Skip to main content

Full text of "Epic Chronology"

See other formats


STOP 



Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 
purposes. 

Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- 
journal-content . 



JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 
contact support@jstor.org. 



Epic Chronology. — By E. "Washbuen Hopkins, Professor in 
Yale University, JTew Haven, Conn. 

This essay is the third installment 1 in the series announced 
in this Journal, vol. xxiii, p. 109. It was presented to the 
Society at its annual Easter meeting in 1902, but in the Journal 
of that year other publications took so much space that the 
editors thought it best to postpone this publication till the next 
year. In the interval I have received two works on the chron- 
ology of India touching directly on epic data. They are of 
very different character. The first is the Chronology of 
Ancient India, by Mr. Velandai Gopala Aiyer, B.A., in which 
are discussed the beginning of the Kali Yuga and the date of 
the Mahabharata war. This is a very ingenious attempt to 
establish the date of the war as beginning Oct. 14, 1194 B.C., 
though "the epic was cast into its present form more than a 
thousand years after the date of the war" (p. 98). The date 
1194 B.C. is reached by a series of eleven converging argu- 
ments, based on (1) the Vedanga Jyotisa, which points to the 
beginning of the Kali Yuga as approximately 1173 B.C.; (2) 
a statement of Garga, which points to the beginning of the 
Yuga as occurring a few years before 1165 B.C.; (3) classical 
historians, whose figures point to 1177-6 as the beginning of 
the Yuga; (4) The Malabar era, which indicates for the same 
event 1176 B.C.; (5) details of the epic which, if the Yuga 
began at the winter solstice preceding 1176 B.C., would indi- 
cate 1194 B.C. as the date of the war; (6) the Rajataraiigini 
tradition, which indicates the dates of the war to be about 
1190 B.C. ; (7) a statement of Aryabhata to the effect that the 
Rsis were in Magha in Kali 1910, i. e. 1192 B.C.; (8) the 
average duration of Hindu reigns, which also would indicate 
about 1193 B.C. as being the date of the war; (9) Garga's 
stanza cited in the Brhat Samhita, which leads (? cf . IA. viii, p. 



1 Compare also the syntactical paper evolved from the same series and 
published separately, AJP. vol. xxiv, p. If.; and the note on the same 
subject at the end of this paper. 



8 E. W. Hopkins, [1903. 

66) ' to the same conclusion; (10) the first year of the Brhaspati 
cycle of sixty years, corresponding to the date as given by 
Garga, i. e. 1194-3 B.C.; (11) a stanza of the epic, fixing the 
day of the winter solstice occurring soon after the war, which, 
in connection with the elements of the Jyotisa, would indicate 
that the war took place in the latter part of 1194 B.C. The 
exact day is then deduced from other epic verses. 

A glance at this array of arguments shows that they fall into 
two divisions, in one of which is sought the date of the Kali 
Yuga, and in the other the date of the epic as based on the date 
of the Yuga. Important as is the general contention, for the 
purpose of the present essay only the latter division comes into 
account, and in this division only the arguments numbered 5 
and 11 above. These points will be briefly considered in 
their proper place in the course of this paper, but I have thus 
outlined Mr. Aiger's contention in advance, that their bearing 
might be understood. In regard to the whole theory I can see 
no objection to ' the conclusion that tradition points to the 
twelfth century as the date of the Bharata war; but it is pos- 
sible that the details of the epic should be considered as based 
on tradition rather than as furnishing it, and that this tradition 
referred originally to a great Bharata war rather than to the 
special Pandu war with which the epic really has to do. That 
the heroes of the present epic lived in the twelfth century B.C. 
seems to be historically impossible, if for no other reason at 
least for this, that the Pandus as such are unknown till long 
afterwards. 

Of a very different sort is the symbolic interpretation of epic 
epochs and eras deduced from a general theory of Hittite and 
Akkadian supremacy in pre-historic times by J. P. Hewitt in 
his History and Chronology of the Myth-making Age, which is 
the second work referred to above.- A few examples will suflice 
to show the character of the ' ' chronology " evolved out of a 
symbolic interpretation of the epic: A year of eleven months 
and another of seventeen months, divided into seven-day 
weeks, are discovered to be latent in the fact that the Kurus 
have eleven and the Pandus seven aksauhinis (armies), inter- 

1 Compare the argument as reported in the Secretary's correspondence 
in the Proceedings for April 1903, at the end of the second half of this 
volume of the Journal. 



Vol. xxiv.] Epic Chronology. 9 

preted as " monthly revolutions of the axle." The eldest Kuru 
■was Duryodhana, who brayed like an ass at his birth, thus show- 
ing him to be the son of the divine (epoch-making) three-legged 
ass. Duryodhana's car was drawn by mules, "thus showing 
him to belong to the race born from the union of the sun-horse 
and ass." The thirteen-month year was brought to India dur- 
ing the rule of Kansa. Kansa is the same as Hansa, the goose- 
god of the Ugro- Altaic Finns (Ugro is Sk. ugra) ; so Su-bhadra 
means the Su-bird, Su is Akkadian-Egyptian Khu (mother-bird). 
The epic shows all the changes from the pole-star epoch to the 
solar epoch of reckoning. The year of seventeen months 
ended and the eighteen-month year began at the epic sacrifice 
of the (sun-)horse, 10,200 B.C., and the eighteen books of 
the epic symbolize the eighteen-month year (of twenty-day 
months), which was the outcome of the Pandus' victory. This 
was the year which was taken from India to Mexico in the 
Bronze Age. The epic is an allegorical history of India from 
the Neolithic to the close of the Bronze Age and represents the 
period of the years of eleven, fifteen, thirteen, and seventeen 
months each. Parthas (sons of Prtha) are Parthians. Despite 
the date of the horse-sacrifice at which he is present, Yudhi- 
sthira himself was born in May, 12,200 B.C. Here, as the 
learned author sorrowfully admits, "there is a difficulty" 
about the exact date ! But that Karna is the " horned lunar- 
solar god of the three-year cycle," and that Gandharl (from 
gan, 'land,' and dhari, 'wetter') is the goddess Dharti, the 
star Vega, in the constellation of the Vulture, now Lyra, 
which was the pole-star from 10,000 to 8,000 B.C., admits not 
even of an interrogation point. 1 The reader will readily see 
why a modest study like mine can dispense with any discussion 
of such conclusions as these, interesting as they are. I turn 
now to a study of epic chronology based not on fancy but on 
facts. 

NEGATIVE TIME; INDEFINITE PERIODS. 

God, as Great Time, Mahakala, a late-epic epithet of Siva, 
and as All-time, is also Not-time, akalas ca Hikalas ca duskalah 
kala eva ca, xii. 285. 143 (after akalah kelikalah kalih ; cf. 

1 Hewitt, op. cit, pp. 309 f., 327, 374, 426, 529, 561 f., 580, 587. 



10 E. W. Hopkins, [1903. 

Mait. Up. vi. 15), or, otherwise, the destruction of time in the 
reabsorption of the universe, pratydhara; though elsewhere 
(loc. cit., Great Epic, p. 182) Time is the destroying Lord. As 
with space, the word antara, interval, when in negative form, 
expresses negative time, "there was no interval," etc. The 
word itself is combined with "winking," in a colloquial form. 
Thus, nimesdntaramatrena, "in the measure of a wink's inter- 
val" (space of a wink), vii. 98. 37, etc. The wink is the 
twinkling of an eye (expressly), caksurnimesamdtrena^. xii. 
321. 11 ; ydvad aksinimesdni, xiii. 1Q0. 41 ; aksnor nimesamd- 
trena, vii. 51. 17; and, as with us, it may be cut in half, 
though the latter phrase is rare, nimesdrdhdt, " in half a wink," 
viii. 25. 13; madhydhne vdi nimesardham {tisthasi tvarh divd- 
kard), "at noon (O sun, thou standest still) half a wink," 
xiii. 96. 6. 1 More common than "half a wink" is muhurtakam, 
which in colloquial language as diminutive of muhurta (nvuhuh 
= mox) has no reference to hour but means a little time, tusnlm 
asm muhurtakam, R. vii. 13. 15, Gorr., but not in Bomb. ed. ; 
Mbh. i. 133. 2, tistha tavan muhurtakam, "stop just a moment." 

In xiv. 48. 2 and 3, a moment is expressed first by a breath- 
ing, uochvdsamdtram, and then by a wink, nimesamdtram, both 
being followed by apt, as marking the shortest time (at death, 
antakale); in v. 79. 20, by lavaiah ksanasas ed 'pi. 

The indefinite non-technical nature of these terms shows 
itself in the exchange of nimesa with unmesamdtrena, for 
example in xii. 313. 6; and in the phrase muhurtam iva, 
"momentarily," compared with muhurtam sahyatdm, "a short 
time," ib. 319. 9. The muhurta in iii. 297. 7 appears with 
veld, period of time, as well as ksana, another indefinite word 
for moment. The eighth muhurta, noon, kutapa, is called 
abhijit and is mentioned by this name in R. Gorr. vi. 112. 70 ; 
but the corresponding passage in Mbh., iii. 291. 66, has only 
the day and asterism. 3 



1 Compare Vas. xi. 36 : divasasya '$tame bhage mandibhavati bhdska- 
rafy, sa kalal), kutapo nama (see the next note). So (epic) vii. 99. 1, 
where the sun "goes slow as it turns in its course (at midday) to the 
west." 

2 In xiii. 64. 37 abhijita (yoga) is mentioned as the twentieth lunar 
asterism ; the same word occurring in i. 123. 6 in the other sense of 
abhijit, the eighth (noon) hour: aindre candrasamayukte muhurte 



"Vol. xxiv.] Epic Chronology. 11 

The flight of an arrow also measures a short indefinite period 
of time (as the stick-cast measures space). Thus in xii. 296. 32: 
isuprapdtamdtram hi sparsayoge ratih smrta, " sensual pleas- 
ure is said to be (short-lived as) the measure of an arrow's flight " ; 
ib. 321. 11, laghvastragatigdmini, "going the pace of a light 
arrow", i. e. in a moment; ib. 328. 30, yathd banarh gunacyu- 
tam (dydntam), (swift) "as a cord-sped arrow." 

But as the indefinite sense of nimesa is lost in the formal time- 
table, so with other small divisions. In v. 109. 4, as typical 
divisions are named the truti and lava; the former being joined 
with kald, portion, ksana, glance, and nimesa, wink, as "hairs 
of Time," in xii. 322. 25. Of these, ksana, 1 etymologically mean- 
ing a "look" or "glance," is, like nimesa, a moment, and so a 
moment of leisure (ksanin, "at leisure," ii. 13. 45), whence 
comes a name for the giver of leisure, Night, ksanadd, a late 
word, found in viii. 1. 8. Characteristic of the later didactic 
epic is the fact that it uses the ending rdtra as an independent 
word, triny rdtrdny upositva tena pdpdd vimucyate, in the jar- 
gon of this period, xiii. 136. 11. 

Time-periods casually mentioned or enumerated in various 
passages of Santi, xii. 137. 21; 227. 97 (repeating, as a section, 
224) and'also xii. 166. 14, do not present the ordered progression 
of the time-table, but juxtapose kasthd, kald, muhurta, diva, 
rdtri, lava (before month, half month, season, aeon, year); or 
ahordtra, month, ksana, kasthd, lava, kald (all ace, followed 
by sampidayati yah kdlo vrddhim vdrdhusiko yatha, "Time 
adds up days etc. as a usurer adds up his increase," 227. 97) ; 
or, in the order of creation, years, seasons, months, half months, 
lavas and ksanas. In ii. 11. 37 (also late), diva is nom., as 
above. 

The lava is a bit (saktuprasthalava, xiv. 90. 115), or minute 
"cut" of time, corresponding loosely to our minute in ordinary 
speech. God is praised as all time in i. 25. 14, 

'bhijite 'sfame, diva madhyagate surye tithau purree 'tipujite. Here 
dindra is the asterism Jyesfha and madhyagate surye is " at midday." 
Compare xii. 326.38 : madhyamgatam iva 'dityam, " like the midday 
sun " (metrically altered). 

1 So too the compound of this word, abhxksnam, every moment, too 
much, i. 78. 9 ; 100. 60 ; xii. 86. 29 (bhrsam va). Compare anvaksam in 
Yaj. iii. 21, for the adverbial use in Jcsanena, " in a glance" (moment). 



12 E. W. Hopkins, [1903. 

tvam muhurtas tithis tvam ca tvam lavas tvam punah ksanah 
iuklas tvam bahulas tvam ca Jcald kastha trutis tathd, 

where bahula is a poetical equivalent of the dark half of the 
month, and truti is a fractional bit of time. The muhurta, in 
ordinary language a moment, in the formal time-table of the 
epic is an hour of forty-eight minutes. " They say that in the 
evening, purvardtra, the twilight hour, muhurta, except for 
eighty lavas, is devoted to demons, the remainder, sesam any at, 1 
to men," i. 170. 8-9. Another passage states that "after mid- 
night " is the time when demons roam about : rdtrdu niilthe tv 
abhlle gate l rdhasamaye, nrpa, pracdre purusdddnam raksasdm 
ghorakarmandm, iii. 11. 4. The former passage is to be com- 
pared with i. 154. 22, which says that the whole twilight, 
samdhya, is rdudra muhurta. Besides the rdudra, after sun- 
set, muhurte ramyaddrune, iii. 1. 45 (both " fair and horrible "), 
the noon hour, abhijit (kutapa), above, and the brdhma mu- 
hurta, the hour before sunrise, are mentioned, xiii. 104, 16 (= 
aparardtresu, "at the end of the night," ii. 5. 29). 

Little can be learned of the relative length of these periods as 
mentioned generally in the epic. They appear to be designations 
of short times as indefinite as twinkling and moment. Nor does 
the order in which they are mentioned in other places help in 
this matter, for sometimes one and sometimes another precedes. 
In xiii. 14. 185, the order is day, half day, muhurta, ksana, 
lava; and ib. 395, naksatrdni, grahdh, 1 mdsdrdhamdsa rtavo 
rdtrih samvatsardh ksandh, muhurtas' ca nimesds ca tathdi , va 
yugaparyaydh. The "year, season, half month, day and night, 
ahordtra, kald, kastha, mdtra, muhurta, lava, ksana," make 
the list of xiii. 159. 32, which brings in the mdtra, mora. This 
is found also in xiii. 17. 141 f., where the list is season, year, 
month, half month, paksa (Siva as "number-effecting," sam- 
khyasamdpanah, is explained by N. as effecting samkrdnti and 
the new and full moon days), kald, kastha, lava, mdtra, mu- 
hurta, day, night, ksana. 

Besides being an astronomical period or course (of the sun), 
as in xii. 51. 15, the kastha (copied from Katha Upanisad, iii. 3, 



1 Compare (tad)ahahsesam, xiii. 19. 101 ; 20. 9. 

2 Compare xii. 285. 128, where also meghakala is mentioned (saihvarta- 
kabalahakah) and the yugavarta, 144-152 (see below). 



Vol. xxiv.] Epic Chronology. 13 

sa kastha sa para gatih) is found in a non-technical sense in xiii. 
16. 57, iyam sa parama kastha iyam sa parama kala . . . iyam sa 
parama gatih. The Upanisads otherwise, it may be remarked, 
have the list, day, night, month, year, kala muhurtdh kdsthds 
ca, but not till Mahanar. i. 8. 

DEFINITE DIVISIONS. 

Nevertheless, the pseudo-epic has its regulated time-table, xii. 
232. 12 f. It is in a lone tristubh stanza, which has been intro- 
duced into the poem at a period later than Manu, whose general 
scheme is followed, but with this important difference, that the 
epic agrees in detail with the later Puranic view rather than 
with Manu; excelling the latter also in exactitude. 1 It is as 
follows (sc. ganayet, "one may reckon"): 

15 nimesas make one kastha 

30 kdsthds " " kala 

30 T V kalds " " muhurta 

30 muhurtas " " day and night 

30 days and nights " " month 

12 months " " year (of two semesters, ayane) . 

Mann's account, i. 64, differs from this in ascribing to the 
kastha eighteen nimesas, instead of the epic and Puranic (VP. 
i. 3. 7) fifteen ; nor does the law-book add to the thirty kalds 
that make a muhurta the epic's one tenth: trinsatkalas ca 'pi 
bhaven muhurto bhdgah kaldyd dasamas ca yah sydt. 

According to this table, the nimesa is about one-fifth of a 
second; the kdsthd, about three seconds; the kald, about a 
minute and a half; and the muhurta, just forty-eight minutes. 
On kala as a fraction, see this Journal, xxiii, p. 135. For a 
Brahmana calculation of time-divisions, see SB. xii. 3. 2. 1-5. 2 

1 But it differs from that later division (not recognized at all in the 
epic) by which the Hindu hours and minutes are exactly inverted as 
compared with ours, that is (instead of a day of twenty-four hours of 
sixty minutes each) a day of sixty hours, nadls, ghafikds, of twenty- 
four minutes each (ahoratrah sasfighatikdbhih, N. to ii. 11. 38). 

5 This Brahmana recognizes half-months of fifteen days, twelve and 
thirteen months; three, five, six, and seven seasons, i. 3. 5. 8 f . ; ii. 2. 3.26 f . , 
etc. The table (referred to above) in the twelfth book (cf . x. 4. 3. 8) has the 
muhurta as above, one thirtieth of a day ; and the nimesa ; but this is cal- 



14 E. W. Hopkins, [1908. 

But, before proceeding with the greater divisions of time 
added to the table, it will be necessary to take up in more detail 
the last three divisions of the scheme already given. 

DAY AND NIGHT. 

Tacitus says of the Germans' view, nox ducere diem videtur. 
The Polynesians and New Zealanders to-day always count by 
nights, and the Babylonians originally made the whole day 
begin with the evening. According to the Vedic views repre- 
sented by the Brahmana period, AB. viii. 15. 2; SB. ii. 4. 2. 3; 
x. 6. 4. 1, and by MS. i. 15. 12, it is not quite certain that 
night was the norm of time. It is true that night generally 
precedes when days and nights are mentioned together, 1 but on 
the other hand, in contrast to space, dydvah, time is reckoned 
as "days" in RV. iii. 32. 9, "nor days, nor months, nor years" 
(harvests). Yet since we find also "nights and years," iv. 16. 
19, and this view prevails, it may, perhaps, be regarded as the 
more primitive Aryan norm of short times. It is the Avestan 
method of measuring, and Dr. Boiling has lately shown that in 
Homer also the day is reckoned from sunset to sunset.* How 
long such a method may continue under favoring circumstances 
was well known a few years ago to the boys of New England, 
whose weekly holiday ceased sharply as the Sabbath began, at 
sundown on Saturday! 

In the great Hindu epic, an inheritance of stereotyped formu- 
las somewhat affects precedence }n the phraseology of the poets, 
who use ahoratra, divaratra, but also ratryahanl. Generally 
speaking, night is the favorite word in compounds such as tri- 
ratra, saptaratra, dasaratra ; but the alternate forms are used 
as well, ekaha, saptadina, etc., and the same passage may give 
precedence to both words, as, for example, in xii. 124. 16, eka- 
ratrena, tryahena, saptaratrena, prthivlm pratipedire. "Sev- 
eral days" is sapta ''py ahani, i. 92. 15, and niianisam and 

culated otherwise, as a subdivision of ksipras, etarhis, idanis, and 
breathings (=nimesas), arranged in multiples of fifteen. Here the year 
has three, five, six, or seven (Vedic) seasons ; twelve or thirteen months ; 
three hundred and sixty days ; ten thousand eight hundred muhurtas ; 
etc. On the sixty-hour division in the Rig Veda, see Zimmer, AIL., p. 
363, and Ludwig's note to RV. 1. 123. 8. 

1 Cf. Zimmer, op. cit., p. 360 f. 2 AJP. xxiii, p. 428 f. 



Vol. xxiv.] Epic Chronology. 15 

divdnUam are used indifferently. So in other phrases, surya- 
gnina ratridivendhanena, ' ' with the sun as fire, night and day 
as kindling-wood," xii. 322. 92; ratrdv ahani samdhydsu, 
"night, day, twilights," xiii. 115. 28; prabhate ca sdyarh ca, 
"morn and eve," iii. 305. 10; sdyam pratas ca, "eve and 
morn," iii. 200. 83; xiii. 78. 9 ; ahnikam cat '«a ndisam ca duh- 
kham, "daily and nightly sorrow," v. 110. 14; avasans tatra 
saptardtram, astame i hani samprdpte, ' ' they stayed there a 
sennight, on the arrival of the eighth day," iii. 158. 22 f. ; astd- 
daSdha, trayovinsatirdtra, astdvinsatiratra, xv. 10. 30 ; xii. 
46. 14; iv. 36. 3, respectively. 

But it is formally stated in xiv. 44. 2, that, as the bright fort- 
night of the moon precedes the dark fortnight, so day precedes 
night : 

ahah purvam tato rdtrir mdsdh SuMddayah smrtdh 

Sravanadlni rksdni rtavah SUirddayah, 

' ' the day comes first, then the night ; the months begin with the 
bright fortnight ; the asterisms begin with Sravana ; the seasons 
begin with (the cool time) Sislra." Moreover, Svah, to-morrow, 
always in practice refers to a day that begins in the morning, 
not at sunset, and "half the day" is measured from the sunrise. 
On the other hand, it is formally stated in xiv. 44. 18, that days 
end at sundown, ahdny astamaydntdni, and night ends at sun- 
rise, udaydntd ca Sarvarl (as "joy ends in sorrow and sorrow 
ends in joy"); but here the "day" is the bright part of the 
whole time. Light ends in darkness as darkness ends in light, 
is the whole meaning ; probably without thought of defining the 
(following) day as beginning with evening. 

Besides the restriction of the border-period, veld, of the twi- 
lights, purvd and pascimd, 1 day and night are divided into fore, 

1 But veld is a general hour ; sarvdsu velasu, "at all hours," iii. 305. 3 
(pascimd veld, iii. 65. 5). 'At dawn' is iarvarydm (or another word for 
night) prabhdtdyam, vyustitdydm, vyatltaydm; or prabhate, prage, 
vimale, umsi, u§asya- (u?asyodhd bhdrateyas ca, xiii. 76. 18). On svo- 
bhute, etc., see the last paper in this series, Journal, xxiii, p. 351. In the 
formal definition of the BS. xlvii. 21, the samdhya is from half-sunset 
(when the sun is half under) till the stars become visible (not yet bright), 
and from their fading till half -sunrise : ardhdstamaydt samdhya, vyale- 
ttbhutd na tdrakd ydvat : tejahparihdnimukhdd, bhdnor ardhodayam 
yavai. See below on the correlation of this division of the day with the 
corresponding centary-samdhyd in the scheme of ages. 



16 JS. W. Hopkins, [1903. 

mid, and after parts, purvdhna, madhydhna (niadhyamdina), 
apardhna; purvardtra, etc., the only measured period being 
the samdhyds, twilights (of one hour and twelve minutes, as 
reckoned later). Compare viii. 91. 51; xii. 207. 29; and xii. 
224. 53 f . : 

rtun nidsdrdhamdsans ca divasani ca ksandns tathd 
purvdhnam apardhnam ca madhydhnam api ca ''pare 
muhurtam api cdi '«» , hur ekam santam anekadhd 
tarn kdlam itijdnlhi yasya sarvam idam vaSe, 

where Time and Fate, bhavitavyam, set at naught the doctrine of 
sin working out in a new birth (ib. 32, parallels, Great Epic, p. 
103). In iii. 65. 6 ; xii. 304. 3, etc., ardhardtra is midnight (also 
nUitha) instead of madhyardtra, as used in the epic, xv. 5. 34, 
where, after prdtar, pradosa, apardtra have been mentioned, 
the king is told to have his vihdra at midnight and midday, 
madhyardtre, madhydhne. Manu, vii. 151 (in the same con- 
nection) uses ardhardtra. 1 Evening has more names than any 
other division of the day, of which sdyarn, sdydhna, is most 
current, as in the phrase, common to both epics, yatra sdyam- 
grha, one whose house is where evening finds him, i. 13. 12, 
etc. ; R. ii. 67. 23. The beginning of evening (when fire-flies 
are out, vii. 15. 18) is currently pradosa (nisamickha) ; "late in 
the afternoon " is mahaty apardhne, i. 190. 47. Compare AB. 
ii. 15. 8; Manu iv. Vl%,mahani&i, and xii. 322. 73: svahkdryam 
adya kurvlta ("do to-day to-morrow's duty") purvdhne cd 
''parahnikam (" do in the morning the work of the afternoon"). 
Compare &B. ii. 1. 3. 9, "put not off till to-morrow; for who 
knows man's morrow ?" 

The three watches of the night are alluded to in a stanza 
which speaks of one night of three watches, triydmd rajani, as 
being so fearful as to seem like a thousand watches, sahasrayd- 
mapratimd, vii. 184. 14. Valmiki expresses the same idea, R. 
ii. 62. 17 (G. 63. 17), and his imitator, G. ii. 10. 17, triydmd 
rdtrih ... varsasatopamd (omitted in the Bomb. ed. 13. 15) ; 
and it is found again in more modern form in the Mbh. viii. 
1. 8, 



1 But a Yogin meditates in the fore-part, purvaratre, and sleeps in the 
middle of the night, madhyardtre, rising within an hour after this, xii. 
326. 43. Compare xii. 229. 39. 



Vol. xxiv.J Epic Chronology. 17 

duhkhena ksanadd rdjan jagdmd 'bdaiatopamd. ' 

The last watch of the night includes the muhurta of Brah- 
man (or Prajapatya, as in Vas. xii. 47), alluded to above. 

The formal rule for the king is that he shall sleep two 
watches and rise in the third ydma, ii. 5. 85. 2 In xii. 53. 1, 
"he woke half a watch before day" is expressed by 

ydmamdtrdrdhaaesdydm ydmlnydm pratyabudhyata, 

where ydma gives the name to ydmini, night. A correspond- 
ing division of the day is apparently alluded to in xiv. 39. 18, 

ahas tridhd tu vijneyam trldhd rdtrir vidhlyate, 

though years and conjunctions of seasons also, varsdni, sam- 
dhayah, are here made three-fold, by virtue of the all-pervad- 
ing gunas. The natural three-fold division of the day, morn, 
or sunrise, suryadariana, noon, hutapa, and evening, of many 
names, is implied in the conventional use of a ritualistic for- 
mula; so that we find not only trivelam, " thrice daily," but 
also saptatrisavanam snatvd, xiii. 136. 18, "bathing three 
times a day for a week." Sunrise gives a number of expres- 
sions indicating that the time for man to be up is already passed 
when the sun is up : utsuryasdyin, abhyuditasdyin (like prage- 
6aya, prageniia), used of lazy people, opposed to early risers, 
kalyam utthdya, Jcalyotthana, prdtar utthdya, etc., as in xiii. 
130. 9; 146. 48; xv. 11. 11. The word sun is not necessary. 
Thus, "at or before sunrise " is simply ndaye or udite i nudite vd 
'pi, xii. 60. 49. An hour after sunrise is muhurtodita dditye, 
i. 126. 12. There is probably no sharp distinction between 
the periods loosely indicated by "brightness." Thus pra- 
bhdte is usually the first dawn (as in nisi prabhdtaydm, "when 
night grows light"), but in i. 21. 1 we find tato rajanydu 
vyustdyani prabhdte 'bhyudite ran du " when day had dawned, 

1 Cf . iii. 164. 13 (grief made every night and day seem like a year). 
Conversely, in iii. 176. 5, four years of joy pass " like one night." 

2 1 have noticed in the epic no "fourth watch," such as is found in 
the classical period, e. g. in the BS. xxx. 3 f.,and Harsacarita, to mark 
the time before dawn. Thus in Harsacar. 166, the fourth watch, before 
dawn : and in 228, at the end of the third watch, eight strokes mark 
the number of leagues in the day's march. The modern Hindus divide 
the day also into watches, pahars, of three hours each. 

vol. xxiv. 2 



18 E.W. Hopkins, [1908. 

and morning had come, and the sun had risen." 1 The sun "sets " 
is usually expressed by "goes home," astarn upditi savitd (at 
sunset, astarn gacchati bhdshare ; just after sunset, surye 'stam 
itesati, "being gone," 2 ix. 29. 64 and 87; the sun sets twice 
on this day !), but this is sometimes filled out with the word 
hill, suryo hy astarn abhyayamad girim, i. 24. 10. 

Of short combination of days, saptardtra, sennight, is col- 
loquial, interchanging with saptdsta divasdh, seven or eight 
days, the former perhaps more common: saptardtrena mrtyu- 
bhalc, "he will die within a week," xii. 318. 13 (amongst the 
aristdni, " death-signs "), etc. But ten days also make a group 
and three and its multiples are more common in the ritual, tri- 
rdtra, tryaha, sanrdtra, dvddasardtra, periods of three, six, and 
twelve days. The fortnight, ardhamdsa, mdsdrdha, pahsa, is 
not regarded as a group of days but as half a moon, or the one 
wing, division, of a month, purva and apara, xiii. 87. 19 = 
Manu iii. 278. 

In accordance with a " Veda- word " (cf. SB. xii. 2. 2. 23) 
there is a formal equation of the year with a day and night in 
iii. 52. 23: 

ahordtram mahdrdja tidyarii samvatsarena ha, 

which may be compared with the ritualistic substitution of a 
month for a year, ib. 35. 33. 3 

MONTHS AND SEASONS. 
The Months : Although the month of thirty days is Vedic, 
yet, to judge by colloquial epic language, the month was a moon's 
length, twenty-seven to twenty-eight days. This is implied in 
the colloquial expressions just referred to, which, like our week of 
seven to eight days, give natural halves of a-half-moon period. 



1 So in Sutras, when "the sun is over the trees," adhivrksasuryam 
adhvanam na pratipadyate, is either noon or late afternoon, Vas. xn. 
43 ; Gaut. v. 40. 

2 Compare the phrases, mryastamanavela (astamana=astamayana) ; 
astarn ydie (or prapte) divahare (or dinakare). Before the evening 
twilight comes on is expressed by anagatayarh sarhdhydyam pa&cima- 
yam. The nooning of the sun is expressed by madhyadeiagate ravau 
(after prapte co 'hnikakdle tu\ xii. 346. 14 (in 18 pitaraJi is accusative) ; 
also by the sun's turning astaSikharam prati, vii. 99. 1. 

3 Compare Mahanar. Up. xxv. 1, ye ahnrdtre te darsapurnamdsdu. 



Vol. xxiv.] Epic Chronology. 19 

But the regular ascription of "ten months" to the period of 
pregnancy (c. 280 days) sets the matter beyond doubt, as other- 
wise we should have an extraordinary duration (300 days) 
assumed as the normal period. Apart from cases of diabolic 
birth, pregnancy of three years, i. 74, or longer, and divine or 
devilish performances of a similar sort, as when "demons con- 
ceive and instantly give birth," 1 birth is usually said to follow 
in ten months (cf. Ch. Up. v. 9. 1), garbhan dasa masdn bibh- 
rati, iii. 134, 17; xii. 7. 14; iii. 128. 7; 132. 14; 205. 10; xii. 
332. 18, etc. So RV. x. 184. 3 ; AB. vii. 13. 9, dasame mast, 
in the course of the tenth month. 2 But. as the solar month 
becomes popular we find in the Sulabha-Janaka episode, inserted 
in xii. 321, that the period of pregnancy is set at the end of the 
ninth month, sampurne navame tndsi jatah, si. 117. So in the 
introduction to the poem, i. 63. 61, Vyasa's mother is ready to 
be born after nine months, as soon as the tenth month arrives, 
mast dasame prdpte; and in the law-book of Yajnavalkya, iii. 
83, birth is said to take place "in the ninth or tenth month." 
Also in iv. 36. 3, " twenty -eight nights or a month to -its end" 
(was the fight) seems to imply a month of thirty days; while 
in xii. 232. 13 a "month" is formally declared to be of this 
length, masah smrto rdtryahani ca trinsat. To distinguish the. 
two kinds of months we find Magha described as sdumya, lunar, 
in xiii. 168. 28. Compare, however, the sense of agreeable or 
moonlighted in v. 142. 16-17 : saumyo (N~. candrikayd abhira- 
mah • this is said a week before the new moon) '■yam vartate 
masah suprdpayavasendhanah . . alpamaksikah, nispankah, nd 
Hyusnasisirah. As the solar month, really civil month, of 
thirty days was thus reckoned from remotest antiquity, it is a 
question of locality or popularity only. In many cases a 
" month," according to long-inherited use and metaphors (below) 
was thirty days; in other cases, however, it was a moon, not 
quite twenty-eight days, though reckoned as full twenty-eight. 

1 So a devil's baby becomes full-grown at once : sadyo hi garbhan 
rdksasyo labharde prasavanti ca (bdlo 'pi yduvanam prdptah), i. 155. 
35 f. 

2 Cf . dasamdsya, dasa masdn, EV. v. 78. 9 ; SB. iv. 5. 2. 4. But !§B. 
xi. 1. 6. 3, samvatsara eva stri vd gdur vd vadavd vd vijayate (" in the 
course of a year," as in SB. iii. 2. 1. 27, samvatsare jayamanah), gives 
only the outer limit, this side of which birth takes place, " in (-side of) 
a year." Cf. ib. xi. 5. 4. 6. 



20 E. W. Hopkins, [1903. 

A stanza cited above, p. 15, shows that the month does not 
begin with the full-moon, as was sometimes the case, 1 but with 
the bright fortnight, mdsdh sukladayah. The moon, it is said, 
is born at the beginning of the bright fortnight (by means of 
drinking the " six essences " of Varuna) , ,/ay«£e tarunah somah 
suklasyd ''ddu tamisrahd, v. 110. 4. The bright half ends with 
the full moon, sukldtyaye pdurnamdsydm, i. 76. 61. The full- 
moon day is par excellence the month-day, e. g. kartikl is the 
full-moon (night) of Kartika. " Fair as the full-moon," paurna- 
masyam ive 'nduh, i. 76. 61, etc., is an epic commonplace; 2 
"ugly as the moon, on the fourteenth day of the dark fort- 
night," drastum na nah prltikarah sail '»a krsnasya paksasya 
eaturdasdhe, xi. 21. 13, refers to a gnawed body, diminished 
and ghastly. 

Only traces remain of Vedic phraseology in naming the sea- 
sons (months) : supuspitavane kale kaddcin madhumadhave 
( = Caitra-Vaisakha) spring-time, i. 125. 2; mcisukrdgama, the 
time of heat, 3 sucimkrdgame kale iusyet toyam ivd 'Ipakam, ii. 
47. 24; "the sun absorbs water with his fierce heat when interven- 
ing between Suci and Sukra," ugrarasmih sucisukramadhyagah, 
viii. 79. 78. But this is also the tempest-time (when the mon- 
soon first blows), janghdvdto vavdu cd 'sya sucisukrdgame 
yathd, i. 151. 2 (jyesthdsadhayoh samaye, N.). The following 
rain-months, Nabhas and Nabhasya, 4 are alluded to in H. ii. 
95. 1 ; but there is no mention of the corresponding Vedic terms 
for autumn, TTrja and Isa; of those for winter, Saha and 
Sahasya; or of those for cool-time, saisirdu, called Tapa(s), 
Tapasya (yad etayor balistharh sydyati) in SB. iv. 3. 1. 19; viii. 
7. 1. 5. 

It is usually in conventional passages that the "four-month " 
season is noticed: darse ca pdurnamdse ca caturmdsye punah 



1 ^B. vi. 2. 2. 18 and Kaus.B.v.l, etc., give the full moon of Phal- 
guna as the beginning of the spring and of the year (paunamasi ha 
vava prathamd vyuvdsa). 

2 Compare Yaj. i. 80, sustha inddu, "when the moon is full" (?), one 
should " avoid Magha and Mula in connubial intercourse." 

3 Cf. SB. iv. 3. 1. 14, Madhu and Madhava are the vdsantikdu months 
of growth ; ib. 15, £ukra and Suci are the grdismau months of strongest 
heat. 

4 SB. toe. cit. 16, vdrsikdu. 



Vol. xxiv.] Epic Chronology. 21 

pwnah, (offered horse-sacrifices) "on the new and full-moon and 
on the thirds of the year," xii. 29. 114. But it is once given as 
the proper term for wages which are to be paid caturmasydva- 
ram, "at least every four months," ii. 5. 118. The seasons thus 
divided are summer, rains, and winter, or spring, rains, autumn, 
according to SB. xii. 8. 2. 33 ; vii. 2. 4. 26, respectively (which 
must embrace the remaining time). The full -moon of Phalguna 
is the regular beginning of the Cdturmasya. See also below. 

According to xiii. 168. 6 and 28, the winter solstice occurs 
near the beginning of the bright half of the month of Magha. 
The saint who dies here is anxious to pass away at an auspicious 
period, i. e. in the northern course of the sun and the bright 
half of the month. He asserts in this passage that the proper 
conditions are fulfilled. The solstice has already taken place 
(xii. 47. 3) and he says: " The lunar month Magha has arrived, 
Yudhisthira, and the bright fortnight must be two-thirds (or a 
quarter) past:" 

magho '■yam samanuprapto mdsah saumyo Yudhisthira 
tribhagasesah pakso '•yam, suklo bhavitum arhati. 

The doubtful meaning of tribhdga (% as well as %) renders 
exactness of translation impossible. The natural agreement of 
the adjective would lead to the meaning given by Mr. Aiyer, 
that the solstice occurred ' ' on the expiry of the fourth part of the 
bright fortnight in the month of Magha, that is, on the fourth 
or the fifth day after new-moon" {op. cit., p. 81). But there 
is no certainty that tribhagasesah does not refer to the word 
month. 1 Mlakantha here says the day is the eighth of the 
month. Mr. Aiyer takes the fifth lunar day after new moon as 
the real meaning (referring to N. on vi. 17. 2). Then, accord- 
ing to the Vedanga, he argues that this would imply that the 
solstice was the fourth of the five winter-solstices of a five-year 
cycle (op. cit., p. 84), and uniting this with the assumed date 
of the Kali Yuga in 1177 B.C., he arrives at 1194-3 as the date 
of the war according to the epic itself. It is at least unfortu- 
nate that a stanza so important for this result should not be less 
ambiguous, for if tribhagasesah refers to mdsah the whole argu- 
ment is invalidated. 



1 Mr. Aiyer's metrical objection (loc. cit.) is inconclusive, as he has 
admitted by letter. 



22 E. W. Hopkins, [1903. 

Though Magha is the solstice-month, neither this nor Caitra 
nor Kartika is the first month, but Margaslrsa. This, as is 
implied in the Gita, stands in the same relation to the months as 
spring stands to the seasons. 1 In connection with the phrase 
agrahayana, this, like the epic list, points to Margaslrsa not as 
the best but as the formal first month (compare agrahayani = 
margasirsi, Vas. xi. 43). The list is given in xiii. 106. 17 f. 
and is as follows (with the alternate names supplied from 
another list at xiii. 109. 3f.). 2 

Margaslrsa (November-December), Pausa, Magha, Bhaga- 
daivata or Phalguna, Caitra, Vaisakha, Jyesthamula or Jyaistha 
(jyestha, • sic), Asadha, Sravana, Prostha- or Bhadra-pada, 
Asvayuja or Asvina, Kartika. 

If, as the epic says, the month begins with the new moon, 
Magha would be from the new moon of December to the new 
moon of January. But this list probably implies that (as usual) 
Magha is counted from the full moon in January and so on, 
Margaslrsa being from the full moon of November. Two pas- 
sages in Virata show that when the " seventh day " and "eighth 
day " are mentioned they refer to the days after the full moon. 
In iv. 47. 10-11, these days are cited merely as saptamyam 
aparahne and astamyam adityasyo 'dayam prati, ' ' on the 
afternoon of the seventh and at sunrise on the eighth," which 
days in 30. 26-27 are referred to as krsnapaksasya saptamlm 
and apare divase, "on the seventh of the dark half and on the 
next day." This is borne out by the fact that the psuedo-epic 
(like the law) gives as marching-months Margaslrsa, Phalguna, 
or Caitra, which would be November, February, March (cf. 
Manu vii. 182, Vishnu, iii. 40), or more particularly the full 
moon of these months, Caitri, Margasirsi ; and in fact Kartika, 
Kaumuda, overlaps autumn and winter, v. 83. 7 : 

Kdurnude masi Revatyarh saradante himagame, 

" on the arrival of cold, at the end of autumn, under the star 
Revati, in the lotus-month;" xiii. 115. 76: 

1 Krs^a says (10.. 35) : mdsdndm marga$ir$o 'ham rtunam kusumd- 
harafy. On this passage and the application of agrahdywrja, compare 
Tilak, The Orion, pp. 67 f. (Phalguni, the first night of the year), 86 f., 
153 ; Jacobi, Beitrdge zur Kenntniss der vedischen Ohronologie, p. 109. 

2 The object of this list is to show that Krsna ought to be adored under 
a different name on the twelfth of each month. 



Vol. xxiv.] Epic Chronology. 23 

pura mansarh na bhaksitam saradam kaumudam ma- 

sam, 

"meat was not eaten of old during the autumnal month Kau- 
muda." 1 

There is, then, a certain discrepancy in the matter of the 
epic months. The data as to marching, etc., ahove, would indi- 
cate that they were the months of the law-books, as follows, 
reckoning from full moon to full moon : Margasirsa, November- 
December; Pausa, December-January; Magha, January-Feb- 
ruary; Phalguna, February-March; Caitra, March-April; Vai- 
sakha, April-May; Jyaistha, May-June; Asadha, June-July; 
Sravana, July-August; Bhadrapada, August-September; As- 
vina, September— October ; Kartika, October— November. But 
this is incompatible with Magha being well on its way by 
December 21st, as above. 

The lunar day, tithi, masculine and feminine, gives rise to 
auspicious and inauspicious days, sutithi, dustithi (but sudina 
and durdina refer to the weather, e. g. i. 190. 46, durdine 
meghasamplute, "on a dull cloudy day"). The form is usually 
feminine, but tithau purne, i. 123. 6. The asterism and mu- 
hiirta are usually mentioned with the lucky or unlucky day, as in 
ii. 25. 4; tithav atha rnuhurte ca naksatre ca , bhipujite ; xii. 
100. 25, tithinaksatrapujitah ; xii. 180. 45, 46, uta jatah 
sunaksatre sutithau swmuhurtajah, naksatresv asuresv anye 
dustithau durmuhurtajdh. Manu's derivation of atithi from 
anityam sthitah is found, with a varied reading, in xiii. 97. 19 = 
M. iii. 102. 

The holiest night is that of the full-moon of Kartika, punya- 
tama ratrih parvasamdhau sma saradi k&rtiM, iii. 182. 16. 
The thirteenth day is a very lucky day, prasasta, iii. 134. 20, 
but not when sun and moon are eclipsed, vi. 3. 28, 32. The 
last stanza refers to the ill-luck of having a new moon on the 
thirteenth lunation, instead of the fourteenth, fifteenth or six- 

1 In i. 209. 80, kdumudl is either the moon or the full-moon night of 
this month : 

akdlakaumudirh cai 'va cakratuh sarvakalikim, 

" they made the moon untimely (rise) at all times" (or the night come, 
v. 1. sarvakdminim). According to the Susruta, Margasirsa is the second 
autumnal month. 



24 M W. Hopkins, [1903. 

teenth, and the ill-omen of the new moon on the thirteenth 
with a synchronous eclipse. But every unusual number anyway 
is ominous of evil. Thus headless trunks are especially por- 
tentous when they have many hands and feet, ix. 58. 56, and 
animals with four eyes and five feet, and women having four or 
five daughters at a birth are grouped together, as of very evil 
omen, vi. 3. 3 f . So the unusual day is the unlucky day apart 
from the eelipse : 

caturdaslm pancadasim bhutapurvdm ca sodailm 
imam tu na 'bhijdne l ham amdvdsydm trayodasim 
candrasurydv ubhdu grastdu ekamdsim trayodasirn. 

It is added that an aparvani eclipse portends disaster. 
With this stanza, vi. 3. 32, compare xvi. 2. 18, 19 : 

evarh pasyan Hrsikesah samprdptam kdlaparyayarn 
trayodasydm amdvdsydrh tan drstvd prdbravld idam 
caturdasi pancadasi krte , yam Hahund punah 
prdpte vai .Bhdrate yuddhe prdptd ca , dya ksayaya nah. 

The "first day" of the month is generally given by the 
moon-day, but sometimes by the number, Kdrtikasya tu mdsa- 
sya prathame i hani, ii. 23. 29. The days, of the new and full 
moon, amdvdsya,' amdvdsi, and pdurnamdsi, are also called 
darsa and purnamdsa. The most prominent and auspicious 
days are those of the new and full moon, the thirteenth, and 
the eighth (end of a week), Sinivali, AnumatI, Kuhii and Raka, 
the first part of the day of the new moon and of the full moon, 
and the latter part of the day of the new moon and of the full 
moon, respectively (AB. vii. 11), viii. 34. 32, etc. The eighty- 
seventh section of Anusasana gives the rules for Sraddhas (com- 
pare Manu iii. 273-276) on each day of both fortnights, the 
fourteenth of each being bad. To live one hundred years, one 
must be chaste on the days of the new and full moon, and on 
the eighth and fourteenth of all lunar fortnights: amdvdsydm 

1 This has the epithet ^akradevata, saptamac ca 'pi divasad amdvdsya 
bhavisyati sariigrdmo yujyatdm tasyam tarn ahuh Sakradevatdm, '.'by 
the end of the week there will be a new moon ; let the fight begin on 
that (new moon, for) they say that has Indra as its divinity," v. 142. 18. 
Aiyer, op. cit. p. 96, interprets as Jyesthd (as if yam stood for tarn). Cf . 
aindrd (p. 32) and sakra-ddivata, BS. vii. 12. 



Vol. xxiv.] Ep>ic Chronology. 25 

pdurnamdsydm caturdasydm . . astamydm sarvapaksdndm, 
xiii. 104. 29; Mann iv. 128. A particularly favorable eighth 
day is known as kdmydstami, for gifts and oblations, xiii. 71. 
49 ; 76. 19 ; 132. 7, the last referring to that of the dark half of 
Kartika : 

Kdrtike mdsi ca , slesd bahidasyd , stdml sivd 
tena naksatrayogena yo daddti guddudanam, etc. 

Special gods have special days. Krsna's day is the twelfth 
(above). Skanda's days are the bright half's fifth and sixth, 
iii. 228. 15; 229. 52 (srljustah pancamlm Skandas tasmdc 
chripancaml smrtd . . sasthl mahdtithih) . The Sun's day is 
the sixth or the seventh, saptamydm atha vd sasthydm bhaktyd 
pujdm karoti yah, iii. 3. 64 (perhaps on account of the title 
Saptasapti, for saptdsva, ib. 63). 

The name of the month is added in noun-form, dvddasydm 
mdghamdse or jydisthe mdsi, etc., xiii. 109, passim, or the 
adjective form of the month is joined to the name of the day, 
cditryd pdurnamdsl, for initiating the king at the beginning of 
the horse-sacrifice ; mdghd pdurnamdsl after dvddasl mdgha- 
mdsiki, xiv. 72. 4; 85. 4-8; or the day is implied, pardm 
cditrlm upasthitdm, ib. 76. 25 {pari cditrlm, 81. 23). 

The Seasons : These are six in number, e. g. iii. 134. 13. 
The group consists of Sisira, cool-time, Vasanta (kusumdkara) , 
Easter-time, Grisma, heat-time, Varsas, rains, Sarad, autumn, 
Hemanta, snow-time. For Grisma is found also Usnani, heats, 
as in i. 222. 14, usndni vartante, " the heated term is at hand." 1 
A favorite epic word for the hot spell is Nidagha (Nidaghakala, 
v. 26. 10) scorching-time: 

meghdv ivd Hapdpdye dhdrdbhih, vii. 98. 14, 
with usnaparydye meghdndm iva vdgurdh, ib. 32. 

niddghavdrsikdu mdsdu lokam gharmdnsubhir yathd 

(sc. ahatdm Pdndavam), vii. 30. 10 (v. 1. C. 1331, gharmdmbii- 
bhir). So iii. 3. 49; vii. 146. 11; R. Gorr. v. 41. 25; and 

1 In xiv. 43. 8, arko 'dhipatir usnanam jyotisam indur ucyate, "the 
sun is lord of hot (things); the moon, of stars," the heated term may be 
specially meant. In the next chapter the sun is called the beginning of 
light, ddityo jyotisam ddir agnir bhutadir ucyate, 44. 5. 



26 E. W. Hopkins, [1903. 

naidagha Huh, AV. ix. 5. 31, etc., all showing that Nidagha 
is the heated term before the rains begin. The first rains are 
called Pravrs, whence the phrase of both epics, yaiha pravrsi 
toyadah, vi. 81. 39; R. iii. 18. 23, etc., though there is no pas- 
sage, I think, indicating that the whole rainy period was for- 
mally divided into two seasons, varsas, sarad, hemanta, vasanta, 
grisma, pravrs, which division occurs first (as just cited) in the 
Susruta (Thibaut, Grundriss, Astronomie, p. 11). 

In the six-season division, the rains take four months and the 
remaining seasons, of which the first is Sisira, must be divided 
between eight months. Compare caturo vdrsikan masdn, i. 62. 
32, "through four rainy months;" and v. 35. 67: 

astamasena tat kurydd yena varsah sukham vaset, 

where the eight months as a group are opposed to "rains," as" 
also in Manu, ix. 304 f . 

Though the earlier literature makes Vasanta the first season, 
that of the epic may begin with its expressed choice, and the 
seasons may be arranged about as follows: Sisira, February; 
Vasanta, March-April ;' Grisma, May till it rains ; Varsas, June- 
September; Sarad, October till cold weather; Hemanta, c. 
November-January. Hemanta is the season when the shadow 
is shortest, muhurtam sukham evdi 'tat tdlacchdye 'va hai- 
manl, ii. 80. 50 (and elsewhere). 2 On the other hand, the 
Asoka blooms at the end of Hemanta, hemantdnte 'soka iva 
raktastabakamanditah (babhdu Rdmah), v. 179. 31. The 
term, probably, is a general one, either ignoring or, according 
to locality, merging into Sisira. 8 In the same way, Vasanta is 
an indefinite period, from the middle of February running 
through March and into April ; as Grisma includes part of June. 
The rains may last four months, but there often are seasons 
when the rains are not more than two or three months, from the 
middle of June to the middle of August or September, instead 

1 The spring-festival comes on the thirteenth of the first half of Caitra. 

2 Compare (at vi. 17. 2) the stanza of the Bharata-savitri : hematite 
prathame mdsi suklapakse trayodaslm pravrttam Bharatam yuddham 
naksatre Yamadaivate (under BharanI). 

3 So in Manu, iii. 281, "thrice in a year ... in Hemanta, Grisma, Var- 
sas," as the three general seasons, though six are known, ib. 217 ; also 
iv. 26 ftvante is at the end of the four-month season. 



Vol. xxiv.] Epic Chronology. 27 

of the "four rainy months," and then Sarad begins earlier. 
The coming of the rains varies by almost a month, so there 
is quite a margin here as well as in respect to the other 
seasons, which are really meteorological divisions shifting 
from year to year like our dog-days and Indian summer. 
They are always numbered as six, rtavah sat, v. 11. 15; 
viii. 34. 47, etc., 1 but they are not enumerated as a group. 
According to iii. 3. 6, when the sun first took pity on 
starving man it began its work in the "northern course," 
and then passed into the "southern course," or in other words 
the year of the sun begins at the winter solstice, which in the 
epic is the time when begins the "northern course of six 
months," sanmdsa uttarayanam, Gita, 8. 24, to reach which 
Bhisma delayed his end, as explained above. The Gita passage 
alone, however, may imply, as Mr. Tilak has suggested, through 
its collocation of fire, flame, day, bright half, and northern 
course, as opposed to smoke, night, dark half, and southern 
course, that the northern course was the fiery, fiaming months, 
or in other words, that the year began not with the solstice but 
with the vernal equinox. This may well be the case, since the 
Gita stanza is merely a recasting of a famous Vedic passage 
(Ch.U. iv. 15. 5, etc.). 2 The other Gita passage, cited above, 
agrees with the older Brahmanas in making spring the first sea- 
son, whereas the pseudo-epic stanza cited above, p. 15, makes 
Sisira the first season ; a discrepancy consonant with the charac- 
ter of the heterogeneous epic. 

A passage in Vana, 163. 34 f., says that the sun, after turn- 
ing on his course, when desirous of making coolness, sisirani, 
favors the southern district. Then the cool-time arrives, sai- 
sirah kalah (when cattle suffer, vi. 118. 8). Then returning, 
nivrttah, he takes to himself the energy of all creatures, which 



1 Three, five, six, or seven in early texts, e. g. SB. ii. 1. 1. 12-13, and 
above, p. 13, note 2. 

2 Compare Tilak, The Orion, p. 33 f . The northern course of the sun 
is the gods' abode, the Manes as opposed to the gods are as dark to light. 
SB. ii. 1. 3. 1 f. Here the gods are represented by the northern course, 
the bright half of the year (spring, summer, rains), the crescent (first 
half of) the moon, the day, and the forenoon, as opposed to the Manes, 
represented by autumn, winter, cool-time, etc., the gibbous moon, the 
night and the afternoon. It is added that spring is the priesthood, as 
the first (best) season and caste. 



28 E. W. Hopkins, [1903. 

thereupon become sweaty, weary, and sleepy. Then, following 
a path incapable of being indicated, anirdesyam rnargam 
avrtya, the sun emits rain. Here the "cool-time" marks the 
beginning of winter, as the time of fatigue indicates summer. 
Instead of bhajate one might expect tyajate, as tatah following 
seems to show that the immediate result is coolness. But the 
words may mean only that after the southern course has been 
completed the cool time comes; though the southern course 
appears to precede the cool time immediately, which would 
require the latter to begin with the early autumn instead of the 
solstice, and the former to begin with the equinox : 

tatha tami&rahd devo mayukhair bhdvayan jagat 
rnargam etad asambddham ddityah parivartate 
sisrksuh sisirany eva daksinam bhajate disam 
tatah sarvani bhutani kalo 'bhyarchati saisirah. 

The rainy season is described in iii. 182 ; it ends some time 
before the full moon of Kartika (16). There is, I believe, no 
passage in the epic grouping the months, beginning with 
Magha, in pairs according to the seasons, as they are both 
defined by native lexicographers and grouped in AB. iv. 26 
(beginning with vasantikau masau); SB. iv. 3. 1. 14 f . ; or as 
ib. viii. 3. 2. 5, dvau hi rndsdv rtuh, limits the season. At 
present the natives reckon three seasons of four months each, 
"hot, cold, and stormy," JASB., 1901, p. 57 of Part III. 

THE ASTER1SMS. 

Though the "star-man," ndksatra, xii. 76. 6, perhaps a mere 
fortune-teller, 1 is a despicable kind of priest, the stars are the 
object of constant study and even the asterisms are frequently 
brought into the epic narrative, sometimes formally, as in the 
lists spoken of below, sometimes incidentally, as when a " Tirtha 
of the Pleiades and Magha " is mentioned, or when " Rohini sur- 
rounded with stars," tdrdbhih, serves the poet as a simile for a 

1 I am not quite sure of this, for the astrologer is also called a samvat- 
sara and he is no better than a ^iidra, xiii. 135. 11, and yet this is the 
very title given to Garga, xii. 59. Ill (other references in my Great Epic, 
p. 15). The asterisms, though admitted into the ritual, are clearly little 
regarded in !§B. ii. 1. 2. 19. Here, by the way, ib. 2, no asterisms have 
more than four stars except the Pleiades. 



Vol. xxiv.J Epic Chronology. 29 

queen at court, ii. 58. 27. The latter star, Aldebaran, is the 
best-beloved wife of the Moon-god, and stirs up jealousy among 
the other asterisms, ix. 35. 47 f. Another story, reported below, 
attributes jealousy to Abhijit, "younger sister" of RohinI, and 
tells how the six Pleiades, with the Fire-god as the seventh, 
replace her, iii. 134. 13; ix. 44. 12, as six; "seven-headed with 
Agni as divinity," iii. 230. 11. The Pleiades are the asterism of 
the Sword as Justice, xii. 166. 82. But apart from the Pleiades 
and Aldebaran, there is little notice taken of the asterisms, except 
when the planets and asterisms give portents, till we come to the 
formal lists of the latter in the late epic tables. Apart from the 
asterisms and planets, however, the Seers, the Great Bear and 
Dhruva, the Pole star, are not infrequently lauded and described: 
A saint-star is located "midway between the Holy Seers (the 
Great Bear) and Dhruva, the son of Uttanapada" (Puranic), xiii. 
3. 15. These Seers rise and set in the royal North (disam udlcl 
raja, xiv. 43. 10) at Mount Meru, Maha Meru, iii. 163. 15, round 
which go daily the sun and moon and other lights of heaven. 
Setting there, astam prdpya, the sun goes north, after the twi- 
light time, samdhyam atikramya . . . udicim bhajate hdstham 
disam (as phrased elsewhere, xiii. 168. 6, drstva [Bhlsmah] 
nivrttam adityam pravrttarh go Htardyanam) ; then eastward 
turning he goes on again. And even so, dividing the months, 
mdsdu vibhajan kale bahudha parvasamdhisu, goes the moon 
with the asterisms, naksatraih saha gacchati, and having gone 
about Meru goes again to Mount Mandara, ib. 27 f . Compare 
v. 111. 14, " Here (in the North) are the Seven Seers, Arundhati, 
and Svati, which rises here, and (the year) Pitamaha " (as sacri- 
fice). 1 

Since the asterisms are known by name even in the Rig Veda, 
i. 24. 9, Satam-bhisajah ; x. (19. 1 ?) 85. 13, Magha(?) and ArjunI; 
while the Atharva Veda, xix. 7, and other Vedic works presuma- 
bly earlier, give lists of them, it may be assumed that, whatever 
the date of the epic, the poets were familiar with all the asterisms, 
and it is to be expected that the later epic will maintain its usual 
character by endorsing both lists, that of the twenty-seven 

1 atra te rsayah sapta devi ad 'rundhati tatha, atra Hsthati vdi svdtir 
atrd 'syd udayah smrtah,, atra yajnam samdsddya dhruvam sthdtd pitd- 
mahah,, i. e. the year. 



30 E. W. Hopkins, [1903. 

asterisms and that of the twenty-eight. Whether the asterisms 
were first counted as twenty-seven or as twenty-eight may still 
be doubtful (twenty-eight is more likely) ;' but the constant later 
view was that they were twenty-seven, and this is the current 
epic view, as represented in the usual ascription of twenty-seven 
wives to Soma. They are naksatrayoginyah, naksatrayogani- 
ratdh, and are counted as seven and twenty, i. 66. 16-17 ; ix. 
35. 45; xii. 207. 24; 343. 57, etc. 

The twenty-seven asterisms are divided into nine-day weeks 
(navaratra) by groups of three (of nine each, beginning with 
Asvini, Magha, Mula, respectively, according to the scholiast), 
and, according to this division, a grdhra, or evil heavenly body, 
papagraha, has a different effect on the fortunes of men, if we 
may further trust this expansion on the part of the scholiast of 
B. vi. 3. 31, . 

trisu sarvesu naksatranaJcsatresu, visdmpate, 

grdhrah sampatate slrsarh ( !) janayan bhayani uttamam, 

which might refer to a three-fold division of the year according 
to the seasons already mentioned. 2 The Rbhus, however, are 
mentioned merely as divinities of the gods, eternal and change- 
less under the aeonic changes, devadevdh sandtandh, iii. 261. 
19-22 : 

na kalpaparivartesu parivartanti te tathd. 

The twenty -eight asterisms are said to "enter the circle of 
the sun " in the West and then, on account of their association 
with the moon, to fall out of the sun again, after twenty-eight 
nights are past, v. 110. 15-16 : 

atah prabhrti suryasya tiryag dvartate gatih 
atra jyottnsi sarvani visanty adityamandalam 
astavihsatiratram ca cankramya saha bhanuna 
nispatanti punah suryat somasamyogayogatah. 

Twenty-eight asterisms are implied at ix. 34. 6, where forty- 
two days pass from Pusya to Sravana (=22 + 20, Pusya being 
sixth and Sravana twenty-first). 

1 But cf. TS. i. 7. 7. 2 ; &B. x. 5. 4. 5 ; Whitney, OLS. ii. p. 360 f., with 
citations. 

8 O. 98 has trisu sarvesu purvesu nfthsatresu . . . sirse. With B. cf . 
RV. x. 61. 10 f., on the Navagvas. 



Vol. xxiv.] Epic Chronology. 31 

The full list of the asterisms, as developed at xiii. 64, is as 
follows. I supply the equivalent forms from other passages, 
chiefly in the same book (giving only epic data) ; sc. naksatram : 

1, Krttikas, Krttikiiyoga, dgneyam, vahniddivatam ; 2, Ro- 
hini, prdjdpatyam, dhruvam / 3, Mrgasiras, Mrgottama, soma- 
ddivatam ; 4, Ardra; 5, Punarvasu, also dual, viii. 49. 28, 
required, where C. 2,328 has sg. ; 6, Pusya, 1 Tisya; 7, Aslesa, 
sg. and pi. ; 8, Magha (N. pitryam), sg. and pi. ; 9 and 10, 
Phalguni, purva, uttard, uttaravisaya ; sg., pi., dual, 2 bha- 
gaddivatam ; 11, Hasta, savitram ; 12, Citra (Mitra? see 
below); 13, Svati, Svati, Svatiyoga; 14, Visakha, sg., dual, pi.; 
15, Anuradha, sg. and pi., mditram ; 16, Jyestha, dindram ; 
17, Mula; 18 and 19, Asadha, purva, uttard; 20, Abhijit, 
Abhijita(-yo^a) ; 21, Sravana, sg. and pi. ; 22, Dhanistha, sg. 
and pi.; 23, Satabhisa(-yo</a), vdrunam ; 24 and 25, Prostha-, 
Bhadrapada, Bhadrapada {-yoga), purva, uttard {-yoga) f 26, 
RevatI; 27, Asvini, sg. and pi.; 28, Bharani, sg. and pi. 

Sravistha, the older name of Dhanistha, does not appear to be 
used. Abhijit is lacking in the alternate list (below). The 
addition of yoga is common, the asterism appearing either as a 
noun (above) or as an adjective, as, for example, in adya pdus- 
yam yogam upditi candramdh, "to-day the moon enters its 
Pusya-conjunction " (a suitable time for a wedding), i. 198. 5. 
In the same way, maghdvisayagah somah, vi. 17. 2. 

Like the naksatriyah p>rajdp>atih, "whose hand is Hasta and 
head is Citra," in TB. i. 5. 2. 2, is the identification made dur- 
ing the moon-rite, candravrata, at xiii. 110. 2f., in which the 
worshipper identifies himself with the moon as anthropomor- 
phized from the feet up in the asterisms. 4 The month is Mar- 
gasirsa, and the moon's feet are Mula; the knees and thighs, 

1 Instrumental asterism, with which, all the others locative. So in the 
list at xiii. 89, except for Hasta and Abhijit (means) : hastena phalabhag ; 
sraddham tv abhijit a kurvan bhisak siddhim avdpnuydt, si. 11. 

2 uttardbhyam phalgunibhydm naksatrabhydm aharh diva, jato hima- 
vatah prsthe, tena mam Phdlguriarh viduh, iv. 44. 16 (cf. Arjuna, 
Arjuni = Phalguni). 

3 V. 114. 3 : nityam prosthapadabhydm ea sukre dhanapatau tathd, 
manusyebhyah samddatte sukrah (dhanam) ; xiii. 89. 13 : purvaprosfha- 
padah kurvan bahun vindaty ajavikan, uttardsu prakurvano vindate 
gah sahasrasah. 

4 Compare also the naksatrapurusaka of BS. cv. 



32 E. W. Hopkins, [1903. 

Asvini and Asadha ; the hands, Hasta, etc. ; while Citra is 
replaced in B. by Mitra (C. has citram, sic), as follows: netre 
mrgasiro vidydl laldte mitram eva tu, si. 8 (probably metrical). 
In this list the double asterisms are Phalguni (sic), Bhadrapada, 
sg. ; Asadhe, dual; while Nos. 8, 15, 21, 27, 28 are singular, 
and Nos. 7, 14, 22 are plural. 

Various periphrases take the place of the names of the aster- 
isms elsewhere in the epic. In xiii. 126. 36, gqjacchdydydm 
purvasydm kutape . . . yada Bhadrapade mdsi bhavate bahule 
maghd, the "elephant's shadow," as in Manu iii. 274, 
Yaj. i. 218, is probably a constellation. According to the 
scholiast at xiv. 63, 18, naksatre i hani ca dhruve, both Rohini 
and the Uttaratraya (cf. BS. loc. cit., PW. s. v.) bear the 
name dhruva (Nos. 2 and 10, 19, 25). No. 1 is vahniddivatam 
also in i. 221. 85; No. 16, Jyestha, called by N. jyesthdnaks- 
atra, is regarded, as Indra's (dindram, i. 123. 6). The 
place of Anuradha, No. 15, is taken by mditranaksatra-yoga 
in ix. 35. 14. In xiii. 89. 12, Satabhisa (epic for -bhisaj), No. 
23, is the naksatram vdnmam ; and in i. 8. 16, Phalguni 
(uttard, No. 10) is (for marriage) bhagaddivatam naksatram. 
The ''five-star" asterism, Hasta, No. 11, is called sdvitram pan- 
cataram, i. 135. 30. On Sakradevatd (not of No. 16, Jyestha, 
but) amdvasyd, see above, p. 24, note. According to xiii. 104. 
127 f . , one should not perform a Sraddha under one's natal 
asterism, nor under the two Prosthapadas, nor under Agneya 
(the Krttikas). The name Tisya for Pusya, No. 6, occurs in 
connection with the planet Brhaspati in the well-known proph- 
ecy at iii. 190. 90 f. : 

yada suryas ca candras ca tatha Tisya- Brhaspati 
ekardsdu samesyanti prapatsyati tadd krtam, 

"the perfect age will come again when sun and moon and the 
asterism Tisya with the planet Jupiter shall meet in one zodiacal 
sign." 1 Cf. brahmardsi below. Possibly the (solar) zodiac 
here referred to may have an earthly counterpart in the process 
of divination known as mrgacakra, alluded to at v. 48. 98 f. : 

1 The epic poet does not hesitate to make the planets Jupiter and 
Saturn stay a year in one asterism (as a bad sign), vi. 3. 27. See under 
Planets, below. Tisya was the birth-place of Jupiter, according to TB. 
iii. 1. 1. 5, cited by Tilak, The Orion, p. 161. 



Vol. xxiv.] Epic Chronology. 33 

sdmvatsard jyotisi cd 'bhiyuktd 

naksatrayogesu ca niscayajnah 
uccdvacam ddivayuktam rahasyam 

divydh prasnd mrgacakrd muhurtah, 

"astrologers and star-seers and those that tell fate by asterisms, 
secret prognostications, oracles, wheels of fortune, sooth- 
sayers^)." I take muhiirta to be a metrical equivalent to 
mduhurla, a soothsayer — astrologer. The latter word is found 
in xii. 121. 46, as in later literature. As to mrgacakra, "ani- 
mal-wheel," it seems to be what the modern Hindus call the 
" wheel of the nine planets " (compare satapada=^cakra), a sooth- 
saying device much in evidence at the beginning of the plague 
in 1896. The commentator explains it doubtfully as a purvani- 
pdta for cakramrga (asking questions with a wheel), according 
to the use prescribed in the Sivaite scriptures; the questions 
being astrological, "what is the asterism, by what planet is one 
injured ?" Perhaps it is a zodiac wheel. 

As in the stanza cited above, the asterisms are frequently 
brought into connection with the planets. Thus, in iii. 281. 
6, " he looked like Saturn entering Rohini." Another example 
occurs in xiii. 25. 22, where one is advised to perform ablutions 
at Great-Ganges, krttikdngdrake, "when Mars is in the 
Pleiades." 

In indicating time, the day and asterism are usually in the loca- 
tive ; the month, locative or genitive : astame i hani RoUinydm 
praydtdh JPhdlgunasya te, "they started on the eighth (day) of 
Phalguna under the asterism Rohini," i. 145. 34; krsnapakse 
caturdasydm rdtrdu, "at night on the fourteenth (day) in the 
fortnight after the full moon," i. 147. 4; Mdrgdslrsasya 
mdsasya candre Mulena samyute, "when the moon of the 
month Margaslrsa is in conjunction with the asterism Mula," 
xiii. 110. 3; Pdusamdsasya sukle vdi yadd yujyeta Bohinl, 
"when Rohini is in conjunction in the fortnight before the 
moon is full in the month Pausa," ib. 126. 48. With the month 
in the locative and the fortnight in the genitive : Kdrtike mdsi 
cd 'slesdbahulasya 'stami, " the eighth (day) of the fortnight 
after the full moon in the month Kartika under the asterism 
Aslesa," ib. 132. 7. With the new (and "full-moon") day, 
"month" is unnecessary (in xiii. 134. 4, somasyo 'ttisthamdnasya 

VOL. XXIV. 3 



34 E. W. Hopkins, [1903. 

paurnamasyam balim haret, the moon is object, "one should 
make an offering to the moon as it is rising on the full-moon 
night"). 

The passage referred to above, p. 15, containing the words 
Sravanadini rksdni, "the asterisms begin with Sravana," is of 
some importance for the date of the epic as well as for the way 
it was put together. In the lists already cited, the Krttikas, or 
Pleiades, begin the series of asterisms, while Sravana is the 
twenty -first in the list, coming just before Dhanistha = Sra- 
vistha. As late as Tajnavalkya's law-book, i. 267, the Pleiades 
hold this position, as opposed to the still later scheme (since c. 
490 A.D.) beginning with Asvini (to indicate the vernal equi- 
nox). The Vedafiga Jyotisa list begins with Sravistha, and 
G-arga says that the Krttikas are the first asterism for the ritual, 
while Sravistha is first for ordinary reckoning : karmasu krttikah 
prathamam (naksatram) Sravistha tu samkhyaydh (cited by 
Tilak, The Orion, p. 30). 

Now we can scarcely believe that the stanza stating that Sra- 
vana is the first asterism refers only to the quality of the aster- 
ism as the best or foremost, since in the same stanza the rela- 
tion of day to night is expressly that of priority and not of 
superiority. Nor is" there here any reference to an "abortive 
attempt" to reform the calendar, as is claimed by Tilak, op,, 
cit., p. 216. The fact is that even in i. 71. 34, where the 
change of the sphere is described, the act is not spoken of as 
abortive, but as one that succeeded. The translation of this 
latter passage, however, is not so certain as Tilak assumes; 
though the change of asterisms is" apparently described as act- 
ually occurring. We are told that Visvamitra, in his anger, did 
several wonderful things. As he caused the river, Kausiki, to 
change its name to Para, so also "he made another world with 
a right arrangement of asterisms," 

cakara ^nyam ca lokam vai kruddho naksatrasampada. 

Then follows: 

pratiSravanapurvani naksatrdni cakara yah, 

the obvious, though rather pointless meaning of which would 
be that "he made asterisms which had a prior promise." By 
separating the compound and giving purva the sense of adi, 
Tilak arrives at the meaning " he made the asterisms begin with 



Vol. xxiv.] Epic Chronology. 35 

Sravana." But even if this be a doubtful rendering, we have 
here the statement that Visvamitra did rearrange the asterisms, 
and in xiv. 44. 2, the statement (but without allusion to this 
story) that Sravana is the first asterism. In another passage, 
which describes how Abhijit, because she was jealous of RohinI, 
her elder sister, retired from the group and became "a star 
fallen -from heaven," naksatram gaganac cyutam, Abhijit is 
replaced by the Pleiades. This is a tale, thus far without his- 
torical meaning (except as showing that Abhijit was regarded 
as originally in the group) ; but in the quandary as to what was 
to be done when Abhijit retired, it is said that 

dhanisthadis tada kalo brahmand parikalpitah, 

"time was arranged by Brahman to begin with Dhanistha," iii. 
230. 10, which can mean only that this asterism was the first of 
the group. 

We thus have a legend peculiar to the later epic describing a 
rearrangement of the asterisms; a decided difference between 
different parts of the epic in regard to the first asterism; and 
the probability that Sravana was made the first asterism because 
the Vedanga system was no longer suited to the seasons, which 
had already receded a fortnight. In other words, the substi- 
tution of Sravana points to a late date (approximating the mod- 
ern substitution of Asvini) for these passages in books i. and 
xiv. 

In the Puranas and classical literature, the naksatrdni are 
called rksani, and this name is found appropriately enough in 
what is most certainly a pseudo-epic passage, xiii. 14. 37, 

stobha rksani pitaro yrahah, 

where the context shows the special meaning to be that of sra- 
vanadlni rksani in the other verse from the pseudo-epic, upon 
which I have just animadverted. 

Of the far-reaching results drawn by Mr. Tilak from a study 
of the stars as affecting the date of the earliest Vedic literature, 
this is scarcely the place to speak, since my study is confined as 
closely as is convenient to epic conditions. But I would sug- 
gest the consideration of two facts. The first is that the loose 
and casual references to the minor heavenly bodies, and the 



36 E. W. Hopkins, [1903. 

indifference with which they were regarded by the earliest 
depositaries of sacred wisdom make it improbable that any care- 
ful astronomical calculations were based upon them at a still 
earlier, pre-Vedic, period. The second is rather a corollary 
than a contradiction of this fact, namely, that though but little 
used as chronological guides, the stars are often regarded in 
their more obvious appearance, and there is no objection to pos- 
tulating a primitive acquaintance with and veneration for bril- 
liant stars, especially groups of stars, marking a seasonal change. 
Thus the savages of the South Sea Islands, though they can 
scarcely be said either to be star-worshippers in general, or to 
measure time by the stars, reckon the year (with thirteen 
moons) as beginning with the rising of the Pleiades just after 
sunset, about the middle of December, and "pay idolatrous 
worship to them." ' 

THE PLANETS. 

The planets, or rather the grahas, which include the planets, 
are reckoned as a group of five (so RV. i. 105. 10 ?) or seven in 
the early epic ; but the later epic makes them nine in number. 
The order in which they are named is interrupted by the intru- 
sion of the additional grahas and even of gods, but it appears 
in xiii. 166. 17 as Venus, Jupiter, Mars, Mercury, Saturn, where 
Bhauma, for Mars, is noteworthy. The eclipse-demon, Rahu 
(whose dimensions, according to the epic, together with those 
of the sun and moon I have given in this Journal, vol. xxiii. 
p. 154), here appears between Budha and Sanaiscara (Mercury 
and Saturn). In iii. 3. 17, the order is Jupiter, Venus, Mer- 
cury, Mars, Saturn. Here Mars is, as usual, Afigaraka. The 
group begins with the moon and Saturn is added after Indra, 
the sun, Agni, and Krsna; the whole group being preceded by 
the group of the five elements : 

sotno brhaspatih sukro budho '"iigaraka eva ca 
indro vivasvdn dlptahsuh sucih saurih ianaiscarah. 



1 Compare Gill, Myths and Songs, p. 317: "The Pleiades were wor- 
shipped [as harbingers of the new year] at Danger Island, and at the 
Penrhyns, down to the introduction of Christianity in 1857. In many 
islands extravagant joy is still manifested at the rising of this constella- 
tion out of the ocean." 



Vol. xxiv.] Epic Chronology. 37 

According to v. 34. 54, the stars, naksatrdni, are affected by 
the grahas (N. sun, etc.). In vi. 17. 2 are mentioned "seven 
great grahas," mahagrahdh, where the scholiast says that there 
are nine in all, " as Rahu and Ketu are to be added as upagra- 
Aas." These upagrahas are recognized late in the epic, being 
mentioned in the Markandeya episode, iii. 227. 1, with the gra- 
has, seers, and mother-goddesses. In vi. 77. 11, where the sun 
is surrounded by krurd mahagrahdh, Rahu, though the number 
is not stated, may be included. The technical name for hos- 
tile planets is here used. So in vii. 137. 23, "the seven Maha- 
rathas oppressed Bhima as the seven grahas (oppress) the moon 
at the destruction of living creatures " (somarh sapta grahd iva; 
compare the "seven suns" active in pralaya). In viii. 37. 4, 

nihsaranto vyadrsyanta surydt sapta mahagrahdh. 

The meaning, according to the scholiast, is that the grahas 
appeared advancing with the sun as the first of the seven. 

In all these cases, when the number is given we find it to be 
not more than seven. But the ascending and descending node, 
Rahu, Ketu, are mentioned together with the sun and moon and 
the planets Saturn, Mars, Jupiter with Venus, Mercury, in xiii. 
17. 38, if we may trust the scholiast, who says that grahapati 
is for Mangala (Mars) and vara is for Brhaspati and Sukra; 
atri being for Budha and Sani (as elsewhere) for Sanaiscara. 
They are forms of God, who is the nidhi, highest number, the 
thousand-eyed soma, the naksatrasddhaka, and 

candrah suryah sanih ketur graho grahapatir varah, 
Atrih, etc. 

While it is doubtful whether the scholiast is right in this case, 
another late passage expressly reckons the grahas as nine, iv. 
2.21: 

yam manye dvddasam rudram dditydndm trayodasam 
vasunam navamam manye grahdndm dasanam tathd, 

where the tenth graha implies nine others. The sun, though 
sometimes not a graha, is expressly called a graha, and is lord 
of grahas, suryo grahdndm adhipo naksatrdnam ca candra- 
mdh, xiv. 43. 6. At iii. 200. 85, are mentioned grahdh suryd- 
dayo divi, "the grahas beginning with the sun," all being reck- 
oned as ddrunah or Sivdh, unfavorable or favorable, according 



38 E. W. Hopkins, [1903. 

to circumstances. Conversely, the planet Saturn is the son of 
the sun and Jupiter is reckoned among the suns, adityesv eva 
ganyate, in i. 66. 39, as Sukra (Venus) is Bhrgu's son, a planet 
meteorologically active, varsdvarse bhaydbhaye, ib. 42, though 
regarded also as the female side of Brhaspati (Venus and 
Jupiter) . 

Sporadic mention of the planets (five in number in vi. 100. 
37, grahah panca, opposed to the sun; and to the moon, ib. 38) 
is common enough. In vi. 101. 59, two heroes are compared to 
Mercury and Venus; and ib. 104. 21, to Mercury and Saturn. 
The last is named also in ix. 16. 10, (rane) candramaso 'bhydse 
sanaiScara iva grahah, Saturn near the moon illustrating &alya 
near Yudhisthira in battle. Such references are found not infre- 
quently: angdraka-budhdv iva, two heroes, " appeared like Mars 
and Mercury," viii. 15. 16; naksatram abhito vyomni sukrd- 
ngirasayor iva (yuddham), a battle such "as in heaven is the 
battle of Venus and Jupiter respecting an asterism," ib. 17. 1; 
vakrdtivakragamandd angaraka iva grahah, a hero storms 
about "like the planet Mars returning in his orbit," ib. 19. 1. 

In the same book, in which occur most of these allusions, viii. 
18. 5, another reference has a pun on the word graha, the 
"seizer:" ' 

sa Mdghaddndm pravaro 'nkusagrahe 
grahe i prasahyo vikaco yathd grahah, 

where vikaca is "the headless one," Rahu, who is here a 
" seizer," but is not grouped with the planets. 

Bad signs are given by the planets. In vi. 3. 12 f., the 
"white graha,'''' Ketu, passing Citra stands still; a great graha, 
a comet, dhumaketu, attacks Pusya (the warrior's asterism) ; 
Mars, angaraka, turns among the Maghas, maghdsu v-akrah ; 
and Jupiter turns in Sravana; the sun's son, Saturn, attacks 
Purva PhalgunI, bhagaih naksatram; Venus shines in Purva 
Prosthapada and going about in TJttara associated (with an upa- 
graha) desires attack ; the white graha attacking Jyestha, din- 
dram naksatram, stands still. The Pole-star (? dhruvd) flames; 

1 Compare S^B. iv. 6. 5. If., where the sun is a graha and " the whole 
Brahmana is a play on the word graha," as Eggeling says (cf. ib. xiv. 1. 
4. 2). 



Vol. xxiv.] Epic Chronology. 39 

sun and moon distress Rohini ; between Citra and Svati is the 
fierce graha; and Mars, lohitanga, turns about Sravana, called 
bere (so N.) Brahma-rasi. So in v. 143. 8 f., and viii. 94. 
49 f., the " son of the moon," Mercury, -goes transversely, being 
fire-colored; while Jupiter, encircling Rohinl, becomes moon- 
colored. In ix. 11. 17, Venus and Mars with Mercury go behind 
the Pandus, portending the fate of their foes. Here Venus is 
" Bhrgu's son" and Mars is the " son of earth," dhardputra, 
while Mercury is again the "son of the moon," sasija. When 
the moon is upside down and the planets encircle the sun to the 
left, something terrible will happen : apasavyaih grahas cakrur 
alaksmanam divakaram, avaksiras ca bhag avail upatisthata 
candramah, vi. 112. 12. 

There is no passage in the epic which gives the Greek order 
of the planets, Sun, Moon, Mars, Mercury, Jupiter, Venus, Sat- 
urn, and this fact, according to Professor Jacobi, ZDMG. xxx, 
p. 307, would imply that the date of composition was not later 
than the third century A.D. This would be valid for the 
greater part of the epic, but would not exclude the possibility 
of still later additions having nothing to do with planets. 
The rare mention of the group of nine grahas, confined to what 
we may unhesitatingly call the later part of the epic, shows that 
the grahas in the early epic were reckoned only as five, or seven 
with the sun and moon added, without recognition of the nodes, 
Ketu and Rahu, as part of the "group of nine planets." 

That the heavenly bodies are sentient creatures needs scarcely 
be observed., The planets have merit, puny a, after losing 
which, kslnapunydh, they fall (as shooting stars) ; and they 
become sad when they see distressing sights. Thus in i. 210. 
26: "moon and sun, the grahas, the stars, the asterisms, (all) 
the inhabitants of the sky, beheld the deed, became despondent." 
For varied views on this point, see my Great Epic, p. 380. The 
formal adoration of planets in connection with Ganesa is recom- 
mended in Yaj. i. 292 f., who recognizes nine grahas, sun, 
moon, son of earth, son of moon, Brhaspati, Sukra, Sanaiscara, 
Rahu, Ketu. Their worship consists in making metal or pic- 
tured models, to which are given clothes, flowers, incense, rice, 
etc. ; and to each planet eight hundred and twenty-eight pieces 
of a special kind of wood (burned for each) ; rites later than 
epic ideas, to judge from silence on a theme so attractive. 



40 E. W. Hopkins, [1903. 

SUN, YEARS, AGES, AEONS, AND CYCLES. 

The extended astronomical phraseology of the late epic may 
he illustrated hy the gift-laud in iii. 200. 125 f., where occurs 
the advanced technicality implied in the word sadasltimukha. 
After stating that a gift at the time of the new moon or the 
full moon has a double value and that a gift at a season-time 
would he of ten-fold value, parvasu dvigunam ddnam rtau 
dasagunatn bhavet, the poet adds that the reward would be 
endless if one gave gifts at the equinox, visuva (tulamesasam- 
krdntyoh, N.), when the sun is sadasitimukhesu, that is (the 
sun's path being divided into arcs of 86°, commencing with the 
autumnal equinox, JAOS. vi. p. 410 f.), entering Gemini, 
Virgo, and Pisces (mithunakanydminasarhkrdntisu, N.), or at 
the eclipse, upardge (eclipsed is upapluta, passim), of the moon 
and sun. This information is conveyed in slokas. Part of it 
is then repeated, amplified, and embellished by being stated 
over again more artistically :' 

rtusu dasagunam vadanti dattam 

satagunam rtvayanddisu dhruvam 

bhavati sahasragunam dinasya JRdhor 

visuvati ca 'ksayam asnute phalam. 

At the same time, though one or two other passages (see above, 
p. 32) point to the recognition of the solar zodiac, I cannot 
believe that this was known in the epic period ; for in that case 
there would inevitably have been references to some samkrdnti, 
which term, however, does not once appear. 

Time is often measured by groups of five divisions, either as 
"months, seasons, semesters, years, ages," as in xii. 47. 66, or 
without ages and with day and night, of which the world is 
made : ahordtramaye loke' . . . mrtyur grasati bhutdni pavanam 
pannage yathd, xii. 299. 29. It' is this latter group which 
Nilakantha thinks has given to Visnu, in xii. 339 (66), his title 
of Pancakalakartrpati, "lord of the five makers of time ;" though 
as Pancaratrika follows and as Nilakantha also gives the five as 
those of Gita, 18. 15, the real application of the epithet remains 
doubtful. In xiii. 149. 60, Visnu is naksatranemir naksatrl. 

1 For the irregular form of the meter, see my Great Epic, p. 344. 

2 Compare the Anugita, xiv. 45. 2 f ., where the " wheel of time " turns 
on day and night, ahordtraparik?epam, but is counted, gaqitam, by 
months and half -months. 



Vol. xxiv.] Epic Chronology. 41 

The period of day and night, the months, and the seasons 
having been discussed, there remain the year and greater periods 
of time. The year {prativatsara, ii. 74. 19, and Sdrada, iii. 
99. 24, are unique; other designations are those current, varsa, 
parivatsara, etc.) is divided most frequently (as in RV. i. 164. 
48, etc.) in metaphors. In the "forest of the great world," 
kdntdra, there is "a black and white elephant, having six faces, 
twelve feet," the year with its dark and light halves of the 
month, seasons, and months, xi. 5. 15; 6. 11. Two men dance 
hand in hand and six men play with golden dice (day and night 
and the six seasons), xiii. 42 and 43. 4f. The whole year is 
measured by the "twelve-fold sun," which is spoken of as be- 
coming "twelve suns," dvddaiddityatdm gatah, iii. 3. 59, 
dvddaidtman, ib. 26; dvddasd 'dityan kathayantl ''ha dhlrah, 
iii. 134. 19. This is God's form at the destruction of the uni- 
verse: "as twelve suns" he destroys, xii. 313. 4. Again, the 
year is a wheel of twelve spokes, turned by six boys, while two 
girls weave black and white threads ; this wheel, however, also 
having three hundred and sixty spokes and twenty-four divisions, 
parvayoga, i. 3. 146 (also xii. 246. 32). Compare iii. 133. 24 f. : 

trinsakadv ddasdnsasya caturv iniat iparv anah 
yah trisastiiatarasya vedd , rtham sa parah kavih 
eaturvinsatiparva . . . sanndbhi dvddasapradhi 
tat trisastisatdram vdi cakram, 

which adds the "group of thirty," trinsaka, as one of the 
divisions, twelve months of thirty days each. In xiii. 159. 23, 
the year as the wheel of time has three naves, seven steeds, and 
three divisions, trinabhi, saptdsvayuktam, tridhdma ; the first 
implying the periods of cold, heat, rain; the last, rain, wind, 
heat; according to the scholiast (compare RV. i. 164. 2). In 
i. 3. 58, the weaving of the year appears again, but a new meta- 
phor follows, that of three hundred and sixty cows having one 
calf, ib. 60; and ib. 61 the wheel again has seven hundred and 
twenty spokes. The wheel of time is analyzed in ii. 11. 37, as 
having divisions of ksanas, lavas, muhurtas, day and night, half 
months, months, seasons (six), years, the cycle of five years, 
pancayuga, and the "four-fold day and night," ahordtras 
caturvidhah, that is, as they belong to men, Manes (whose day 
is a moon-month), to gods (measured by years), and to Brah- 



42 M W. Hopkins, [1903. 

man (measured by ages). This is the sole passage in the epic 
recognizing the five-year cycle by that name (Brah. pancakam 
yugam). It may be inferred from the attempt made to bring 
the solar and lunar year into line in iv. 52. 3, 

pancame pancame varse dvdu mdsdv upajdyatah, 

where, in Vedic phrase, RV. i. 25. 8, there is a calculation of 
the months "born after;" though here an estimate is made of 
the difference in the course of thirteen years between the lunar 
and solar years. The epic here uses the technical term, abhyadh- 
ika, and says that in this period five months and twelve days 
would be in excess, 

esdm abhyadhikd masah panca ca dvddasa ksapah 
trayodasdnam varsdndm. 

The difference is caused by " excess of time," kaldtirekena, 
and "the transit of luminaries," jyotisdm ca vyatikramdt. 1 In 
this passage the "wheel of time" suffers "partition" into the 
elements already mentioned, kald, kasthd, muhurta, dina, in- 
cluding asterisms and grahas as time-recorders, after fortnight, 
month, season, and year. 

Two other passages may possibly refer to the five-year cycle 
by implication and suggestion. One of these is that containing 
the pseudo-epic name of Visnu, Yatsara, xiii. 149. 63, as this 
is the name of the year of a cycle ; and the other is the passage, 
i. 124. 22, likening the (group of) five Pandus to years : anu- 
samvatsaram jdtdh . . Pdnduputra vydrdjanta panca sarhvat- 
sard iva, "like (the group of) five years." The sixth-year 
intercalated month of thirty-six days (SB. ix. 1. 1. 43 ; x. 5. 4. 
12) is not recognized in the epic. 

After the time-table given above, p. 13, which is virtually 
that of the later first book of Manu and of the Puranas, the epic 
poet, like the law-giver, continues with an account of greater 
periods of time. The sun, as is often said, is the chief divider of 
time. God is "the fruit in the acts accomplished in the moments 
and other (time-divisions) of the sharp-rayed sun," xiii. 14. 419. 
The sun divides the day and night into work-time and sleep- 
time. In the day and night of the Manes, the bright fortnight 



1 In five years there are sixty days over ; in thirteen, one hundred and 
fifty-six days, five (lunar) months and twelve days (156— 12=144-s-5 = 28f). 



Vol. xxiv.] Epic Chronology. 43 

is their day and the dark is their night ! Manu inverts the 
order, 1. 66; hut here, krsnah svapndya sarvari, etc., there is 
a confusion of men and Manes. A year of man is a day and 
night of the gods. The northern course of the sun in their 
day; the southern, their night. Their are four ages, Krta, of 
four thousand years, etc., as in Manu, with the "twilight 
periods of just as many hundreds;" each later age losing a quar- 
ter, ekapddena hiyante, in thousands and hundreds. In vi. 10. 
3, the name of the fourth age, Kali, is Tisya. The length of 
the ages is stated again, as just given, in iii. 188. 22 f., hut 
without aeonic speculations added. The moral qualities of each 
age are often described; at length, for example, in iii. 149. 
According to the usual later view, the Kali age begins with the 
death of Krsna; but according to v. 142. 8 f., at the very 
beginning of the great war, though probably the moral and not 
the chronological side is emphasized in Krsna's repeated words, 
na tadd bhavitd tretd na krtam dvdparam na ca. That " the 
Raj makes the age " is an epic truism that discounts all chro- 
nology. As to how the ages got their twilights, see Mr. Aiyer's 
Chronology of Ancient India, p. 129, where it is shown that 
one-tenth of the age makes the twilight, as the twilight of 
a day is one-tenth of a day of twelve hours, measuring 3 
ghatikds, 1 h. 12 m. 

The sum of the thousands and hundreds (to continue the time- 
table already cited) is twelve thousand (years). Both seers and 
mathematicians, mmkhydvidah, recognize this age, yuga, of 
twelve thousand (years); and one thousand such ages (12,000,- 
000 human years) are equal to a day of Brahman, whose night is of 
the same length, sahasrayugaparyantam, xii. 232. 15 f. Noth- 
ing is said here of divine Yugas. In xii. 343. 3, the period of 
creation lasts till the end of a thousand caturyugas. The day of 
Brahman is again recognized as a thousand Yugas in iii. 3. 55 
and vi. 32, 17, with no intimation that the Yuga is other than 
that of the twelve thousand human-year Yuga. The divyam 
varsasahasram is a commonplace in tales, as in the account at 
iii. 173. 7 of Daitya austerities. According to xii. 227. 70, the 
(Vedic) gods live only a thousand (divine) years, varsasahasrd- 
ntam. 

The aeons, Kalpas, mark a greater period. At the end of a 
Kalpa the creative eighth of God changes, parivartate, xii. 281. 



44 E. W. Hopkins, [1903. 

63. The Kalpa is thus one day of the creator-god (1000 X 12000 
years) and forms a new unit. It is in such units that the day 
of Brahman is reckoned in the later epic, xii. 312. If. The day 
of the TJnmanifest is "twice five thousand Kalpas," paflca 
kalpasahasrani dvigunany ahar ucyate, and his night is the 
same. He creates, when he wakes, the demiurge creator, Brah- 
man, and the latter's day is the same length less a pada, dasa 
kalpasahasrani padondny ahar ucyate. Thus Brahman's day 
is now reckoned as one quarter less than that of the TJnmanifest, 
or as seven thousand five hundred Kalpas, " and his night is of 
the same extent." 1 

According to Manu, 1. 71 f., the total of four ages, consist- 
ing in all of twelve thousand (human) years, is one age of the 
gods, and a day of Brahman is a thousand such divine ages, 
the expressions being caturyugam (dvadasasdhasram) as deva- 
nam yugam, and ddivikanarh yuganam sahasrdm as brahman 
ekam ahah. Now in xii. 208. 9, Soma is represented as a 
paryupasita, ascetic reverer (?), during a thousand divine ages, 
sahasram divydndm yuganam, which should be a day of Brah- 
man ; and in xii. 328. 24, Mahadeva stands, as an ascetic, on one 
foot during a divyam varsasahasram, or thousand years divine 
(of the gods). But in xii. 303. 14, as in the Kalpa enumera- 
tion above, the day of Brahman is reckoned not in ages, Tugas, 
but in aeons, Kalpas, albeit not of the same sort: 

yugam dvadasasdhasram kalpam vidhi caturyugam 
dasakalpasatdvrttam ahas tad brdhmam ucyate, 

"know that twelve thousand (years) are an age; a total of four 
ages, an aeon; the day of Brahman is said to be ten hundred 
times an aeon." Nilakantha interprets the years and ages as 
divine, and the Kalpa as a thousand caturyugas, the Kalpa thus 
being a day of Brahman. This certainly cannot be extracted 
from the text, though it is the orthodox view. Nilakantha at 
this place reckons out the year of Brahman in divine days and 
years, which gives the usual Puranic creative period. But a 
thousand caturyugas being the usual epic view of a day of 
Brahman, the text above is uncombinable with other epic data, 



1 The third creation (Ego-creation) and sense-creation have, respect- 
ively, days of five and three thousand Kalpas. ib. 11 and 15. 



Vol. xxiv.] Epic Chronology. 45 

and must be twisted out of its real meaning to be in accord 
with them. 

Two passages give the duration of a single spirit's reincarna- 
tions, in which the jiva is supposed to wander. In xii. 304. 44 ; 
305. 1 (continuation of the passage above) the spirit passes 
through sargakotisahasrani or thousands of crores of crea- 
tions; and in xii. 281. 36 and 43, every jiva, in a Buddhistic 
passage, passes through fourteen hundred thousand courses on 
its way to perfection; eight hundred periods of samhara- 
viksepa in man's estate alone. 

Only the later epic knows the Maha-kalpas by name. Thor- 
oughly Puranic are the passages; xii. 337. 1, tato Hite maha- 
kalpe ; 340, 115, rnahdkalpasahasrdni mahdkalpasatdni ca 
samatitani rdjendra sargai ca pralayds ca ha. So in xiii. 107. 
77, phalam pddmaiataprakhyam mahdkalparh dasadhikam • 
but never any such allusion in the real epic. 

After the specimen of time-measures given in my Great Epic, 
p. 206, I need offer no further examples of the epic's tendency 
to count time by "oceans," sdgara, and other similar terms, 
padma, pataka, ianku, nidhi, etc. They will be found, years 
rolled up to countless billions, in xiii. 107. 21 f., to select only 
one short passage from the numerous instances afforded by these 
arithmetical jugglers. As the poets come to consider the extent 
of time in aeons, creations, visarga and samhara, the imagina- 
tion is stretched to its utmost to devise parallels illustrative of 
the periods. A particle of sand removed daily from the Hima- 
laya till all the mountain is reduced to the plain; a drop of 
water daily drawn from thousands of league-long mile-deep 
lakes till all are drained ; such are the images that describe these 
(pseudo-epic) creations. One will suffice : xii. 281. 30 f. : 

samhdraviksepasahasrakotis 

tisthanti jivdh pracaranti ca 'wye 1 
prajdvisargasya ca pdrimdnyam 

vapisahasrani bahuni, daitya, 
vapyah punar yojanavistrtds tali 

krosarh ca gambhtrataya, ''vagadhdh 

1 The inanimate and animate world. 



46 E. W. Hopkins, [1908. 

dydmatah pancaiatds ca sarvdh 

pratyekaio yojanatah pravrddhdh 

vdpydjalam ksipyati vdlakotyd 1 tv 

ahnd sakre ca , py atha na dvitiyam 

tasam ksaye viddhi param visargam 

samhdram ekam ca tathd prajdndrh. 

The passage cited above, p. 32, from iii. 190. 90, which speaks 
of sun, moon, Jupiter and Tisya as being together, implies the 
recognition of the sixty-year Brhaspati cycle, as the sun, moon, 
and Brhaspati are in Pusya once only in this cycle. 2 

The doctrine of Manvantaras is implied (according to the 
commentator) in Bali's prophecy at xii. 225. 31. When the sun 
shining in the meridian, madhyarhdine, ceases to shine from all 
directions, then there will be war again between the gods and 
demons. Indra repudiates the idea, however, saying that the 
sun will never depart from his ordained course. But the 
Manvantara-theory may well be implied here, for, though for- 
eign to the early epic, the Manvantaras, not only in their earlier 
form but even in their later fourteen-Manu form, are known to 
the pseudo-epic. In the early epic, only the Mihira hymn, a 
late intrusion, recognizes these periods of time. Here, iii. 3. 
55-56, a day of Brahman is defined (as above) and the Manus 
and Manvantaras are referred to. The periods are referred to 
next in xii. 59. 115 and the first group of Manus is implied at 
xii. 285. 1, Vdivasvate l ntare, which phase appears again in 
337. 56, " when the Treta Yuga shall have replaced Krta in the 
(Manv)antare of Vivasvat." From here on, the later epic is 
full of allusions to the Manvantaras: purve ca manvantare 
Svdyambhuve, 343. 26; Manoh svdyambhuve 'ntare, 350. 42; 
manvantaresu, ib. 43; Manu Svarocisa, 349. 36 f . ; Sandisca- 
rah (Saturn) suryaputro bhavisyati Manur mahdn, tasmin 
manvantare cdi '«)« Mativddiganapurvakah (tvam eva bhavitd, 
vatsa), 350. 55. In xiii. 14. 38-39, for the first time in the 
epic, the Manvantaras appear (as protectors, in connection with 
the wives, maids, and mother of the gods) along with seasons, 

1 The water of the lake as flung out drop by drop with " the end of a 
hair " seems to be a play on the chronological meaning of Jco{i, the ' ' end " 
of arithmetic thought. 

2 Compare Aiyer, Chronology, p. 183. 



Vol. xxiv.] Epic Chronology, 47 

years, ksanas, lavas, muhurtas, nimesas and Yuga-changes. 
Finally, a Manu of the second group (of seven) appears in xiii. 
18. 43, Savarnasya Manoh svarge saptarsis ca bhavisyati. 

The real epic knows nothing of Great Kalpas and secondary 
Manvantaras. They belong to the Puranic period including 
the later epic, but even the first (Puranic) chapter of Manu 
falls short of the extravagance reflected in Santi and Anusasana. 
In regard to the cogency of this relation used as an argument 
historically, it does not appear to me that the remark of M. 
Barth, Btdletin, 1902, p. 30, poses the question properly. 
Here, in a resume of a similar argument, M. Barth says: " ces 
mentions sont rares dans les portions narratives, frequentes dans 
les parties didactiques, ce que, a premiere vue, parait assez 
naturel." The light sarcasm would be justified if the preceding 
words gave the whole situation, but they do not. The narra- 
tive portions of the poem are not quite sundered from didactic 
material, and the point is that such didactic material, though 
treating of the same matter, treats it in less modern fashion ; 
whereas the treatment of Santi and Anusasana is, in contrast, 
rather that of the later Puranas. Fear of being thought "ex- 
pert in cutting up the poem " need deter no one from the admis- 
sion that epic chronology represents an earlier point of view in 
the early books, and the Puranic point of view in what I call 
the pseudo-epic. Kay, rather, one might ask, is he an expert 
historian who thinks that such a difference of view is quite with- 
out historical significance ? 

EXCURSUS. ANALYSIS OF EPIC DATES. 

The Pandus were born a year apart, i. 124. 22, and when 
they first went to Pancala they were all proficient in the use of 
arms. The youngest must therefore have been at least 16 years 
old, and Arjuna 17 at this time, when they ravaged Drupada's 
kingdom, i. 61. 31-35; 135-138 (Karna made king of Anga, 
136). A year after this, Yudhisthira was installed, 139. 1, and 
Drupada sought a son to avenge his overthrow, 167. 14 (Arjuna 
now 18). Some time was spent in conquering the world, 139, 
but, regarding this as a matter of a few weeks (!), the next 
stage is marked by the expedition to Varanavata, where the 
Pandus spent a year (148. 1, Arjuna was now 19), before they 
traversed the woods and, after spending quite a long time, cira- 



48 E. W. Hopkins, [1903. 

ratrositdh, 168. 3, at Ekacakra, went to Pancala again, 168. 11. 
Here they remained one year, parisamvatsarositah, 1. 61. 31, 
after getting Draupadi (she was born grown up, like her twin 
brother), and Arjuna must then have been 20. After this they 
visited Hastina, went to Khandava, and remained there "many 
years," samvatsaraganan bahun, before Arjuna was exiled, 
i. 61. 35. 

Arjuna on being exiled "lived a whole year and one month 
in the wood " and then sought out Krsna at DvaravatI and took 
Subhadrfi, i. 61. 42, 

sa vai samvatsaram, purnam masam cai 'karh vane vasan 
tato '■gacchad Dhrslkesam, etc. 

Then, after the burning of Khandava and the gambling at 
Hastina, the Pandus were exiled for 13 years; and the war 
began on the 14th year, i. 61. 50, etc. 

If the " many years " at Khandava be reckoned only as three, 
Arjuna would be 23 when exiled for circa two years, aet. 25, 
returning 4 years before the banishment of all the Pandus, at 
the beginning of which he would be 29, and at the end of the 
13 years, when his son was 16, Arjuna himself would be 42. 
Part of the years reckoned by Abhimanyu's age is included in 
the 1 year 2 months of the building of the Sabha, ii. 3. 37, 
rnasaih paricattordasaih. 

But discrepancies occur. There is no record of a year spent 
with Drupada in the full account of the wedding. On the con- 
trary, i. 199-207 imply that the Pandus return to Hastina soon 
after the wedding. Nor is it consistent that the Pandus, who 
have already ravaged Drupada's kingdom, should regard it as 
apurvadrsta on their second journey thither, i. 138 and 168. 6. 
In i. 141, moreover, Yudhisthira is clearly not yet installed as 
heir-apparent, although he had already been installed in i. 139. 
1, a year after Drupada's defeat. Even the year spent in Var- 
anavata (i. 149. 1) seems in i. 146 and 147 to be regarded as a 
term of a month or a few days. There is no inconsistency in 
the timeless birth and growth of Bhima's son, for it is ex- 
pressly declared to be such; though the period of wandering, 
i. 156, should occupy some reasonable time omitted in the 
account above. But the difference between the clear statement 
of the first book, that Arjuna lived a year and a month in the 



Vol. xxiv.] Epic Chronology. 49 

wood and then raped Subhadra, and the subsequent prolonga- 
tion of Arjuna's exile to twelve years i. 212-221, three years of 
which are spent at Manipur, i. 215. 26, and one each in Dvaraka 
and Puskara, 231, is important for the critique of the epic. 
Draupadl's five children were born "at intervals of a year," 
ekavarsdntardh, i. 221. 66, 78, 86, but though Arjuna's son was 
necessarily born nearly a year after his return, and was a 
young warrior in the great war, his age at death is not specified. 
There is a difficulty, however, even here, for Draupadl's sons 
born one and two years later are also young fighters and the 
youngest would be but 14, whereas Abhimanyu at 16 is always 
celebrated as the youthful warrior par excellence, and 16 is the 
youngest age at which boys were considered equal to war. 

Another discrepancy which gives a comical effect is found in 
iii. 33. 12, where Abhimanyu, who should now be about four, 
is solemnly said to be one of those who did not approve of 
Yudhisthira's life! At this time the Pandus had lived 13 
months in the wood. In iii. 36-37 they still live " some time " 
before Arjuna starts on his trip. Five years Arjuna passed in 
heaven, iii. 44. 5; 141. 7; 164. 17, while the Pandus waited 
five years for him, iii. 50. 12. In iii. 158. 3, the time of wan- 
dering is now four years, Arjuna is to be expected about the 
fifth, pancanlm abhitah samdm, and in iii. 165-174. 9 the five 
years end with Arjuna's return. In iii. 176. 5-8 they "lived 
six years before and four years with Arjuna," ten samah alto- 
gether, and it is now the eleventh year in the wood, ekddasam 
varsam idam vasdmah. After this they lived in Visakhayupa 
forest one year, 177. 17 (with some preliminary marching), and 
in 177. 20, the somewhat belated twelfth year arrives, dvddasam 
varsam upopaydtam. The twelve years in the forest are re- 
ferred to again in 183. 39 ; 239. 18 as not yet over. In iii. 243. 
15, Yudhisthira cannot rescue Duryodhana personally because of 
his vow, Tcratu, but he urges his brother to do this ! In this 
(twelfth) year, Karna (already the king of Afiga) conquers the 
Afigas (and Drupada), and all the north, east, west, and south 
country (254. 19), including the Yavanas and other foreigners 
all "in a short time," Tcdlena nd Hidlrghena, 254. 33. 

A year and eight months now elapse (after the twelfth year 
has begun) and the Pandus are still in the wood, eating deer 
(after the Ghosayatra), sdstamdsam hi no varsam yad enan 

VOL. XXIV. 4 



50 E. W. Hopkins, [1903. 

upayuksmahe (eating the deer here), iii. 258. 12 (after the lib- 
eration of Duryodhana in 246 ; here Yudhisthira has a conver- 
sation with Duryodhana after rescuing him). 

By this time about fourteen years must have passed since the 
Pandus were banished, but in iii. 259 the narrative reverts with 
the opening statement that while they dwelt miserably in the 
woods " eleven years passed away," as if the twelfth had not 
yet come. The thirteenth year, however, finally comes, iii. 
315. 5, sesam varsam trayodaiam, the completion of the 
twelfth being announced in 310. 41, "what did they do when 
the twelfth year was over ?" 

In the fourth book occurs the most glaring inconsistency in 
the poem. The bow of Arjuna is here said to have been 
already carried for 65 years by that hero (who is now 40 or 50 
years old, according to the contradictory data already furnished), 
iv. 43. 6. The 13 years of waiting are paralleled by the 13 days 
which DraupadI begs to be allowed still to remain in the town, 
24. 29, the agreement and the 13th year expiring, iv. 31. 2 and 
4, on the Trigarta expedition, though in 47. 4 the 13th year is 
not yet over, vartate tu trayodaiam, in Duryodhana's opinion. 
But it becomes now a question of years reckoned as lunar or 
solar, five months and twelve days being the difference (see 
above, p. 42). In 26. 3, Duryodhana says that most of the 
time is past and very little remains, alpdvaSistarh kdlasya gata- 
bhuyistham antatah; in 21. 17, only a month and a half remain. 
In 48. 5, Arjuna is represented as having been samahitah (and 
therefore out of practice in fighting) for 13 years, varsdny 
astdu ca panca ca ; in iv. 49. 6-8, he is said to have learned 
arms from Sakra for 5 years, as he practiced brahmacarya for 5 
years and then stole DraupadI. In iv. 49. 18 Arjuna is said by 
Krpa to be freed, having been deceived, nikrtah, by the Kurus 
for 13 years, as in 62. 14. The exile in the wood for 12 years 
is alluded to in 60. 7. Finally, in iv. 72. 14, on the end of the 
13th year, trayodase varse nivrtte, being at Upaplavya, Abhima- 
nyu is married (cf. 23). 

In v. 1. 11-13, varsdni sat saptd ca trayodaSai cat 'va sudus- 
taro '■yarn, the 13th year is now ended. In v. 20. 9, the expres- 
sion vasitas ca maharanye varsani 'ha trayodasa seems to imply 
that the 13 years are passed in a forest ; but the sequence recog- 
nizes the Virata episode. In v. 21. 13, Kama says the 13 
years are not yet past, and puts it as if they still had to stay in 



Vol. xxiv.] Epic Chronology. 51 

the wood, yathdpratijnam kalam tarn carantu vanam dsritdh. 
Although Virata is recognized in v. 22, yet no notice is taken 
of the sojourn there in v. 26, where indeed it is said (25) that 
the Kurus live only because they have not yet heard Arjuna's 
bow, a curious statement in view of iv. 55, etc. The thirteen 
years (one incognito) are recognized, however, in v. 48. 92, 
and elsewhere (below). 

In v. 52. 10-11 occurs the following sloka, spoken just before 
the war, at the end of the thirteenth year (the subject is 
Arjuna) : 

trayastriniat samahuya khdndave i gnim atarpayat 
jagdya ca surdn sarvdn nd 'sya vidmah pardjayam. 

G has suta, vocative, for huya, and N. interprets " thirty -three 
years, samah, are past." But this is impossible. B's reading 
is evidently correct (N.'s comment fits only C!) and the 33 are 
not years but the gods challenged by Arjuna, samahuya being 
a common epic word in these circumstances. 

Another year appears to be added in v. 79. 19, where the 
cattle-lifting foray of Virata is alluded to as occurring samvat- 
saragate '■dhvani, "on the expedition of a year past " (N. gata- 
samvatsare ; cf. xi. 3. 16, samvatsaragata, "a year old"). But 
in v. 82. 40 it is still only 13 years that Draupadi has waited 
for revenge, and Prtha says in v. 90. 47, caturdasam idam 
varsam yan nd , paSyam (Drdupadlm) , so ib. si. 60 and 70; 
and in 129. 47, the queen says, alam afiga nikdro '■yam trayo- 
dasasamdh krtah. The ' ' fourteenth year " merely implies that 
the thirteenth is ended. 

An apparent discrepancy occurs at v. 141. 13. Karna was 
made king of Afiga at the tournament, which according to the 
narrative already given occurred several years before the gam- 
bling. Yet in this passage Karna declares that through his 
fidelity to the Kaurava prince he has enjoyed a kingdom " with- 
out thorns " for thirteen years, mayd trayodasa samd bhuktam 
rdjyam akantakam. This may be made to mean that his king- 
dom has been thornless only since the retirement of the Pandus ; 
but the natural interpretation is that the kingdom has been his 
only for the time mentioned, for Karna himself is reviewing 
his life and this is the only allusion in his speech to the kingdom 
given to him by Duryodhana. The explanation, however, lies 
rather in the assumption of a poetic lapsus, for the words are 



52 K W. Hopkins, [1903. 

almost identical with those employed by Duryodhana himself in 
v. 160. 110, where he says to Arjuna, trayodasa sama bhuktam 
rajyam vilapatas tava, and the thirteen years of kingly enjoy- 
ment on the part of Duryodhana are contrasted with the weep- 
ing of his foes (161. 8 and 28 repeat this in the herald's words). 

The exact time of the battle is given as to occur on the 
seventh day from the interview in v. 142. 17, at the time of the 
new moon. In Mr. Aiyer's little book, The Chronology of 
Ancient India, the statement in v. 83. 7, that Krsna set out on 
his mission "in Kartika, under the star Revati, at the end of 
autumn," is united with this, which in turn is interpreted to 
mean that the new moon will happen in Jyestha Naksatra (in 
seven days, emended by Mr. Aiyer to "ten"). It was, how- 
ever, under Pusya Naksatra that the Kuru army took the field, 
v. 150. 3, pusyo '■dya, as did the Pandus, ix. 35. 10 and 15, 
and the armies were prepared, according to vi. 17. 2, magha- 
visayagah somah under Magha. The asterism should be Citra 
(v. 143. 10; vi. 3. 12, 28, etc.). Eighteen days of battle are 
recognized, save in the interpolation of Balarama, whose jour- 
ney can be interpreted only to mean that the battle lasted forty- 
two days (below). 

During the battle, the only point to be noticed is the age of 
the combatants, the leaders Duryodhana and Yudhisthira being 
now over forty or fifty (as above). Arjuna, two years younger, 
is taruna and yuvan, Nakula is sukumdro yuvd surah, but 
Drona is eighty-five; vii. 12. 22; 83. 23; 110. 81; 125. 73; 
126. 39. Despite these epithets applied to the Pandus, which 
imply middle-age strength or even youthful delicacy, in vii. 
196. 44, Arjuna says that the short remnant of their days will 
be affected by Drona's unrighteous death, yadd gatarh vayo 
bhuyah sistam alpataram ca nah, " gone is the greater part of 
life, it is the lesser part remains to us," though vayah (cf. pra- 
vayah) may imply strength of life more than life (yet the con- 
clusion does not favor this, tasye 'danim vikdro 'yam adharmo 
'■yarn krto mahdn). Karna also is yuvan, viii. 8. 11. This is 
not middle-age, however, according to the antithesis of yduvana, 
madhya, vrddha, or yduvana, madhya, jara (see the citations, 
in the last article of the series in this Journal), nor can it be 
interpreted as fool's age as in x. 3. 11, for it is intended, as in 
the citations above, for a compliment. The time-term for fool 
is bdla, as in xi. 17. 20, applied to Duryodhana. 



Vol. xxiv.] Epic Chronology. 53 

After the battle comes the inconsistency of Balarama's expe- 
dition. He started out just before the battle and returned at 
its close, making forty-two days in all, from'Pusya to Sravana, 
catvdrinsad ahany adya dve ca, etc. , ix. 34. 6 ; 54. 12. 

It is quite impossible to reconcile this with the statements in 
regard to the length of the battle (eighteen days) found else- 
where in the epic. In ix. 35. 14, Balarama starts under Maitra 
Naksatra, i. e. Anuradha. Mr. Aiyer, op. cit. p. 101, emends by 
changing 42 to 24, catvarinsat to caturvinsaty , and Sravane 
to Mohinyam ; but this is merely a confession of inability to 
reconcile the conflicting statements except by changing the text 
completely. At the same place, Mr. Aiyer endeavors to recon- 
cile the appearance of the moon in the night battle in vii. 185 f . 
on the fourteenth day, with the previous account of the new 
moon. Mr. Aiyer's conclusion that the war ended on the 51st 
day before the winter solstice, and began on Oct. 14th, 1194 
B.C. (or that at least the war took place in the latter half of 
this year), does not depend altogether on the rectification of 
these obvious errors, but is based to some extent on the inter- 
pretation of the doubtful verse xiii. 168. 28 (27, "for 58 nights 
Bhisma lies on his couch"), as already explained. 

In xi. 17. 21, "he who has enjoyed undisputed royalty for 
13 years now lies dead," the fourteenth year implied in the early 
account is pointedly ignored, as it is elsewhere, notably in viii. 
68. 9, "there are now these 13 years in which we have lived in 
the hope of Arjuna," and the battle takes place immediately on 
the end of the 13th year. 

But as to the assumption that the Pandus were originally ban- 
ished for only twelve years and that the thirteenth year is a later 
addition, it must be proved by the content, style, and metrical 
form of Virata rather than by the discrepancies in the texts 
that refer to the years of banishment. I used to think that the 
thirteenth year was interpolated on the further ground that such 
discrepancies revealed a prior stage in which the thirteenth year 
was actually unrecognized, as in iii. 24. 2 ; v. 72. 9 ; but a care- 
ful survey of all the cases now leads me to the conclusion that 
this may be due merely to the poetic point of view. An 
example as good as any other is found in vii. 137. 47, in 
which a reference is make to the fire of rage lasting 13 years, 
and 197. 7, "the impatience of 13 years," as compared with 
ib. 145. 93, where "the sorrow of 12 years" is mentioned. 



54 M W. Hopkins, [1903. 

In one case the anger during the whole period, in the other the 
wretchedness during the life in the wood, is emphasized. So in 
viii. 9. 58, it is said that Yudhisthira did not sleep for thirteen 
years because of his fear of Kama, a statement repeated in 
different words in 66. 15 and again in xi. 21. 7, and this is pre- 
sented, in viii. 74. 47, as the grief acquired in thirteen years, 
duhkham trayodasasamdrjitam, whereas in viii. 11. 27 mention 
is made of the grief (arrow) of twelve years, salyo mama dvd- 
dasavdrsikah. Other references in this book are found in viii. 
91. 4, in which an extra year is recognized besides twelve in the 
wood, and 96. 45, "we shall sleep well to-day after being 
awake in sorrow for 13 years." The next book too recognizes 
only 13 years. In ix. 33. 4, which is repeated in 58. 19 with 
var. lee, this section repeating the substance of 33 after the 
Tirtha episode, a long interpolation (ch. 33-51), it is said that 
Duryodhana has been practicing on an iron statue of Bhima for 
13 years (this iron statue reappears in xi. 12. 15 f.). Also in 
xv. 4. 15 the thirteenth year is recognized. In the earlier books, 
the thirteenth year is recognized, besides passages already cited, 
in ii. 46. 11; 74. 18 f. -76; 77. 30; 80. 34; iii. 3. 74; 8. 3; 
46. 58; 49. 11; 51. 33 f. ; 176. 10 f . ; 252. 43; 256. 14; 261. 
50; v. 61. 19; 95. 41; 160. 89; all referring either to the thir- 
teenth year as being completed, or, what amounts to the 
same thing, to what will happen in the fourteenth year, after 
the thirteenth, e. g. ii. 77. 30 and iii. 261. 50. I think now, 
therefore, that the thirteenth year must be regarded as belong- 
ing to the original conception of the present poem and that the 
late characteristics of Virata are due to subsequent working- 
over of the delectable scenes embodied in it. Possibly the 
original form was simply an extra year "in concealment" 
(incognito). The time-discrepancy is of no more weight than 
in the application of thirteen years to the wood-life exclusively. 
This curious statement, that the Panda va& lived not only in 
banishment but in the wood for 13 years, is found three times, 
once as cited above, p. 50 ad fin., again in vii. 197. 10, 

vanam- pravrdjitas ca sma valkaldjinavdsasah 
anarhamdnds tam bhdvam trayodasasamdh pardih, 

" we were exiled by our enemies to the woods, clothed in bark 
and skins, undeserving of that condition, for thirteen years ;" 
and in xv. 11. 23, 



Vol. xxiv.] Epic Chronology. 55 

yatra trayodasasama vane vanyena jtvatha, 

(that condition) ' ' when for thirteen years you lived in the 
wood on forest products." But as the latter is easily explained 
as a phrase (also in the Ramayana, see my list of parallels, 
Great Epic, p. 433, No. 242), and is preceded by an explicit 
reference to the "secret residence" in contrast to the "twelve 
years hate," ib. 20, so in the former case, "that condition" 
carries the thought over to the end of the period during which 
the Pandus were treated badly for thirteen years. So also the 
fourteenth year in the wood, logically to be extracted from the 
narrative as sketched above, is probably merely a poetic lapsus. 
Abhimanyu is killed at 16 years of age, after having been mar- 
ried for six months, i. 67. 117, etc. ; xi. 20. 29. 

In the tenth and eleventh books the data carry us forward to 
the end of the Pandus. Krsna is slain on the 36th year after 
the war; xi. 25. 44, repeated in xvi. 1. 1. During this time 
the Pandus defer to Dhrtarastra for 15 years, and the latter, 
xv. 20. 32, lives three years more. It is 16 years after the war 
in xv. 29. 37, at which time Draupadi is "just about touching 
middle age" (!), xv. 25. 9. Two years more pass, xv. 37. 1, 
after more than a month's visit on Dhrtarastra, masah samadhi- 
Icah, xv. 36. 11. Three of the eighteen years after the war 
were passed by the old king in the wood and fifteen in town, 
xv. 39. 25. These form explicit denials of the fact (inferred 
from the circumstance that Pariksit was a baby at the time of 
the visit) that Pariksit was crowned about sixteen years after 
the war, as shown by Mr. Aiyer. The epic in this regard con- 
tradicts itself and can scarcely be taken as a safe guide for its 
own date as far as these data are concerned, x. 16. 7; xv. 15. 
10; 25. 10. Pariksit reigns 60 years, according to x. 16. 15, 
though in a final extravagance the epic declares that Yudhi- 
sthira's reign alone embraces "thousands of years," xv. 10. 22, 

tatha varsasahasrani kuntiputrena dhimata 
palyamana dhrtimata sukhath vindamahe nrpa. 

Altogether the epic is as fairly consistent in its dates as was 
to be expected of so huge a compilation. Some of the incon- 
sistencies, however, are so decided as to admit of no reasonable 
doubt that the poem has been largely interpolated. 



56 JEJ. W. Hopkins, Epic Chronology. [1903. 

SYNTACTICAL NOTE ON THE ABLATIVE OF TIME. 

It is not worth while to make a separate article out of this 
note, so I append it here, though scarcely in place. In the Am. 
Journal of Philology, xxiv, p. If., I have tried to show that, in 
epic and earlier Sanskrit, the ablative of time-words does not mean 
"after" but "up to" or "within" the time named, my general 
conclusion in respect of all the grammatical cases being that in 
early Sanskrit no grammatical case expresses temporal posteriority 
any more than it does temporal priority, though "time after" 
may be implied by any case (except the vocative), even by the 
nominative and accusative. The ablative in particular approxi- 
mates to a true indication of posteriority, yet only in serial time, 
measured from a starting-point. This note illustrates the use of 
the ablative in the later literature of the Brhat Samhita. Here is 
strikingly shown how such an expression as sanmasat regularly 
includes the period and does not mean "after a semester," but 
within it. Good examples are found in xxx, xxxii, and xlii. 
In xxx. 12 and 31, saptahat is "within a week;" in xxxii, "in 
four fortnights," and "in a week" are expressed by the instru- 
mental and by saptahat respectively; in xlii. 7, it is said that a 
rise in price will take place sasthe mast, varsardhat, and (sthi- 
tvd) mdsam, all alike giving the limit. So in xlvi. 14, 30, 39, 
53, ' time within which ' is expressed by instrumental and abla- 
tive, sanmasat, and so elsewhere. But in this later literature, 
BS. lxxviii. 20, appears (as noticed in my article) an innovation 
in tryahat samnivartate, in the apparent meaning "ceases after 
three days." BS. has a, murdhatah in the sense "from the 
head onward," lii. 10, as well as a varsat, " within a year," 
xlv. 16 (also the antam construction, dbrahmakitdntam, lxxiv. 
20). The nominative of duration, as illustrated in the article 
referred to above, has here a still more striking illustration : (yo 
l dydt), sdi 'Icani vinsatir (the reading of all MSS. ; see Kern's 
note) ahdni, lxxvi. 3: (whoso eats), "days a twenty (nom. !) 
and one."