As to the honesty of his associates--before the death of Henry VIII. (January 28, 1547), the Castilians had promised him not to surrender the place without his consent, and to put Arran's son in his hands, promises which they also made, on Henry's death, to the English Government; in February they repeated these promises, quite incompatible with their vow to surrender if absolved. Knox represents them as merely promising to Henry that they would return Arran's son, and support the plan of marrying Mary Stuart to Prince Edward of Wales! {26a} In March 1547, English ships gathered at Holy Island, to relieve the castle. Not on June 21, 1547, as Knox alleges, but before April 2, the papal absolution for the murderers arrived. They mocked at it; and the spy who reports the facts is told that they "would rather have a boll of wheat than all the Pope's remissions." {26b} Whatever the terms of the papal remission, they had already, before it arrived, bound themselves to England not to accept it save with English concurrence; and England, then preparing to invade Scotland, could not possibly concur. Such was the honesty of Knox's party, and we already see how far his "History" deserves to be accepted as historical. Next, what is most surprising, Knox's account of the month of ineffectual siege by the French, while he was actually in the castle, rests on a strange error of his memory. The contemporary diary, Diurnal of Occurrences dates the _sending_ (the arrival must be meant) of the French galleys, not on June 29, as Knox dates their arrival, but on July 24. Professor Hume Brown says that the Diurnal gives the date as _June_ 24 (a slip of the pen), "but Knox had surely the best opportunity of knowing both facts" {27a}--that is, the number of the galleys, and the date of their coming. Despite his unrivalled opportunities of knowledge, Knox did not know. It is not quite correct to say that "Knox in his 'History' shows throughout a conscientious regard to accuracy of statement." Whatever the number of the galleys (Knox says twenty-one; the Diurnal says sixteen), on July 13-14, they are reported by Lord Eure, at Berwick, as passing or having just passed Eyemouth. {27b} They did not therefore suffer for three weeks at the garrison's hands, or for three weeks desert the siege, but probably reached the scene of action before the date in the Diurnal (July 24), as, on July 23, the French Ambassador in England heard that they were investing the castle. {27c} Allowing five or six days for transmission of news, they probably began the attack from the sea about July 16 or 17, not, as Knox says, on June 30. Perhaps he is right in saying that the French galleys only fired for two days and retreated, rather battered, to Dundee. Land forces next attacked the hold, which surrendered on July 29 (as was known in London on August 5), that is, on the first day that the _land_ battery was erected.