They are:—for Belvoir and other houses of the Earls of Rutland, Rutland MSS. (Hist. MSS.), iv. 260; for the house of Richard Bertie and his wife the Duchess of Suffolk at Grimsthorpe, Ancaster MSS. (Hist. MSS.), 459; for Wollaton, the house of Francis Willoughby, Middleton MSS. (Hist. MSS.), 446; for Maldon and Saffron Walden in Essex, A. Clark's extracts in 10 Notes and Queries, vii. 181, 342, 422; viii. 43; xii. 41; for Newcastle-on-Tyne, G. B. Richardson, Reprints of Rare Tracts, vol. iii, and 10 N. Q. xii. 222; for Reading, Hist. MSS. xi. 177; for Oxford, F. S. Boas in Fortnightly Review (Aug. 1913; Aug. 1918; May 1920); for Stratford, J. O. Halliwell, Stratford-upon-Avon in the Time of the Shakespeares, illustrated by Extracts from the Council-Books (1864); for Weymouth, H. J. Moule, Weymouth and Melcombe Regis Documents (1883), 136; for Dunwich, Various Collections (Hist. MSS.), vii. 82; for Aldeburgh, Suffolk, C. C. Stopes, William Hunnis, 314. References for a few other scattered items are in the foot-notes. The warning should be given that the dates assigned to some of the provincial performances are approximate, and may be in error within a year or so either way. For this there are more reasons than one. The zealous antiquaries who have made extracts from local records have not realized that precise dates might be of value, and have often named a year without indicating whether it represents the calendar year (Circumcision style) or the calendar year (Annunciation style) in which a performance fell, or the calendar year in which a regnal, mayoral, or accounting year, in which the performance fell, began or ended. When they are clearly dealing with accounting years, they do not always indicate whether these ended at Michaelmas or at some other date. They sometimes give only the year of a performance, when they might have given, precisely or approximately, the month and day of the month as well. But it is fair to add that the accounts of City Chamberlains and similar officers, from which the notices of plays are generally derived, are not always so kept as to render precise dating feasible. Some accountants specify the days, others the weeks to which their entries relate; others put their entries in chronological order and date some of them, so that it is possible to fix the dates of the rest within limits; others again render accounts analysed under heads, grouping all payments to players perhaps under a head of 'Gifts and Rewards', and in such cases you cannot be sure that the companies are even entered in the order of their visits, and if months and days are not specified, cannot learn more than the year to which a visit belongs. Where, for whatever reason, I can only assign a performance to its accounting year, I generally give it under the calendar year in which the account ends. This, in the case of a London company and of a Michaelmas year (much the commonest year for municipal accounts), is pretty safe, as the touring season was roughly July to September. Some accounting years (Coventry, Marlborough, Stratford-on-Avon) end later still, but if, as at Bath, the year ends about Midsummer, it is often quite a toss-up to which of two years an entry belongs. In the case of Leicester performances before 1603, I have combined the indications of Michaelmas years in M. Bateson, Leicester Records, vol. iii, with those of calendar years in W. Kelly, Notices Illustrative of the Drama (1865), 185, and distinguished between performances before and after Michaelmas. I hope Kelly has not misled me, and that he found evidence in the entries for his dating. After 1603 he is the only source. I do not think that the amount of error which has crept into the following chapter from the various causes described is likely to be at all considerable. I have been as careful as possible and most of Murray's own extracting is excellently done. I should, however, add that the Ipswich dates, as given both here and by Murray, ii. 287.
