Stapleton, of course, knew better than this, and assigned the survey at one time to _circ._ 1108,[3] but in his _Rotuli Scaccarii Normanniæ_[4] to 1106-20. It was subsequently investigated and analysed with great care by Mr Eyton, whose note-books, now in the British Museum, show that he adopted the sound method of comparing it in detail with Domesday Book. After his death Mr Chester Waters issued (1883) an annotated translation of the text, with an introduction, analysis, etc., in which the place-names were carefully identified, and the same system of comparison with Domesday adopted.[5] It is, unfortunately, necessary to explain that Mr Waters in the table of contents described his translation as 'from the Cotton MS., Claudius C. 5', and wrote on the opposite page: This MS. engaged the attention of Thomas Hearne, the antiquary, who has printed it amongst the additaments to his edition of the _Liber Niger Scaccarii_; but Hearne was one of those industrious but uncritical antiquaries who had no conception of the duties of an editor of the importance of accuracy. Knowing the high opinion entertained of Mr Waters' works,[6] I accepted his translation in all good faith as 'from the Cotton MS.' and was, I confess, not a little startled to discover from Mr Greenstreet's facsimiles that it was made not from the Cotton MS., but from that inaccurate edition by Hearne, which Mr Waters had mentioned only to denounce. On fo. 4_b_ a whole line, containing three entries, was accidentally omitted by Hearne, and is, consequently, absent also from Mr Waters' version. On collating the two, however, I found, to my great surprise, that matters were even worse than this, and that Hearne's text was far less inaccurate than Mr Waters' own, the erroneous figures found in the latter being almost always correctly given by the 'uncritical' Hearne. As for the version given by Mr Waters, even in the very first Wapentake, there are three serious errors, five carucates being given as three, nine as seven, and eleven as two! And for Bradley Wapentake (p. 27), his figures are so erroneous that, according to him, 'Radulf Meschin alone had 42 cars. 6 bovs. in this Wapentake', though his real holding was only fifteen cars. three bovs. With another class of resultant errors I shall have to deal below. To the enterprise of Mr Greenstreet scholars were indebted for an _édition de luxe_ of the record in facsimile, which made its appearance shortly after the treatise of Mr Waters. Unfortunately, no attempt was made in the appended literal translation to identify the names of places or persons, while such a word as '[ap]pendiciis', which occasionally appears in the survey, is mistaken for a place-name 'Pendicus'. The book enjoys, however, the great advantage of an index.