on a Earn in Kent, &c. 121

gazine [/] of a curious ancient barn at Calcot in the parifh of Newington Bagpath in the county of Gloucefter, that has on its porch an infcription noticing its being built in the year 1300. In an end wall of this barn there is a window that plainly exhibits a coetaneous mark, there being fo many windows of that pattern {till difcoverable in churches of the fame age. A fearch will be fruitlefs in buildings of an earlier age for a window framed after a model refembling that of the Prefton Hall barn. The characters on the Calcot barn, both letters and numerals, are alfo coeval ; not fo 3, 3, in the Helmdon infcription, with the form of that figure in the twelfth century, and an unequivocal infcription carved on ilone in Arabic figures previous to the fifteenth century, I fufpecl:, may be ftill a dcfideratum. Mr. North and others have proved that many of the fpecimens produced will not Hand the teft ; and even Profeflbr Ward allows that fome which were confidered as nume- rals were really letters fpecifying initials or abbreviations of names. Mr. North, in his letter to the earl of Morton [] already cited, ventured to foretell that a different account from that fuggefted in the Philofophical Tranfactions would hereafter appear, if the original piece of wood remained at Helmdon Parfonage undefaced. Though not apprized of this circumftance, I have attempted to fulfil in part this prediction on a clofer view of the delineation and defcription given by Dr.Wallis, who had feen the mantle-tree ; and mould it have been preferved a century beyond his time, it is to be wifhed that it may be re-examined. But if the reafons adduced will warrant my reading of the date, the principal bafis on which the hypothefis of a very priftine introduction of Arabic numerals is built muft be fo much weakened, that it will be extremely difficult to find any fubflantial proofs to uphold it ; none afTuredly from [*] For the Year 1795, May. PI. III. , [] Archaeologia, V. X. p. 371. VOL. XIII. R the