गुप्त अभिलेख संग्रह (जे. एफ. फ्लीट - गुप्त सम्राटों एवं परवर्ती नरेशों के ऐतिहासिक पाषाण व ताम्रलेख)

THE GUPTA INSCRIPTIONS INTRODUCTION 127
whether the era was actually established by the Early Guptas themselves, we must refer
the rise of the Early Gupta power to somewhere about A.D. 319, instead of
placing the period of their supremacy anterior to that year, and their downfall
in it.
A few concluding words, however, seem necessary as to the exact years of the
Christian era, which represent respectively the epoch or year 0, and the commencement
or first current year, of the Gupta-Valabhî era.
Taking the years quoted in the records without qualification, as current years,
we have obtained the following results,¹—by the Êraṇ pillar inscription of Budhagupta,
Gupta-Saṁvat 165 current = A.D. 484-85 current;² by the Parivrâjaka grants, 156 current
= A.D. 475-76 current;³ 163 current = A.D. 482-83 current;⁴ 191 current = A.D. 510-11
current;⁵ and 209 current = A.D. 528-29 current;⁶ by the Nêpâl inscription of Mânadêva,
386 current = A.D. 705-706 current;⁷ and by the Verâwal inscription of Arjunadêva,
945 current = A.D. 1264-65 current.⁸ And all these equations give the uniform result of
Gupta-Valabhî-Saṁvat 0 = A.D. 319-20 current, or more precisely, by the Śaka
year, the period from the 9th March, A.D. 319, to the 25th February, A.D. 320;⁹ and
Gupta-Valabhî-Saṁvat 1 current = A.D. 320-21 current, or more precisely, by the
Śaka year, the period from the 26th February, A.D. 320, to the 15th March, A.D. 321.
The results obtained from the Kaira grant of the year 330 and the Verâwal inscription
of Valabhî-Saṁvat 927, differ slightly from the above, and are—Gupta-Valabhî-Saṁvat
330 current = A.D. 648-49 current;¹⁰ and Valabhî-Saṁvat 927 current = A.D. 1245-46
current.¹¹ In these two instances the difference is due to a local alteration of the true and
¹ I exclude the Bhumarâ pillar inscription, as proving nothing definite, because the Gupta year is
not given in it.
² Or, more precisely, by the Śaka year, as commencing with Chaitra śukla 1, the period (see
Indian Eras, p. 153) from the 14th March, A.D. 484, to the 2nd March, A.D. 485.—The dates given
in these notes are quoted as approximately correct; they may, or may not, be the exact dates.
³ Or, in the same way, the period from the 21st February, A.D. 475, to the 11th March, A.D. 476.
⁴ Or, in the same way, the period from the 6th March, A.D. 482, to the 22nd February, A.D. 483.
⁵ Or, in the same way, the period from the 25th February, A.D. 510, to the 15th March, A.D.
511.
⁶ Or, in the same way, the period from the 8th March, A.D. 528, to the 24th February, A.D. 529.
⁷ Or, in the same way, the period from the 1st March, A.D. 705, to the 20th March, A.D. 706.
⁸ Or, in the same way, the period from the 1st March, A.D. 1264, to the 19th March, A.D. 1265.
⁹ Here I owe the initial and ending dates, which it was desirable to have exactly, to Mr. Sh. B.
Dikshit.
¹⁰ Or, more precisely, by the southern Vikrama year, as commencing with Kârttika śukla 1, the
period (see C. Patell's Chronology, p. 122) from the 24th September, A.D. 648, to the 12th October,
A.D. 649.
¹¹ Or in the same way, the period from the 23rd October, A.D. 1245, to the 12th October, A.D.
1246.

128 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
original scheme of the Gupta year; made in such a way that each subsequent year com-
menced with the Kârttika śukla 1 immediately preceding the true commencement of the
year with Chaitra śukla 1. And for these two dates, and any that may be found
hereafter to belong to the same class, we have to apply the equations of Gupta-Valabhî-
Saṁvat 0 = A D. 318-319 current, or more precisely, by the southern Vikrama year, the
period¹ from the 12th October, A.D. 318, to the 30th September, A.D. 319; and Gupta-
Valabhî-Saṁvat 1 current = A.D. 319-20 current, or more precisely, by the southern
Vikrama year, the period from the 1st October, A.D. 319, to the 18th October, A.D. 320.
These two instances, however, are purely exceptional ones. And, in the case of all
dates in the era referable to the true and original scheme of its years, we have to apply the
epoch of A.D. 319-20; and to treat the years of the era as northern years, commencing
with Chaitra śukla 1.
The equation between the epoch of the Gupta-Valabhî era and the Christian era, is
not intrinsically dependent on any reference to the Śaka era; and it could be established
directly by European Tables. In this inquiry, however, it has been established through
results that have been worked out from Hindu Tables which are arranged for the Śaka
era according to expired years: and, in order to use those Tables, the given Gupta-Valabhî
years had to be converted into expired Śaka years. The process, however, has not converted
the given Gupta-Valabhî years themselves into expired years But what has been done
has simply been, first, by the addition of a uniform running difference, to obtain the current
Śaka year corresponding to each given current Gupta-Valabhî year; and then, in the usual
way, to take the immediately preceding Śaka year as the expired year that is required as
the basis of the calculation. Thus, the details of the date in the Êraṇ pillar inscription of
Budhagupta, which really belong to Gupta-Saṁvat 165 + 242 = Śaka-Saṁvat 407 cur-
rent, have been calculated with the basis of Śaka-Saṁvat 406 expired; and the details of
the other dates in the same way.
Now, in the case of an era used specially by astronomers for their technical pro-
cesses, as the Śaka era was, since we have to work with expired years, it is natural
enough that the Tables should be arranged accordingly. And possibly, after a certain
period, and in certain parts of India, we may have to interpret any given year of such an
era as an expired year, whether it is expressly denoted as such or not.² But the same
rule does not hold good in the case of eras that are not actually used for astronomical
processes, though they are quoted in connection with details fixed by such processes.
¹ Here, again, I owe the exact dates to Mr. Sh. B. Dikshit.
² A clear instance of this is furnished by the date in the Dêôgaḍh inscription of Bhôjadêva (see
page 107 above, note 1). By the literal rules of translation, the given Śaka year, 784, has to be
interpreted as a current year; but, for the calculation, it has to be applied as an expired year.

[Header] THE GUPTA INSCRIPTIONS; INTRODUCTION. 125

adding, according to his most explicit statement, two hundred and forty-one years to
the Gupta-Valabhî dates. This fixes the starting-point of the era, approximately, as
having occurred when Śaka-Saṁvat 241 had expired, and, by the epoch of the well-known
Śaka era, when A.D. 319-20 was current; leaving only the determination of the exact
epoch by the calculation of recorded dates. And, as regards a special point in his state-
ments, of extreme importance, with the help of Prof. Wright, we have now obtained a
translation which,—if it does not actually mean only that the Early Gupta kings had
exercised so powerful a sway that, even when their dynasty came to an end, the era that
had been used by them still continued in use,—is yet fully capable of that interpretation.
At any rate, this translation frees us from the obligation under which we lay, by reason
of M. Reinaud's rendering of the same passage, of connecting the establishment of an era
with the extermination of the dynasty, and of placing the period of the Early Gupta
supremacy anterior to A.D. 319, and the termination of it in that year. And the most that
can be said against it, is, that it is the literal rendering of an ambiguous original, the real
meaning of which must be determined by extraneous considerations.
The Mandasôr inscription of Mâlava-Saṁvat 529 expired shews that we must look
to somewhere about A.D. 319 for the starting-point of the era in which are recorded the
dynastic dates of Kumâragupta and the other kings of the Early Gupta dynasty, and
any others that are to be referred to the same uniform series with them.
The dates in the records of the Early Guptas themselves, as far as the time of
Skandagupta, do not afford details for computation. But, clearly belonging to the same
uniform series of years, is the date contained in the Êraṇ pillar inscription of Budha-
gupta. And converting this date into a Śaka date, in the manner indicated by Albêrûnî,
we have found that, the resulting Śaka year being taken as an expired year, the details
work out quite correctly.
Such dates, also, are those contained in the records of the Parivrâjaka Mahâ-
râjas; which, moreover, include a specific declaration that the Gupta sovereignty was then
still continuing. And, calculating them in precisely the manner that is indicated by the
results for the date in the Êraṇ pillar inscription, we have obtained equally correct and
uniform results. Also, the latest of these records, the Khôh grant of the Mahârâja
Saṁkshôbha, No. 25, page 112, shews that the Gupta sovereignty continued for at least
two hundred and nine years. And this fact is amply sufficient to explain why,—whatever
may have been its historical origin,—the era used in all these records should eventually
come to be popularly known as the Gupta era.
Such dates, again, are those contained in the inscriptions of Śivadêva I. and
Mânadêva of Nêpâl. And, that the first of them is recorded in the era in question, is
shewn by the dates, in the Harsha era, for Aṁśuvarman, the contemporary of Śivadêva I.;
while, with the same treatment, the details of the second of them work out quite correctly.

126 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Such another date is that contained in the Môrbî grant of Jâïnka. And the details
of this, again, work out correctly with the same treatment.
Such a series of dates, too, is that contained in the records of the Valabhî family.
And, with a slight modification, due to a change in the scheme of the year, easily explain-
able, the same treatment gives correct results for the date in the Kaira grant of Dharasêna
IV. of this family, of the year 330; the only one, at present, that affords exact details for
calculation. Also, these records give us a succession of twelve generations, commencing
with the Sênâpati Bhaṭârka, and ending with king Śîlâditya VII., with dates ranging from
the year 207 to the year 447. For the first six or seven generations, the members of this
family were only feudatory Nṛipatis and Mahârâjas, without the authority to establish an
era of their own. And as a matter of fact, the date of the year 207 for the Mahârâja
Dhruvasêna I., in the second generation, proves that the era did not run from the rise to
power of his father Bhaṭârka, the founder of the family, but must have been adopted from
some outside source. While, on the other hand, the long duration of this family, coupled
with the fact that several of their charters were issued from the city of Valabhî itself,¹ and
all of them belong either to that vicinity or to the neighbouring parts of Gujarât, is amply
sufficient to explain why the era used by them should eventually come to be popularly known,
in those parts, as the Valabhî era.
And, finally, undoubted instances of the actual use of an era known as the Valabhî era,
as late as the thirteenth century A.D., are furnished by the Verâwal inscriptions dated
in Valabhî-Saṁvat 927 and 945. For the details of the earlier of these two dates, correct
results can be obtained by applying the same slightly anomalous treatment that applies to
the date in the Kaira grant of Dharasêna IV. of the year 330. The latter of them, however,
goes far beyond this. Not only does it fix the epoch of the era approximately, and in accord-
ance with Albêrûnî's statement, through the concomitant mention of the equivalent
Vikrama and Hijra years; but also the details of it are such as to prove that the epoch of
the era was exactly when Śaka-Saṁvat 241 had expired, and A.D. 319-20 was current.
And it has furnished, in fact, the exact analogy, in accordance with which all dates in the
Gupta-Valabhî era, that follow the true and original northern scheme of its years, have to
be tested.
All this uniform agreement of results cannot be attributed to mere coincidence. But
we must take it now, as a settled matter, that all the dates in question belong to one and
the same era, running from the epoch of A.D. 319-20. And, irrespective of the question
¹ e.g. the grants of Dhruvasêna I., of the year 207 (Ind. Ant. Vol. V. p. 204); of Guhasêna, of
the year 248 (id. Vol. V. p. 206, and Archæol. Surv. West. Ind. Vol. III. p. 93); of Dharasêna II., of
the year 252 (Ind. Ant. Vol. VII. p. 68: Vol. VIII. p. 301; Vol. XV. p. 187; and No. 30 below,
p. 164); of Śîlâditya I., of the year 286 (Jour. Bo. Br. R. As. Soc. Vol. XI. p. 359, and Ind. Ant.
Vol. XIV. p. 327), and of the year 290 (Ind. Ant. Vol. IX. p. 237); of Dhruvasêna II., of the year 310
(id. Vol. VI. p. 12); and of Dharasêna IV., of the year 326 (id. Vol. I. p. 14, and Jour. Bo. Br. R.
As. Soc. Vol. X. p. 66).

THE GUPTA INSCRIPTIONS; INTRODUCTION. 129
Such an era is the Vikrama era.¹ And,—though the expired years of this era might
be quoted, as is shewn, for instance, by lines 19 and 21 of the Mandasôr inscription
of Mâlava-Saṁvat 529 expired, No. 18, page 79, and by line 21 of the Kaḍi grant of
Jayantasiṁha of Vikrama-Saṁvat 1280 expired,²—yet, that, occasionally at least, the
current years were used, is proved by the Gwâlior Sâsbahû temple inscription of Mahî-
pâla,³ in which we have first in words, the number of years expired, 1149, and then,
partially in words and fully in figures, the number of the current year, 1150. Such an era,
again, is the Gupta-Valabhî era; or, at least, we have not as yet obtained the slightest
indication of its ever having been used by astronomers as the basis of calculations.
And, in the absence of the use of any word meaning "expired" in connection with the
year in a Gupta-Valabhî date, it is only reasonable that we should follow the ordinary rules
of interpretation, and render the original passage as denoting a current year.
In one instance only, among the Gupta-Valabhî dates at present known, is a word
meaning "expired" used in connection with the year. This exceptional instance is the
Môrbî grant of Jâiṅka, in which an eclipse of the sun is recorded as having occurred when
the year 585 had passed by. Unfortunately, the month and tithi, in and on which the
eclipse occurred, are not specified; nor even the week-day. And, as we have seen at page
100 above, it might be possible to identify the solar eclipse of this record with that
of the 10th November, A.D. 904. In that case, the given year 585 expired, and the
indicated year 586 current, would be equivalent to A.D. 904-905 current. It would then
be as an expired year, not a current one, that the year 165 of the Êraṇ pillar inscription
of Budhagupta is equivalent to A.D. 484-85 current; and so on with all the other dates.
And we should have to apply, in the case of all dates in the era referable to the true
and original scheme of its years, the epoch of A.D. 318-19 current, or more precisely,
by the Śaka year, the period⁴ from the 18th February, A.D. 318, to the 8th March, A.D.
319; and, in the case of dates belonging to the same class with those of the Kaira
grant of the year 330 and the Verâwal inscription of Valabhî-Saṁvat 927, the epoch of A.D.
317-18 current, or more precisely, by the Vikrama year, the period from the 23rd Sep-
tember, A.D. 317, to the 11th October, A.D. 318. But we have also seen that the solar
eclipse in question can be far more satisfactorily identified with that which occurred
on the 7th May, A.D. 905; to do which, we have to take the given year 585 expired, and
the indicated year 586 current, as equivalent to A.D. 905-906 current. And this record,
therefore, furnishes strong and instructive corroboration of my view that, in the absence of

¹ The present Tables of this era, however, seem to be arranged, like those of the Śaka era,
according to expired years. And some of the almanacs quoted in Appendix I. below, give them in
the same way.
² Ind. Ant. Vol. VI. p. 197.
³ For the full reading and translation of the date see Texts and Translations, page 22, note 5.
⁴ Here, again, I owe the exact initial and ending dates to Mr. Sh. B. Dikshit.

130 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
any distinct specification to the contrary, we must interpret the years in Gupta-
Valabhî dates as current years.
The Origin of the Era.
I have shewn, at page 33 ff. above, that the so-called Gupta era is not one which,
due originally to some event occurring only in approximation to A.D. 318, 319, or 320, had its
exact epoch determined, for convenience of comparison with the Śaka era, by adopting the
expiration of an even number of cycles of the planet Jupiter, either of the Twelve-Year or
of the Sixty-Year System. And no other chronological or astronomical considerations sug-
gest themselves, for the selection of the exact epoch that has been proved. Its origin,
therefore, must be found in some historical event, which occurred actually in
A.D. 320, or so closely to that time that, when the scheme of the northern Śaka year was
applied, the reckoning of the era was not affected to any appreciable extent. And here,
though the point is not conclusive either way, we must bear in mind that, as has been
shewn at page 19 ff. above, in the epigraphical references to the era there is nothing at
all, at any early period, to connect the name of the Early Guptas with it, especially as
the founders of it; and nothing to connect the name of Valabhî with it, until at least nine
centuries after its establishment.
We must also bear in mind that it is certain that the era cannot have been
established by any member of the Valabhî family; the reasons for this being
—(1) that, for the first six or seven generations, the members of this family were mere
feudatory Sênâpatis and Mahârâjas, without the authority to establish an era of their own ;
—and (2) that the date of the year 207 for the Mahârâja Dhruvasêna I., in the second
generation, proves that the reckoning runs from long before the first rise to power of
his father, the Sênâpati Bhaṭârka, by whom the family was founded.
In the same way, the first two members of the Early Gupta family, Gupta and Ghaṭôt-
kacha, held only the feudatory rank of Mahârâja, and had not the authority to establish
an era. The first paramount sovereign in the family was Ghaṭôtkacha’s son, Chandragupta
I. And, if a Gupta era, truly and properly so called, was devised in his time, then as its
starting-point there would have been selected the commencement of his reign, not the date
of the rise to power of his first recorded ancestor, the Mahârâja Gupta ; as was done in
the case of the Harsha era, which disregards, not only two generations of Mahârâjas at the
commencement of the genealogy, but even the reigns of two kings, Prabhâkaravardhana
and Râjyavardhana II., and runs from the commencement of the reign of the third para-
mount sovereign, Harshavardhana himself. So, also, when the Western Châlukya king
Vikramâditya VI. established a new era under the name of the Châlukya-Vikrama-Kâla,¹
he disregarded the reigns of all his ancestors, and made the era date from his own accession
¹ See Ind. Ant. Vol. VIII. p. 187 ff.

THE GUPTA INSCRIPTIONS; INTRODUCTION. 131
to the throne. The dates in the Early Gupta records shew clearly that the Gupta era, cannot,
under any circumstances, run from the accession of any member of the dynasty later than
Chandragupta I. And there are essential difficulties, under any normal conditions, in the
way of making the era date from the commencement of his reign; i.e. of taking A.D. 320-21
as his first current year. For his great-grandson, Kumâragupta, we have dates in the era,
ranging from the year 96 to the year 130 odd;¹ of which we may take, as the latest certain
one,² that of the year 129, recorded in the Mankuwâr inscription, No. 11, page 45. And,
as we ought to assume that Chandragupta I. was at least twenty years old when his reign
commenced, this gives us a period of a hundred and forty-nine years, which, spread over
four generations, gives to each a duration of thirty-seven years and a quarter, or nearly
half as much again as the usually accepted average maximum rate of twenty-five years for
a Hindu generation. This, too. is only dealing with the question of generations. If we
take the period of a hundred and twenty-nine years only, from the commencement of the
reign of Chandragupta I. to nearly the end of that of Kumâragupta,—which gives an
average of thirty-two years and a quarter for each of the four reigns,—then, as compared
with the average duration, twenty years at the outside, of a Hindu reign, the excess is still
more remarkable. And almost exactly the same results are obtained, if, instead of con-
sidering four generations and reigns, down to the end of the time of Kumâragupta, we
take the latest certain date³ of Chandragupta II., viz. the year 93 given in the Sâñchi
inscription, No. 5, page 29, and spread the period of ninety-three years over three
reigns, or, on the same assumption as regards the age of Chandragupta I., the period of a
hundred and thirteen years over three generations. On the question of generations, I will
not base any particularly special objection. An analogy for an abnormal average rate
might be deduced from the Western Châlukya genealogy,⁴ in which we have Śaka-
Saṁvat⁵ 930 for the commencement of the reign of Vikramâditya V., and Śaka-Saṁvat
1060 for the end of the reign, and it may safely be assumed the death, of Sômêśvara III.
in the third generation after him. If we take it that Vikramâditya V. was twenty years
old in Śaka-Saṁvat 930, we have one hundred and fifty years for the four generations; or
an average of thirty-seven years and a half for each. But, from Śaka-Saṁvat 930 to
1060, there were six reigns; with an average of twenty-five years, or seven less than we
¹ See page 67 above.
² And it must be very nearly his latest date; for, he had then been reigning for at least thirty-
three full years, and we have the date of the year 136 for his son and successor, Skandagupta.—The
selection of any later date would, of course, only intensify the force of the argument.
³ The silver coins (see Ind. Ant. Vol. XIV. p. 65 f.) seem to give the year 94 or 95; but the latest
absolutely certain date is the one that I quote. Here again, the adoption of a later date would only
strengthen the argument.
⁴ See my Dynasties of the Kanarese Districts, p. 18, Table.
⁵ The exact year was doubtful when I wrote the book referred to in the preceding note; but it
has now been established by the Kauthêm grant (Ind. Ant. Vol. XVI. p. 15 ff.)
R 2

132 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
should have to allot to each of the four Early Gupta kings in question. And even this
result is due chiefly to the extraordinarily long reign of Vikramâditya VI., for fifty-two
years, from Śaka-Saṁvat 997 to 1048. If we take the whole period of the Western
Châlukya dynasty, covering one hundred and ninety years, from Śaka-Saṁvat 895, the
first year of Taila II., down to Śaka-Saṁvat 1084, as the end of the reign and the death
of Taila III.,¹ we have ten reigns, with an average duration of just nineteen years each.
An average of thirty-two years for four successive reigns of Hindu fathers and sons, seems,
from every point of view, an impossibility. And this prevents our making the Gupta era
run from the commencement of the reign of Chandragupta I. We must, therefore,
accept it as certain that the Early Guptas only adopted the era of some other
dynasty. And we must look for its origin to some extraneous source.
Now, it is evident that the Early Guptas rose to power first as feudatory Ma-
hârâjas, the third of whom, Chandragupta I., while holding that same rank, established his
independence; so that, his successors maintaining the same position, the paramount titles,
and not his original feudatory title, are always coupled with his name in the genealogical
passages in their records. And, from the Mahârâja Gupta down to Kumâragupta, we
have two feudatory governments and four reigns; which, at the average rate of twenty
years, almost fill up the period indicated by the latest certain date for Kumâragupta, and,
by a coincidence, place the commencement of the government of the Mahârâja Gupta
very near to A.D. 320. If, then, we could determine the paramount sovereign of whom
the Mahârâja Gupta was a feudatory, we should have in him the founder of the era; pro-
vided we could only shew that his successors also dated their records in it. And the only
difficulty then remaining, would be,—When Chandragupta I. and his descendants had
asserted themselves as independent sovereigns, by rebellion against their masters, why
should they continue to use a purely dynastic era, which had only been running for a short
time and had certainly not become an astronomical era, and which would always remind them
of the originally subordinate status of their ancestors; instead of establishing a new era
of their own, or instead of adopting some well-known era, of general use, which could
evoke no reminiscence of a humiliating kind?² The Early Gupta records, however, throw
no light on this point; nor can we expect any, unless we obtain inscriptions of the time of
the Mahârâjas Gupta and Ghaṭôtkacha, or of the early years of Chandragupta I. And at

¹ I exclude the short reign of Sômêśvara IV., from Śaka-Saṁvat 1104 to about 1111, because there
had been meanwhile an interruption of the Western Châlukya power by the Kalachuris of the
Dekkan.
² An objection of this sort does not apply to the use of the Gupta era by the Valabhî family. The
Sênâpati Bhaṭârka drove out the invaders who had overthrown the Gupta sovereignty in Western
India; and may possibly have been himself the feudatory of some descendant of the original Gupta
stock. And when Dharasêna IV. became a paramount sovereign, it was on the disruption of the
Kanauj kingdom. At neither point was there any reason for the members of this family to feel any
ersion to the Gupta era.

THE GUPTA INSCRIPTIONS; INTRODUCTION. 133
present, in connection with India itself, we know of no king the commencement of whose
reign can with any certainty be referred to A.D. 320; and of no historical event to which
we can safely allot that date. Nor, while the Early Gupta sovereignty continued, is there
any indication of the Gupta era having been used, in India, by any other independent
dynasty. The nearest approximation to the year in question that we have, is in the case
of the Kalachuri dynasty of Central India; in respect of which certain points in the
records of the Parivrâjaka Mahârâjas and the Mahârâjas of Uchchakalpa do tend to
support the actual existence, in the Early Gupta period, of a Kalachuri era, and,
consequently, of Kalachuri kings under some earlier name.¹ The Kalachuri dates,
however, certainly cannot be referred to the Gupta epoch. And circumstances indicate
that the dominion of the Kalachuri kings, at that time, was confined entirely to the more
eastern parts of Central India; so that they were only contemporaries of the northern
dynasty of which the Early Guptas were at first the servants. Mr. Fergusson's opinion,²
again, was in the direction of the era being established, with the foundation of Valabhî as
a new capital of Western India, by the Andhra king Gôtamiputra, whom he placed³
between A.D. 312 and 333; the Mahârâja Gupta being a feudatory of him or of one of
his immediate successors. But the chronology of the Andhras,—who, at the best, seem
to have been too essentially a western and southern dynasty to be concerned in any lead-
ing way with the history of Northern India,—still remains to be finally determined. And
Dr. R. G. Bhandarkar, who has given more consideration to the subject than any one else
as yet, places Gôtamiputra about two centuries earlier,⁴ in the period A.D. 133 to 154;
and, according to his view of the early chronology, we should have to refer the establish-
ment of the Gupta era to some event connected with either the downfal of the Kshatrapas
of Saurâshṭra or the history of the Râshṭrakûṭas of the Dekkan. The Kshatrapas, how-
ever, certainly did not use the Gupta era. And there is not the slightest particle of
evidence that the Râshṭrakûṭas ever had an era of their own. There can be but little doubt
that the real paramount lords of the Mahârâjas Gupta and Ghaṭôtkacha, and at first of
Chandragupta I. himself, were some of the later Indo-Scythian kings of Northern India,
whose duration is certain at any rate up to the time of Samudragupta. These Indo-
Scythian kings must have used the Śaka era. But this era, again, had not then become
an astronomical era;⁵ and there was, therefore, no special inducement for the Early Guptas
to adopt it; but, on the contrary, there was an objection of the kind already indicated.
Further, the Vikrama era was not an astronomical era; and the use of it, in those days, under
the name of the Mâlava era, was probably confined to the different sections of the Mâlava
tribe, and to territories of which no part was brought under the Early Gupta sway until the
time of Samudragupta. And, finally, the Kaliyuga era in all probability was used only
¹See page 8 ff. above.
²Jour. R. As. Soc. N. S. Vol. IV. p. 128 f. ³id. p. 122.
⁴Early History of the Dekkan, p. 27. ⁵See Appendix I. below

by the astronomers of Ujjain, for purely technical purposes; and was not known at all in
the territories in which the Early Guptas first rose to power. In fact, in India itself
there was no already existing era which would recommend itself to the Early
Guptas. And we have next to inquire whether there may have been any such era,
beyond the limits of India proper.
By a comparison of the dates of Sivadêva I. and Aṁśuvarman, at page 95 f. above,
I have already shewn, in a general way, that the Gupta era was in use beyond the
north-eastern frontier of India, in Nêpâl; a fact which is duly corroborated by the
results for the date in the inscription of Mânadêva, of the year 386. We must, therefore,
now see what more particular information can be gathered from the epigraphical records
of that country.¹

¹ And here we may note that the Kings of Valabhî can have had nothing to do either with the
introduction of an era into Nêpâl, or with the borrowing of an era from that country. As I have
already had occasion to remark, the members of the Valabhî family, for the first six or seven genera-
tions inclusive of Bhaṭârka, were mere feudatory Sênâpatis and Mahârâjas; and these members of the
family, at any rate, cannot possibly have conquered Nêpâl, or even have extended their territory up
to the confines of that country. The first of the family who claimed to be a paramount sovereign is
Dharasêna IV., with the dates of 326 and 330; and with the titles of Paramabhaṭṭâraka, Mahârâjâ-
dhirâja, and Paramêśvara, in common with all his successors, and also with that of Chakravartin,
which, not being assumed by any of his successors, may perhaps indicate that his power was more
extensive than theirs ever was. Now, in passing; if we refer his first date of 326 to the epoch of A.D.
319-20, the result, A.D. 645-46, brings us to a very suitable period indeed for him to assume the posi-
tion and titles of a paramount sovereign; viz. to the commencement of the anarchy which, as Ma-
twan-lin tells us (Ind. Ant. Vol. IX. p. 20) attended the death of Harshavardhana, "the warlike lord
of all the region of the north." It ended in the complete disruption, for the time, of the kingdom of
Kanauj. Aṁśuvarman became paramount in Nêpâl, and Âdityasêna in Magadha; and the opportu-
nity was of course taken advantage of by Dharasêna IV., to assert his independence in the west of
India. But, to say nothing of the improbability of the thing on other grounds, the fact that
Aṁśuvarman became king of Nêpâl, is in itself enough to prevent our admitting the possibility of a
conquest of that country by Dharasêna IV. Referring the same date of 326 to the three earlier
proposed epochs, we have respectively A.D. 403, 492, and 516. For these periods there is, perhaps,
no particular objection to our assuming, for the sake of argument, that Dharasêna IV. may have
extended his power over a considerable portion of Northern India, in the parts nearer to Kâṭhiâwâḍ
and Gujarât. But the Valabhî charters, in which a conquest so extensive as that of the whole of
Northern India up to Nêpâl, or inclusive of that country, would most certainly have been recorded,
give not the slightest hint of any such event at any time in the history of the family. In fact, with
the exception of the allusion to the overthrow of the Maitrakas by Bhaṭârka, they give absolutely no
detailed information at all in connection with any of the successes claimed by the members of this
family; which tends to shew very plainly that, from beginning to end, the Valabhî power was purely
local. And, in connection with the earlier proposed epochs, even if Dharasêna IV. did conquer
Nêpâl, or Northern India up to the frontier of Nêpâl, and did introduce there the era of A.D. 319-20,
the question still remains, and cannot be answered,—Why should he act with such extreme inconsist-
ency as to introduce there this era, which, according to those who have sought to establish those
epochs, was not brought into actual use in his own territory; instead of the Gupta era which he him-
self, and his successors, continued to employ for all the official purposes of their own kingdom?

THE GUPTA INSCRIPTIONS, INTRODUCTION. 135
In Appendix IV. below, I give an account of such of the inscriptions from Nêpâl
as have any bearing on the question now under consideration. The actual dates of them
range from A.D. 635 to 854; and give a fairly clear idea of the history of the reigning
families of the country during that period. They shew two separate houses, ruling con-
temporaneously, and mostly on equal terms; and each preserving certain distinctive cha-
racteristics of its own. One of them was a family, the name of which is not mentioned
in the inscriptions, but which in the Vaṁśâvali is called the Ṭhâkurî family, issuing its
charters from the house or palace called Kailâsakûṭabhavana, and uniformly using the
Harsha era. The other was the Lichchhavi family, distinctly so named in the inscrip-
tions, and in the Vaṁśâvali allotted to the Sûryavaṁśa or solar lineage, issuing its
charters from the house or palace called Mânagṛiha, and uniformly using an era with
the Gupta epoch.
That the Lichchhavi clan or tribe was one of great antiquity and power,
in the direction of Nêpâl, is shewn by the writings of Fa-Hian and Hiuen-Tsiang.¹ which
connect them with events that preceded the nirvâṇa of Buddha. No exception, there-
fore, need be taken to the general outlines of the long account in one of the inscriptions,
which, so far as the Nêpâl branch of the tribe is concerned, gives us the first really
historical member of it in the person of Jayadêva I., who, by the ordinary allowance of time
for each Hindu generation, must be referred to the period A.D. 330 to 355.
Proof of friendly relations between the Early Guptas and the Lichchhavis,
at an early time, is given by the marriage of Chandragupta I. with Kumâradêvî, the
daughter of Lichchhavi or of a Lichchhavi king. And, that the Lichchhavis were then at
least of equal rank and power with the Early Guptas, is shewn by the pride in this alliance
manifested by the latter; exhibited in the careful record of the names of Kumâradêvî, and
of her father or her family, on some of the gold coins of Chandragupta I., and by the uni-
form application of the epithet, “daughter’s son of Lichchhavi or of a Lichchhavi,” to
Samudragupta in the genealogical inscriptions. Again, the Allahâbâd pillar inscription
shews that, even if Samudragupta did not make Nêpâl a tributary province, his kingdom
extended up to the confines of that country.
There can be no doubt that the Early G kings must have known the
nature and origin of whatever era was being used by their Lichchhavi connec-
tions in Nêpâl. And the period established for Jayadêva I. approximates so closely to
A.D. 320-21, that it needs but little adjustment to place the commencement of his reign
actually in that year. This arrangement would give a perfectly intelligible reason for the
origin of the era, which was clung to so persistently by his descendants that they conti-
nued the use of it for at least two centuries after the introduction of the Harsha era into
¹See Legge's Travels of Fâ-Hien, pp. 71, 76; Beal's Buddhist Records of the Western World,
Vol. I. pp. xiii, lii. lv. and Vol. II. pp. 67 note, 70, 73, 77 note, 81.

136 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Nêpâl, and its acceptance by their immediate neighbours, the Ṭhâkurî family of Kailâsa-
kûṭabhavana. And no objection could be taken by the Early Gupta kings to the adoption
of the era of a royal house, in their connection with which they took special pride. I think,
therefore, that in all probability the so-called Gupta era is a Lichchhavi era,
dating either from a time when the republican or tribal constitution of the Lichchhavis was
abolished in favour of a monarchy ; or from the commencement of the reign of Jayadêva I.,
as the founder of a royal house in a branch of the tribe that had settled in Nêpâl. But
the question of the origin of the era is one, of course, on which further discoveries,
especially if any can be made in Nêpâl, may be expected to throw more light.
J. F. FLEET.
4th November 1887.

THE GUPTA INSCRIPTIONS; APPENDIX I. 137
APPENDIX I.
A Note on the Epoch and Reckoning of the Śaka Era.
Gen. Sir A. Cunningham's Tables,¹ and Ganpat Krishnaji's and Kero
Lakshman Chhatre's almanacs, shew the period from the 5th April, A.D. 1886, to the
24th March, A.D. 1887, as corresponding to Śaka-Saṁvat 1808. So, also, the Sâyana-
Pañchâṅg shews the same period, as corresponding to Śaka-Saṁvat 1808, as the ordinary
nirayaṇa luni-solar year; the period covered by Śaka-Saṁvat 1808, as a sâyana year, being
from the 6th March, A.D. 1886, to the 22nd February, A.D. 1887. For all purposes of cal-
culation, however, this Śaka year has to be taken as "the year 1808, expired;" even if²
we were working out the very first day of it, Chaitra śukla 1. The Tables are intended for
this application of the number; and there are, in fact, no grounds for doubting that the
above period really is equivalent to Śaka-Saṁvat 1808 expired, and 1809 current. But
it is quoted, for all ordinary purposes, simply as Śaka-Saṁvat 1808. And, if a Hindu
were converting "Saturday, the 1st January, A.D. 1887," into its corresponding Hindu
date, he would write down, as the result, "Śakê 1808 Pausha śukla saptamî Śanivâra;" in
which, not only does he abstain from including any word meaning "expired," but he
actually uses, instead even of the crude form Śaka, the Sanskrit locative Śakê, which literally
means "in Śaka (1808)," i.e. "while Śaka 1808 is current;" and this is the meaning
which the mention of the year presents to any Hindu who is not an astronomer, and
¹ The proper method of applying his Tables may perhaps be inferred from his remarks (e.g.
Indian Eras, pp. 5, 48, 52) that the numbers of the years in Hindu dates refer to years actually elapsed;
and that the Hindus count only by completed years. But I am speaking of the meaning which the
Tables present to a general reader, at first sight. Thus, anyone turning to his Table XVII. p. 199,
—to which there is not attached a note that the Hindu years given therein are expired years,—in
order to ascertain the Śaka equivalent of A.D 1886-87, finds Śaka-Saṁvat 1808; and naturally takes
it as a current year. So, also, with any similar Tables; e.g. those in Mr. C. Patell's Chronology.
Such Tables would be much more useful for general purposes, if they shewed the current Hindu years
opposite the current Christian years, as is done in the case of the saṁvatsaras of the two cycles of
Jupiter; leaving it to anyone who has to make a particular calculation, to take the preceding year as
the basis of his work. And, in ordinary writing, the current Hindu years should certainly be quoted
with the current Christian years.
² Unless with Tables based on the Mêsha-Saṁkrânti, or entrance of the sun into Aries, as Prof.
K. L. Chhatre's Tables are; in which the Śaka year is practically treated as commencing with the
day of the Mêsha-Saṁkrânti. With such Tables, for any tithi connected with Śaka-Saṁvat (1809
current and) 1808 expired, up to the tithi that coincided with the solar day on which the Mêsha-Saṁ-
krânti occurred, we must work with the basis of even one year still earlier, viz. Śaka-Saṁvat 1807
expired.

138 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.

who is not acquainted with the technical application of the number of the year. So, also,
the same expression is used in the almanacs themselves; thus, in the first two almanacs
mentioned above, "Śakê 1808 Vyaya-nâma-saṁvatsarê," on the title-page; and "Śakê
1808 Chaitra-śukla-pakshaḥ," on the top of the page which exhibits the bright fortnight
of the month Chaitra; and, in the Sâyana-Pañchâṅg, "Śâlivâhana-Śakê 1808 Vyaya-
nâma-saṁvatsaraḥ," on the title-page, and elsewhere "amântaḥ Chaitra-śukla-pakshaḥ
Śâlivâhana-Śakê 1808 Vyaya-nâma-saṁvatsaraḥ." In the same way, I find, for the same
period, "Śakê 1808 Vyaya-nâma-saṁvatsarê" on the title-page of an almanac pub-
lished at Pandit Umacharan Muhatmim's Press at Gwâlior; and, for the period
from the 17th March, A.D. 1885, to the 4th April, A.D. 1886, "Śâlibâhana-Śakê 1807,"
on the title-page of the Jôdhpur Chaṇḍû-Pañchâṅg for that year, and "Śrî-Saṁvat
1942 Śakê 1807 Chaitra-śukla-pakshaḥ," in Bapu Deva Shastri's almanac, prepared
at Benares and published at Lakhnau.
Again, in the preliminary passages that introduce the saṁvatsara-phala or '(astrolo-
gical) results for the year,' and other similar matter, Ganpat Krishnaji's and K. L.
Chhatre's almanacs, contain the passage — atha gata-Kaliḥ 4987, śêsha-Kaliḥ 427013;
Svasti; śrîman-nṛipa-Vikramârka-samay-âtîta-saṁvat¹ 1942, Hêmalamba-nâma-saṁvatsarê;
tathâ śrîman-nṛipa-Śâlivâhana-Śakê 1808, Vyaya-nâma-saṁvatsarê; asmin varshê râjâ
chandrah,—"now the expired (portion of the) Kali (age) (is) 4987 (years); (and) the
remainder of the Kali (age) (is) 427013 (years). Hail! In the year 1942 expired from the
time of the glorious king Vikramârka,² (and) in the Hêmalamba saṁvatsara; so also in
the Śaka (year) 1808 of the glorious king Śâlivâhana, (and) in the Vyaya saṁvatsara; in
this year, the king (is) the Moon." And, for the nirayaṇa year, the Sâyana-Pañchâṅg
for Śaka-Saṁvat 1808 has — Kaliyugasya gata-varshâṇi 4987; śrîman-nṛipa-Vikra-
mârka-saṁvat³ 1943 Vilambi-saṁvatsaraḥ; śrîman-nṛipa-Śâlivâhana-Śak-âbdaḥ 1808
Vyaya-nâma-saṁvatsaraḥ; ath=âsmin varshê râjâ chandrah,—"the expired years of the
Kaliyuga (are) 4987; in the year 1943 of the glorious king Vikramârka, (there is) the
saṁvatsara named Vilambin; (and there is) the year 1808 of the Śaka of the glorious
king Śâlivâhana, (and) the saṁvatsara named Vyaya; now, in this year, the king (is)
the Moon." In these passages, these three almanacs again treat the Śaka year,

¹ i.e. saṁvatsarê, or saṁvatsarêshu.
² It is curious that here the Vikrama year should be distinctly specified as expired, while the Śaka
year is not qualified in the same manner; as if a distinction were being made in the methods of
reckoning the two eras.
³ i.e. saṁvatsarê, or saṁvatsarêshu.—The figures here, and in the Gwâlior almanac, 1943, differ
from those in Ganpat Krishnaji's and K. L. Chhatre's almanacs, 1942, because the latter quote the
southern reckoning, by which each Vikrama year commences with the month Kârttika, seven
lunations later than the same year in the northern reckoning; consequently, at the commencement
of Śaka-Saṁvat 1808 (expired), on the first day of the bright fortnight of Chaitra, Vikrama-Saṁvat
1402 was still running, by the southern reckoning.

THE GUPTA INSCRIPTIONS; APPENDIX I. 139

apparently, as a current year. The Gwâlior almanac, however, which I have quoted
above, has — gata-Kaliḥ 4987, śêsha-Kaliḥ 427013 ...........; tan-madhyê gata-
Śakaḥ 1808, śêsha-Śakaḥ 16192 .......... Svasti; śrî-Vikramârka-râjya-samayâd
atît saṁvat¹ 1943, Śaka-gata-varshêshu 1808, chandra-mânêna Vyaya-nâma-saṁvat-
sarê; Bârhaspatya-mânêna, Śakê 1807 Âśvina-kṛishṇa-7 Śukrê sûry-ôdayâd gata-ghaṭî-
shu 47 palêshu 24 tad-avadhi, Śakê 1808 Âśvina-kṛishṇa-14 Bhaumê ghaṭî[shu*]
46 palê[shu*] 3 tâvat-paryantaṁ, Vilambi-saṁvatsar-ôllêkhaḥ vidhêyaḥ, tad-agrê Vikâri-
saṁvatsar-ôllêkhaḥ kâryaḥ; Chaitr-âdau râjâ chandraḥ,— “the expired (portion of the)
Kali (age) (is) 4987 (years), (and) the remainder of the Kali (age) is 427013 (years)
..........; in it, the expired (portion of the) Śaka (era) (is) 1808 (years); and
the remainder of the Śaka (era) (is) 16192 (years) .......... Hail! In the
year 1943 expired from the time of the reign of the glorious Vikramârka, (and) in the
expired Śaka year 1808, (and), by the lunar reckoning, in the (current) saṁvatsara
named Vyaya,²—by the reckoning of Jupiter, the Vilambin saṁvatsara is to be used in
writings from the expiration of 47 ghaṭîs, 24 palas, from sunrise on Friday, the seventh
lunar day of the dark fortnight of Âśvina, in Śaka 1807, up to (the expiration of) 46
ghaṭîs, 3 palas, (from sunrise) on Tuesday, the fourteenth lunar day of the dark fortnight
of Âśvina, in Śaka 1808; after that, the Vikârin saṁvatsara is to be used in writings,—at
the beginning of Chaitra, the king (is) the Moon.” Passages of a similar kind with those
quoted above, occur at the end of each almanac, in connection with the Saṁkrântis.
In the same passages for the sâyana year, Śaka-Saṁvat 1808, the Sâyana-Pañchâṅg
does not confine itself to any indefinite expression, but explicitly quotes the Śaka year as
a current year; thus — Kaliyugasya saṁdhyâyâ âditaḥ, Śâlivâhana-Śak-ârambhakâla-
paryantaṁ, Nand-âdr-indu-guṇa-(3179)-mitâni saura-varshâṇy=atîtâni; pravartamâna-
Śâlivâhana-Śak-âbdaḥ ashṭ-ôttar-âshṭâdaśa-(1808)-mitaḥ; amuṁ saṁvatsaraṁ Narma-
dâyâ dakshiṇa-bhâgê Vyaya-nâmnâ vyavaharanti, uttara-bhâgê cha Vilambi-nâmnâ;
ath=âsmin varshê râjâ Śaniḥ,—“from the commencement of the _saṁdhyâ_³ of the
Kaliyuga, up to the time of the commencement of the Śâlivâhana-Śaka, there expired
solar years which are measured by the (nine) Nandas, the (seven) mountains, the (one)
moon, and the (three) qualities, (3179); (and) the current year of the Śâlivâhana-Śaka
is measured by eighteen hundred, increased by eight, (1808); on the south side of the

¹ i.e. atîtê saṁvatsarê, or atîtêshu saṁvatsarêshu.
² The context is “at the beginning of Chaitra, the king (is) the moon,” a little further on. The
intervening matter is by way of a parenthesis.
³ saṁdhyâ, which is usually translatable by ‘morning or evening twilight,’ means, as applied to
any of the four ages, a long period that runs at the commencement of each, before the full develop-
ment of the age itself. The saṁdhyâ of the Kali age is one hundred divine years, equivalent to
36,000 years of men; so that we are still only in this period. The age itself will run for 360,000
years of men. And it will end with a saṁdhyâṁśa of 36,000 years of men. These figures make up
the total of 432,000 years in the age.
s 2b

140 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Narmadâ, they distinguish this samvatsara by the name of Vyaya; and, on the north
side, by the name of Vilambin; now, in this year, the king (is) Saturn." But, in the
corresponding passage in the same almanac for the preceding year, Śaka-Saṁvat 1807,
after giving in the same words the number of the solar years that had expired from the
commencement of the saṁdhyâ of the Kaliyuga up to the commencement of the Śaka
era, the text runs — tatô vartamâna-vatsar-ârambhakâla-paryantaṁ sapt-ôttar-âshṭâdaśa-
śata-(1807)-mitâni varshâṇi gatâni; amuṁ vartamâna-saṁvatsaraṁ Narmadâyâ dakshiṇê
bhâgê Pârthiva-nâmnâ vyavaharanti, uttarê bhâgê cha Hêmalamba-nâmnâ; ath=âsmin
varshê râjâ Bhaumaḥ,—"from then, up to the commencement of the current year, there
have expired years which are measured by eighteen hundred, increased by seven, (1807);
on the south side of the Narmadâ, they distinguish this current saṁvatsara by the name
of Pârthiva; and, on the north side, by the name of Hêmalamba; now, in this year, the
king (is) Mars."¹ In passing, therefore, from Śaka-Saṁvat 1807 to 1808, a verbal dis-
tinction, at least, was made between expired and current years; and the phraseology
adopted in the almanac for Śaka-Saṁvat 1808, has been repeated in the almanac for the
next year, 1809. In one instance, A.D. 85-86, the Pârthiva or Hêmalamba samvatsara,
each current, was treated as equivalent to Śaka-Saṁvat 1807 expired; while, in the
other, A.D. 1886-87, the Vyaya or Vilambin samvatsara, each current, and each the next
in the cycle after respectively Pârthiva and Hêmalamba, is treated as equivalent to
Śaka-Saṁvat 1808 current. What were the reasons for this change, I do not know; and
I will leave it to Mr. Sh. B. Dikshit, one of the editors of the almanac, to explain them
elsewhere. But, by the literal interpretation of the phraseology for A.D. 1885-86, and
in accordance with the principles of the Tables, that period was equivalent to Śaka-
Saṁvat 1807 expired, (and 1808 current); and A.D. 1886-87 should have been described
as being represented by Śaka-Saṁvat 1808 expired, (and 1809 current).
I have now to quote the fact that, in Madras, the same English period, A.D. 1886-87,
is actually called Śaka-Saṁvat 1809, with the same saṁvatsara of the Sixty-Year
Cycle, Vyaya, attached to it. There are, it is true, two somewhat varying practices to
be found in Southern India. Thus, the Telugu Siddhânta-Pañchâṅgam, published,
I think, in the Arcot District, gives the luni-solar period from the 5th April, A.D. 1886, to
the 24th March, A.D. 1887, as being the Vyaya samvatsara, and as corresponding to
Śaka-Saṁvat 1808 expired; and at the commencement, it quotes the expired years
throughout; thus — "Kaliyuga-gat-âbdâḥ 4987; Śâlivâhana-Śaka-gat-âbdâḥ 1808; Vikra-
mârka-Śaka-gat-âbdâḥ 1943." But, on the other hand, the Telugu Calendar, published
at Madras, gives the same luni-solar period, from the 5th April, A.D. 1886, to the 24th
March, A.D. 1887, as being the Vyaya samvatsara, and as corresponding to Śaka-Saṁvat
¹ Except for reading amuṁ vartamâna-vatsaraṁ, and omitting atha before asmin varshê, the
text is exactly the same in Bapu Deva Shastri's almanac for Śaka-Saṁvat 1807. I have not been
able to obtain a copy of his almanac for the next year.

1809, Kaliyuga-Saṁvat 4988, and Vikrama-Saṁvat 1944, which are not specified as either
current or expired, but can only be intended as current. And, in the same way, the
Tamil Siriya-Pañchâṅgam, for the following year, published at Madras, gives the solar
period from the 12th April, A.D. 1887, to the 11th April, A.D. 1888, as being the Sarvajit
saṁvatsara, and as corresponding to Śaka-Saṁvat 1810, Kaliyuga-Saṁvat 4989, and
Vikrama-Saṁvat ¹ 1935, which, similarly, are not specified as either current or expired, but
can only be intended as current.² And, from other indications, there seems to be no
doubt that, of these two practices of Southern India, thus illustrated, the popular and
generally current one is the latter one, by which the period A.D. 1886-87 is quoted as
Śaka-Saṁvat 1809; the reckoning, in this and the other eras, being thus, at first sight,
one year in advance of the customary reckoning of Northern and Western India.
The difference, however, is only an apparent one ; and is due to the evident fact that
the Madras reckoning has preserved the system of current years, while the other
is regulated by expired years. But it is almost always the reckoning of Northern and
Western India that is now quoted. And the years of it, though really expired years, are not
distinctly and habitually quoted as such. And hence there is a general understanding that,
as between the Śaka and the Christian eras, the additive quantity, to be applied to the
former, is 78-79;³ and that the epoch or year 0 of the Śaka era, is the period from the
3rd March, A.D. 78, to the 20th February, A.D. 79, both included ; and its commence-
ment, or first current year, the period from the 21st February, A.D. 79, to the 10th
March. A.D. 80, both included.⁴ This, however, really gives current Christian years,
¹ Here the 3 in the tens place must be a mistake for 4.
² It would appear that the Tamil Vâkya-Pañchâṅgam, published at Madras, gives the solar
period from the 12th April A.D. 1887, to the 10th April, A.D. 1888, as being the Sarvajit saṁvat-
sara, and as corresponding to Śaka-Saṁvat 1809, Kaliyuga-Saṁvat 4988, and Vikrama-Saṁvat 1945];
all of which are distinctly specified as current. But this cannot possibly be correct, in respect of the
Śaka and Kaliyuga years.
³ Thus, even Dr. R. G. Bhandarkar, — through whose "Note on the Śaka Dates and the
Years of the Bârhaspatya Cycle, occurring in the Inscriptions" (Early History of the Dekkan,
p. 105 ff.) my attention was first drawn to the desirability of examining the details of the alma-
nacs, — has written (id. p. 99; the italics are his) "191 Gupta past + 242 = 433 Śaka current

• 78 = 511 A.D. current ......... 209 Gupta past + 242 = 451 Śaka current + 78 = 529
  A.D. current." I myself had the same view, till not very long ago. Other writers could easily be
  shewn to have lain under the same misconception. And Dr. Burnell even went so far as to say
  (South-Indian Palæography, p. 72, note) "the rough equation for converting this era into the Chris-
  tian date, is + 78¼. The beginning of the year being at the March equinox ; if the Śaka atîta (i.e.
  expired) "year be mentioned, the equation is + 79¼.'"
  ⁴ I owe these four dates to Mr. Sh. B. Dikshit. Gen. Cunningham (Indian Eras, p. 139) gives
  from the 14th March, A.D. 78, to the 17th February, A.D. 79; and from the 18th February, A.D. 79,
  to the 8th March, A.D. 80. But a comparison of his initial days for the epoch and the first year,
  shews at once that there is some mistake. The 18th February is twenty-four days earlier than the
  14th March; whereas the difference should be only eleven days. Mr. C. Patell (Chronology, p. 96)
  does not give the initial day of the epoch; but gives, in the same way, the period from the 18th Feb-
  ruary, A.D. 79, to the 8th March, A.D. 80, for the first year.

142 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
equivalent to expired Śaka years. It is evident from the details given above regarding
Śaka-Saṁvat 1808 and 1809, that, according to the reckoning of the era as fixed by the
early astronomers, and as preserved to the present day, the true epoch is A.D. 77-78,
and the period from the 3rd March, A.D. 78, to the 20th February, A.D. 79, is in reality the
commencement, or first current year; and that, to obtain current Christian years,
equivalent to current Śaka years, the true additive quantity is 77-78. But, of
course, there is always the possibility that, if ever we obtain a date, with full details for
calculation, in a very early Śaka year, or in one of the very earliest of the regnal or
dynastic years which afterwards developed into the Śaka era, this exact equation may
not hold good; in consequence of the date belonging to a period anterior to the adoption
of the era by the astronomers.
The Śaka era is emphatically one of the eras that originated in an exten-
sion of regnal or dynastic years. The chief Hindu tradition about it, is, that it was
founded in celebration of a defeat of the Śaka king by the king Vikrama or Vikramâditya
who is also the supposed founder of the Vikrama era, commencing a hundred and thirty-
five years earlier.¹ This tradition is mentioned by Albêrûnî ;² but he saw through it so far
as to remark "since there is a long interval between the era which is called the era of
Vikramâditya and the killing of Śaka, we think that that Vikramâditya from whom the
era has got its name is not identical with that one who killed Śaka, but only a namesake
of his." And the tradition has now been quite exploded by the Bâdâmi cave inscrip-
tion of the Chalukya king Maṅgalîśa,³ which is specifically dated "when there have
expired five centuries of the years of the installation of the Śaka king (or kings) in the
sovereignty." It is certain, from this record, that the real historical starting-point of the
era, is the commencement of the reign of some particular king, or kings, of the Śaka
¹ Another tradition (e.g. Prinsep's Essays, Vol. II. Useful Tables, p. 154) is that the era dates
from the birth of Śâlivâhana, king of Pratishṭhâna, who opposed Vikramâditya, king of Ujjayinî.
But the introduction of the name of Śâlivâhana in connection with the era, is of comparatively
modern date, the earliest instance that I have succeeded in obtaining, being one of the thirteenth cen-
tury A.D. ; and the epigraphical instances speak of the year as having been 'established, settled,
or decided' (nirṇîta) by Śâlivâhana, but not as running from his birth (see Ind. Ant. Vol. XII. p.
214 f.).—A passage to the latter effect has been quoted by Prof. Max Müller (India ; What can it
teach us? p. 300 f.) from the Muhûrtabhuvanônmârtaṇḍa of Nârâyaṇa, which means "in the year
measured by three, the (nine) numerals, and the (fourteen) Indras, from the birth of Śâlivâhana (i.e.
in Śaka-Saṁvat 1493), in (the month) Tapas (Mâgha), this Mârtaṇḍa was composed."—As Prof.
Max Müller has pointed out, in his comments on this passage, it is not exactly wrong to speak of
the era as the Śâlivâhana-Śaka or Śâlivâhana era ; for there are ample instances in which the Hindus
give it that name, in epigraphical records of authority and of some antiquity. At the same time,
those instances shew that it was only in comparatively modern times that the name of Śâlivâhana
came to be connected with the era. And, in all discussions respecting early dates, it is an anachronism
and a mistake, to call the era by his name.
² Sachau's Albêrûnî's India, Translation, Vol. II. p. 6.
³ Ind. Ant. Vol. VI. p. 363 f., and Vol. X. p. 57 ff.

THE GUPTA INSCRIPTIONS; APPENDIX I. 143
tribe; and, therefore, that the years were originally regnal or dynastic years. Now,
such years must run on for a considerable time, before they can develop into a recog-
nised era; and this is undoubtedly the reason why we find the earlier years of every
such Hindu era quoted simply by the term varsha or saṁvatsara, 'a year,' without any
dynastic appellation. Again, such regnal or dynastic years can only come to be quoted
as expired years, when they have actually developed into an era which has become recog-
nised, or is sought to be applied, by astronomers for astronomical processes; up to that
point, the years, being wanted only for quasi-private dynastic purposes, would certainly
be quoted as current years. It is impossible to believe that the first Śaka king decreed,
immediately after his coronation, that a new era had been established from that event;
that it was to come at once into general use; and that, for the convenience of astrono-
mers, the first year, then running, was to be quoted as an expired year, which, in fact, it
would be rather difficult to do. If it were sought to fix the exact chronological position of
any public act performed in that first year, it might be referred to the expired years of an
earlier era; e.g. of that of the Kaliyuga. 'But, for any reference to the regnal year alone,
that act would be recorded as being performed "in the year one," "in the first year," or
"while the first year of the reign is current;" as, for instance, "in the first year; while
the Mahârâjâdhirâja, the glorious Tôramâṇa, is governing the earth," in line 1 f. of
the Êraṇ inscription, No. 36, page 158. This custom would continue as long as the
years were simply dynastic years; and perhaps, during the whole of that period, the
years might remain purely dynastic years, each of them having for its initial day the
anniversary of the original coronation from which they started, irrespective of the initial
day of the years of the astronomical era still continuing in use. When, however, astro-
nomers came to adopt them as an astronomical era, they would establish an exact epoch
by reckoning back from the dynastic year then current to the last year of the Kaliyuga
that had expired when the first current dynastic year commenced; in the course of
which, they would simplify matters by allotting to the dynastic years the same scheme,
as regards the starting-point of each year, and the arrangement of the fortnights of the
months, which belonged to the Kaliyuga in their part of the country. Thus they would
fix all the data necessary to enable them to use the new era for astronomical purposes.
All that would remain, would be to use its expired years, in accordance with the custom
and necessities of their science. The substitution of the Śaka era for the Kaliyuga, for
astronomical purposes, seems to have taken place after the time of Âryabhaṭa (born A.D.
476),¹ who used the Kaliyuga, and in or just before the time of Varâhamihira (died A.D.
587),² who used the Śaka era; and probably the apparent difference of one year in the
reckonings of the Śaka era will be found to have originated not far from Śaka-Saṁvat
500. Let us assume that this adoption of the Śaka era was made in Śaka-Saṁvat 500,
equivalent to A.D. 577-78. The astronomers would take it, at starting, as "Śaka-
¹ Jour. R. As. Soc. N. S. Vol. I. p. 405. ² Jour. R. As. Soc. N. S. Vol. I. p. 407.

144 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Saṁvat 499, expired;" and, in quoting it and several subsequent years, would probably
be careful to connect with each year a word distinctly meaning "expired." In course of
time, however, such precision of expression would come to seem superfluous to them;
and, in issuing their almanacs, they would drop the word "expired," and would write,
for instance, simply "Śakē 510 Chaitra-māsa-śukla-pakshaḥ." It would cause no
difference or inconvenience to them; because any initiate would know that this really
designated the bright fortnight of the month Chaitra of Śaka-Saṁvat 511 current after
Śaka-Saṁvat 510 had expired. The people at large, however, including persons who
would use the almanacs for practical purposes without being properly initiated into the
application of them, would be thrown back in their reckoning by a year; and doubtless
at first a good deal of inconvenience and confusion might result. But this would soon
be forgotten; or might, for the sake of convenience, be intentionally put aside. And
thus they would very quickly arrive at the understanding, by which, in Northern and
Western India, Śaka-Saṁvat 1808 answers, popularly, as a current year, not as an expired
year, to the period from the 5th April, A.D. 1886, to the 24th March, A.D. 1887.
Postscript.
In connection with my remarks, on page 140 above, regarding the Sāyana-Pañchāṅg
for Śaka-Saṁvat 1807 and 1808, I would add that I had drawn Mr. Sh. B. Dikshit's
attention to the circumstances of the case; and I now find that, in their almanac for
Śaka-Saṁvat 1810 (expired) (A.D. 1888-89), the editors have reverted to the phraseology
used in their almanac for Śaka-Saṁvat 1807 (expired.)