« السابقةمتابعة »
Bäcklund may have been part of the result of this attempt. Mr. Bendall wrote to me on the 15th July, 1898, "I have been comparing your pamphlet about the xylographs from Central Asia with a block-print recently acquired by the British Museum from Lieut. Cobbold. What is curious is that it is a duplicate of the book figured on your first Plate, but does not contain the writing between the two columns of print to which you call attention." This observation of Mr. Bendall very possibly gives the key to the situation. If there exist any forgeries, they are, in all probability, duplicates of genuine books that have been discovered. The preparation of a duplicate is probably well within the capability of a modern Khotanese forger, but the hypothesis that he is capable of inventing not only one but several scripts, and of intricate, but self-consistent systems of their arrangement in books, and finally of binding them after a method, quite unknown in Khotan at the present day, contains more elements of improbability than the hypothesis of the genuinenesses of the books.
The manufacture of duplicate block-prints postulates the existence of old blocks from which new ones may have been prepared, and from which (or from their new facsimiles) the modern reprints (if there are any) must have been made. I have shown in the description of the First and Second Sets, how utterly improbable it is that the blocks of type can have been invented by the forger. The overwhelming probability is that sets of old blocks of type have been discovered in the Takla Makan, and from these reprints may have been made. But moreover, actual old books printed from those blocks and representing each of the nine Sets must have been found. For the systems of printing and binding which are used in the books are unknown in Khotan in the present day, and imitations could not have been made, unless models had been found. Add to this not only that most of the books, though printed (as I believe) on Khotanese paper, are printed on varieties of it (viz., IIIa bc) which are not known in Khotan at the present day; but also that there are others (as those in the Eighth and Ninth Sets) which are printed on paper of a kind which is not Khotanese at all. That some of the block-print books are printed on paper of the variety IIId is quite true; but this fact, by itself, does not prove forgery; for it cannot be doubted (considering Oriental conservative habits)
6 Published in the Proceedings of the Asiatic Society of Bengal, for 1898.
7 An alternative hypothesis would be that no blocks have been found, but only books; and that from these books new blocks have been prepared, and then employed to print new books. The prints, however, as shown by measurements, are so accurate facsimiles, that considering the inveterate inaccuracy of Orientals this hypothesis of the imitation of new blocks from old prints seems excluded.
that the particular process of paper manufacture which is still followed in Khotan may have been in vogue there for centuries before. At all events, it cannot well be supposed that those books, which are printed on old paper of a kind never known, or no more known in Khotan, are modern Khotanese forgeries. If they are modern forgeries, they must have been forged somewhere else than Khotan; and this complicates the theory of forgery with additional improbabilities. Further, some of the books, admittedly or probably written on paper of Khotanese manufacture, exhibit peculiarities which it may safely be said, would not have occurred to a forger to introduce. I refer, for example, to the sketches of heads, which are found in books No. I of the Fourth Set and No. V of the Seventh Set, to the occurrence of the recensions Ig and Ih in book No. VIII of the Fifth Set, and to the Pōthi with its entirely Indian arrangement. Such books cannot well be forgeries.
Further, forgeries may be admitted to be quite possible in the case of block-prints, in the reprinting of which from genuine old blocks there is no serious difficulty. But it is different with manuscripts; and let it be noted, that there are not only fragments of manuscripts, but whole books-some of them fairly large books-of manuscript. The difficulties of forging these would be enormous. In this case there are no duplicates. There are, indeed, a fair number of them in the collection; but they are all different from one another. It would mean that they had all been forged, within a comparatively short time, from no models whatsoever. Some are written on paper which is not Khotanese at all; others are on paper, similar to that of some of the block-prints, but of a variety now obsolete (viz. IIIa). Some are bound in the Indian fashion of a Pōthi; others in the Khotanese fashion with copper pegs or twists of paper. These manuscripts cannot be forgeries; and pro tanto they make against the hypothesis of forgery in the case of the block-prints.
The mystery of the scripts-so many, and so intricately arrangedis, no doubt a difficulty. But to solve it by the hypothesis of forgery is only to substitute one riddle, and a harder one, for another. How can Islām Akhūn and his comparatively illiterate confederates be credited with the no mean ingenuity necessary for excogitating them? Moreover the riddle of one of the scripts, which occurs in two of the manu
8 These sketches are not easily observable. The books were some months in my hands, before I discovered them, and I did so only on carefully examining them page by page. Their existence does not appear to have been known either to Mr. Macartney or to Islam Akhun who sold them to him. It does not seem probable that a forger would have omitted to draw the buyer's attention to the existence of such a valuable peculiarity in his own handiwork.
script books written on Khotanese paper (variety IIIa), has been solved. In January 1898 I showed these books to Sir Charles J. Lyall, and he agreed with me that the script seemed to resemble Pahlavi and to be in verse. In December last, when I had an opportunity of showing them to Dr. Aurel Stein, who has made Iranian scripts and languages a special study, he at once recognized the Pahlavi script in verse. He even read some portions of it, though, of course, as will be readily understood by those who know the difficulties of reading unknown texts in Pahlavi, it was not possible for him, at such short notice, to determine what the purport of the text might be.
Finally to add a minor point, book No. VJ of the Second Set, is a mere fragment. One cannot easily conceive why a forger should sell a portion of a forged book of a kind, of which he could with comparative ease fabricate a large number of complete copies; while it is perfectly natural that he should dispose of a genuine old book, even if he had found or secured only a portion of it.
To sum up, the conclusion to which, with the present information, I have come, is that the scripts are genuine; and that most, if not all, of the block-prints in the Collection also are genuine antiquities; and that if any are forgeries, they can only be duplicates of others which are genuine, and must be found among the books of M. 8 and M. 9 which are written on the IIId variety of Khotanese paper. By duplicates, I do not mean such in point of size or variety of paper, but with reference to the arrangement of the impressions of the block, or blocks, on the pages. The determination of whether or not there are any such duplicates in the British Collection, must remain over for a future opportunity of examination of the block-prints.
In addition to the block-printed text, two books (No. IV of the. First Set and No. III of the Second Set) conWritten Legends. tain additional small legends, inserted in blank intervals between the repetitions of the text. Their letters are larger than, and their direction sometimes different from, that of the surrounding text. They have every appearance of not being printed but written by hand. Their shape is irregular, and their ink is darker and does not look as if imprinted from type.
On the whole the block-print books are in a fairly good state of
preservation. Some of them are much torn or otherwise damaged. Many of them bear stains of oil or other fatty substance; some also seem to have been exposed to the action of fire or water, as their leaves are found more or less strongly singed or rotten.
FIRST SET. (Plates V-VIII, XI and XIII.)
This set comprises eight copies, namely, one pōthi and seven books. Its text is also found on three pages of book No. II of the Second Set; and some portions of the text of it are also found on one page of book No. VIII of the Fifth Set and in book No. III of the Sixth Set.
The text consists of five portions, which comprise 5, 2, 7, 2 and 4 lines respectively, and which I shall distinguish as the formulas A, B, C, D and E. The total text, therefore, comprises 20 lines of type. Each line appears to contain from 10 to 13 letters.
The entire text of 20 lines, with the five formulas, arranged in the order above given, is found in four books (Nos. I, III, IV, V). One book (No. VI) and the pōthi (No. II) contain only the formulas A, C and E; with this difference, however, that in book No. VI the three formulas are arranged connectedly in the same order as in Nos. I, III, IV, V, only leaving blank intervals in the place of B and D; while the pōthi gives the three formulas A, C and E independently of one another, in no connected series and on different pages. Of the remaining two books, one (No. VII) gives only a portion of formula A, viz., lines 1, 2, 4 and 5, omitting line 3; while the other (No. VIII) gives only a portion of formula C, viz., lines 9, 11, 13 and 14, omitting lines 8, 10 and 12. In book No. III of the Sixth Set the same three formulas are found, but not in any connected series; viz., (1) the formula A, mutilated as in No. VII, (2) the formula C, as in No. VIII, (3) a portion of formula E, viz., lines 17, 18, 19, omitting line 20. The two formulas B (lines 6 and 7) and D (lines 15 and 16) have never been found by me separately from the entire text. These formulas, therefore, are found only in the books Nos. I, III, IV, V.
Accordingly the text of the First Set exists in the following eight forms or recensions, denoted by Ia, Ib, Ic, Id, Ie, If, Ig and Ih.
(1) The full text (Ia) of 20 lines, consisting of the formulas A, B, C, D and E, in a connected series. This is found in books Nos. I, III, IV, V; also in book No. II of the Second Set.
(2) The shorter text of 16 lines (Ib), consisting of the formulas A, C and E, with proportionate blank spaces for B and D; the whole in a connected series (as in Ia). This is found in book No. VI.
(3) The three formulas A (Ic), C (Id), and E (Ie), given separately, and forming no connected whole. In pōthi, No. II, also in book No. VI. (4) The formula A (If), in a mutilated form, viz., lines 1, 2, 4, 5. In book VII; also in book No. III of the Sixth Set.
(5) The formula C (Ig), in a mutilated form, viz., lines 9, 11, 13, 14. In book No. VIII; also in book No. VIII of the Fifth Set and in book No. III of the Sixth Set.
(6) The formula E (Ih), in a mutilated form, viz., lines 17, 18, 19. In book No. III of the Sixth Set.
I have carefully measured these various recensions of the text. The measurements were made vertically, horizontally and diagonally across the prints. I measured, in this manner, the whole text in all its forms, as well as groups of lines, single lines, half-lines, groups of letters and single letters. The result was invariably the same; the corresponding measures in the several books exactly agree, in whatever variation they may be taken. Thus measuring the recension I a, from the top of the last letter of the first line to the bottom of the last letter in the last line (both on the left-hand side of the column, see Plate V), the distance is exactly 6 inches. Measuring similarly, the distances in the formulas A, C, and E are 13, 1% and 1 inches respectively; and again measuring similarly, the width of each of the intervals between A and C, and between C and E (within which the two formulas B and D are placed) is of an inch. Measuring the recension Ib, exactly the same result is obtained. This recension, as has been already explained, consists of the formulae A, C and E, with blank spaces for B and D. The distance from the top of A to the bottom of E, including the blank spaces, is exactly 6 inches; the width of each blank space is of an inch; and the widths of A, C and E are 13, 15% and 1 inches respectively. The widths of the latter three formulas or the three recensions Ic, Id, Ie (see Plates VI and VII), when they occur separately in the Pōthi (No. II), are precisely the same; and the same is the case with them in their mutilated forms If, Ig, Ih (see Plates XI and XIII). Thus in the recension If, a blank space is left, for the omitted line 3, between the lines 2 and 4; and measuring from the top of line 1 to the bottom of line 5, the distance is, as before, exactly 13 inches. It is evident, therefore, from these comparative measurements that the xylographs of the first set were all printed either from the very same block, or from a number of blocks with facsimile type.
It can easily be shown that the latter alternative is the true one, and that in all probability eight blocks have been employed in printing the xylographs of the first set. A block was prepared in this wise. From the smoothened surface of a rectangular (oblong) piece of wood the type was cut out in relievo by counter-sinking the background. Along the edge of the piece of wood a thin ridge was also left in relievo, enclosing the type and the counter-sunk background on all four sides. In order to take an impression the surface of the type was inked. Of course, the ridge was also inked, but this appears to have been done very imperfectly, for in no case did it give more than a very intermittent J. 1. 13