^{page 127 note 1} About 10 miles S.S.W. from Allahabad, on the south bank of the Jumnā.

^{page 127 note 2} General Cunningham (A.S.R., iii, 47) held that Bhīṭā represented the ancient Bitbhaya-paṭṭaṇā, but his arguments are not convincing. On the other hand, the place-name Vichhī occurs on a seal-die which I found at Bhīṭā, and the longer form Vichhīgrāma on a sealing. The former, at any rate, is unlikely to have been brought from elsewhere, and it probably gives us, therefore, the name of the ancient town.

^{page 128 note 1} I have attached this name to the house for the sake of distinguishing it, though the seal is probably older than the house. In the case of the other houses there is every reason to suppose that the names given to them were the names of their actual occupiers.

^{page 128 note 2} Pl. III, 2; the reading is doubtful. The lettering on this seal is raised, not countersunk. A full account of the many seals found on this site is being prepared by my assistant, Pandit Daya Ram, and edited by Dr. Vogel. To their notes I am indebted for the transcripts of the seals mentioned here.

^{page 129 note 1} The bricks of the chambers on the south - east side average 18½″ × 9½″ × 2½″, while in the rest of the house they average 21″ × 10½″ × 2½″. The two sizes were, no doubt, used at one and the same time for purposes of economy.

^{page 131 note 1} Some of these beads are unique, being composed of glass laid on in thin layers, with gold leaf between. Among other pieces of jewellery found on this site were: the gold medallion of Gupta date figured in Plate III, 3; a crystal pendant with face in relief, of the same age; a gold serpentine finger ring with vajra device, of the Mauryan epoch; and a beaded earring.

^{page 132 note 1} Lying against these steps was found a coping-stone of a railing with a line of writing in Brāhmī characters of about the second century b.c., to the following effect: na(?) Sēliyā-putreṇa Gōmitrēn(ṇ) a kāritā bhagavatō Nāgasa . . .

^{page 134 note 1} As to the character of the foundations themselves, they are very similar in buildings of this date to the earlier foundations of the Mauryan epoch, the chief points of difference being as follows: (1) In the earlier foundations, where lcankar is used, it is laid with broken pottery or brick in alternating courses, while in the later the kankar is mixed indiscriminately with broken brick; (2) broken potsherds are almost entirely absent in the later; (3) in the earlier the walls are generally carried deeper under ground than in the later; (4) heavy stones are used to protect the corners in both periods, but in the later they project above the ground-level, whereas in the earlier they are completely buried.

^{page 135 note 1} The engines of war used for propelling these balls must have been of considerable power, the balls, which are of stone, measuring as much as 8½ inches in diameter. A large number of them were found on the site, the finest, though not the largest, specimens being those of the Mauryan period, which are perfectly spherical and finished with that wonderful precision which characterizes all the stone-work of the Mauryan period.

^{page 136 note 1} The name of a Śiva-liṅga.

^{page 137 note 1} Examples of such conservatism are to be found among the ancient Egyptians and the Mexicans. The Jews, too, it will be remembered, continued to use stone knives for circumcision in a metal age (Ex. iv, 25 and Josh, v, 2), while the Romans used them for sacrifice; whence the proverb inter sacrum saxumque stare. But I do not know that the true neolithic types of implements were preserved in any of these cases.

^{page 137 note 2} In some other houses they are over 20 feet deep.

^{page 139 note 1} This sealing has the most perfectly cut legend of any yet found in India.

^{page 141 note 1} Arrian, it will be remembered, states (Ch. X) that cities situated on the banks of rivers or on the sea-coast were built of wood, while those in higher and dryer situations were built of brick and mud.

^{page 141 note 2} In the absence of Dr. Spooner, who is on leave in the South Sea Islands, I am indebted to his Assistant, Wasi-ud-din, for the information contained in this note.

^{page 146 note 1} JRAS., October, 1909, pp. 1080–1.

^{page 147 note 1} In another note on this find Mr. Cousens wrote that these coins were ⅜″ square and ⅛″ thick, but that, owing to corrosion, it had not been possible to identify them.

^{page 151 note 1} Cf. Cunningham, , A.S.R., vol. xx, pp. 48 f.Google Scholar, pi. v, No. 4. The designation “well-inscription” is misleading, as the slab does not appear originally to have belonged to a well.

^{page 155 note 1} [These tablets are of a type frequently met with on Buddhist sites in India, and there can be little doubt, I think, that they were stamped with dies brought from India. They belong to the tenth and eleventh centuries a.d.—J. H. M.]

^{page 156 note 1} [The temple in which these inscriptions occur was found by Mr. Bhandarkar in the jahāgīr of Khaṇḍelā, in the midst of a thick jungle. Only the sabhāmaṇḍapa and parts of the outside shrine wall are now preserved of the old structure. Mr. Bhandarkar states that the pillars of the hall are “deeply and elegantly carved in the pot and foliage style, and, though perhaps not earlier than the ninth, are certainly not later than the tenth century A.D.”—J. H. M.]

^{page 157 note 1} This inscription resembles closely the Haidarābād plate of the same king in the historical portion. The latter is, however, not dated.