THE GUPTA INSCRIPTIONS; APPENDIX III. 169
In the preceding remarks, my object has been to explain concisely the system of the
Twelve-Year Cycle of Jupiter that is based on his heliacal risings; and the three methods
for determining the nakshatra with which the heliacal rising takes place. I have now to
make a few observations of a more general kind; and to introduce the eleven
other authorities of which I have spoken at page 162 above. Before doing so, however,
I must refer more fully to another system, which also has been incidentally mentioned, of
naming the saṁvatsaras of the Twelve-Year Cycle according to Jupiter's passage from one
sign of the zodiac into another, with reference to his mean longitude.
The rule for that which I have named the mean-sign system, is thus given by
Âryabhaṭa in his Ârya-Siddhânta or Âryabhaṭîya, Kâlakriyâpâda, verse 4,—Guru-bhagaṇâ
râśi-guṇâs tv=Âśvayuj-âdyâ Gurôr abdâḥ,—"the revolutions of Jupiter, multiplied by the
signs (twelve), (are) the years of Jupiter, the first of which is Âśvayuja." And it is given,
in very similar words, by Brahmagupta, in his Brahma-Siddhânta, adhyâya xiii. verse 42,—
Guru-varshâṇy=Âśvayujâd dvâdaśa-guṇitâ Gurôr bhagaṇâḥ. In this rule, the revolutions
(bhagaṇa) are meant to be taken from the beginning of the Kalpa, or of a Mahâyuga.
But, for practical purposes, we need not go back so far. One Twelve-Year Cycle is com-
pleted in one revolution. And, therefore, we can determine the saṁvatsara for a given
year, or for any given date in it, by taking Jupiter's signs, including the current sign, of the
current revolution, and counting from Âśvayuja. In this system, the signs are intended to
be taken, and in practice are taken, according to Jupiter's mean longitude. Now, suppose
that on a certain day Jupiter's mean longitude is 9 signs and 12 degrees; i.e. that he is in
the tenth sign. Then, counting from Âśvayuja, we have Âshâḍha, as the current saṁvatsara
for the given day.¹
The names of the saṁvatsaras of the Sixty-Year Cycle also,—according to that
which is usually called the northern system, but which is, in reality, the only truly astrono-

¹ In his treatment of the Twelve-Year Cycle (Indian Eras, p. 26 ff.), Gen. Sir A. Cun-
ningham opens the subject by quoting Kern's translation (see page 161 above, note 1) of the same
verse in the Bṛihat-Saṁhitâ which I use for the same purpose. But, for the rest, he altogether
ignores the force of the reference to Jupiter's rising, and treats the subject as if the heliacal-rising
system, described by me above, did not exist at all. What he has deduced from the verse, is only the
mean-sign system, which the verse does not really refer to at all; and in respect of which he is wrong
in making the saṁvatsaras of both the Twelve-Year and the Sixty-Year Cycle, begin and end with the
luni-solar years. The rules that he applies give, at the best, only the saṁvatsaras of the Twelve-Year
Cycle by the mean-sign system, and of the Sixty-Year Cycle, that are current at the commencement of a
given solar year; they do not provide for the essential point in both the cycles, viz. the determination
of the exact day on which a given saṁvatsara of either cycle commences—I have not had an
opportunity of seeing the writings of Davis and Warren on this subject. But, with all due deference,
I may safely say that the Twelve-Year Cycle by the heliacal-rising system, has hitherto remained
quite unknown to European scholars.—I would take this opportunity of remarking that I can find no
authority for the introduction, in Kern's translation, of the bracketed words "(during which Jupiter
completes a twelfth part of his revolution)." Jupiter completes a twelfth part (râśi) of his revolution
in about three hundred and sixty-one days; while the interval between two of his heliacal risings is
about four hundred days.
A b

mical system of it, and was current, as is shewn by many epigraphical quotations of it,
in Southern India also,—are determined by Jupiter's mean place. And the rule is thus
given in the Sûrya-Siddhânta, i. 55,—dvâdaśa-ghnâ Gurôr yâta-bhagaṇâ vartamânakaiḥ ।
râśibhiḥ sahitâḥ śuddhâḥ shashṭyâ syur Vijay-âdayaḥ ॥,—" the expired revolutions of Jupi-
ter, multiplied by twelve, (then) increased by the current signs (of the current revolution),
(and then) divided by sixty, are (i.e. the remainder is) (the saṁvatsaras counted from)
Vijaya as the first." In fact, the Sixty-Year Cycle, and the Twelve-Year Cycle of the
mean-sign system, are quite identical with regard to the day of the commencement of each
saṁvatsara; and, consequently, with regard also to its duration, which, according to
Jupiter's mear motion as given in the Sûrya-Siddhânta, is 361 days, 1 ghaṭî, 36 palas.
The Twelve-Year Cycle of the mean-sign system, therefore, with regard to its other details,
may be most properly described with the Sixty-Year Cycle, which I shall treat of fully on a
separate occasion. And I have introduced the present brief mention of it, only because
references to it are unavoidable in the following observations.
The eleven authorities for the heliacal-rising system, in addition to the Bṛihat-
Saṁhitâ, of which I have spoken, are as follows;—(1) Parâśara says—Kṛittikâ-
Rôhiṇîsh=ûditê kshuch-chhastr-âgni-vṛishṭi-vyâdhi-prâbalyaṁ ..... Chitrâ-Svâtyôr uditê
nṛipa-sasya-varsha-kshêm-ârôgya-karaḥ.—(2) Garga says—pravâs-ântê sah=arkshêṇa
hy=uditô yuga-pach charêt । tasmât kâlâd ṛiksha-pûrvô Gurôr abdaḥ pravartatê ॥—(3)
Kaśyapa says—saṁvatsara-yugê ch=aiva shashṭy-abdê¹-ṅgirasas sutaḥ । yan-nakshatr-
ôdayaṁ kuryât tat-saṁjñaṁ vatsaraṁ viduḥ ॥—(4) Ṛishiputra says—yasmin tishṭhati
nakshatrê saha yêna pravardhatê । saṁvatsaras sa vijñêyas tan-nakshatr-âbhidhânakaḥ
॥— (5) Also, in the following, Ṛishiputra quotes Vasishṭha and Atri, in addition to
Parâśara, who is separately referred to above,—Tishy-âdika-yugaṁ prâhur Vasishṭh-Âtri-
Parâśarâḥ । Bṛihaspatês tu Saumy-ântaṁ sadâ dvâdaśa-vârshikaṁ ॥ Udêti yasmin mâsê
tu pravâs-ôpagatô=ṅgirâḥ । tasmât saṁvatsaraḥ.—(6) In the Samâsa-Saṁhitâ, a
short work by Varâhamihira, we have²—Gurur udayati nakshatrê yasmin tat-saṁjñitâni
varshâṇi.—(7) Bṛihaspati, as quoted in the Kiraṇâvalî, a commentary by Dâdâbhâî
on the Sûrya-Siddhânta, says—yadâ Gur-ûdayô bhânôr Gurôr abdas tad-âditaḥ.—
(8) In the Nârada-Saṁhitâ, Guruchârâdhyâya, we have—yad-dhishṇy-âbhı uditô Jîvas
tannakshatr-âhva-vatsaraḥ.—(9) In the Muhûrta-Tattva, Guruchâra, verse 7, we have—
dvy-ṛikshô=Gnêḥ Kârttikât try-ṛiksha ishu-ravi-Śivô=bdaḥ sa yên=ôdit=Êjyaḥ.—(10)
In the Jyôtishadarpaṇa, adhyâya v., we have — yasminn abhyuditô Jîvas tan-

¹ From this, it appears that the saṁvatsaras of the Sixty-Year Cycle also, were originally determin-
ed by the heliacal risings of the planet. And corroboration of this is afforded by a verse of Varâhamihira,
quoted further on. The above verse of Kaśyapa, evidently applies to the Twelve-Year Cycle also.
² I have taken this, and the preceding five quotations, from Utpala's commentary on the
Bṛihat-Saṁhitâ.

THE GUPTA INSCRIPTIONS: APPENDIX III. 171
nakshatrasya vatsaraḥ. — It is unnecessary to translate these ten quotations; as the
general purport of all of them is the same; viz. that "a saṁvatsara is to be named
after that nakshatra in which Jupiter attains his rising."—(11) In the remaining au-
thority, the Sûrya-Siddhânta, the rule appears to be a little different from that
of all the other authorities; though the general features are the same. And, for
this reason, though the work is one of the earliest, I have quoted it last. The rule
is given in the verse—Vaiśâkh-âdishu kṛishṇê cha yôgaḥ pañchadaśê tithau l Kârttik-
âdîni varshâṇi Gurôr ast-ôdayât tathâ ll,—"and the years Kârttika and others (that
follow) (are to be named) from the rising, after setting, of Jupiter,¹ according to the occur-
rence (of Kṛittikâ and other nakshatras)² on the fifteenth tithi in the dark fortnight of
Vaiśâkha and other (months that follow)." The application of this rule seems to be
thus:—The years are to be named, according as Kṛittikâ and other nakshatras,—
evidently those in which the sun and the moon stand,—occur on the amâvâsyâ of Vaiśâ-
kha and others, immediately preceding or following³ the day on which Jupiter rises;
that is, on whatever day Jupiter may rise, the nakshatra on either the preceding, or
the following, amâvâsyâ, gives the name to the saṁvatsara. This rule seems to have
been alluded to, though not in the name of the Sûrya-Siddhânta, by Utpala; but it is
rejected by him. It may possibly have been occasionally in use; but it certainly does not
apply to the Gupta records.
Now, all these quotations distinctly refer, in some form or another, to the rising of
Jupiter; which it is impossible to understand as meaning anything but his heliacal-
rising. And, if the rising referred to is the heliacal-rising, then no astronomer can deny
that, in the period of about twelve solar years, there are only eleven conjunctions of Jupiter
¹ I should remark that Raṅganâtha, one of the best commentators on the Sûrya-Siddhânta,
explains this verse with the understanding, into which he seems to have been led by Utpala's com-
ments on the Bṛihat-Saṁhitâ, viii. 1, that ast-ôdayât means "from the setting or rising." He adds,
however, — idânîm udaya-varsha-vyavahârô gaṇakair gaṇyatê,—"at the present time, the practice
of (naming) the year by the rising, is taken into account by astronomers." For the sake of consist-
ency, it is justifiable to translate the expression ast-ôdayât as I have translated it above; and the
use of the compounded base in the singular, supports that translation. I also find that Dâdâbhâî,
in his commentary on the Sûrya-Siddhânta, explains it in the same way; his words being —
tath=âstâd udaya-kâlê Gurôs tad-yukta-nakshatra-saṁjñô Gurôr abdô jñêyaḥ,—"so the year of
Jupiter is to be known as having the appellation of that nakshatra with which he is joined at the
time of (his) rising after setting."—[There can be no doubt that ast-ôdayât is used here in the sense
of "from the rising after setting;" compare such compounds as supt-ôtthita, 'having arisen from
sleep;' lit. 'having slept and then having risen.'—J. F. F.]
² These words, in connection with yôgaḥ, have to be supplied from the preceding verse, in which
the result exhibited in Table VIII. page 163 above, is given.
³ Which of the two, cannot be determined from the context: and I can find no commentator
who has explained the point properly. I myself think that the following amâvâsyâ is intended. But
seems that Utpala, in his allusion to this rule, takes the preceding amâvâsyâ.

• 2b

172 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
and the sun, and consequently eleven heliacal risings of the planet. And, this fact being
established, the interval of about four hundred days between two successive risings, —the
same period, for the duration of each saṁvatsara,—the omission of one saṁvatsara in
each cycle of twelve years,—and all the other points described by me,—follow of necessity
It will, doubtless, have been noticed that I have not been able to give any quotations
from the first Âryabhaṭa (born A.D. 476) or Brahmagupta (born A.D. 598), in support
of the heliacal-rising system. And it might be sought to base some argument against
its existence, on the grounds that these two early authorities, who moreover may be said to
be the originators of two of the three schools of astronomers in India, are silent about it,
though they do give the mean-sign system. But the facts only prove the early existence
and use of the mean-sign system; which I do not seek to deny. They do not prove, either
that the heliacal-rising system did not exist; or that, having existed, it had gone out of use
in their time. To take another instance, Âryabhaṭa and Brahmagupta give no rule for
finding the saṁvatsaras of the Sixty-Year Cycle; but, to say that this cycle was not
known to either of them, would hardly be sensible. The mean-sign system for the
Twelve-Year Cycle is undoubtedly early. But the heliacal-rising system is earlier still.
Among the authorities quoted above, the Sûrya-Siddhânta¹ is as early a work as the
Âryabhaṭîya, if not earlier. And, that Parâśara, Garga, and Kaśyapa, are earlier than
Âryabhaṭa, cannot be denied. Utpala quotes a verse of Garga, which, with some proposed
emendations, I give as it stands in my manuscript copy,—êvam Âśvayujaṁ ch=aiva Chaitraṁ
ch=aiva Bṛihaspatiḥ | saṁvatsarô(?raṁ) nâma(?śa)yatê saptaitêda (? saptaty-abda)-śatê=
dhikê ||.² This verse seems to refer to the mean-sign system; though I can say nothing
definitely about it, without seeing the context; of which I have not the opportunity at
present. But, supposing that it does refer to the mean-sign system, still it is not un-
favourable to the heliacal-rising system; for the same Utpala also quotes a verse of the
same Garga, given at page 170 above, in which Garga says—"when Jupiter, after dwelling
together (with the sun),³ rises and walks along with a nakshatra, then commences the year
of Jupiter, of which the first part (of the name) is the (name of that) nakshatra."
Not only that the heliacal-rising system is very early, but that it alone is the
original system of the Twelve-Year Cycle, is self-evident. The heliacal rising of Jupiter
is a natural phenomenon. No scientific apparatus is needed for the observation of it; nor are
¹ To say that the Sûrya-Siddhânta is a work of Varâhamihira, is simply a mistake. I cannot
enter iuto this point at present; but would refer any who hold that view, to Varâhamihira's Pañcha-
Siddhântikâ, and to Kern's Preface to his edition of the Bṛihat-Saṁhitâ.
² [Gen. Cunningham also has quoted this verse (Archæol. Surv. Ind. Vol. X. p. 114); and in
such a way, his authority giving nâśyatê, where Mr. Sh. B. Dikshit's MS. gives nâmayatê, as to shew
that, in the second half of it, we certainly have to read nâśayatê saptaty-abda-śatê.—J. F. F.]
³ pravâsantê may be translated by "after having completed his journey;" i.e. having completed
a revolution of some kind or another.

THE GUPTA INSCRIPTIONS ; APPENDIX III. 173
any calculations required.¹ But such is not the case with the passing of Jupiter from one
sign of the zodiac into another. No kind of apparatus can determine Jupiter’s mean longitude.
It is to be found only by calculations, the rules for which must have been established
after observations carried on during a very long time ; to lay down the means of determining
Jupiter's mean yearly or daily motion, is not a thing that can be accomplished in a few
years. It seems, therefore, quite evident that the system of naming the saṁvatsaras after
the nakshatras in which Jupiter rises, i.e. the heliacal-rising system, is the only origi-
nal one.
So much, as regards the antiquity of the heliacal-rising system. Let us now look into its
somewhat later use. Varâhamihira is later than Âryabhaṭa. And his verse with which
I have opened my account of the Twelve-Year Cycle, undoubtedly refers to this system.
So, also, another of his verses, in the Brihat-Saṁhitâ, viii. 27, points to the same thing.
It runs—âdyaṁ Dhanishṭhâṁ samabhiprapannô² Mâghê yad=âyâty=udayaṁ Surêjyaḥ |
¹ The phenomenon, moreover, is one to which much attention has always been paid in India.
Even in the present day, almost all the Hindu Pañchâṅgs give the dates of the heliacal rising and
setting of Jupiter. This is for religious purposes ; since, when Jupiter is invisible, some duties and
ceremonies, such as investiture with the sacred thread, marriage, pilgrimages, &c., are not to be per-
formed ; and the dates in question are necessary, in order to know when he is visible, and when he
ceases to be so. In Pañchâṅgs printed in Mahârâshṭra, the custom of naming the saṁvatsaras after
the months is not in use ; in fact, the Twelve-Year Cycle is now almost unknown to the people at
large of the Dekkan. But in a Pañchâṅg prepared by Jwâlâpati Siddhânti, and printed in the
Jñânasûryôdaya Press at Madras, I find that the author, in the saṁvatsara-phala, after giving the two
saṁvatsaras of the Sixty-Year Cycle, according to the northern and the southern reckoning, goes on
to say—sarvatra Gur-udaya-vaśât (Chaitr)-âbdô grâhyaḥ,—" all over (India) a (certain) year (i.e.
such saṁvatsaras as Chaitra, &c.), which depends on the rising of Jupiter, is to be taken ;" from
which it is evident that some such rising-system, as that described above, is followed in this Pañchâṅg
from Madras. Also, in the Chaṇḍu-Pañchâṅg, prepared for the meridian of Jôdhpur in Mârwâḍ, and
used there and by Mârwâḍîs in other parts of India, the saṁvatsaras are named as Chaitra-Varsha,
Vaiśâkha-Varsha, &c. ; the system adopted in that Pañchâṅg, however, for naming the saṁvatsaras,
is the mean-sign system.
² The reading that I give, is from my manuscript copy. But, s and ś being very liable to confu-
sion, the reading âdyaṁ Dhanishṭh-âṁśam abhiprapannô, which is given in some printed copies, may
also be of some authority. By calculation, however, I find that Jupiter does not always stand just at
the beginning of Dhanishṭhâ, at the time of his rising at the commencement of the Prabhava saṁvat-
sara. And I think, therefore, that the reading which I give, is the author's original reading. I take
âdyaṁ to qualify the word nakshatram understood. And Varâhamihira calls the nakshatra in ques-
tion "the first," because his opinion was, I think, that Dhanishṭhâ was the first nakshatra of the
Sixty-Year Cycle, as it was of the Five-Year Cycle of the Vêdâṅga-Jyautisha [see its verse—svar
âkramêtê sôm-ârkau yadâ sâkaṁ sa-Vâsavau syât tad=âdi-yugaṁ Mâghaḥ,—" when the moon and
the sun, being in (the nakshatra) Vâsava (Dhanishṭhâ), go together in the heavens, then there is the
beginning of the Yuga (and of the month) Mâgha"], which two he intends to reconcile.—[Kern would
seem to have known only the reading âdyaṁ Dhanishṭh-âṁśam. His translation (Jour. R. As. Soc.
N. S. Vol. V. p. 49 f.), which follows also the reading of pravartatê for prapadyatê, is—" when Jupi-
ter, on reaching the first quarter of Dhanishṭhâ, rises in the lunar month of Mâgha, then begins the
first year of the cycle of sixty, named Prabhava, a year salutary to all beings." And he adds the

174 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
shashṭy-abda-pûrvaḥ Prabhavaḥ sa namnâ prapadyatê bhûta-hitas tad=âbdaḥ ||.—
“when Jupiter, having come to the first (nakshatra) Dhanishṭhâ, attains (his) rising in
(the month) Mâgha, then commences that year, beneficial to beings, which is named
Prabhava, the first of the sixty years.” Now, in eighteen centuries from the beginning
of the Śaka era, the Prabhava saṁvatsara has occurred thirty times. And, by approxi-
mate calculations, I find that Varâhamihira’s assertion, though not quite correct, is
nearly so by the heliacal-rising system. Out of the thirty occasions, Jupiter rose, accord-
ing to the Amânta southern arrangement of the fortnights,¹ twenty-six times in the month
of Mâgha ; three times early in Phâlguna ; and once in Pausha, nearly at the end of
it. And, on many occasions he rose in the Dhanishṭhâ nakshatra ; and on some, in
Śravaṇa. The assertion, however, is not correct for the mean-sign system. According
to that system, at the commencement of the Prabhava saṁvatsara, which always begins
when Jupiter, by his mean longitude, enters Kumbha or Aquarius, he evidently always
stands in the middle of Dhanishṭhâ.² But the Prabhava saṁvatsara may begin, and
as a matter of fact I find by calculation that it has begun, in any of the twelve months ;
as will be shewn when I treat of the Sixty-Year Cycle. It is evident, therefore, that the
verse in question supports the heliacal-rising system. Moreover, the word udaya itself is
used in it. It might be argued that Varâhamihira’s rule (Bṛihat-Saṁhitâ, viii. 22 ;
êk-aikam abdêshu nav-âhatêshu, &c.) for finding the nakshatra of Jupiter, has reference
to his mean longitude ; and, consequently, that it gives the mean-sign system. But it
is absurd to suppose that so learned an astronomer as he, stating the rule that a saṁvatsara
is to be named from the nakshatra in which Jupiter rises, did not know that the nak-
shatra, at the time of his rising, can be determined only from his apparent longitude.
The reason why he gives a rule for finding the mean longitude only, is, in my opinion,
this : — Anyone conversant with Hindu mathematical astronomy, knows how tedious a
task it is to find the mean longitude of a planet from the ahargaṇa, or number of elapsed
days, counted from the beginning of the Kalpa or of a Yuga. But, when once it has been
found, it is much less tedious to calculate, from it, the apparent longitude by the general
rules. So, Varâhamihira, I think, gives a simple rule for finding the mean longitude of
Jupiter on a given day, and leaves to the astronomer the calculation of the apparent longi-
tude. Or, it may also be said, as the difference between the mean and the apparent
longitudes of Jupiter, on the day of his heliacal-rising, is not very much,—sometimes about
five degrees, and sometimes nearly nil, as I have found from several calculations,—
Varâhamihira gives the simple rule for rough practical purposes only ; leaving it to the
note—“cf. Davis, Asiatic Researches, Vol. III. p. 220. The word aṁśa also means ‘degree,’ but is
here rendered by ‘quarter,’ according to Utpala, who explains it by pâda.”—[J. F. F.]
¹ See page 148 above, note 1.
² The sign Kumbha commences in the middle of Dhanishṭhâ, and ends at the end of the third
quarter of Pûrvâ-Bhâdrapadâ.

THE GUPTA INSCRIPTIONS; APPENDIX III. 175
astronomer to make more detailed calculations, when absolute accuracy is required. But,
even supposing that the said verse of Varâhamihira implies the mean-sign system, it will
only prove that he gives both systems.
Let us now turn to the more modern period. Of the authorities quoted above in
support of the heliacal-rising system, the Muhûrta-Tattva and the Jyôtishadarpaṇa are
comparatively modern. The author of the first is the father of the well-known Gaṇêśa
Daivajña, the author of the Grahalâghava; and, therefore, its date is about Śaka-Saṁvat
1420 expired (A.D. 1498-99); and it was written at Nandgaum on the western coast,
about forty-five miles south of Bombay. The date of the second, as I have determined
from certain data in it, is Śaka-Saṁvat 1479 expired (A.D. 1557-58); and it was written
at Koṇḍapalli, somewhere in the Karṇâṭaka country. And, in addition to these two, a
third reference to the heliacal-rising system has been given above, at page 171 note 1,
from Raṅganâtha's commentary on the Sûrya-Siddhânta. The date of this commentary
is Śaka-Saṁvat 1525 expired (A.D. 1603-1604); and it was written at Benares. These
details suffice to shew that the heliacal-rising system is referred to in works of a modern
period, and belonging to different parts of the country.
It is true that the use of the Twelve-Year Cycle is rare. This is shewn by the
fact that some astronomical works do not refer to it, in either system, though it would not have
been beyond their sphere to do so; and by the fact that, out of many hundreds of inscrip-
tions, only seven have been found, mentioning the saṁvatsaras of it.¹ But, if the use of
the heliacal-rising system is rare, the use of the mean-sign system is still more so. The saṁ-
vatsaras mentioned in the four records of the Early Gupta period, which give full details
for calculation, are proved to belong to the heliacal-rising system. And, while I have
been able to quote no less than eleven authorities referring to that system, a distinct reference
to the mean-sign system is to be found in only two; viz. Âryabhaṭa and Brahmagupta,
as quoted above. If to these we add an interpolated verse in the Siddhânta-Śirômaṇi,
and also hold that Garga and Varâhamihira give the system, even then the number comes to
only five.² And, in about eight of such works as would be expected to refer to the system,
of a date later than A.D. 1478, which I have examined, I find a reference to it in none at all.
Looking to the present time, if the mean-sign system is used in one part of the
country, the heliacal-rising system is used in another;³ while, in others, both systems are
almost, if not quite, unknown. The mean-sign system is doubtless the more convenient

¹ viz. the five records of the Mahârâjas Hastin and Saṁkshôbha; and the two grants of the
Kadamba chieftain Mṛigêśavarman, referred to at page 106 above, note 1.
² A verse containing a reference to the system, is given, as belonging to Śrîpati, in Pandit
Bapu Deva Shastri's edition of the Siddhânta-Śirômaṇi, p. 13, note; but I do not find it in the
Ratnamâlâ of Śrîpati.
³ See page 173 above, note 1.

176 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
of the two; since the duration of the saṁvatsara, roughly three hundred and sixty-one
days, approaches so near to that of the solar year; and the omission of a saṁvatsara
occurs only once in about eighty-five years. And this point of convenience seems to be the
reason for which the system was invented by the astronomers. But, as both the systems
are given in astronomical works, both of them must certainly have been actually used.
And such will always be the case, as long as those works exist.

THE GUPTA INSCRIPTIONS; APPENDIX IV. 177

APPENDIX IV.

The Chronology of the Early Rulers of Nêpâl.

The inscriptions which give the matter for this Appendix all belong to Khâṭmâṇḍu
and its vicinity, and consist of fifteen collected by the late Dr. Bhagwanlal Indraji,
and published by him and Dr. Bühler in the Ind. Ant. Vol. IX. p. 163 ff.; and of four
obtained by Mr. Bendall, of which the most important one was published originally in
the Ind. Ant. Vol. XIV. p. 97 f., and has been reprinted in his Journey in Nêpâl and
Northern India, p. 72 ff. and Plate viii., with the first publication of the other three,
p. 74 ff. and Plates ix. x. and xi.
The historical results of the former set were discussed at length by Dr. Bhag-
wanlal Indraji in his paper entitled "Some Considerations on the History of Nêpâl," edited
by Dr. Bühler, and published in the Ind. Ant. Vol. XIII. p. 411 ff. His results, how-
ever, are unfortunately vitiated by a radical error; viz. the reference of one series of the
dates to the Vikrama era, instead of to the Gupta era nearly four hundred years later.
This was due, partly to the misinterpretation of an important verse in the inscription of
Jayadêva II. of Harsha-Saṁvat 153; and partly to the want of the key-note supplied by
Mr. Bendall's inscription of Gupta-Saṁvat 316. And it was, of course, the publication of
this last inscription that led me to look carefully into the whole matter, and at length to hit
upon the fundamental mistake, without a recognition of which it might still be argued that
Mr. Bendall's date of 316, for Śivadêva I. and Aṁśuvarman, stands alone in belonging to
the Gupta era, and that, in spite of it, Dr. Bhagwanlal Indraji was right in referring the
other larger dates to the Vikrama era.
Taking the inscriptions in question in regular chronological order, the dates and
other important points are as follows :—
A.—Mr. Bendall's inscription No. 1; Ind. Ant. Vol. XIV. p. 97 f., and Journey in
Nêpâl, p. 72 ff. and Plate viii.—The charter recorded is issued from the house or palace
called Mânagṛiha (line 1). The inscription is one of the Bhaṭṭâraka and Mahârâja, the
illustrious¹ Śivadêva I. (l. 2 f.), the banner or glory of the Lichchhavikula.² And it
records a grant made by him on the advice, or at the request,³ of the Mahâsâmanta
Aṁśuvarman (l. 6 f.). The Dûtaka is the Svâmin Bhôgavarman (l. 15).⁴ The date

¹ śrî. ² Lichchhavi-kula-kêtu, line 2.
³ mahâsâmant-Aṁśuvarmmaṇâ vijñâpitêna mayâ; line 6-7.
⁴ This person would seem to be Aṁśuvarman's sister's son, who is mentioned in inscription E.
below.—He must not be confused with Śivadêva II.'s father-in-law, the Maukhari Bhôgavarman, who
is mentioned in inscription R. below, and was at least a full generation later.—See page 180 below,
note 3, on another point.

178 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
(l. 15), in numerical symbols here and throughout the series, is (Gupta)-Saṁvat 316,
Jyaishṭha-śukla-divâ-daśamyâm; which, with the epoch of A.D. 319-20, answers, approxi-
mately, to the 3rd May, A.D. 635.
B.—Dr. Bhagwanlal Indraji's inscription No. 5; Ind. Ant. Vol. IX. p. 168 f.—The
charter is issued from Mânagṛiha (line 1). The inscription is one of the Bhaṭṭâraka
and Mahârâja, the illustrious Śivadêva I. (l. 2 f.), the banner of the Lichchhavikula.
It recorded some act, the details of which are broken away and lost, done by him, as in
inscription A., on the advice, or at the request, of the Mahâsâmanta, the illustrious
Aṁśuvarman (l. 8 f.).¹ The date, and the name of the Dûtaka, are broken away and
lost in line 11 ff.
C.—Dr. Bhagwanlal Indraji's inscription No. 6; Ind. Ant. Vol. IX. p. 169 f.—The
charter is issued from the house or palace called Kailâsakûṭabhavana (line 1). The
inscription is one of the Mahâsâmanta, the illustrious Aṁśuvarman (l. 2). The
Dûtaka is the Mahâsarva[daṇḍanâ]yaka Vikra[masêna] (l. 14).² The date (l. 14)
is (Harsha)-Saṁvat 34, Jyaishṭha-śukla-daśamyâm; which, with the epoch³ of A.D.
605-606, answers, approximately, to the 18th May, A.D. 639.
¹ As in page 177 above, note 3.
² This is the name as completed by Dr. Bhagwanlal Indraji. But, if we accept it as correct, we
must be careful not to confuse this person with the Râjaputra Vikramasêna who was the Dûtaka of
inscription S. below, more than two hundred years later.—I have not seen the original rubbing of
this inscription.
³ This epoch is taken from Mr. Sh. B. Dikshit's calculation of the date in the Dighwâ-Dubaulî
grant of the Mahârâja Mahêndrapâla. At the end (Ind. Ant. Vol. XV. p. 113, line 14) we have, for
the assignment of the charter, the year 155, Mâgha śu di 10; and from line 12 we learn that, when
making the grant, the Mahârâja had bathed on the occasion of the Kumbha-Saṁkrânti or entrance
of the Sun into Aquarius. The epoch of the era is fixed approximately by a statement of Albêrûnî
(Albêrûnî's India, Translation, Vol. II. p. 5), to the effect that, in a Kaśmîrî almanac, he had read
that Harshavardhana was six hundred and sixty-four years later than Vikramâditya. Subject to some
doubt as to whether Albêrûnî is speaking of the epoch of the era, or of its first current year, and as to
whether the Vikrama year indicated by him is to be taken as current or as expired, this statement
indicates, approximately, Śaka-Saṁvat 528 expired, and 529 current (A.D. 606-607), as the epoch or
year 0 of the era. And this is the epoch selected by Gen. Cunningham (Indian Eras, pp. 64 f., 157 ff.),
on the authority of Albêrûnî's statement. By calculations based on the Sûrya-Siddhânta, however,
Mr. Sh. B. Dikshit finds that an epoch one year earlier than this is required, in order that the details
of the date in the Dighwâ-Dubaulî grant may work out correctly. Thus, with the basis of Harsha-
Saṁvat 155 + Śaka-Saṁvat 527 expired = Śaka-Saṁvat 682 expired, the Kumbha-Saṁkrânti
occurred at 43 ghaṭîs, 40 palas, on Monday, the 19th January, A.D. 761; and the ninth tithi of the
bright fortnight of Mâgha ended about four hours later, at 53 ghaṭîs, 15 palas, on the same day. Or,
by the Ârya-Siddhânta, the times were, for the Saṁkrânti, 42 ghaṭîs, 30 palas; and for the tithi,
52 ghaṭîs, 41 palas. In both cases, the ghaṭîs and palas are from mean sunrise at Ujjain. And he
adds the note—"the puṇyakâla or meritorious time for performing religious duties in connection
with a saṁkrânti, is, according to some authorities, from sixteen ghaṭîs before, to sixteen ghaṭîs after,
the time of the saṁkrânti; and, according to others, in the case of some saṁkrântis, from ten to

THE GUPTA INSCRIPTIONS; APPENDIX IV. 179
D.—Mr. Bendall's inscription No. 2; Journey in Nêpâl, p. 74 f., and Plate ix.—
The charter is issued from Kailâsakûṭabhavana (line 1). The inscription is one of an
illustrious Mahâsâmanta, whose name is effaced, but may with tolerable certainty be
restored as Aṁśuvarman (l. 2). And it records the grant of two fields, for the purpose
of providing for the repairs of some building. The Dûtaka (l. 17) is the Mahâbalâdhyaksha
Vindusvâmin. The date (l. 16) is (Harsha)-Saṁvat 34, prathama-Pausha-śukla-
dvitîyâyâm; which should answer,¹ approximately, to the 3rd December, A.D. 639.
E.—Dr. Bhagwanlal Indraji's inscription No. 7; Ind. Ant. Vol. IX. p. 170 f.—The
charter is issued from Kailâsakûṭabhavana (line 1). The inscription is one of the
illustrious Aṁśuvarman (l. 5), to whom no formal titles are allotted here. It mentions
sixteen or even forty ghaṭîs before the saṁkrânti, up to the time of the saṁkrânti, and, in the case
of the other saṁkrântis, from the time of the saṁkrânti, up to from ten to sixteen or even forty
ghaṭîs after it. But generally, when a saṁkrânti occurs at midnight,—as it did, roughly, in the
present instance,—the puṇyakâla is held to be on the following day. And, in the present case, there
is every probability of the bathing &c. having been performed on the following day; i.e. on Mâgha
śukla daśamî."—With other years, the given saṁkrânti cannot be brought into accordance with the
given day and tithi. Thus, with Śaka-Saṁvat 680 expired, the saṁkrânti occurred six or seven days
after the given tithi, on the Pûrṇimânta Phâlguna kṛishṇa 2; with 681 expired, on the Pûrṇimânta
Phâlguna kṛishṇa 13; with 683 expired, on the Pûrṇimânta Phâlguna kṛishṇa 5; with 684 expired,
on Mâgha śukla 1; and with 685 expired, on Mâgha śukla 12. It is certain, therefore, that the
correct English equivalent is the one obtained with Śaka-Saṁvat 682 expired. And, taking the given
Harsha year as a current year, and treating it as having the same scheme with the northern Śaka
year, the epoch is Śaka-Saṁvat 527 expired, or A.D. 605-606 current; and 528 years have to be
added, to convert current Harsha years into current Śaka years, and 527, to convert them into expir-
ed Śaka years as the basis of calculations. The commencement or first current year of the era, is A.D.
606-607 current.—I should remark, as I did when editing the grant, that a slight doubt may be felt as to
the value of the third numerical symbol, in the units place, which I interpret as 5. And the exact epoch
would be varied, by allotting a different value to this symbol. But the choice lies only between 4, 5, and
8. I do not think that any authority can be found for interpreting it as 4. And, if it were interpreted
as 8, then the epoch of the era would be A.D. 602-603; and its first current year would be A.D. 603-
604, for which year, as the commencement of the reign of Harshavardhana, there seems to be no
authority whatever.—I should also point out that the next inscription, D. below, of Harsha-Saṁvat 34,
refers to an intercalation of the month Pausha; and that Prof. Adams of Cambridge, and Prof. Schram
of Vienna, as reported by Dr. Bühler in the Ind. Ant. Vol. XV. p. 338 (see also Journey in Nêpâl,
p. 76), obtained a mean intercalation of Pausha in A.D. 640; which would necessitate the adoption
of A.D. 606-607 as the epoch. But this intercalation remains to be examined in detail; and, for the
present, I have only a note by Mr. Sh. B. Dikshit that, according to the verse Mêsh-âdi-sthê savitari
&c. (see page 88 above, note 5), the intercalated period should be called Mârgaśîrsha; and that, in
calling it Pausha, exactly the contrary is done of the process by which we obtain Mârgaśîrsha as the
name of the intercalated period that is referred to in the Kaira grant of Dharasêna IV. of the year
330 (see page 93 ff. above).—The question will, I hope, be treated more fully by Mr. Dikshit hereafter.
Meanwhile, for present purposes, I adopt A.D. 605-606 as the epoch.
¹ As regards the equivalent English date, however, see a remark, in the preceding note, on the
indicated intercalary nature of this month Pausha.
y 8

180 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Aṁśuvarman’s sister Bhôgadêvî, who was the wife of the Râjaputra Sûrasêna,¹ and the
mother of the illustrious Bhôgavarman² and Bhâgyadêvî. It records certain orders ad-
dressed by Aṁśuvarman to the officials of the Western Province (paśchim-âdhikaraṇa-
vṛittibhujah, lines 5-6; and paśchim-âdhikaraṇa, line 14), in connection with three liṅga
forms of the god Îśvara or Śiva. The Dûtaka is the Yuvarâja Udayadêva.³ The date
(l. 22) is (Harsha)-Saṁvat 39, Vaiśâkha-śukla-divâ-daśamyâm; answering, approximately,
to the 24th April, A.D. 644.
F.—Dr. Bhagwanlal Indraji’s inscription No. 8; Ind. Ant. Vol. IX. p. 171.—This is not
a formal charter, issued from any specified place. It simply records that, by the favour of
the illustrious Aṁśuvarman (line 2), a praṇâlî or ‘conduit’ was caused to be built by
the Vârtta⁴ Vibhuvarman, for the increase of his father’s religious merit. The date (l. 1)
is (Harsha)-Saṁvat⁵ 44 or 45, Jyêshṭha-śukla . . . . . . . . . . . ., which has its equivalent in
the month of May or June, A.D. 649, or May, A.D. 650.
G.—Dr. Bhagwanlal Indraji’s inscription No: 9; Ind. Ant. Vol. IX. p. 171 f.—The charter
is issued from Kailâsakûṭabhavana (line 3 f.). The inscription is one of the illustrious
Jishṇugupta (l. 4). It records that, at the request of the Sâmanta Chandravarman, a
tilamaka or ‘water-course,’ which had been constructed by the Bhaṭṭâraka and Mahârâjâ-
dhirâja, the glorious Aṁśuvarman (l. 6 f.), was made over by Jishṇugupta to Chandra-
varman, to be repaired; &c. The Dûtaka (l. 21) is the Yuvarâja, the illustrious Vishṇu-
gupta. The date (l. 21) is (Harsha)-Saṁvat 48, Kârttika-śukla-2; answering, approxi-
mately, to the 30th September, A.D. 653.—Lines 1 and 2 of this inscription mention
Mânagṛiha,⁶ and, in connection wich it, the Bhaṭṭâraka and Mahârâja Dhruvadêva;
but the passage is much mutilated, and does not in itself suffice to explain the connection
between Dhruvadêva and Jishṇugupta.⁷
¹ The original is somewhat damaged, but it has undoubtedly s in the first akshara of this name;
not ś as given by Dr. Bhagwanlal Indraji in his text and translation. And the spelling is the same
in line 3 of an inscription of the Sûrasêna family published by him in the Ind. Ant. Vol. X. p. 34 ff.
But very probably ś is the more correct spelling; especially as line 8 of the present inscription men-
tions the liṅga named Śûrabhôgêśvara.
² See page 177 above, note 4.
³ The dates seem to prevent the identification of this person with the Udayadêva of the Ṭhâkurî
family, mentioned in inscription R. below.—He was perhaps a Lichchhavi, as suggested by Dr. Bhag-
wanlal Indraji. If so, in this Lichchhavi Dûtaka of a Ṭhâkurî charter, we have an instance parallel to
the Ṭhâkurî Dûtaka of a Lichchhavi grant, suggested in page 177 above, note 4.
⁴ Dr. Bhagwanlal Indraji took Vârta as a family or tribal name. It seems more likely, however,
tnat, as suggested by himself, it is an official title, synonymous with the vṛittibhuj of inscription E.
above; and that the correct form is vârtta (with the double t) from vṛitti.
⁵ The second symbol is doubtful; but it is either 4 or 5.—I have not seen the original rubbing of
this inscription.
⁶ The two syllables mâna, immediately after svasti, are clear enough in the original rubbing,
though they hardly shew in the lithograph, and were not given by Dr. Bhagwanlal Indraji in his text.
⁷ The purport of it, however, must have been the same as in the next inscription H.

THE GUPTA INSCRIPTIONS; APPENDIX IV. 181
H.—Dr. Bhagwanlal Indraji's inscription No. 10; Ind. Ant. Vol. IX. p. 173 f.—The
charter is issued from Kailâsakûṭabhavana (line 6). The inscription is one of the illus-
trious Jishṇugupta (l. 7). The details are much mutilated; but the subject is a tilamaka
or 'water-course,' which had been constructed by the Mahâsâmanta, the illustrious¹......
dêva. The name of the Dûtaka, and the date, have peeled off and are lost.—Lines 3 and
4 of this inscription again mention Mânagṛiha, and, in connection with it, the Bhaṭṭâra-
ka and Mahârâja, the illustrious Dhruvadeva, the banner of the Lichchhavikula. Be-
tween mânagṛiha and dita-chitta in line 4, there are four (or perhaps five) aksharas
which are much damaged, and are quite uncertain even in the rubbing; but the ha (not
hâ) is distinct enough; and this, and the whole construction, shews that we have here,—
not Mânagṛihât, the formal ablative of issue,—but the base Mânagṛiha, as the first mem-
ber of a compound, ending with santati, that qualifies śrî-Dhruvadêva, and means some-
thing like "who belonged to a lineage which had its thoughts [gladdened by residing] at
Mânagṛiha." As regards the connection between Dhruvadêva and Jishṇugupta,—in line
5, after śrî-Dhruvadêva, Dr. Bhagwanlal Indraji read purassaré sakala, and interpreted
the passage as shewing that Jishṇugupta acknowledged Dhruvadêva as his lord paramount.
The interpretation is possibly correct; though the text may equally well mean nothing
more than a courteous reference by Jishṇugupta to one whose position was equal to his
own. But, from the rubbing, the proper reading is certainly purassara-sakala, in compo-
sition. The whole passage, in fact, from Mânagṛiha, line 4, down to mânasa, lines 5-6, is
one continuous compound, meaning, in the latter part, that Jishṇugupta's mind had been
brought into a state of contentment by devising the means of freeing from calamity all the
people headed by Dhruvadêva, &c. And the expression Dhruvadêva-purassara-sakala-
jana is analogous in its purport to the sa-nṛipatêr=jjagatô (hitâya), "(for the welfare) of
the world (i.e. of all the people), together with the king," of line 2 of inscription M. below.
I.—Dr. Bhagwanlal Indraji's inscription No. 11; Ind. Ant. Vol. IX. p. 174.—This is
not a formal charter, issued from any specified place. It simply records certain grants
made in the augmenting victorious reign of the illustrious Jishṇugupta (line 9). The
era-date, if any was recorded, is lost with lines 27 ff.
J.—Mr. Bendall's inscription No. 3; Journey in Nêpâl, page 77 ff. and Plate x.—
The charter is issued from Kailâsakûṭabhavana (line 1). The king's name is lost,
somewhere in lines 2 to 7. The inscription provides for the worship of the god Vajrêśvara
(l. 17), and other religious matters. The Dûtaka (l. 28) is the Bhaṭṭâraka, the Yuvarâja
Skandadêva. The date (l. 29) is (Harsha)-Saṁvat 82, Bhâdrapada śukla-divâ.......;
which has its equivalent in the month of August, A.D. 687.

¹ In the original rubbing, śrî is very distinct at the beginning of line 14. Two aksharas are
illegible between it and dêvêna.

182 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
K.—Dr. Bhagwanlal Indraji’s inscription No. 1; Ind. Ant. Vol. IX. p. 163 ff.—This
furnishes the following short genealogy:—Vṛishadêva (line 3):—his son Śaṁkaradêva
(l. 12), with a reference in lines 9 and 10 to other sons, who, however, are not mentioned
by name;—Śaṁkara’s son, Dharmadêva (l. 16), who with justice [governed] a great
hereditary kingdom (kula-kram-âgatam . . . . . . . . . . râjyaṁ mahat, line 17), and whose
wife was Râjyavatî;—and Dharmadêva’s son Mânadêva (l. 20), who was appointed
to reign by his mother on his father’s death. The inscription then records that
Mânadêva made an expedition to the East, and reduced to obedience some “roguish,”
ill-behaved, or rebellious Sâmantas there; and then went back to the West, where also
he heard of misdoings on the part of a Sâmanta. Here, unfortunately, the available
part of the inscription terminates, the rest being buried underground. The date (l. 1 f.)
is (Gupta)-Saṁvat 386, Jyêshṭha-mâsa-śukla-pakshê pratipadi 1 Rôhiṇî-nakshatra-yuktê
chandramasi muhûrtê praśastê=Bhijiti; the exact equivalent of which, as we have seen at
page 95 ff. above, is Tuesday, the 28th April, A.D. 705.
L.—Dr. Bhagwanlal Indraji’s inscription No. 12; Ind. Ant. Vol. IX. p. 174 ff.—The
charter is issued from Kailâsakûṭabhavana (line 1). The inscription is one of the
Paramabhaṭṭâraka and Mahârâjâdhirâja, the glorious Śivadêva II. (l. 3). The Dûtaka
(l. 23) is the Râjaputra Jayadêva. The date (l. 23) is (Harsha)-Saṁvat 119, Phâlguna-
śukla-divâ-daśamyâm; answering, approximately, to the 20th February, A.D. 725.
M.—Dr. Bhagwanlal Indraji’s inscription No. 2; Ind. Ant. Vol. IX. p. 166 f.—This
is not a formal charter, issued from any specified place. It simply records that, by the
favour of the illustrious ‘king’ Mânadêva (line 1), a certain Jayavarman erected a liṅga
named Jayêśvara, for the welfare of the world, (i.e. of all the people), together with the
king,¹ and endowed it with a permanent endowment.² The date (l. 1) is (Gupta)-
Saṁvat 413, equivalent to A.D. 732-33, without any further details.
N.—Dr. Bhagwanlal Indraji’s inscription No. 13; Ind. Ant. Vol. IX. p. 176 f.—This
inscription is very much damaged.³ The name of the palace whence the charter was
issued, is lost. And the king’s name is illegible in line 3; but, as it is preceded by the
titles of Paramabhaṭṭâraka and Mahârâjâdhirâja, the name was probably that of Śivadêva
II., as in inscription L. above; and Dr. Bhagwanlal Indraji filled up the lacunæ in that
way. This is a Buddhist inscription. The Dûtaka (line 36) is the Bhaṭṭâraka, the
illustrious, or perhaps the venerable, Śivadêva. The date (l. 37) is somewhat doubtful;⁴ but
¹ sa-nṛipatêr=jagatô hitâya; line 2. ² akshaya-nîvî; line 2.
³ I have not seen the rubbing of this inscription.
⁴ In Śaka-Saṁvat 672 current, or 671 expired, the month Jyêshṭha was intercalary (see Indian
Eras, p. 161, and Chronology, p. 126). And the absence of any hint of this in the present record,
would, if the reading of the second numerical symbol as 40 could be established, be a point in favour of
the year being Śaka-Saṁvat 671 current, or 670 expired; which is in accordance with the epoch of
A.D. 605-606.

[Header] THE GUPTA INSCRIPTIONS; APPENDIX IV. 183

Dr. Bhagwanlal Indraji read it as (Harsha)-Saṁvat 143, Jyaishṭha-śukla-divâ-trayôdaśyâm ;
answering, approximately, to the 25th May, A.D. 748. There is, however, the possibility
that the second numerical symbol is 20 or 30, instead of 40.
O.—Dr. Bhagwanlal Indraji's inscription No. 14; Ind. Ant. Vol. IX. p. 177 f.—All the
introductory part of this inscription, recording the palace whence the charter was issued
and the name of the king, is broken away and lost. The Dûtaka (line 17) is the Yuvarâja
Vijayadêva. The date (l. 17) is (Harsha)-Saṁvat 145, Pausha-śukla-divâ-tritîyâyâm ;
answering, approximately, to the 6th December, A.D. 750.—Dr. Bhagwanlal Indraji, tak-
ing Vijayadêva as a "vicarious" name of Jayadêva II., allotted the inscription, chiefly on
this account, to Śivadêva II. As far as the date goes, the inscription may be one either
of him, or of Jayadêva II. But Vijayadêva cannot be a "vicarious" name of Jayadêva
II. Occasional instances may be cited, in which the special terminations of names vary;
thus, the Vasantasêna of inscription P. below, appears in R. line 10, and in the Vaṁśâvali,
as Vasantadêva ; and Jayadêva I. of R. line 8, perhaps appears in the Vaṁśâvali as Jayavar-
man. But even this is rare enough. And, apart from the substitution of birudas or secon-
dary titles, the only epigraphical instance that I can quote, of any variation in the first and
really distinctive part of a king's name, is to be found in line 46 ff. of the Kauṭhêm grant¹
of the Western Châlukya king Vikramâditya V., of Śaka-Saṁvat 930, where the Daśavar-
man of other inscriptions is called Yaśôvarman. But that instance is not a very satisfac-
tory one.² And in the present case there is not any metrical exigency, as there is in the
places where the name of Daśavarman occurs, to justify any variation. If the present in-
scription is one of Śivadêva II., then Vijayadêva was another son of his; if, as seems to
me more probable, it is one of Jayadêva II., then Vijayadêva was a son of the last-mentioned
person.
P. —Dr. Bhagwanlal Indraji's inscription No. 3; Ind. Ant. Vol IX. p. 167.—The char-
ter is issued from Mânagṛiha (line 1). The inscription is one of the Mahârâja, the illus-
trious **Vasantasêna.**³ The Dûtaka (l. 21) is the Sarvadaṇḍanâyaka and Mahâpratîhâra
Ravigupta. The date (l. 20 f.) is (Gupta)-Saṁvat 435, Âśvayuji śukla-divâ-1; answering,
approximately, to the 23rd September, A.D. 754.
Q.—Mr. Bendall's inscription No. 4; Journey in Nêpâl, p. 79 f. and Plate xi.—This is
not a formal charter, issued from any specified place; nor does it give any king's name.
It only records a grant by a private person to a Pâñchaka or Committee. The date
(line 1) is (Harsha)-Saṁvat 151, Vaiśâkha-śukla-dvitîyâyâm; answering, approximately, to
the 8th April, A.D. 756.
R.—Dr. Bhagwanlal Indraji's inscription No. 15; Ind. Ant. Vol. IX. p. 178 ff.—This
is not a formal charter, issued from any specified place. The inscription, which gives a

¹ Ind. Ant. Vol. XVI. p. 23.      ² See id. p 19 f.
³ In inscription R. below, line 10, he is called Vasantadêva.

184 CORPUS INSCRIPTIONUM INDICARUM, VOL III
good deal of genealogical information, to be commented on below, is one of Jayadêva II.
(line 14), who also had the second name or biruda of Parachakrakâma (l. 18). The object
of it is to record that he caused a silver water-lily to be made, for the worship of Śiva under
the name of Paśupati; and that it was worshipped and installed by his mother Vatsadêvî.
The date (l. 35) is (Harsha)-Saṁvat 153, Kârttika-śukla-navamyâm; answering, approxi-
mately, to the 16th October, A.D. 758.
S.—Dr. Bhagwanlal Indraji's inscription No. 4; Ind. Ant. Vol. IX. p. 168.—The
commencement of the inscription, recording the place whence the order was issued and the
king's name, is broken away and lost. And the inscription, therefore, is of importance
only as indicating the period during which the era that is referred to in it, continued to be
used. The Dûtaka (line 17) is the Râjaputra Vikramasêna.¹ The date (l. 18) is
(Gupta)-Saṁvat 535, Śrâvaṇa-śukla-divâ-saptamyâm; answering, approximately, to the 1st
July, A.D. 854.
When Dr. Bhagwanlal Indraji summed up, according to his view of them, the histori-
cal results of his inscriptions, Mr. Bendall's inscription No. 1, A. above, was not known;
and the only available clue was the mention, in B. C. E. F. and G., of Aṁśuvarman, who
was evidently to be identified with the king of that name who was reigning during, or
shortly before, Hiuen Tsiang's visit to Northern India, in or about A.D. 637. Acting on
this identification, he very properly referred the dates of (C.) 34, (E.) 39, (F.) 44 or 45,
(G.) 48, (L.) 119, (N.) 143 (?), (O.) 145, and (R.) 153, to the era dating from the com-
mencement of the reign of Harshavardhana of Kanauj in A.D. 606 (or 607).
Mr. Bendall's discovery of the Gôlmâḍhitôl inscription, furnished the key-note for the
interpretation of the whole matter. For, mentioning Aṁśuvarman as the contemporary of
Śivadêva I., and giving for the latter the date of the year 316, it shews that this date, and
all belonging to the same uniform series, must be referred to an era commencing just about
three centuries before the Harsha era; i. e. to the so-called Gupta era, with the epoch of
A.D. 319-20; for then the year 316 + A.D. 319-20 = A.D. 635-36, which is in the
closest possible accordance with the first recorded date of Aṁśuvarman, in A.D. 639.
But, when Dr. Bhagwanlal Indraji wrote, nothing was certain about the other larger
dates of (K.) 386, (M.) 413, (P.) 435, and (S.) 535; except that all the circumstances of
the case shewed that they were not of the same series with the smaller dates belonging to
the Harsha era. And, after examining the theory of the Śaka era, commencing A.D. 78,
and rejecting it on the grounds that even this was not early enough, Dr. Bhagwanlal Indraji
finally referred them to the Vikrama era commencing B.C. 58.
Curiously enough, the Nêpâl Vaṁśâvali states that Vikramâditya came to Nêpâl, and
established his era there. And, as we have seen at page 74 f. above, an offshoot of the
Vikrama era certainly was introduced into Nêpâl in A.D. 880. But the statement of the
Vaṁśâvali refers to the time of the predecessor of Aṁśuvarman; i.e. to the end of the
¹ See page 178 above, note 2.

THE GUPTA INSCRIPTIONS; APPENDIX IV. 185
sixth, or the beginning of the seventh, century A.D. And Dr. Bhagwanlal Indraji
shewed¹ that the statement is certainly quite wrong as regards the name of the king and
the particular era intended; and that what it probably contains, is a reminiscence of the
conquest of the country by Harshavardhana of Kanauj, and the adoption of the Harsha
era as the result. This statement, therefore, can have influenced him but little, if at all,
in his assignment of the above-mentioned dates to the Vikrama era.
The real grounds for his doing so,—grounds which, in spite of the existence of
the date of (Gupta)-Saṁvat 316 for Śivadêva I. and Aṁśuvarman, still require to be
cleared away, in order that we may properly adjust the whole of the early chronology
of Nêpâl,—are to be found in his erroneous treatment of R., the inscription of
Jayadêva II. of (Harsha)-Saṁvat 153.
Starting with a mythological genealogy, this inscription carries the descent from the
god Brahman (line 3), through Sûrya or the Sun,² Manu, Ikshvâku, and others, down to
Raghu, Aja, and Daśaratha (l. 6). After Daśaratha, there were eight kings, in lineal suc-
cession of sons and sons' sons, who are passed over unnamed; and then there was the
illustrious Lichchhavi (l. 6). Then comes verse 6, which records that "even still," at
the time of the writing of the inscription, "there exists a family which bears the pure second
name of Lichchhavi;"—svachchhaṁ Lichchhavi-nâma vi(bi)bhrad=aparaṁ vaṁśaḥ, (l. 7).
Dr. Bhagwanlal Indraji, in his lithograph, text, and translation, gave aparô vaṁśaḥ "a new
race ................ which bears the pure name Lichchhavi." But, on examining his
original rubbing, I find that the real reading is aparaṁ (nâma), "another name, a second
name;"—not aparô (vaṁśaḥ), "another race, a new race." This verse, therefore, is of some
interest, in shewing that,—in addition to the appellation of Lichchhavivaṁśa or Lichchha-
vikula, the latter of which actually occurs in inscriptions A. B. and H. above,—the family
had another original name, which, however, is not recorded. After Lichchhavi there came
some kings, who again are passed over unnamed, and the number of whom is
illegible³ at the end of line 7 and the beginning of line 8; and then the illustrious king
Supushpa (l. 8) was born at the city of Pushpapura.⁴ After him, "omitting in the interim
(to mention the names of) twenty-three kings," there was another⁵ king, the famous
¹ Ind. Ant. Vol. XIII. p. 421 f.
² In accordance with this, the Vaṁśâvali enters the historical Lichchhavis of this inscription as
members of a Sûryavaṁśi family.
³ But the number was probably twelve, as hitv=âpa[vân=dvâdaśa] seems to suit best the metre
and such traces as are discernible in the rubbing.
⁴ i.e. Pâṭaliputra, the modern Paṭna in Bihâr; Ind. Ant. Vol. IX. p. 180, note 44.—This part of
the inscription, therefore, relates to a period antecedent to the settlement of the Lichchhavis in
Nêpâl.
⁵ As no previous Jayadêva. with whom this one could be contrasted, is mentioned in the inscrip-
tions, apara seems to introduce another branch of the Lichchhavi family, not directly descended from
Lichchhavi and Supushpa.

186 CORPUS INSCRIPTIONUM INDICARUM, VOL. III.
Jayadêva I. (l. 8), who was treated by Dr. Bhagwanlal Indraji as the first really historical
member of the family, and the founder of the Nêpâl branch of it,¹ and on the Vikrama-
Saṁvat theory, was placed about A.D. 1. After this “victorious” Jayadêva I., and again
“omitting in the interim (to mention the names of) eleven ...... kings,” the inscription
gives the first unbroken succession of names that it contains; viz.—Vṛishadêva (l. 9);—
his son, Śaṁkaradêva;—his son, Dharmadêva;—his son, Mânadêva (l. 10);—his
son, Mahîdêva;—and his son, Vasantadêva. The first four of these names have
already been made known by inscription K. above; and the sixth, in the form of Vasanta-
sêna, by inscription P. And these six kings plainly belong to the family, the genealogy
of which forms the subject of all the preceding part of the inscription; viz. the
Lichchhavikula.
Then follows, in lines 10 and 11, the eleventh verse, which was entirely misunder-
stood by Dr. Bhagwanlal Indraji, and which, through that misunderstanding, led to the
erroneous reference of the larger dates of his inscriptions to the Vikrama era. He read this
verse—asy=ântarê=py=Udayadêva iti kshitîśâj=jâtâs=trayôdaśa [tata]ś=cha Narêndradêvaḥ
mân-ônnatô nata-samasta-narêndra-mauli-mâlâ-rajô-nikara-pâṁśula-pâdapîṭhaḥ; and tran-
slated — “afterwards came thirteen (rulers), sprung from king Udayadêva (l. 10),² and
then Narêndradêva (l. 11), who was proud, and whose footstool was covered with the dust
from the row of diadems worn by numerous prostrated kings.” With the exception that
the original has pâṁsula, not pâṁśula, the reading and translation of the second half of the
verse represent the original correctly. It is the first half that has not been properly treat-
ed. In the first place, asy=ântarê does not mean ‘afterwards,’ or ‘after him.’ The literal
meaning of antara is ‘an interval;’ and in anantaram it helps to make up the equivalent
of ‘afterwards,’ only from its meaning in the first place ‘no interval after.’ But, standing
without the negative particle, antara can have only the meaning of ‘interval;’ and it is
used twice before by the composer of this inscription, in lines 8 and 9, distinctly in that
sense. Asy=ântarê can only mean “in an interval of this;” and, supplying vaṁśasya, from
the vaṁśaḥ of line 7, in apposition with asya, we have “in an interval, i.e. at an intervening
point, of this (lineage that has just been detailed).” The expression plainly introduces
some names, in respect of which it is intended to be conveyed that they are of another
family or branch; and that the last of them comes contemporaneously with or immediate-
ly after the name of Vasantadêva, the last mentioned of the immediately preceding suc-
cession, and the first comes at some unspecified point intervening between Vṛishadêva and
Vasantadêva. In the second place, Dr. Bhagwanlal Indraji’s reading of kshitîśâj-jâtâs
cannot be maintained. In the two syllables jjâtâ, the rubbing shews distinctly the double
jj, of which, in the lithograph, the lower one is only sketched in faintly, as if it were doubt-

¹ Ind. Ant. Vol. XIII. p. 424.
² Or, again, — “after him (Vasantadêva), followed thirteen, sprung from the lord of the earth
called Udayadêva;” Ind. Ant. Vol. XIII. p. 425, note.

THE GUPTA INSCRIPTIONS; APPENDIX IV. 187
ful; but the second is as distinctly ta, not tâ, as is in fact shewn in the lithograph. There
is a slight abrasion mark between the ta and the following akshara, which does not appear
in the lithograph; but the rubbing shews most distinctly that this mark is no remnant of
a half-obliterated â, and that the stroke for â, for which in fact there is no room, never was
engraved here The reading of the original, in short, is jâtas, the nominative singular;
not jâtâs, the nominative plural. This is the first objection to the more serious error, the
introduction of trayôdaśa, 'thirteen,' immediately after jâtas. The second is, that, of the
six following aksharas which are read s=trayôdaśa [tata], the only parts that can be pro-
nounced on with any certainty, are s, as part of the first, and, at the beginning of line 11,
da, the third, which is well preserved and unmistakable. The others are hopelessly injured
and unrecognisable; and,—with the exception that the first probably had a t below the s;
and that the second looks more like thâ (tathâ), or dâ (tadâ), or pô (tapô), than yô,—it is
quite impossible, even from the rubbing, to say what they may be. But the culminating
and final objection to the reading of trayôdaśa [tata]ś=cha is, that, in the passage, as thus
read, there is no such word as vyatîtya, vihâya, hitvâ, or tyaktvâ, "having passed over or
omitted (to mention the names of)," which we have in lines 4, 6, 7, 8, and 9, where, in each
case, a specified number of kings is passed over; and that the metre does not allow of our
introducing any such word. The lacunæ at the end of line 10 and the beginning of line 11
render it, as I said, impossible to decide what the original reading may have been. But the
whole structure of the passage leaves no doubt whatever that the original contained nothing
but an epithet, or perhaps two, of Narêndradêva; and that, so far from thirteen rulers
having intervened between him and Udayadêva, he was the son of Udayadêva.
Narêndradêva's son was Śivadêva II. (line 12), who married Vatsadêvî, of the family
of the Maukharis who abounded in strength of arm,¹ the daughter of the illustrious
Bhôgavarman, and the daughter of the daughter of "the great" Âdityasêna, the lord of
Magadha (l. 13). And their son was the Râja, the illustrious Jayadêva II. (l. 14), also
called Parachakrakâma (l. 18), whose wife was Râjyamatî, of the family of king Bhagadatta
or of the Bhagadatta kings (l. 16), the daughter of Harsha, king of Gauḍa, Ôḍra; &c., and
Kaliṅga, and Kôsala (l. 15). The rest of the inscription details the beauty of the silver
water-lily which Jayadêva II. caused to be made, and how it was worshipped and installed
by his mother Vatsadêvî; and then concludes with the date.
Dr. Bhagwanlal Indraji's acceptance of Udayadêva as the successor of Vasantadêva,
and insertion of thirteen rulers between him and Narêndradêva, led of necessity to the
reference of the dates for Mânadêva of (K.) 386 and (M.) 413, and for Vasantadêva of
(P.) 435, to the Vikrama era. It is unnecessary to repeat his calculations in full. But,
starting with fifteen names between Vasantadêva and Śivadêva II., or nineteen from Mâna-
¹ The original has dêvî vâ(bâ)hu-va(ba)l-âḍhya-Maukhari-kulâ śrîvarmma &c.; not kula-śrî-
varmma, in composition, as in the published text.
3 2b

188 CORPUS INSCRIPTIONUM INDICARUM, VOL III
dêva to Śivadêva II. (both included), all of which denote "generations of kings, not reigns
of collaterals,"¹ and taking twenty-one years as the smallest possible average for these
generation-reigns, he found that no era later than the Vikrama era would meet the
requirements of the case ; and that that era would meet them. For, on the Vikrama-Saṁvat
theory, Mânadêva's first date represented A.D. 329 ; the interval from this to A.D. 759,
the date of Jayadêva II., was 430 years; and this, divided by nineteen gave about
twenty-two years and three-quarters as the average for each generation-reign. This was
all right enough from his point of view.
But let us now take the matter from the correct point of view; viz. that Udayadêva
did not come after Vasantadêva. This frees us at once from the necessity, under which
Dr. Bhagwanlal Indraji laboured, of forcing Vasantadêva and his ancestors back to such
early times ; and leaves us at liberty to follow the analogy of inscription A., and to refer
his date and Mânadêva's to the Gupta era. The result is, that we have for Vasantasêna
the date of (P.) A.D. 754, just synchronous with the dates of perhaps (O.) A.D. 750,
and certainly (R.) A.D. 758, for Jayadêva II., exactly what inscription R. seeks to
convey ; and we have for Mânadêva, the grandfather of Vasantasêna, the dates of (K.)
A.D. 705 and (M.) A.D. 732-33, just about one generation before Jayadêva II.'s father
Śivadêva II., for whom we have the dates of (L.) A.D. 725 and probably (N.) A.D. 748 (?).
Here the question naturally suggests itself :—As Udayadêva and his descendants were
not successors and descendants of Vasantasêna, who were they ? I think the answer is
perfectly plain :—That they were successors of Aṁśuvarman ; and, though not his direct
lineal descendants, belonged, like him, to the family which in the Vaṁśâvali is called the
Ṭhâkurî family
Inscription R., in fact, furnishes another instance of the double government of
Nêpâl, to which Dr. Bhagwanlal Indraji drew attention in the case of Śivadêva I. and
Aṁśuvar-man ; and which is illustrated in the most pointed way throughout these
inscriptions.² We have two separate families, ruling contemporaneously and mostly on
equal terms, but each preserving certain distinctive characteristics of its own.
On the one side (see Table XI. page 189 below), we have the Lichchhavikula of
the inscriptions,—the Sûryavaṁśi family of the Vaṁśâvali,—issuing its charters from
the house or palace called Mânagṛiha ; and using the Gupta era. To this family
belong inscriptions A. B. K. M. P. and S. ; and it was represented in A.D. 635 by
¹ Ind. Ant. Vol. XIII. p. 425.
² So also in later times. See, for instance, the colophon of a manuscript of the Ashṭasâhasrikâ-
Prajñâpâramitâ (Bendall's Catalogue of Buddhist MSS. p. 4), dated in Nêpâl-Saṁvat 128 (A.D.
1007), which speaks of the double rule of Nirbhaya and Rudradêva ; and another manuscript bearing
the same title (id. p. 151), dated in Nêpâl-Saṁvat 135 (A.D. 1015), which speaks of the government
being enjoyed, half by Bhôjadêva and Rudradêva, and half by Lakshmîkâmadêva.