Sec. i] RUIN N. t AND FIRST FINDS OF INSCRIBED TABLETS 321
falls within the first three centuries of our era. Hence, while still engaged in gathering the remarkable documents that were coming to light that day in such surprising numbers, I felt absolutely assured as to their exceptional value. At first the apprehension obsessed me, that these strange records, strikingly similar in their outward form, and almost all showing when complete an identical short formula at their commencement, might prove to be mere replicas of the same text ; perhaps a prayer or an extract from some Buddhist sacred text. But such hurried comparison and study of the tablets as I was able to make during the evening fol. lowing their discovery, quickly showed that their text varied greatly in both extent and matter ; and as I steadily continued it during the scanty hours of rest of the next few days, evidence gradually accumulated in support of a far more encouraging supposition.
The care taken about the sealing of most of the tablets from the first suggested that Prakrit
their contents were of a more practical nature—letters, perhaps, or documents of official character. language of
I could not fail to realize how much the historical interest of the finds would be increased thereby ; but only decipherment could establish the fact, and for the exceptional difficulties of this the experience furnished by the Dutreuil de Rhins fragments, and more than one Kharosthi inscription from the Indian North-West Frontier, had prepared me. I could not hope to attempt on the spot what, owing to the very cursive character of the writing and all the uncertainties of language and contents, has since proved so serious a task even for expert epigraphists working in their study. Hence, I had all the more reason to feel gratified when a series of philological observations bearing on the phonetic character of the characters, single or compound, and on the recurrence of particular inflexional endings, together with the tentative identification of a few detached words, proved that the language was an early Prakrit, probably of a type akin to the dialect found in the legends of the Sino-Kharosthi coins of Khotan and in the Dutreuil de Rhins codex.
As regards the character of the contents in the largest portion of the finds, I could remain Official no longer in doubt when, a day or two later, I succeeded in definitely deciphering the brief documents. initial formula which appeared invariably at the head of the text on all wedge-shaped under-tablets, and which, curiously enough, had at first been a cause of misgivings 4. Its wording mahanuava maharaya lihati, ` His Excellency the Maharaja writes,' at once established that the documents thus prefaced conveyed official orders. But even without this explicit assurance as to the contents, there seemed enough in the first day's discoveries to justify the conclusion that, with the Kharosthi script transplanted from the extreme north-west of India, an early form of Indian speech had also been brought into use within the territories of ancient Khotan, probably from the same region. Such a fact could be accounted for only by historical events of far-reaching importance, or else by ethnic movements little suspected hitherto. The hope of recovering evidence that might help to elucidate the problem thus raised made me look with intense interest for additional epigraphical finds at the site.
My expectation of such finds soon proved well founded, when I began next morning Clearing of (January 29) the clearing of the southern wing of the ruined building. Owing to the scanty room N. iii. cover of sand, and progressive erosion, the walls in this part of the building had suffered even more than in the row of small rooms excavated on the previous day. As shown by the plan} (Plate XXV III), it adjoined at right angles the eastern end of this row, and communicated with it by a door leading at first into a small room (N. iii.), only to ft. broad, which might
4 See for this formula, always separated from the body of 137, xv. 24, N, ix. i, N. xv. 71, reproduced • in Plates
the text by a considerable interval, the under-tablets of N. xv. XCVIII—C.