532 THE ANCIENT BUDDHIST SHRINES OF MÏRAN LChap. XIII
Of the inscriptions it was easy to secure exact copies by means of tracings. But for the frescoes I found it practically impossible at the time to obtain a record worthy of their artistic and archaeological importance. Even a professional photographer, working with special plates and appliances, might have found his skill taxed in the attempt to do justice to the harmonious, but often faded or partially effaced, colours of these paintings under the conditions prevailing. For an amateur like myself they were almost prohibitive. It was difficult enough to squeeze myself, in the bulky fur kit rendered necessary by the bitter cold, into a position low and distant enough to photograph a painted dado just above the floor and on the curving wall of a passage barely seven feet wide. The violent winds rendered photography impossible for days, or else by the thick dust haze raised made the light so poor that prolonged exposure was necessary, with increased risk of the camera shaking in the gusts. The difficulty of securing satisfactory negatives was much increased by the intense cold, which prevented development being done at night except at the risk of the plate freezing in the tent. In order to reduce the risk of total failure, I laboriously took several complete rounds of the frescoes with varying light and exposure. But when development of the plates became possible some four months later, it showed that the record secured was far from being adequate. As a result of this partial failure, it became necessary to strengthen in places the photographic prints shown in Figs. 134-43 before they could be reproduced satisfactorily through the half-tone' process. I was fortunately able to entrust this delicate task to the qualified hand of my friend Mr. F. H. Andrews. Combining as he does the artist's eye and the critical archaeologist's accuracy, his help affords full assurance that this ` touching up ' was effected with the utmost caution and has not interfered with the faithfulness of the reproductions.
It was a matter of keen regret for me at the time, and has become still more so since, that I had no means of reproducing those fine paintings in colour. 'l'he alternative was to remove them bodily from the walls. In spite of the reluctance I felt to the quasi-vandal proceeding of cutting up a fresco composition like the frieze into panels of manageable size, I should have felt bound to attempt it, had not a carefully-conducted experiment convinced me that, with the means and time then at my disposal, the execution of such a plan would have implied grave risks of virtual destruction. The plaster of the cella wall here differed materially from that in M. iii. It consisted of two distinct layers, of which the outer one was remarkably well finished and smooth, but only about a quarter of an inch thick and exceedingly brittle. The inner layer, about an inch thick and softer, had very little admixture of straw, and in consequence broke far more easily than the plaster surface of M. i11, which, uniform in surface and full of chopped reed straw, possessed more cohesion. When I was, with the help of Naik Ram Singh, very carefully removing the Phrygian-capped putto head M. v. x and the male portrait head M. v. vi (Plate xLiv) from the dado, I found it impossible to prevent portions of the thin outer layer of plaster breaking off in fragments, as its backing of soft mud plaster was being loosened from the wall.
So 1 reluctantly realized that there was here no hope of safely detaching any larger fresco panels unless I could first have the brick wall behind systematically cut away by a kind of sapping, and special appliances made for first strengthening, and then lifting off, the curving panes of mud stucco. It was certain that, even if somehow I succeeded in improvising appliances in my desert camp devoid of all resources, it would necessitate the sacrifice of weeks to carry through such difficult operations and to assure safe packing for transport over huge distances. A variety of practical considerations, connected with the physical difficulties to be faced on the journey through the Lop desert to Tun-huang and with the tasks ahead, made it clear that I could not safely delay the start of my caravan, for the new field of work awaiting me far away in the east, beyond the latter half of February. In my Personal Narrative I have fully explained the cogency of these