^{page 237 note 1} But rare exceptions occur in the ancient buildings and city walls of early cities of Pegu, especially on the eastern coast of the Gulf of Martaban, where laterite has been to some extent employed. As regards hewn stone, only two buildings at Pagān are constructed with it, the quality being sandstone.

^{page 238 note 1} Ancient manuscripts are extremely scarce, and the oldest ones known are on palm-leaf and do not, as a rule, go back more than three centuries. No coins with inscriptions or monograms dating earlier than the fourteenth century have as yet come to light.

^{page 239 note 1} See supplementary note with plate of this inscription in the appendix to the present paper.

^{page 239 note 2} See this Journal for July, 1897, pp. 572, 573, and Table IV, No. 79.

^{page 240 note 1} Gupta era in Burmā (fifth century a.d.) ; also Buddhist era (from a.d. 1084 downwards), and Śakarāj (Culla Śaka) era at about the same period (from a.d. 1017 downwards). In Siām the Buddhist era occurs at times on purely religious inscriptions, but not before a.d. 1357, when it is but cursorily mentioned in the Thai inscription from Wat Sī-C‘hum at Old Sukhôthai. Its first direct employment is in the Pāli inscription on the model of Buddha's footprint from Sukhôthai (now in the former ‘ Second King's’ temple, Bāngkōk), dated in the year 1970 from Buddha's Nirvāṇa = a.d. 1426. In Kamboja the practice of dating documents, whether epigraphic or otherwise, in the Buddhist or Culla Śaka eras is even more recent, while the Mahā Śaka is still employed in historical literature.

^{page 240 note 2} Asiatic Quarterly Review for 10, 1900, pp. 375–376 and 379–381.Google Scholar

^{page 241 note 1} On these Swankhalôk wares see my articles in the Asiatic Quarterly Review for 04, 1902 (pp. 361–368), and October of the same year (pp. 391–395).Google Scholar

^{page 241 note 2} See on this industry, as well as on the bronze castings of the period, my remarks in the Asiatic Quarterly Review for 10, 1902, pp. 396–397 and 404–405.Google Scholar

^{page 243 note 1} The plate (now preserved with the statue in the royal palace, Bāngkōk) is circular, in the form of a cakra; and the inscription in Pāli is a mere repetition of the famous Yē dhammā stanza, the characters being probably of the third century a.d.

^{page 244 note 1} See “Essays relating to Indo-China,” first series, vol. i, p. 234 and pl. iv.Google Scholar

^{page 245 note 1} As regards the considerable antiquity of the statues, I may mention that two gigantic trees, locally known as Tòn Bë (probably Lagerstroemias), have grown round the images, and so completely enfolded them as to make their disentanglement and removal impossible unless the trees themselves are cut down to the root. It appears that the three statues were brought down to their present site from an old shrine which, according to local tradition, stood on the summit of the hill. No traces of such a structure have, however, so far, been discovered, owing doubtless to the thick jungle that covers the hill and hides the remains from view.

^{page 246 note 1} “Five Years in Siam,” London, 1898, vol. ii, pp. 24–25 and 128.Google Scholar

^{page 247 note 1} See this Journal for July, 1897, pp. 572–573, and Table IV, No. 79. The question has been more fully dealt with in my forthcoming monograph on the Ptolemaic geography of Indo-China, now in the press. The alternative suggestion of the Pāk-chãn inlet is justified from the fact of this estuary lying within the limits of the region of Takũa (Takōla), which includes the three districts now distinguished under the denominations of Takũa-thùng (the southernmost, facing Junkceylon), Takũa-p (the central one), and Takūa-thai (the northernmost, bordering upon the Pāk-chān inlet). Moreover, the rectified Ptolemaic data as regards the position of the mart of Takōla argue a site near the southern point of entrance to the Pāk - chān inlet, close by the present Ranōng, and therefore on Takūa-thai territory. (See the above-cited Table IV.)