The Origin of the Bengali Script/Chapter 2

by Rakhaldas Bandyopadhyay

 

CHAPTER II

The Northern Indian Alphabets (B.C. 350—A.D. 600).

A.The Older Maurya Alphabet.

Leaving aside the various theories about the origin of the ancient Indian alphabet, we turn to examine it as it has been found to exist at the beginning of the historical period. It is sufficient for the purpose of the present article that Dr. Bühler recognised the antiquity of the Indian Alphabet in Aśoka's time. "The existence of so many local varieties, and of so very numerous cursive forms, proves, in any case, that writing had had a long history in Aśoka's time and the alphabet was then in a state of transition."^[1] The alphabet is also recognised to be "a script framed by learned Brāhmaṇs for writing Sanskrit."^[2] The earliest Indian inscription is the record on the Piprāwā vase discovered in 1898. It can be proved on palæographical grounds that the forms of Brāhmī letters used in incising this record are older than those of Aśoka's inscriptions. The vases found in the Stūpa at Piprāwā contained according to one authority the relic (Śarīra) of Buddha himself,^[3] and according to another, those of his kinsmen of the Śākya clan.^[4] It has been surmised that the stūpa was raised over the relics of the Śākyas, who were slain by Viruḍhaka, King of Kośala, during the life-time of Buddha himself. Consequently the date of the Piprāwā inscription must lie either in the 5th or the 4th centuries B.C. Palæographical evidence fully supports this conclusion: the archaic forms of the Brāhmī alphabet found on the Persian sigloi, which went out of the general use in Aśoka's time, are found to have been used in the inscription. An analysis, of the characters of this inscription, would be out of place here, as it does not properly belong to the Eastern variety of the Maurya alphabet. It serves to indicate the upper limit of the use of the alphabet of this period. The lower limit has been fixed by Bühler at 200 B.C.^[5] The seals, found by Cunningham at Pātnā,^[6] which according to Bühler belong to the period when Brāhmī was written boushophedon (βονστροφπδον), were really seal-matrices, like the Rohṭāsgaḍh Rock seal-matrix of the Mahāsāmantādhipati Śaśāṅka.^[7]

B.Varieties of the Older Maurya Alphabet.

In 1896, Bühler admitted the existence of two distinct varieties of this alphabet, viz:—

(i) the Northern: to be found in the rock-edicts at Kālsi, the pillar-edicts at Allahabad, Rādhiā, Māthiā, Niglivā, Paḍeriā and Rāmpurwā, the minor rock-edicts at Bairāṭ, Sahasrām, the inscriptions of the Barābār caves and Sāñci and Sārnāth pillars;

(ii) the Southern: to be found in the-rock edicts at Girnār, Dhauli and Jaugaḍa and the minor rock-edicts at Siddapura.

Bühler already noticed the existence of varieties, at this period, in the Northern Maurya alphabet. "Even the writings in the northern versions are not quite homogeneous. The pillar edicts of Allahabad, Māthiā, Niglivā, Paḍeriā, Rādhiā and Rāmpurwā form a very closely connected set, in which only occasionally minute differences can be traced, and the edicts of Bairāṭ No. I, Sahasrām, Barābār and Sāñci, do not differ much. A little further off stands the Dhauli separate edicts (where Edict VII has been written by a different hand from the rest), the Delhi-Mirāṭ edicts and the Allahabad Queen's edict, as these show the angular da. Very peculiar and altogether different is the writing of the rock-edict of Kālsi, with it, some letters on the coins of Agathocles and Pantaleon (but also some in the Jaugaḍa separate edicts), agree. Perhaps, it is possible to speak also of a North-Western variety of the older Maurya alphabet."^[8]

Thus Bühler distinguishes three different sub-varieties in the Northern Maurya alphabet. According to their geographical distribution, they may be classified as follows:—

(a) The North-Eastern—found in the Allahabad, Rādhiā, Māthiā, Rāmpurwā, Niglivā, Paḍeriā and the Sārnāth pillar edicts. The Earthen seals found at Pātnā^[9] (seal matrices bearing the inverted inscriptions Naṁdāya and Agapalaśa) as well as that found by Cunningham at Bodh-Gayā^[10] (Mokhalinam) belong to this period.

(b) The North-Central—found in the rock-edicts at Bairāṭ and Sahasrām, the pillar-edicts at Sāñci and Delhi and the cave-inscriptions at Barābār.

(c) The North-Western—represented by the characters of the Kālsi rock-edicts and the letters on the coins of the Greek kings Agathocles and Pantaleon.

In this paper we are concerned only with the North-Eastern variety, of the older Maurya-alphabet, and such inscriptions of the Northern Central variety as are to be found in North-Eastern India. A detailed description of the older Maurya alphabet would also be out of place here, as it is not yet possible to improve upon Dr. Bühler's admirable description of it. Consequently, one has to remain content, simply with the noting of the peculiarities of the alphabet as found in different inscriptions. Among vowel signs the only letter to be noted is the initial ī which has been found in one of the inscriptions on the railings around the great temple at Bodh-Gayā, where Bühler reads Īdāgimitasu for Iṁdāgimitasa read by Cunningham.^[11] But in reality, the characters of this inscription belong to the younger Maurya alphabet, as shown by Bloch. Among the consonants the form of kha found in one of the inscriptions at Bodh-Gayā,^[12] with a triangle as its base, should be noted, but this inscription also, belongs to the younger Maurya alphabet. The only instance of ṇa, among the inscriptions of this period, is to be found in the mason's marks on the pillars of Buddha's walk, inside the temple enclosure at Bodh-Gayā. Cha with two loops, one on each side of a vertical straight line, instead of a circle divided into two unequal parts, have also been found among the mason's marks on the pillar-bases of Buddha's walk. The usual form of ja, is the Northern form with a loop or a dot. Other letters do not call for special attention but forms of the test letters ya, la, sa and ha may be noted. The form of ya is essentially the Northern one, which Bühler calls the "notched ya." The form of la is generally cursive. One important exception is to be found, in the extremely cursive form, used in the Jaugaḍa separate edicts, which is essentially the same to be found in the Eastern variety of the Early Gupta alphabet of the 4th and 5th centuries A.D. The position of the Jaugaḍa edict is somewhat peculiar. The edicts of Dhauli and Jaugaḍa, though relegated to the Southern variety of the older Maurya alphabet, stand in an intermediate position. "The Southern variety is most strongly expressed in the Girnār and Siddapura edicts, less clearly in the Dhauli and Jaugaḍa edicts by differences in the signs for a, ā, kha, ja, ma, ra, sa, the medial i, and the ligatures with ra."^[13] Most probably, the cursive forms of ha and la, found in the Jaugaḍa separate edicts, were imported from Northern India, as will be seen later on. In the North-Eastern variety, the usual form of ha is also cursive. The extremely cursive form of this letter, in the Jaugaḍa separate edicts, is peculiar and an importation from the North.^[14] This statement is corroborated by the discovery of a slightly different cursive form in the Allahabad separate edicts, line 1 in the word mahāmāta.

C.The Younger Maurya Alphabet.

The last eight columns, of Plate II of Bühler's tables, represent the younger Brāhmī alphabet of Northern India. The letters are taken from six series of inscriptions—

(i) The Nāgārjunī cave-inscriptions of Daśaratha, ca. 200 B.C.

(ii) The inscriptions on the Toraṇas, railing-pillars and cross-bars of the Bhārhut Stūpa, ca. 150 B.C.

(iii) The cave inscriptions at Pabhosa in the United Provinces, ca. 150 B.C.

(iv) The oldest inscriptions from Mathurā. These letters are principally taken from the oldest inscriptions discovered by Dr. A. Führer during the excavations at Kaṅkālī Ṭīlā, but the most ancient inscription from the district of Mathurā was discovered by Cunningham at Parkham. This inscription is incised on the base of a mutilated image of Yakṣa, at present in the Archæological Museum at Mathurā^[15]. Most probably its characters belong to the younger Maurya alphabet.

(v) The Hāthigumphā inscription of Khāravela of Kaliṅga, ca. 160 B.C.

(vi) The Nānāghāṭ inscriptions of the Andhras, ca. 150 B.C.

Among these, only the Nāgārjunī cave-inscriptions of Daśaratha can be said to belong to the North Eastern variety. During subsequent years one other group has been added to the above list:

(vii) The inscriptions on the railing-pillars around the great temple at Bodh-Gayā. The late Dr. Theodor Bloch drew attention to the fact that "the older part of the Bodh-Gayā railing was put up in the middle of the 2nd century B.C., about 100 years after the time of Aśoka".^[16] The cave-inscriptions of Daśaratha are about half a century older than those on the railing pillars at Bodh-Gayā. The following points are worth noting on the alphabet of the cave inscriptions:—

(1) the form of la closely resembles, that of the extremely cursive one, found in the Jaugaḍa separate edicts (see ante p. 14);

PLATE II.

Fragmentary Kusana Inscription-Rajgir-Patna (I. M. No. 6283).

(2) the form of lingual ṣa is peculiar and resembles the form found in the Kālsi edicts, probably, it was the precursor of the looped lingual ṣa found in Eastern India in the 4th or 5th centuries A.D.;

(3) the form of ha is primitive and resembles that of the Siddapura edicts;^[17]

(4) the form of ṣa shows an advance—the upper hook has been lengthened to form a slightly slanting second horizontal line.

The form of the remaining letters in Column XVII of Plate II of Bühler's work does not call for remarks. The inscriptions on the railing-pillars and cross-bars at Bodh-Gayā exhibit further changes, though they were incised only about fifty years after Daśaratha's time:—

(a) a shows two forms. In the word Amoghas, the first letter is decidedly of southern appearance^[18] (e.g. Pl. II, Col. VIII, 1.); the other form is to be found in the various inscriptions of the noble lady Kuraṅgi and resembles that used in the Hāthigumphā inscriptions (Pl. II, Col. XXI, 1);

(b) ka has invariably the dagger-shaped form which was current up to the end of the 6th century A.D. and was formed by the elongation of the vertical line of the older Maurya form, cf. ka in Tabapanaka^[19], Kuraṅgiye^[20], Sakaputrasa^[21], Cetika^[22];

(c) kha occurs once only, in Bodhirakhitasa^[19] where it resembles the form used in the oldest inscription in Mathurā (Pl. II, Col. XX, 10); there is a very slight difference between these two forms, the Eastern variety form as found in the Bodh-Gayā inscription, being slightly longer than that of the Western variety;

(d) ga occurs several times in the name Kuraṅgi^[20] where it has two varieties:—(1) cursive as in Pl. X, p. 4 and (2) the angular as in Pl. X, 6-7;

(e) gha also occurs only once in Amoghasa^[23]; its appearance shows great change, though it resembles one of the forms used in the Kālsi edicts (Pl. II, Col. 3-12); it is, on the whole, different from the form to be found in the North-Eastern variety of the early Maurya alphabet;

(f) ca occurs twice in Cetika,^[24] but its form does not show much difference from that of the older Maurya one;

(g) two forms of ja are to be found in these inscriptions:—(i) one form resembles the ja in Bühler's Pl. II, Col. X, 15, while (ii) the other form is the usual older Maurya one with a dot in place of the central loop;

(h) ta resembles the southern form in Bühler's Pl. II, Col. VII, 23 and the usual form of later Brāhmī inscriptions;

(i) da occurs in all of the inscriptions discovered on the pillars, copings and cross-bars of the Bodh-Gayā railing, and resembles the angular form of the older Maurya alphabet (Bühler Pl. II, Col. V-VI, 23);

(j) dha occurs once only in Bodhirakhitasa; there is no change in the form of this letter from the 3rd century B.C. till the 10th or 11th century A. D.;

(k) na also occurs in all of the inscriptions from Bodh-Gayā and its base line shows no curvature at all, proving that these inscriptions cannot be placed later than the 2nd century B. C.;

(l) pa shows a greater degree of change; in all cases of its occurrence, it shows two well-formed right angles, at its lower extremeties; cf. Tabapanakasa,^[25] Sakaputrasa, Jivāputrāye,^[26] Pājāvātiye, Jivāputrāye and Pāsādā^[27].

(m) the form of ba shows no change;

(n) two forms of ma have been found in these inscriptions:—(1) ma with a circle at the lower part and a semicircle over it, as in Amoghasa^[28] and (2) ma with a triangle at the lower part and a right angle over it, as in Mitrasa^[29];

(o) two forms of pa also are to be found: the first form is the notched one, which is to be found on the coping inscriptions only,^[30] and the second form, that with the curve below, is to be found in pillar-inscriptions^[31];

(p) ra is always represented by a curved line;

(q) va shows the formation of a triangle at its base in the place of the circle;

(r) two forms of the dental sa are to be found: on one of the inscribed cross-bars, we find a slight curve to the left, attached to the lower extremity of the lower hook, cf. sa in Amoghasa; the other form is the usual older Maurya one, where in some cases, the elongation of the lower hook, marks a slight modification;

(s) ha has been found only once in the inscription recently discovered by the late Dr. Bloch, where it occurs in a ligature. The form of this letter, in the word Brahmamitra^[32] is extremely cursive and shows that this hooked form is peculiar to the eastern variety of the older alphabets of Northern India;

(t) The newly discovered inscription has supplied a new letter ña which is to be found in the first word in rāño, and resembles the form in the Bhārhut and the Pabhosa alphabets with a downward elongation of the left vertical line.

No inscription, which can safely be assigned to the 1st century B. C. or A. D., has been found anywhere in North-Eastern India, except at Sārnāth. The records which can be assigned to the 1st century B. C. are very few in number:

(i) Inscription on the upper side of the lower horizontal bar of the stone-railing surrounding the old stūpa in the south chapel of the main shrine^[33]. The second half of the inscription only, is of earlier date, the first half belonging to the second century A. D. (not the 3rd or 4th as Messrs. Konow and Marshall imagine). The date of the second half also has not been correctly given. It is impossible to assign it to the 2nd century B. C. The shortening of the verticals in pa and ha, as well as the curvature in the base line of na, indicates that the record must be ssigned to the 1st century B. C.

(ii) "When clearing the south chapel, the top of a stone railing became visible above the floor * * * a short votive inscription on one of the stones, places the erection of the railing in or before the 1st century B.C."^[34] Here also the second part of the inscription only can be referred to the first century B.C. This part consists of the word "Parigahetāvaṁ".

(iii) Inscriptions on the pillars of a railing around a votive stūpa.^[35] The first of these inscriptions (No. III) probably belongs to the 2nd century B.C. The probable reading is:—Sihāye Sāhijāteyikāye thabho. The second inscription (No. IV) has been very badly preserved. The fac-simile shows:—

1. ...niya Sonade (va).

2. Thabho dāna (ṁ).

"The pillar-gift of Sonadeva (Svarṇadeva) of........." Inscription No. II which ends with the word "Dānaṁ thabho undoubtedly belongs to the early Maurya period of the 3rd century B. C.

(iv) Inscription on a rail stone (? cross bar):— Bhariniye Sahaṁ Yateyikā (ye)^[36]—the gift of Yateyikā with Bharini. This inscription also belongs to the 1st century B.C., as indicated by the form of medial i and the shortening of the verticals in ya.

(v) Inscription of the king Aśvaghoṣa, the year 40. incised on the pillar of Aśoka..............." "...............rparigeyhe rājña Aśvaghoshasya chatariśe Savachhare hematapakhe prathame divase dasame."^[37] Certain words following the above record, have been read by Dr. Venis as follows: Sutithage 4, 200, 9.^[38] Drs. Fleet and Venis hold that this date should be referred to the Mālava-Vikrama era and arrive at 111-151 A.D. as the date of Aśvaghosa. If Drs. Fleet and Venis be correct, then it shall have to be admitted that, Kaṇiṣka, Huviṣka and Vāsudeva reigned in the latter half of the second and third centuries A.D., because in a treatise on Palaeography, it is impossible to admit, that the group of Kuṣāna inscriptions, came before those of Aśvaghoṣa, the Kṣatrapas Nahapāna and Śoḍāsa, and the archaic inscriptions from Mathurā.

(vi) Fragmentary inscriptions of the time of Aśvagho a:—

1. Rājña Aśvaghora (sya)......... 2. Upala he ma (ṁtapakhe?)^[39]………

The principal characteristics of the above inscriptions from Sārnāth are:—

(i) total absence of any difference from the forms of the characters of the 1st and 2nd centuries B.C. found in North-Western India;

(ii) consequently we find the general shortening of vertical lines, angularisation of curved strokes, and in the case of medial vowel signs, cursiveness of the angular forms of the older Maurya Brāhmī.

 

D. Kuṣāna Inscriptions.

Under the above title the inscriptions of the great Kuṣāṇa Kings, Kaṇiṣka, Huviṣka and Vāsudeva are to be considered, the dates in whose inscriptions are generally taken to be Śaka dates.^[40] At present two theories are current about the dates used in the inscriptions of the Kuṣāṇa kings mentioned above.

(i) That the dates in the Kuṣāṇa incriptions should be referred to the Mālava-Vikrama era which was established by Kaṇiṣka in the year 57 B.C. The expounders of this theory hold that the inscriptions of the Satraps Śoḍāsa and Rañjuvula fall after those of Kaṇiṣka, Huviṣka and Vāsudeva in the chronological order. This fact cannot, for a moment, be considered to be true, in a paper on Palæography.

(ii) That the dates in the Kuṣāṇa inscriptions should be referred to the Śaka era, which was founded by Kaṇiṣka in the year 78 A.D. In the following pages I have adopted this theory, which was started by Oldenberg and Fergusson, adopted by Bühler and Rapson, defended by myself and finally accepted by Mr. V.A. Smith. The inscriptions of the Ku āṇa period (1st and 2nd centuries A.D.) are more abundant in North-Western India. On this point Bühler says: "The next step in the development of Brāhmī of Northern India is illustrated by the inscriptions from the time of the Kuṣāṇa kings Kaniṣka, Huviṣka and Vāsuska-Vāsudeva, the first among whom made an end of the rule of the older Śakas in the Eastern and Southern Punjab. The inscriptions with the names of these kings which run from the years 4 to 98 (according to the usually accepted opinions, of the Śaka era of A.D. 77-78, or of the 4th century of the Selukid era) are very numerous in Mathurā and its neighbourhood, and are found also in Eastern Rājputānā and in the Central India Agency (Sāñei)."^[41]

In subsequent years a number of inscriptions have been discovered in North-Eastern India, which can without doubt be referred to this particular period:—

(i) the Bodh-Gayā Fragmentary inscription on the diamond throne (vajrāsana);^[42]

(ii) the Sārnāth Umbrella-staff inscription of the 3rd year of Kāṇiṣka;^[43]

(iii) the inscription on the base of the Bodhisattva Image dedicated in the 3rd year of Kaṇiṣka;^[44]

(iv) the inscription at the back of the Bodhisattva image of the 3rd year of Kāṇiṣka;^[45]

(v) inscription on the pedestal of an image of Bodhisattva from Sāhet Māhet (the ancient Śrāvastī);^[46]

(vi) inscription on an umbrella-staff, now in the Indian Museum, probably found in the ruins of Sāhet Māhet;^[47]

(vii) inscription on the pedestal of an image of Bodhisattva found at Sāhet Māhet;^[48]

(viii) fragmentary inscription on a fragment of a sculpture discovered at Rājagṛha (Rājgir), in the Pātnā District;^[49]

(ix) fragmentary inscription on the pedestal of an image discovered at Rājagṛha;^[50]

The records of the 1st century A.D. fall into two distinct and separate classes.—

I. The Eastern variety of the North-Indian Alphabet of the Kuṣāṇa period, earlier variety. All the inscriptions enumerated above belong to this class. Six years ago, I stated, that inscription No. VIII belongs to the class of Epigraphs known as inscriptions written in the Northern-Kṣatrapa alphabet, but now I agree with Dr. Vogel in calling them by the new name "Early Kuṣāṇa." Inscription No. I. is by far the oldest inscription of the Kuṣāṇa period, discovered up to date, in North-Eastern India. It was incised on the edge of a slab of stone, which is at present lying under the Bodhi tree, at Bodh-Gaya.^[51] It was in a very bad state of preservation

PLATE III.

Inscription of the time of Mahendrapala-Ramgaya-Gaya.

at that time and has since suffered much from the weather. When I examined the stone in 1906, I found that the fragmentary inscription, was almost illegible. The use of the broad-backed śa, the shortening of the verticals in pa, and the ma in which the lower part is invariably triangular in form, show that the inscription belongs to the early Ku āṇa period. Yet, the doubtful ya in the opposite corner of the inscription, which is archaic in form, proved that the record must be referred to a period slightly earlier than those, in which the later, fully developed tripartite form of ya is found to be used.

II. The Eastern variety of the North-Indian Alphabet of the Kuṣāṇa period, later variety. No inscription, which can be safely referred to this class, has been discovered as yet in any part of North-Eastern India.

The principal characteristics of the earlier variety of the North-Eastern Kuṣāṇa alphabets are:—

(i) the use of the broad-backed śa:—daṇḍaśca and Śāvastiye (L. 2 Śrāvastī image-inscription, Indian Museum, daṇḍaśca (L. 7), Śāvastiye (L. 8) of the Indian Museum umbrella-staff inscription, Śīvadharasya, Śiāvāsta (L. 1), kuśalā, bhuyakuśalaṁ, and Śivamitrena (L. 3) of the new Bodhisattva image-inscription from Sāhet Māhet, Śakyamuni, on the fragmentary sculpture from Rājgir; Indraśiri and Parahaśalika (L. 2) in the inscription on the newly discovered pedestal from Rājgir;

(ii) the lingual ṣa, angular in form in which the cross-bar does not reach the left vertical line: Kaniṣkasya (L. 1), bhikṣusya, Puṣya (L. 2), yaṣṭi and pratiṣṭhāpito (L. 4), kṣatrupena (L. 8), pariṣā (L. 9) of the Sārnāth Umbrella-staff inscription, pratiṣṭhāpito (L. 1), kṣatrapena, mahākṣatrapena and Vanaṣparena in (L. 2) of the inscription on the pedestal of the Sārnāth Bodhisattva image; Kaniṣka (L. 1), bhikṣusya (L. 2), yaṣṭi (L. 3) of the inscription on the back of the Bodhisattva image from Sārnāth; bhikṣusya and Puṣya (L. 1) bhikṣusya (L. 2) in the inscription on the pedestal of the Bodhisattva image in the Indian Museum found at Sāhet Māhet; kṣatriyānaṁ, veliṣṭanaṁ (L. 1): vīcakṣaṇa (L. 2) on the inscription on the pedestal of the new image from Sāhet Māhet. It should be noted in this connection, that the form of the subscript lingual ṣa, as found in kṣatriyānaṁ (L. 1) and vīcakṣaṇa (L. 2), is still more archaic, having the cursive form of the older Maurya alphabet;

(iii) the cursive form of ha, which seems to have been derived from the cursive forms of the Jaugaḍa separate edicts and the Kauśāmbī edict on the Allahabad pillar: this form occurs on one inscription only, viz. on the pedestal of the new Bodhisattva image from Sāhet Māhet; Bohisatva (L. 1—3),^[52] but in all other cases the angular form of ha is found to have been used;

(iv) in the majority of cases, the subscript ya has the tripartite form. The only exceptions being Puṣya in (L. 1) of the inscription on the pedestal of the Bodhisattva image from Śrāvastī, now in the Indian Museum and in Śakyamuni on the fragmentary sculpture from Rājgir, which is also in the Indian Museum. The dearth of inscriptions, written in characters of the later variety of the Northern Kuṣāṇa alphabet, in Eastern India has already been noticed above. Inscriptions of the 3rd and 4th centuries A.D., are also very rare in the whole of the Northern India. With the exception of two inscriptions from Mathurā, which I hold to belong to the 3rd century A.D.^[53] and which others hold to belong to the 6th century A.D.^[54], no inscriptions are known which can be said to belong to the pre-Gupta period.

At the beginning of the Gupta period, we are confronted with three distinct varieties of the alphabet, used in Northern India. Inscriptions belonging to the first-half of the 4th century A.D., are unknown unless the Allahabad pillar-inscription of Samudragupta^[55] be referred to that period. The second inscription in the chronological order, which can be safely referred to this period, is the Bodh-Gayā Image-inscription^[56] of the Gupta year 64=383-84 A.D. Scholars are divided in opinion about the date of this inscription also. Prof. Lüders of Berlin holds Cunningham's theory and says that it is a Śaka date^[57], inspite of Dr. Bühler's clear statement on the point.^[58]

E. The so-called Gupta Alphabet of the 4th and 5th centuries A.D.

Dr. Bühler recognises three different varieties in the Northern Indian alphabet of the 4th and 5th centuries A.D.:–

• • (i) the Eastern variety—distinguished by the peculiar forms of la, ha, ṣa and sa,^[59]
  • (ii) the Western variety—cursive roundhand type,^[60]
  • and (iii) the Western variety—angular monumental type.^[61]

In the light of later discoveries, especially the important finds of the British and Prussian expeditions into Central Asia under Sir Marc Aurel Stein, Grünwedel and others, the Northern Indian alphabet of the 4th and 5th centuries A.D., should be divided into the following varieties:

• 1. The Eastern variety: specimens—
  • (i) the Allahabad pillar-inscription of Samudragupta,
  • (ii) the Udayagiri cave-inscription of Candragupta II,
  • (iii) the Gaḍhwā fragmentary inscriptions of the times of Candragupta II and Kumaragupta I,
  • (iv) the Dhānāidaha grant of Kumāragupta I,
  • (v) the Mānkuwār inscription of Kumāragupta I,
  • (vi) the Bihār pillar-inscription of Skandagupta,
  • (vii) the Kosām image-inscription of Bhīmavarman,
  • (viii) the Kahāuṁ pillar-inscription of Skandagupta.
• 2. The Western variety: specimens—
  • (i) the Mathurā inscription of Candragupta II,
  • (ii) the Sāñei inscription of Candragupta II,
  • (iii) the Bharaḍi Ḍiḥ or Karamdaṇḍa inscription of Kumāragupta I,
  • (iv) the Bhiṭāri pillar-inscription of Skandagupta,
  • (v) the Indore grant of Skandagupta,
  • (vi) the Erān pillar-inscription of Budhagupta.
• 3. The Southern variety: specimens—
  • (i) the Bilsaḍ pillar-inscription of Kumāragupta I,
  • (ii) the Gangdhar inscription of Viśvavarman,
  • (iii) the Mandaśor inscription of Kumāragupta I, and Bandhuvarman,
  • (iv) the Vijayagaḍh inscription of the Yaudheyas
  • (v) the Vijayagaḍh pillar-inscription of Visṇuvar-dhana,
  • (vi) the Girnār (Junagaḍ) Rock inscription of Skandagupta.
• 4. The Central Asian varietyy: specimens—
  • (i) the Bower Manuscript,
  • (ii) numerous other manuscripts written in the Central Asiatic variety of the Gupta alphabet discovered by the British and German expeditions.

I. The Eastern Variety.

Twenty-one years ago, five years before the publication of Dr. Bühler's work on Indian Palaeography, Dr. A. F. R. Hoernle recorded the following observations on the Indian script of the 4th and 5th centuries A.D.: "There existed at the time of the Gupta period two very distinct classes of the ancient Nāgarī alphabet, North Indian and the South Indian. The test letter for these two great classes is the character for m. The Northern class of alphabets, however, is again divided into two great sections which, though their areas overlapped to a certain extent, may be broadly, and for practical purposes sufficiently, distinguished as the Western and Eastern sections. The test letter in this case is the cerebral sibilant sha (ṣa)"^[62] This classification was also adopted by the late Dr. Bühler, who added two more test letters: la and ha. "The differences between the Eastern and Western varieties of the so-called Gupta alphabet appear in the signs of la, ṣa and ha. In the Eastern variety, the left limb of la is turned sharply downwards: cf. the la of the Jaugaḍa separate edicts. Further the base stroke of ṣa is made round and attached as a loop to the slanting central bar. Finally the base stroke of ha is suppressed, and its hook, attached to the vertical, is turned sharply to the left, exactly as in the Jaggayyapeta inscriptions. In the Western variety these three letters have the older and fuller forms." Another test letter, of the Eastern alphabet of this period, is the dental sibilant sa. In the inscriptions of the Eastern variety, this letter always has a loop at the end of its left vertical line instead of the customary curve or hook, cf. the form of the letter in the Allahabad pillar-inscription of Samudragupta. This form of sa has also been found in the inscriptions of the Kuṣāṇa period, discovered in Mathurā. The Kaṅkālīṭīlā inscription of the 25th year, shows that, in that inscription, all cases of sa, have this form.^[63]

The characteristics of the epigraphic alphabet of the 4th and 5th centuries A.D. have already been discussed at length by Dr Bühler.^[64] It will only be necessary to trace the history of the development of the Eastern variety in the following pages. In 1891, Dr. Hoernle perceived that, "in India proper, the North-eastern alphabet gradually came to be entirely displaced by the North-western alphabet, in comparatively very early times. This displacement must have been in progress during the earlier part of the sixth century A.D. and must have been completed about 580 A.D., for in 588 A.D., we already find inscriptions in Bodh-Gayā (inscription of Mahānāman, Fleet, p. 274), which show an exclusive North-Western character. There is not a single inscription known, so far as I am aware, about and after 600 A.D., which show the distinctive marks of the old North-Eastern alphabet."^[65] This statement will have to be examined in the light of later discoveries made during the last two decades—

• • (i) The Dhānāidaha grant of Kumāragupta I, G.E. 113 = 432 A.D.^[66]
  • (ii) The Mathurā Jaina image-inscription of the time of Kumāragupta I, G.E. 113 = 432 A.D.^[67]
  • (iii) The Karamdaṇḍa image-inscription of Kumāragupta I, G.E. 117 = 436 A.D.^[68]
  • (iv) The Amaunā plate of the Mahārāja Nandana, G.E. 232 = 531 A.D.^[69]
  • (v) The Paṭiakellā grant of the Mahārāja Śivarāja; G.E. 283 = 602 A.D.^[70]
  • (vi) The Gañjām grant of the time of Mahārājādairāja Śaśāṅka, G.E. 300 = 619 A.D.^[71]
  • (vii) The Muṇḍeśvarī inscription of Mahāsāmanta Mahāprātīhāra Mahārāja Udayasena, the Harṣa year 30 = 686 A.D.^[72]
  • (viii) The Purī grant of Sainyabhīta-Mādhavarāja II.^[73]
  • (ix) The Parikuḍ grant of Madhyamarāja, the Harṣa year 88 = 694 A.D.^[74]

The Eastern variety of the epigraphic Alphabet of Northern India of the 4th and 5th centuries A.D. did merge, as Dr. Hoernle has observed, into the Western variety. Inscriptions, discovered after the publication of Dr. Hoernle's article, show the gradual changes in the epigraphic alphabet of the 5th and 6th centuries A.D., and tend to prove that this change is already in evidence in the first half of the 5th century. This displacement of the Eastern variety of the alphabet of this period by the Western must have been completed before the end of the first half of the 6th century.

The Allahabad pillar-inscription of Samudragupta shows the fully developed form of the Eastern variety and the test letters can be observed here to their best advantage. The next inscription, in the chronological order, in which the Eastern alphabet has been used, is the Udayagiri cave-inscription of Candragupta II, on which Dr. Bühler observes "The fact that Fleet's No. 6 is found far west, near Bhilsā in Mālva, may be explained by its having been incised during an expedition of Candragupta II, to Mālva, at the command of his minister, who calls himself an inhabitant of Pāṭaliputra."^[75] Next we come to two new inscriptions both of which were incised in the year 113 of the Gupta era = 432 A.D.—

• • (i) The Mathurā Jaina image-inscription.
  • (ii) The Dhānāidaha grant.

PLATE IV.

Dinajpur pillar inscription—Saka 888 (?).

As both of these records mention Kumāragupta I by name, so there cannot be any doubt as to their proper date. The Mathurā inscription shows the typical forms of the Western variety.^[76] In the Dhānāidaha grant of Kumāragupta I, we find—

(i)	in all cases, the looped form of the dental sibilant śa, has been used,
(ii)	in all cases, the looped form of the lingual sibilant ṣa, has been used,
(iii)	in all cases, the hooked form of ha, has been used,
(iv)	in the majority of cases the hooked form of la has been used. In one solitary instance the Western variety form has succeeded in replacing the older one, viz:—Vakkralana (?) in L. 8.[77]

But in a stone-inscription incised sixteen years later, we find Eastern variety forms of sa, ṣa and ha in all cases. In the Mānkuwār inscription of Kumāragupta I, we see that ṣa, sa and ha have not changed in the year 129 G.E. = 448 A.D. The discrepancy may be explained thus. The current script of a country, as found on copper-plates, generally shows a more advanced form than that of the Epigraphic alphabet, found in stone-inscriptions. Copper-plates, in ancient India, should be taken to belong to the same class of records, as paper manuscripts or papyri of other countries. The forms of the alphabet used in them should be distinguished from the forms used in epigraphs proper. Twelve years later than the date of this inscription, we find the Eastern forms of ṣa, sa, la and ha still persisting in all cases, in the Kahāuṁ pillar-inscription of Skandagupta, of G.E. 141 = 460 A.D. But in an undated inscription of the same king, we see that the Western forms are gradually taking the place of Eastern ones. In the Bihār pillar-inscription of Skandagupta, the first half of the record shows uses of Eastern forms in the majority of cases:—

I.	la—(i) tulya, (ii) atulyaḥ (L. 1), (iii) atulya (L. 3), (iv) maṇḍalaṁ (L. 5), (v) vyālamba (L. 7), (vi) lokān (L. 9), (vii) kāla (L. 11)
II.	ha—(i) hi havya (L. 4), (ii) gṛhaṁ (L. 8);

Only in one instance we find a Western variety form, viz:—agrahāre (L. 13). But in the second half of the record, we find that the Western variety form of ha has invariably been used, in all cases. In the second half of this record there are two instances of la:—(i) kulaḥ (L. 28) and saulkika (L. 29), but as the facsimile given in Dr. Fleet's work is incomplete and does not contain these lines, it is not possible to compare the forms of la used in the second half of the inscription with those of the first half. I have found that the Bihār pillar-inscription has suffered much from exposure in the weather, after the publication of Dr. Fleet's work, and at present it is not possible to get a clearer, and more complete, inked impression than the one taken for Dr. Fleet. On the clear evidence of the Bihār inscription of Skandagupta, we have the fact that Western forms were replacing the Eastern ones in the alphabet of North-Eastern India in the first half of the 5th century A.D. The Pāli grant of Lakṣmaṇa, of the Gupta year 158^[78] = 477 A.D., shows no form, in the alphabet used, which has any resemblance to those of the Eastern variety. The Pāli grant should be included among North-Eastern inscriptions, instead of North-Western ones, as it was found about thirty miles from Allahabad It may be mentioned that the Kosām image- inscription of Bhīmavarman, of the Gupta year 139 = 458 A.D., shows the use of eastern forms and the findspot of this record is close to Pāli. In this inscription, we find that, all the test letters, ṣa, sa, ha and la, have assumed Western forms. The evidence of the Pāli grant of Lakṣmaṇa is further borne out by the alphabet used in the Amaunā grant of Nandana, of the Gupta year 232^[79] = 551 A.D. This inscription was discovered in the Gayā District of Bihar and Orissa and cannot be referred to any other class of inscription but the North-Eastern. In this inscription we find that ṣa, sa, ha and la are of the western variety. Consequently we are now in a position to reconsider the statement made by Dr. Hoernle twenty-one years ago: "This displacement must have been in progress during the earlier part of the 6th century A.D., and must have been completed about 580 A.D., for in 588 A.D., we already find inscriptions in Bodh-Gayā (Inscription of Mahānāman, Fleet, p. 274) which show an exclusive North-Western character."^[80] We are now in a position to state definitely that the movement towards the adoption of Western variety forms in North-Eastern inscriptions was already in evidence in the 4th decade of the 5th century A.D. So early as the days of the Gupta emperor Skandagupta, the change had already affected the epigraphic alphabet of the time. The displacement was completed before the eighth decade of the 5th century and all traces of Eastern variety forms or characters had disappeared from the plains of Northern India, before the beginning of the 6th century A.D. We should now proceed to the Palæographical examination of a class of records, about which there is much difference of opinion. I refer to the four copper-plate inscriptions, which have been discovered at various times during the last three decades. The first three was published by Mr. F. E. Pargiter in 1910^[81] and the last one was published by myself^[82] as well as by Mr. Pargiter^[83] in 1911. In size, script and composition the four records indicate that they belonged to the same variety. These four grants differ from all other copper-plate inscriptions discovered in India on the following points:—

(i) they are not grants of lands, made by any paramount sovereign, nor by any feudatory chief, with the sanction of his suzerain,

(ii) they purport to be deeds of transfer of property, made by certain local officials, to a private person, as well as deeds of grants, made by those private persons to certain Brāhmaṇas;

(iii) they mention a number of officials by their proper names, and not merely by designations, as usual.

The facts, quoted above, would alone go to prove that the records were spurious. But in addition to them, we have the palæographical evidence, which shows that the alphabets of two different periods and in the case of the last one, of three different periods, have been used in the composition of these inscriptions. In these records we find that, (1) ṣa, la and ha have two forms and often three; and are used in conjunction with forms of the sixth or even of the seventh or ninth centuries A.D. In the first grant: the grant of Dharmāditya of the year 3, we find that two different forms have been used, in the case of three test letters ṣa, la and ha.

I. Ṣa:—

(i) Eastern variety.

1. Ambarīṣa (L.I), 2. viṣayapati (L.3), 3. viṣayamahattara (L.4), 4. Ghoṣacandra (L.5), 5. kṣettra (L.7), 6. viṣaye (L.8), 7. viṣaye (L.10), 8-9. kṣettrāṇi and kṣettra (L.11), 10. dṛṣṭi (L.12), 11. ṣaḍ-bhāgaḥ (L.13), 12. abhilāṣa (L.14), 13. ṣalaṅga, (L.19), 14. modaneṣu (L.21).

(ii) Western variety. Strictly speaking, the forms of the letter, used in the following words, are much later in date than the North- Western Gupta alphabet. In all cases, the letter is found in the ligature kṣa and we find that peculiar curvature before ka denoting the presence of the ṣa, which we see for the first time in the inscriptions of Ādityasena and those of the Gāhaḍavāla princes of Kanauj^[84], in the 11th and 12th centuries A.D. There are five instances of this later form in the first grant—

1. kṣettra (L.16), 2. anugrahākāṁkṣiṇā (L.18), 3. kṣepa (L.21), 4. dakṣiṇeṇa (L.23), 5. kṣeṇī. (L.25).

II. La:—

(i) Eastern variety.

1. lavdha (L.2), 2. kālasakha (L.5-6), 3. durllabha (L.6), 4. lābhah (L.18), 5. saṁkalpābhiḥ (L.14), 6. Śīlakuṇḍaś=ca (L.214).

(ii) Western Variety.

1. kāle and 2. vārakamaṇḍale (L.3), 3. āluka (L.5), 4. kuṇḍalipta and 5. kulasvāmi (L.6) 6. mūlyaṁ (L. 8), 7. pustapāla (L. 9), 8. kutlya and khaṇḍala (L. 11), 10. kalanā (L. 12), 11. abhilāṣa (L. 14), 12. nalena (L. 16), 13. dhrurvilāṭyāṃ (L. 16), 14. kulya (L. 16), 15. kāla (L. 18), 16. ṣal-aṅga (L. 19), 17. uparilikhita (L 20), 18. anupālana (L.21), 19. pratipālanīyaṁ (L. 22), 20. liṅgāni (L. 28).

III. Ha:—

(i) Eastern variety.

1. vṛhac-caṭṭa (L. 4), 2. icchāmy-ahaṁ (L. 7), 3. brāhmaṇasya, 4. gṛhītvā (L. 8), 5. aradhṛtam=astī=ha (L. 10), 6. hastena (L. 15), 7. paratr=ānugraha (L. 18), 8. himasena (L. 23).

(ii) Western variety,

1. mahārājādhirāja and 2. mahārāja (L. 2), 3. mahattara (L. 4), 4. tad-arhatha (L. 8), 5. mātā-pitror-anugraha (L. 19), 6. haret (L. 26), 7. himase'a (L. 25), 8. saha (L. 26).

Similarly in the second grant from Faridpur we find that—

I. In all cases the Western variety form of ha has been used.

II. The Eastern variety form of la has been used in one case only e.g. in maṇḍala (L. 4). In all other we find the Western variety forms—

1. lavdha (L. 3), 2. kāle (L. 4), 3. gopāla (L.5), 4. khaṇḍalakaiḥ (L. 9), 5. lauhitya (L. 11), 6. kulya (L. 14), 7. akhila (L. 15), 8. satpālāni (L. 17), 9. khandalaka (L. 17), 10. pustapāla (L. 18), 11. dharmmaśīla and 12. nalena (L. 19), 18. liṅgāni (L. 20), 14. ślokāni (L. 24).

The form in the last example is very late. It is the 9th century form, found for the first time in the Dighwā-Dubhauli grant of Mahendrapāla^[85].

III. In the case of the lingual ṣa we find eastern variety forms in:

1. Nahuṣa (L. 1), 2. Ambarīṣa (L. 2), 3. viṣaya (L. 5), 4. Jyeṣra (L. 7), 5. Somaghoṣa and 6. viṣayānāṁ (L. 8) 7. ṣaṣṭha (L. 16) 8. vṛkṣa (L. 21), 9. ṣaṣṭiṁ and 10. Varṣa (L. 24), 11. śva-viṣṭhāyāṁ (L. 27) and western variety ones in:

1. kṣettra (L. 9), 2. kṣet rāṇi (L. 14), 3. ākṣeptā (L. 25); another indistinct form is to be found in hastāsṭaka. The late seventh or eleventh century form of kṣa is found in kṣettra in L. 17.

The third plate is in a very bad state of preservation and the facsmile published with Mr. Pargiter's article is very indistinct; the reverse or the second side of the plate only, is capable of being analysed for palaeographical purposes. In it, we find, that in all recognisable cases, the lingual ṣa is of the Eastern variety of the early Gupta alphabet. Both forms of ha have been used. Only one instance of the Western variety is legible:—maha in L. 3. In all other instances where the record is legible we find the use of the Eastern variety:—(1) mahattaraḥ (L. 8-9), (2) hastāṣṭaka (L. 10), (3) agrahāra (L. 22), (4) har-ta (L. 24), (5) saha (L. 25).

So also in the case of la we find that the Eastern variety form has been rarely used while the Western variety form is common:—

I. Eastern variety:—(i) Vatsapāla (L. 5), (ii) liṅgāni (L. 21).

II. Western variety:—(i) mūlyam (L. 14), (ii) kulavārān (L. 18), (iii) prakalpya (L. 18), (iv) dharmaśīla (L. 19), (v) nalena (L. 19), (vi) Vatsapāla (L. 19), (vii) kulya (L. 20), (viii) Dhruvilāṭa (L. 22), (ix) Śīlakuṇḍa (L. 23).

It should be noted in this connection that in the majority of cases we find the bipartite form of ya. In the fourth grant we find, that in all cases the bipartite form of ya, the Western variety form of the lingual ṣa and la have been used. With the exception of three instances, ha also has the Western variety form. These three instances are: (1) vrāhmaṇ-opāya (L. 11), (2) vrāhmaṇa (L. 14), (3) sahasrāṇi (L. 20-21). In addition to these, we find later forms, in the word parkkaṭṭi (L. 18) and svāminaḥ (L. 17), in case of ka and na respectively. In conclusion, we may freely say, that all four copper plates are forged. It may be asserted that, the plates belong to the transitional period, when Eastern variety forms were gradually being displaced by Western ones. But, the use of mediaeval forms, precludes such a possibility:—(1) I have already commented on the form of the ligature kṣa in the first plate. (2) Another, much later form, is that of śa, in the date of the first plate, which occurs for the first time in the Aphsaḍ inscription of Ādityasena and Dighwā-Dubhauli grant of Mahendrapāla, the Pratīhāra, of V.E. 955=898 A.D. The form of ka in parkkaṭṭi and na in svāminaḥ in the fourth grant had already been commented upon. Consequently we find that the four copper-plate inscriptions, being forgeries, are of no use in a palaeographical discussion.

 

---

PLATE V.

Bodhicaryavatara (Ms. Ga 8067) Fol. 65, obv. A. S. B.

1. Bühler's Indian Palæography (Eng. Ed.), p. 7.
2. Ibid, p. 17.
3. J.R.A.S., 1898, p.388.
4. J.R.A.S., 1905, p. 680.
5. Indian Palæography (Eng. Ed.), p. 33.
6. Cunningham's Arch. Survey Report, Vol. XV, Pl. III.
7. Fleet's Gupta Inscriptions, p. 383, Pl. xliii B.
8. Ibid, p. 34.
9. Cunningham's Archæological Survey Rep., Vol. XV, Pl. III. 1, 2.
10. Cunningham's Malabodhi, Pl. XXIV, p. 1.
11. Mahābodhi, Pl. X, Nos. 9 and 10.
12. Ibid, Pl. X, No. 5.
13. Ind. Palæo. (Eng. Ed.), p. 34.
14. Burgess, Stūpa of Amarāvatī, p. 125.
15. Cunningham, A. S. R., Vol. XX, p. 41, Pl. VI.; Vogel, Cat. of Arch. Museum at Mathurā, 1910, p. 83, C. I.
16. Annual Rep. Arch. Survey of India, 1908-9, p. 147.
17. Bühler's Indian Palæography, p. 36.
18. Cunningham's Mahābodhi, Pl. X, 2.
19. Ibid, Pl. X, 3.
20. Ibid, Pl. X, 4—7, 9—10.
21. Ibid, Pl. X, 10.
22. Ibid, Pl. X, 9, 10.
23. Ibid, Pl. X, 2.
24. Ibid, Pl. X, 9, 10.
25. Ibid, Pl. X, 3.
26. Ibid, Pl. X, 9.
27. Ibid, Pl. X, 10.
28. Ibid, PI. X, 2.
29. Ibid, PI. X, 9—10.
30. Ibid, PI. X, 9—10.
31. Ibid, PI. X, 4—7.
32. Annual Rep. Arch. Survey of India, 1908—09, p. 247.
33. Annual Report of the Archæological Survey of India, 1906—07, p. 96, No. IV.
34. Ibid, 1904-5, p. 68, PI. XXXII, No. IX.
35. Ibid, PI. XXXII, Nos. III and IV, p. 102.
36. Ibid, 1906-7, p. 95, No. II, PI. XXX.
37. Ep. Ind. Vol., VIII, p. 171.
38. J. R. A. S., 1912, pp. 701-707.
39. Ep. Ind., Vol., VIII., p. 172.
40. Bühler's Indian Palæography (Eng. Ed.), p. 40 and Ind. Ant., Vol. XXXVII, p. 25.
41. Ibid.
42. Cunningham's Mahābodhi, p. 58.
43. Epi. Ind., Vol. VIII, p. 176.
44. Ibid, p. 179.
45. Ibid.
46. Arch. Survey, Rep., Vol. I, p. 339 f.; J.R.A.S., N.S., Vol. V, p. 192 J.A.S.B., 1898, p. 274 and Ep. Ind., Vol. VIII, p. 179.
47. Epi. Ind., Vol. IX, p. 290.
48. Annual Rep. Arch. Survey of India, 1908-9, p. 133.
49. Ind., Ant., Vol. XXXVIII, p. 49.
50. Annual Rep. Arch. Survey of India, 1905-6, pp. 105-6.
51. Cunningham's Mahābodhi, Pl. X, ii; XIII and XIV.
52. Annual Rep. Arch Survey of India, 1908–09, p. 135.
53. Ind. Ant., Vol. XXXVII, p. 29.
54. Fleet's Gupta Inscriptions, pp. 262, 273; Kielhorn's List of Inscriptions of Northern India, Ep. Ind., Vol. V, App. p. 63, No. 445 and p. 65, No. 463.
55. Fleet's Gupta Inscriptions, p. 1.
56. Cunningham's Mahābodhi, Pl. XXV.
57. Ind. Ant., Vol. XXXIII, p. 40.
58. Bühler's Indian Palæography, Eng. Ed., p. 46 and note 10.
59. Ibid, Eng. Ed., p. 46.
60. Ibid, p. 47.
61. Ibid.
62. J. A. S. B., 1891, Pt. I., p. 81.
63. Epi. Ind., Vol. I, p. 384, No. v.
64. Bühler's Indian Palæography, Eng. Ed., p. 47.
65. J.A.S.B., 1891, Pt. I, p 82.
66. J.A.S.B. (N.S.), Vol. V, p. 459.
67. Epi. Ind., Vol. II, p. 210, No. XXXIX.
68. J.A.S.B. (N.S.), Vol. V, p. 457; and Ep. Ind., Vol. X, p. 70.
69. Ibid, p. 49 and J.A.S.B. Vol. V, N. S. p. 164.
70. Epi. Ind., Vol. IX, p. 285.
71. Epi. Ind., Vol. VI, p. 141.
72. Ibid, Vol. IX, p. 289.
73. J.A.S.B., 1904, Pt. I, p. 284, Pl. VI.
74. Baṅgīya-Sāhitya-Pariṣad-Patvikā, Vol. XVI, p. 185, also Epi. Ind., Vol. XI, p. 281.
75. Bühler's Indian Palæography, Eng. Ed., p. 46.
76. Epi. Ind., Vol. II, p. 210, No. XXXIX.
77. J.A.S.B. (N.S.), Vol. Pl. XX, p. 461.
78. Epi. Ind., Vol. II, p. 363.
79. Ibid, Vol. X, p. 49.
80. J. A. S. B., 1891. pt. I, p. 82
81. Ind. Ant., Vol. XXXIX, p. 193.
82. J. A. S. B., Vol. VI, p. 435.
83. Ibid, Vol. VII, p. 476.
84. Bühler's Indian Palæography, pl. IV, XVIII, 45 & pl. V, XII, XX, 44.
85. Ind. Ant., Vol. XV, p. 112.