brief analysis of the most instructive of such names found in the Vitastāmāhātmya.

This text claims to furnish an account of all the Tirthas along the course of the holy river and is designated as a portion of the Bhṛngisasamhita. Notwithstanding this pretended antiquity we find the famous Nilanaga introduced to us by the name of Viranāga (i. 58; ii. 33). This form is wholly unknown to the Nilamata, Rājatarangiņi or any old text. It is nothing but a clumsy rendering of the modern name of the village Vērnāg near which this fine spring is situated.1 The ancient site of Jayavana, mentioned by Bilhana and Kalhana, the present Zevan, is metamorphosed into Yavani (vi. 4). The village of Pandrethan which derives its name from Purāṇādhiṣṭhāna, 'the old capital,' and bears the latter designation even in Srivara's Chronicle, figures as Padadṛṣṭika (!), xii. 24. That, Mākṣikasvāmin (Māy sum)* and the Mahāsarit (Mār) ↳ appear as Māyāsīmā and Mārī, can after this specimen of fancy nomenclature scarcely surprise us.

But we must all the same feel somewhat startled when we find that this text which claims to be revealed by Siva, refers repeatedly to the modern village of Shadipur, at the confluence of the Vitasta and Sind, by the name of Sāradāpura. Shādipūr, an abbreviation for Shahabuddinpur, was, as Jonaraja's Chronicle shows, founded only in the 14th century by Sultān Shahābu-d-din. Quite on a level with the knowledge of old topography here displayed are many other references to localities, e.g., the mention of the modern garden Shalimar, a creation of the Mughals (Sālamāra), xxi. 39; of the ancient Huşkapura as Uṣaḥkarana (for Uşkür!), xxix. 103, etc.

In several cases these fancy renderings of modern local names are explained by whimsical etymologies which again in due turn give rise to new-fangled legends quite in the style of the old nidānakathās.

Similar proofs of modern origin can be traced in several other popular Mahatmyas, though perhaps not with equal frequency. Thus we find in the Haramukuṭagangāmāhātmya the name of the sacred mountain itself transformed from Haramukuța into Haramukha (the

1 The name Vērnāg is probably derived from the name of the Pargana Vēr, mentioned by Abu-l-Fazl, ii. p. 370.

2 See below, § 105.

8 See Rajat. iii. 99 note and below, § 89; also Srīv. iv. 290.

4 See Rajat. iv. 88 note and below, § 99.

5 Compare Rajat. iii. 339–349 note and below, § 65.

6 See Jonar. 409. A popular etymology accepted in good faith by more than one European writer, sees in Shadipur the village of the marriage,' scil. between the Vitasta and Sind Rivers!

7 Compare Rājat, i. 168 note and below, § 124.

present Haramukh), the ancient site of Bhutesvara (Buth'sēr) so wellknown to the Rājatarangiņī and all old texts, turned into Bodheśvara, etc. In the Amaranāthamāhātmya of which there is a comparatively old copy in the Poona collection, we are also treated to Padṛṣṭi as the Sanskrit name of Pandrethan, to Suśramanaga (for Ks. Suśramnag) as the name of the lake where the Naga Susravas of the old legend took up his abode, and the like. Examples of local names similarly perverted in other Māhātmyas will have to be mentioned passim in our account below.

It is important to note that by the side of texts like those just mentioned, there are others which on the whole show close conformity with our genuine old sources both in matter of legend and local names.3 And even in the Māhātmyas which in their present form we have every reason to consider as recent compositions, there is often abundant evidence of the use of earlier materials and traditions. It will be easier to understand the singular discrepancies in the value and character of these texts on examining the peculiar conditions under which they have originated.

Origin and purpose

of Māhātmyas.

30. The Mahatmyas are in the first place hand-books for the Purohitas of the particular Tirthas who have the privilege of taking charge of the pilgrims. They serve the priests as chief authorities for the claims they put forth on behalf of the holiness of their Tirtha, and for the rewards they promise for its visit. They are also intended to support their directions as to the rites to be observed by the pilgrim, and the route to be taken by him on the journey. It is usual for the Purohitas to recite the Māhātmya for the benefit of their clients in the course of the pilgrimage tour. At the same time its contents are expounded to them by a free verbal rendering in Kaśmiri.

1 See below, § 57. The kh at the end of the modern name is due to a phonetic law of Kasmiri which requires the aspiration of every final tenuis; see J. A. S. B., 1897, p. 183.

2 Compare Rajat. i. 267 note. The modern Kś. form Suśramnāg is the regular phonetic derivative of Susravanāga by which name the lake is designated in the Nilamata, Haracaritacintāmaņi, etc.

8 Among such the Mahatmya collection known as the Sarvāvatāra (No. 213 in my list of MSS.), the Martāṇḍamāhātmya (No. 219), the Vijayesvaramāhātmya (No. 220), may be particularly mentioned. None of these, however, are now known to the local Parohitas, more recent and inferior texts having taken their place.

4 Thus e.g., the Mahatmya of the present Isabar (Iseśvara; see Rajat. ii. 134) shows plainly its very recent origin by calling the Tirtha Isavihara (a garbled Yet it knows correctly reproduction of Isabrōr < Iseśvara), and by similar blunders. the sacred spring of Satadhārā already mentioned by Kṣemendra,

As but very few of the priests have enough knowledge of Sanskrit to follow the text intelligently, these translations are more or less learned by heart. Often as my manuscripts show, interlinear Kasmiri glosses are resorted to in order to assist the reader's memory.

These local priests known now in Kasmir as than pat (Skr. sthānapati), are as a rule quite as ignorant and grasping as their confrères, the Pujaris, Bhōjkis, etc., of India proper. They are held deservedly in very low estimation by the rest of the Brahman community. That their condition was more or less the same in earlier times too, though their influence and numbers may have been greater, can be safely concluded from more than one ironical allusion of Kalhana. These are the people to whose keeping the Mahatmya texts have always been entrusted. Their peculiar position and calling explain, I think, most of the curious changes which the latter have undergone.

Tenacious as local worship is, there is the evidence of concrete cases to show that not only the route of pilgrimage, but the very site of a Tirtha has sometimes been changed in comparatively recent times. In proof of this it will suffice to refer to the detailed account I have given of the transfers that have taken place in the case of the ancient Tirthas of Bheḍā and Sāradā.2 Minor modifications must naturally have been yet far more frequent. The visit of a principal Tirtha is regularly coupled with bathings, Sraddhas and other sacrificial functions at a series of other sacred spots. The choice of these subsidiary places of worship must from the beginning have depended on local considerations. As these changed in the course of time, variations in the pilgrimage route must have unavoidably followed.

To bring the text of the Mahatmya into accord with these successive changes was a task which devolved upon the local Purohitas. The texts we have discussed above bear, in fact, only too manifestly the traces of their handiwork. Sound knowledge of Sanskrit and literary culture are likely to have been always as foreign to this class of men as they are at present. When it became necessary for them to introduce the names of new localities into the text of the Mahatmya there was every risk of these names being shown not in their genuine old forms, but in hybrid adaptations of their modern Kaśmiri equivalents. This risk naturally increased when Sanskrit ceased to be the official language of Kaśmir, and the knowledge of the old local names was gradually lost even among those maintaining scholarly traditions in the country.

1 Compare Rajat. ii. 132 note and v. 465 sqq.; vii. 13 sqq.; viii. 709, 900 sqq., 939. 2 Compare Notes A (Rājat. i. 35) and B (Rājat. i. 37).

Popular etymology

in local names of Māhātmyas.

31. Another potent cause seems to have co-operated in this vitiation of the local nomenclature of the Mahatmyas. I mean popular etymology.' We have already referred to the tendency displayed throughout these tracts of making the names of localities, rivers, springs, etc., the starting-point for legendary anecdotes. For men of such very scant knowledge of Sanskrit as the than pat's invariably are, it was naturally far easier to explain such etymological stories when they were based on the modern local names.

It is undoubtedly this reason which has, e.g., led the compiler of the present Haramukuṭagangāmāhātmya to substitude the name Karankanudi for the old Kanakavāhinī. By the latter name the stream coming from the Haramukuta lakes is designated in all our old texts, as explained in my note on Rajat. i. 149-150. By turning Kankanai, the modern derivative of this old name, into Karankanadi, 'the skeletonstream,' the compiler of the Māhātmya gets an occasion to treat his readers to a legend likely to appeal to their imagination. The river is supposed to have received this appellation, because Garuda had dropped at its Samgama with the Sindhu the skeleton (karanka) of the Rși Dadhici which Indra before had used as his weapon, etc.1 This story, it is true, is wholly unknown to the Nilamata or any other old text. But, on the other hand, it has got the great merit of being easily explained and proved to any Kaśmiri pilgrim. He cannot fail to realize the manifest connection between Karanka and his familiar kranz, skeleton.'

An exactly similar case of popular etymology' has been noticed in the analysis of the Saradāmābātmya as contained in my Note B (i. 37). There the name of the village Sun-Drang is reproduced as Suvarṇārdhāngaka and explained by a legend, how the Muni Sandilya had at that spot half his body (ardhāngaka) turned into gold (suvarṇa), etc. In reality the village name is derived from the old term Dranga, 'watch-station,' by which the place is mentioned by Kalhana. The distinguishing prefix Sun?-, meaning 'gold' in Kaśmīri, was given to it, because it lay on the route to the old gold-washing settlements in the Kisanganga Valley.3

1 The story is spun out at great length in Patala iii. of the Haramukuṭagangā. mahatmya, MS. No. 221.

2 See viii. 2507, 2702.

8 For other examples of local names in Mahatmyas metamorphosed for the above reason, compare my notes Rājat. vi. 177 (Bhimadvipa in the Mārtāṇdamāh., for Bumaza); i. 267 (Seṣanāga in the Amaresvaramāh., for the older Suśramanāga, recte Susravonaga); Note C, i. 124 (Jyesthesvara, the present Jyether, turned into a site of Jyestha), etc.

It would be easy to multiply examples showing the strange vicissitudes to which old topographical names are exposed at the hands of the local Purohita. But the explanations already given will suffice to prove that the topographical data found in Māhātmyas can only then be used safely when they are critically sifted and supported by our more reliable

sources.

A critical examination of these data is, however, much impeded. by the difficulty we experience in fixing the exact age of particular Māhātmyas and their component portions. Even in the case of apparently old texts modern additions and changes may be suspected, while again the most recent concoctions may preserve fragments of genuine tradition. In view of these considerations I have not thought it safe to crowd my maps with hundreds of names of petty Tirthas as found in the Mahatmyas, but have marked only those pilgrimage sites the ancient names of which can be established with certainty.

1 The difficulty here indicated is increased by the fact that no really old manuscripts of Māhātmyas seem to be preserved in Kaśmir. MSS. written on birch-bark, i.e., earlier than the 17th century, are quite unknown at present. Of the numerous paper MSS. I have examined, none seem to me older than two centuries at the utmost. It is probable that this absence of older copies is due to the rough usage to which Mahatmya MSS. are exposed when carried about on the pilgrimage tours.

2 I am glad that chance gave me an opportunity of gaining some personal experience of the manner in which Māhātmyas are occasionally produced. Some ten years ago the Purohitas or Bachbaṭṭas of the Ganapatyar quarter in Srinagar recovered an ancient Linga from a Mosque and began to erect a small shrine for it near the river Ghat of Malayār. Guided by a local tradition which, as far as I can judge, may be genuine, they believed this to have been the site of the shrine of Siva Vardhamȧnesa mentioned already in the Rajatarangiņi (see note ii. 123). The Linga was re-consecrated accordingly by this name.

In 1891, when examining old sites in this part of the city, I also visited the temple of Vardhamaneśa then under construction. The interest I showed in the old Linga and in the tradition regarding it, coupled with an appropriate Dakṣiņā, soon secured me the confidence of the head-Purohita of the little shrine. Pandit' T.R., a man more intelligent than the average of his fraternity, was not slow to confess to me that the Māhātmya of the Tirtha in spe was as yet under preparation. Some weeks later when in camp near Srinagar, I received the visit of my Purohita from Vardhamāneśa's shrine. He brought me the draft of the new Mahatmya and asked my assistance in revising it.

I found it to consist chiefly of extracts from the Vitastāmāhātmya. The passages dealing with Vardhamāneśa and the neighbouring Tirthas within the city had been suitably amplified with laudatory verses in the usual Mahatmya style culled from other texts. The vested interests of other local shrines had received due recognition by being included in the Yatra of Vardhamāneśa. I did what I could to indicate the genuine names of these localities. This quasi-antiquarian co-operation does not seem to have detracted from the popularity of the new Mahatmya among the Bachbaṭṭas of Ganapatyār.