfkla pakfa, with Sun fillingup the phases by his surum/Jiiray, in daily sequence.

bhakrartham alI/rtam somal] pour1]amfisyam upasate I

ekam ratrim surail] sarvail] pitrbhil] sar~'ibhil]saha II

CONNECTIONSBETWEENVJANDOTHERVEDICLITERATURE 363

soma5ya kr~'1a-pak~adau bhaskarabhimukhasya tu I

prak~lyantepitr-devai~ plyamanal} ka/a-kramat

trayasca trimsatascaiva trayal} -tri/!ziat tathaiva ca I

trayasca tri-sahasrCiScadeval} somam pibanti vai II

ityetai~ plyam[masyakpp}avardhantivai kalal} I~ayanti tasmat suk/CiScakr~}a apyayayantica II

(Brah.Pu.part I; 23.66-69)

Moonis approachedby all gods,alongwithmanesand sagesfor a nighton Full Moon, for partakingam[fa. From the beginningof the dark fortnight,phases of Moon facingSun, decreasebeing drunkby pitr-devatasdigitby digit. Three and three hundred,then thirty-threeand againthree and-three thousand(3339) gods drink soma.Beingdrunkthis way,the darkdigits increasewith conespo.ndingdecreasein the bright digits.

This is a clear enunciation of the model behind the 3339 Vedic gods andwhat their cosmic role was. Their count was sequential, in the order of thedecreasing phases of Moon (kalii-kramiit) adding to 3339. It follows thesymbolism of gods drinking away the digits of Moon, which obviously refers tothe dark fortnight and their total number being 3339 has its origin in the ~gveda.For this characteristic number the above Pural).icmodel has to be accepted as theproper explanation, being a typical example of ancient naturalism. The countstarted on a Full Moon to proceed till amiiviisya and stopped till the next FullMoon, to repeat again in the same fashion with gaps in the bright fortnights. Inother words, this number is clearly the number of tithi in the dark fortnightscounted as 3339 sequentially for a special purpose. Ifboth the fortnights were tobe considered, this number would add to 6678 tithi. Now, at the rate of thirtytithi per lunation, this number is equal to 222.6 which in round figures is theeclipse period of223 synodic months. Hence, the RV number 3339, which is halfof 6678, is a proxy for the so called Saros period of 18 (solar) years, mistakenlyattributed to the Chaldeans. RV Siikta X.55 is inspired by a total lunar eclipsewith the moon turning red. Thus, 3339 is the Vedic long count linked with similarlunar eclipses occuning at the same nakIjatraposition in the sky.The characteristicnumber mentioned here is neither the number of lunation nor the total length oftithi, but the number of tithi counted only in the dark fortnights. The countingmethod embeds an uncanny and original symbolic imagery of cosmic gods drinkingsoma that unambiguously refers to the waning moon.


VJ PARAMETERS

With the above long count of 3339 tithi, we can understand how thebasic VJparameters might have been arrived at. The eclipseperiod would havebeen taken equal to 18nominal solar years.This was a consequenceof the olderconcept of 37-38/synodicmonths being equal to three solar years, consisting of1110tithi alreadydiscussedabove.If the solaryearwereto be takenequalto370 tithi, one would directly get 18.04years as the eclipseperiod. On the otherhand to get a round figure; 18-yearwas taken as a gauge number leadingto 371tithi per year, which is an important VJ parameter. Since we know that thecorrect solar day count would be 6585.32, dividing this by 18 gives the lengthof the nominal solar year to be 365.851 days rounded off to 366 by VJ.lfonetakes 223 synodic months as equal to 18 years, the first four convergent of thefraction 223/18 are 12/1,25/2,37/3,62/5. The last one namely (62/5) is the VJapproximation.This perhapswas an improvementover a previousapproximationof(37/3). Similarly,since 223 synodicmonths are equal to 241 siderealmonths,we can approximate the fraction (241/223) as 13/12,27/25,40/37,67/62. VJuses the last approximationof 67 siderealmonthsas equalto 62 synodicmonths,which is better than the previous one of 40/37 corresponding to three yearsmentioned in the Nid[ma Sutra.

Y AJURVEDICTEXTS

There are several instances of numbers adding to 17, 18 or 19 as speciallength of years embedded in the Yajurveda texts. -In the Viijasaneya Saf[lhitii(XVIII.24-28) the number sequences 1 to 33 of odd integers and 4 to 48 of evenintegers increasing in steps of four, are given followed by a list of symbolic animalswith their ages. The ages mentioned are 1Y2,2, 2Y2, 3, 4 and 6 adding to 19years. This is followed by offerings to seasons and months showing the contextto be part of Time worship. The same Sa1phitii at (XXI 12-17) repeats yearnumbers adding to 19 associating them, respectively with meters giiyatri, u~1Jik,anu~{up, brhati, pmikti, and tri~{up. Similar statements occur in the Kiil}VaSamhitii (30.24-28), and in the Taittirfya Sm/milii (TS: 4.7.10), where theanimal-ages add to either 17Y2or 18 years. The difference arises depending onthe interpretation of the word pa~!avaham, taken to be 412 or 5 or 6.

The interesting fact here is the number of syllables in each of the abovenamed meters increase by four and the total adds to 204 corresponding nearly

CONNECTIONSBETWEENVJANDOTHERVEDICLITERATURE 365

to the othertotalnamely 17years.Evenin tpe ritualisticcontextthe hymnappearsto embed some type of equivalencebetween the animal-ages and the meters. Ifthe length of the year is taken as 360 tithi, we have 17x 360 = 204 x 30. Onthe other hand if it is taken as 354 (solar) days,we get the length ofthe synodicmonth to be (17x354)/204 = 29.5 days, which is the value adopted by VJ. Asalreadynoted the nak§atrayear of324 nights/dayswith 27 units per month wasalso in vogue in ancient times. It is observed that 17 x 324 = 204 x 27. Suchinterestingpropertiesofthe number 17basedon observationof Moon couldhaveleadto the earlyadoptionof thisas Prajapati'snumberin the Vedas.The immediatenext hymn of the TaittirfyaSalphita(4.7.11-12)supports this inference. This isthe famous sequenceof seventeenodd integers 1 to 33, increasing in steps of2,adding to 289, equal to square of 17. This is followed by a sequence of evenintegers 4 to 48 increasing in steps of 4 adding to .312.The implied time-wiseequivalenceof two numbers in the previoushymnmakesus wonderwhether thenumber pair (289, 312) also has some useful astronomical property. Quitesurprisingly289 synodicmonths are very nearly equal to 312 sidereal months.

Vedicliterature,ITomthe verybeginning,is preoccupiedwith the 27 lunarmansions (nak~atra)along the ecliptic. Most probably in the R Vhymns,-manydeities are neither imaginarynor mysticalbut simplyrefer to their correspondingnak§atrain the form of inspiredpoetry,whichhave been later used for liturgicalpurposes. Quite categorically the Taittirfya BrahmalJa (TB) declares that nak~atraare the abodes ofthe deities (devagrha vai nak§atra1}iTB 1.5.2.6). The ritualisticcorrespondence is conspicuous in the TS and the TB wherein special rites aredescribed invoking the 27 nak§atrasand their presiding deities. It is generallyaccepted that the Satapatha BrahmalJa(SB) reference to star Krttika (Pleiades)not slipping from the east is an observation of the vernal equinox in the thirdmillennium BCE. This position ofKrttika at the head of the nahjatra list hasremained unchangedall through the Vedictexts covering a time span of severalcenturies.Neverthelessan evolvingtimeline is recognizablein the TB text. Thisbook contains two sets of long hymns devoted to the nahjatra rites. Both startwith KrttikaQ as the first stares).The first set of hymns refers to Sravi~tha asapproaching from the south (dahji1}ato 'bhiyantu sravi~!hlilJ II TB III.1.2.7).Indicationof a directionfor the star group Sravi~thashould have been a specificsky observation introduced into the ritual. The second set of hymns is morespecific about its calendar leanings.It invokes Sravi~thaas arriving at thefrontof the equals (that is years)and offersan oblationto this star naming it Forefront


(agram ha vai samiiniiniim paryeti agriiya sviihii II TB III. 1.5.8).This hymnappears to be in hannony with the astronomy of VJ when Sravi~tha or Dhani~tha

was the star heralding the yearwith the winter solstice.

Sengupta(1947)pointsout severalinterestinginformationavailableaboutsolstices in the Vedic literature. The Kau~ftaki Briihamw]a (XIX.3) mentionsthat the northward motion of the sun starts on the New-moon of the Maghamonth, which is same as the one mentioned in VJ;There are indications in theMaUri Upani~at(Chapter 6) about times when the northern course of the sunstarted at the middle ofthe Dhani~thadivision.This would have happeneda fewcenturies before the VJ period. Even if the above texts are not exactly dateable,there are several evidencesto infer that VJtradition representsconscious effortson the part of its originatorsto provide a computationaltool, incorporatingpastobservational knowledge and beliefs ofthe Vedic community, for conductingprescribedrituals.

DISCUSSION

The roots ofthe VediingaJyoti~a are traced to other Vedic texts to theextent evidences are available in the form of numbers that can be related to VJ

parameters.It turns out that the lengthof the yearof371 tithi or 366 (solar)daysof VJ is closelylinked to the Vediclong count of3339 tithi first appearingin theRV.This number is the count of tithi in the dark fortnightsmaking up nearly 18(solar) years, which is the so called Saros period of (1unar)eclipses of the sametype.Abhyankar(2007)pointsout how our ancientscouldhave used this 18yearlength for adjustingtheir VJcalendar.Sometimesopinionsare expressedthat VJowes its inspirationto foreignsources(pingree 1978).Internalevidence,however,from the Vedic texts indicates that VJ is an indigenous concept of the Vedicpeople based on observing the sky over long periods.of time. While no archivalevidence as in the case of Babylonia exists, the special numbers inbuilt into therituals almost always carry some celestial imagery showing that such numbersshould have had a countableobservationalsubstratum.

Even though there is a case for the VJparameters to have come out ofthe observed 18or 19 eclipseyears, lunar eclipsesare not mentioned in VJ. Thissituationlookssurprising.However,there is mentionof moon's lateralmovementacross the ecliptic, denoted as ayana similar to the seasons associated with the

CONNECfIONSBETWEENVJANDOTIIER VEDICLITERATURE 367

north-southmovementof the sun.This has been ignoredin the past (Sastry 1984)as being of no astronomicalsignificance.But, a littlereflectionwill show that thisindeed is the evidence for VJto have been based on observations of moon and

not on any borrowed material. VJ mentions that there are 134 ayana or north-south movementsof moon in 67 nak~atramonths. Even though this has nothingto do with the ~tu or seasonsassociatedwith the Sun, the lateral lunarmovementis real to an observer on earth. Moon stationed with a known nak~atra sayMagha will come back to the same star after nearly 28 tithi, but not with thesame phase. Thus starting with MaghapourlJimiione sidereal month later, thenahjatra will be Magha but the tithi will not be POtl11}.ima.During the course ofthis month, every night moon can be observed to occupy different nahjatrapositionin a sinuousfashion.This happensall throughthe (solar)yearwith nearly27 ayana for moon. VJ recognizes the similarity between sun and moon in thesense of what happens to sun in one (sidereal) y~ar happens to moon in one(sidereal) month. Further as the year evolves, moon continuously wanders oneither side of the 27 nahjatra band closelyrepresentingthe ecliptic. During thisserpentinemovementwhenevera FullMoon occurson the ecliptic,a lunareclipseis possible. We have already seen how the symbolism of the 3339 VisvedevaQ(Cosmic-deities)is connectedwith lunar eclipses.This imagery is reinforcedbythe statement devii vai sarpii/}in TB (II.2.6.2) which denotes these deities asserpents.

SUMMARY AND CONCLUSION

The connection of the VediiJigaJyoti~a with other Vedic texts is thesubject matter of this study.It is pointed out that the assumptions and the basicparanleters used in VJwere alreadyknown to the Vedicpeople in some form orother for use in the rituals. The approximation of 62 synodic and 67 siderealmonths in five solar yearsand the lengthof the solar.yearbeing taken as 366 daysare the outcome of the Vediclong count of3339 tithi in the dark fortnights.Thisappears first in the ~gveda representingthe 18-yeareclipseperiod, the so calledSaros. Medieval Hindu astronomers knew this period but their source for thisknowledge has remained obscure. Comparative study of R V hymns with laterPural}.iclegendsassociatedwithAgni, Riihu and Ketu will ilirow fhrther light onIndian astronomy prior to the Siddhanticperiod.


REFERENCES

Abhyankar, K.D. (1993), "A Search for the Earliest VedicCalendar, lJHS, 28.1, pp.I-14.

Abhyankar, K.D. (2002), Pre-siddhiintic Indian Astronomy, ISERVE, Hyderabad.

Holay, P.v., (1989) VedicAstronomy, B.Apte Smarak Samiti. Nagpur.

Iyengar, R.N., (2005) "Eclipse period number 3339 in the Rgveda ", lJHS,40.2, 139-152.

Pingree, D (1978) "Indian Astronomy", Proc. Am. Phil. Soc., 122,6, pp.361-364..

Roy, S.B. (1976) Prehistoric LUl/arAstronomy. Inst. of Chronology; N.Delhi

Sastry, T.S.K (1984) (Ed.) VediingaJyouti~a ofLagadha, IJHS, 19.4. Supplement, ppl-74.

Sengupta, P.C. (1947) Ancient Indian Chronology, Univ. of Calcutta, Calcutta.

Shamasastry, R.(1936) Vedangajyoutisha,Text, Translation and Commentary. Mysore.

Shamasastry, R. (1938) Drapsa. the Vedic Cycle of Eclipses. Mysore.

Siirya-prajiiapti-Candraprajnapti(1989) (Ed.) Kanhaiyalal,JinagamaGranthamala,29 Beawar.

Documents

Connections between Vedanga Jyotisha and other Vedic Texts