Whilst this work of demolishing was being carried out, an interesting discovery was made, Roman tiles and mortar being found, worked up into the materials of which these walls were built. At the south-east corner, and adjoining the remains of the Wardrobe Tower, a portion of Roman wall was disclosed, having three courses of bonded tiles showing above the surface of the _débris_. This piece of wall is in a direct westerly line with the old city wall, shown in a plan of the Tower made in 1597, the demolished buildings likewise appearing on this plan, which can be seen in the office of the Commissioners of Works. Two inferences are possible from the discovery of this Roman work; either it is part of the old city wall or the remains of a Roman building, and if it is satisfactorily proved to be Roman, it will practically settle the contested point as to whether there was ever a Roman fortress on the site of the White Tower or not. Holinshed, in the third Book of his history of England, quoting both Leyland and Fabyan, says, that Belins, who began to reign conjointly with Brennus as King of Britain, which was “about the seventh year of Artaxerxes, the seventh king of the Persians, builded a haven with a gate within the city of Troinovant, now called London. This gate was long after called Belins gate, and at length, by corruption of language, Billingsgate. He builded also a castle westward from this gate (as some have written) which was long time likewise called Belins Castell, and is the same which we now call the Tower of London.” It was pointed out in the first volume of this work that Fitzstephen declared the White Tower to have been built by Julius Cæsar, and that the mortar used in the building was “tempered with the blood of beasts,” but the Roman habit of mixing powdered tiles with their mortar, may have given rise to this theory. Stowe, in his survey of London about 1076, says, that William the Conqueror caused the present White Tower to be erected at the south-east angle of the city wall, which would be the actual spot where the fragment of the recently discovered Roman wall now stands. On removing the southern wall of this building, it was found that it was built up to, and not bonded into the south-east turret of the White Tower, which forms the apse of St John’s Chapel. When it was taken down, the original stone-work of the White Tower was laid bare. It is quite honeycombed by age, Sir Christopher Wren having, of course, been unable to reface it as he did the exposed portions of the Tower.