Sec. vii] GENERAL OBSERVATIONS ON SITE 381
scant prospect of interesting results. A change in the arrangements already made for the, morrow's start eastward would have implied fresh hardship to the men, exhausted by close on three weeks' continuous labour, and to the camels, equally hard-worked in the transport of our supplies of ice and water. Hence I decided to leave those structures unopened.
On their tramps between camp and the various ruins the men, always on the look-out for ` treasure ', had occasion to pick up from the sand or eroded loess ground a variety of small objects in metal, stone, or other hard substance, which under my standing orders were shown to me, and acquired for suitable compensation when of any antiquarian value. Many of the miscellaneous relics thus obtained (all marked with N. oo and entered at the end of the descriptive list) have been reproduced in Plate LXXIV, some also in Plates IL and L ; and it will be convenient briefly to note here those of special interest. Among metal objects the fragments of circular mirrors cast in bronze, apparently of Chinese make (N. oo9, 12. f, g, h, see Plate LXXIV) are conspicuous. The obverse bore in each case ornamentation in bold relief. The only object in gold is the small ornament N. 0017 (Plate LXXIV), which may have formed part of a pendant or earring. It is made of thin gold-plate filled with lac or wax, and has small jewels of pebble or glass set at one end. Of bronze rings we have several bearing devices cut into the metal and still recognizable, such as N. 0014. f (Plate LXXIV), and N. 0014. g, 15, 16 (Plate IL). With the exception of the well-preserved signet-ring (N. 0015), which shows a lion with a bird clinging to his breast, the devices are very crude, consisting of a few strokes only. The square bronze seal N. ooi 1. b (Plate L) is decorated with a geometrical design closely resembling those on some seals acquired from the Ak-sipil site. The two bronze arrowheads, N. oo5, 14. c (Plate LXXIV), are of superior workmanship, and still show the arrangement by which they were secured to the shaft. Owing to oxidization few of the small iron objects found still retained recognizable shape, one of these being the buckle N. oo8 (Plate LXXIV). Beads in pebble, glass, shell, and pottery were picked up plentifully, see N. 0014. d, 20. a, 21. b in Plate LXXIV. Among them a number of small gilt glass beads (N. 0020. a) were found at one spot near N. ix, and had evidently once formed a string. Glass was used also for seals, as shown by N. oo6 (Plate L). The fragment of good white glass, N. 0019 (Plate LXXIV) was apparently the foot of a vessel. Whereas the glass objects may be supposed to have come from the West, it is more difficult to determine the origin of the pyramidal seal with scroll design N. 0018 (Plate L), and the Amalaka-shaped bead N. oo5. b (Plate LXXIV), both made in a hard ceramic substance, green or blue, which resembles porcelain, but is correctly described as a kind of frit or celadon. According to information kindly supplied by Mr. C. Read, of the British Museum, its manufacture was known from a very early period not only in China, but also in some parts of Western Asia. Beads of exactly similar shape are found also among Roman work.
None of the small objects thus picked up can be ascribed to a date appreciably later than that which, as ascertained above, saw the final abandonment of the site. But it is well to bear in mind that there is no chronological limit in the opposite direction determined by equally clear evidence. Objects which may have got lost and hidden in the soft loess ground during widely different periods would be found lying side by side after the strata containing them had been successively eroded. The importance of this observation becomes evident when we remember that ground unprotected by buildings can be proved to have undergone erosion to a depth of 15 to 25 ft. below the ground-level as it existed about the last quarter of the third century A. D. The worked flint flake (N. 003) probably of Neolithic make, picked up on the eroded ground immediately to the south of my first camp (C. 93) supplies tangible proof of the need of such