^{page 145 note 1} Thousands of loose brick tesserae scattered about, chiefly to the south and east, show that at least one other room was tessellated, but that ploughs had reached it.

^{page 145 note 2} See above, p. 144.

^{page 149 note 1} The 2nd and 3rd Houses presumably both come under the designation nubilaria.

^{page 149 note 2} Further details of this clay water-tank are as follows: I ft. from north wall of house; 1 ft. 10 in. from east wall; west side 2 ft. thick; south end probably 1 ft. thick; bottom 6 in. thick, 1 ft. 5 in. deep, or 3 ft. 11 in. from existing top of wall.

^{page 151 note 1} I am indebted to Messrs. Stephenson and Hayter for a reference to a furnace of this type at Silchester, block 3, Insula xxxiii (Arch., lix, 336, fig. 1). which was somewhat larger than the present example, except the width of the top portion, which was less. Other examples are at Caerwent,

^{page 155 note 1} I took this to be the impression of a tree trunk which had formed the bottom frame of a wattle-and-daub wall, dividing the house into two compartments.

^{page 156 note 1} Beginning rather before A. D. 50, and ending about 410. Three occasions or periods of disturbance may be faintly hinted at by this site: first, a gap in the series of coins, between the death of Septimius Severus, A. D. 211, and early in the reign of Gallienus, A. D. 253. This includes the probable date of burial of five persons in Pit 6 (p. 149) who, however, may have been merely victims of an epidemic, and not necessarily of any violence. Secondly, the apparent throwing away into Pit 3 of unbroken, expensive, and beautiful sigillata bowls of Drag, forms 43, 45 (two), 31, etc., at about the end of the third century. This may have taken place before the end of the gap in the coins. Thirdly, the hiding in the 2nd House of 294 coins between the years 317-26.

^{page 159 note 1} Homer, , Iliad, xiii, 588; xx, 495; xxi, 77Google Scholar; Cato, , Scriptores De re rustica, 91, 139Google Scholar; Varro, , De rustica, i, 13, 51, 52Google Scholar; Columella, , De re rustica, i, 6; ii, 9, 19Google Scholar; Pliny, , Historia naturalis, xiii, 29, 30, etc.Google ScholarVirgil, , Georgics i, 178Google Scholar; Palladius, , De re rustica, i, 36; viii, 1Google Scholar.

^{page 159 note 1} See p. 150.

^{page 159 note 1} See p. 150.

^{page 159 note 1} I think Rolleston has given the fullest analysis of the Romano-British type of skull. See Scientific Papers and Addresses of George Rolleston, M.D., F.R S., edited by Win, , Turner, and Tylor, Edward, 2 vols., 1884.Google Scholar

^{page 162 note 1} Estimated by Pearson-Lee formula.

^{page 162 note 2} Measured between lateral borders of second molar teeth.

^{page 162 note 3} Measured from point between middle incisor, from crown, and a point midway between hinder borders of third molar teeth.

^{page 165 note 1} V. C. H. Bucks., i, 146.Google Scholar

^{page 165 note 2} B. 0. U. List, 2nd ed., 1915, p. 307.Google Scholar

^{page 175 note 1} W. E. Cobbold, C.E., was the actual discoverer.—A. H. C.

^{page 185 note 1} Abbreviations: L = Lezoux; B = Banassac; La Grau. = La Graufesenque; R = Rheinzabern.

^{page 185 note 2} Mr. Atkinson, D. in Cat. Rom. Pottery in Tidlie House, Carlisle, p. 65,Google Scholar note, states that some authorities hold that this date for the Pudding-pan Rock pottery is too late, and that its period is rather 150 or even 140-90.

^{page 1 note 1} Abbreviations: L = Lezoux; B = Banassac; La Grau. = La Graufesenque; R = Rheinzabern.

^{page 187 note 1} Abbreviations: L = Lezoux; B = Banassac; La Grau. = La Graufesenque: R = Rheinzabern,

^{page 197 note 1} If read the reverse way up, these marks would pass for the early Runes equivalent to UL ø [or Y].