Sec. iv] ANCIENT DOCUMENTS ON WOOD AND LEATHER 345
smooth sheepskin, with the Kharosthi text written on one face only and in lines parallel to the longer side, as seen in the specimens reproduced in Plates XCI—XCI II. The length of the extant specimens varies from I Os- in. in N. xv. 319, to 68 in. in N. xv. 333 (see Plate XCI I) ; N. xv. 305 (see Plate XCIII) shows the greatest width, 54 ins., while N. xv. 42, with a width of 2 in., is the narrowest. N. xv. 88 (see P1. XCI I) and N. xv. 35o show the longest text, each with eleven lines. Separated from the text and near the bottom edge of the obverse to the R. proper, there appears in the majority of the complete documents a brief date entry, mase ... divase ... , with numerical figures indicating the month and day 2, as seen in the complete specimens which Plates XCI—XCIII reproduce. In no case did I note a record of the year. Almost all the documents which have preserved the topmost portion of the text show there the same initial formula, mahanuava maharaya lihati, indicating official origin, which we have mentioned before as regularly introducing the text of wedge-shaped under-tablets s. Here, too, it is invariably separated from the commencement of the text proper by a considerable interval, as seen in Plates XCI, XCI I. For the few documents like N. xv. 333 (see Plate XCII) which do not open with this formula, we may assume a non-official character.
The great majority of the documents on leather were found folded up into little rolls, and in the case of the rest previous folding was indicated by the leather being worn or otherwise marked where the folds had been. The well-preserved documents still found in their folded condition could in most cases be opened out by me without serious difficulty ; but it required expert treatment at the British Museum to make these pieces of leather, bent by the folding of so many centuries, assume the flat appearance now shown in their reproductions. In Plates XCI—XCIII the lines of the folding can still clearly be seen, even where they have not, as in the case of N. xv. 88, led to the leather on the edges of the folds getting worn and broken. The reproduction in Plate XCI of the document N. xv. 310, both before and after its opening, best illustrates the method of folding usually adopted. The piece having been folded longitudinally into half its size, was next folded a second time in the same direction, in such a way that the lowest quarter of the piece, containing ordinarily only the date entry, would form one of the outside faces of the narrow roll, but with its written surface turned inwards. Thus the whole of the writing of the obverse was protected. Finally the roll thus formed was doubled over vertically, as seen in the folded state of N. xv. 310. In the latter document the bottom edge towards the left proper is seen cut off into a kind of half-attached strip, the extreme end to the left proper being turned down so as to expose just the initial formula of the topmost line. Remains of a similar cut-off strip are noticeable also, though less clearly, in some other leather pieces (e. g. N. xv. 319). This has led me to conjecture that this strip might in some way have been utilized in connexion with the fastening of such documents. What the exact manner of fastening was in the case of leather documents I have not been able to ascertain fully, but there are some observations bearing on this point which I think it useful to record here.
In view of the elaborate and ingenious arrangement which, as I shall presently have occasion to explain, was adopted in order to secure the contents of communications written on ` double-wedge ' or double rectangular tablets against unauthorized inspection, it appears a priori probable that some device must have been available for similarly protecting the contents of leather documents, especially when the latter, as the presence of the initial formula in the case of most of the extant specimens proves, conveyed official orders. An envelope after our fashion would,
2 See N. xv. 88 (Plate XCII), 112, 164, 201, 304, 3057 310 (see Plate XCI), 319, 333.