ji 
i 
47 
The 58 parchment rolls of hundreds with index are supposed 
to be in the hands of the transcribers. Internal evidence, f. 82 
and elsewhere, shows that Reginald (probably de valle torta) took 
the oversight, and himself copied. He was assisted by a Norman 
clerk, and by at least two Saxon clerks. The Saxon clerks are 
easily known, as Sir William Ellis has shown, by their symbol for 
“et.” In the Analysis of the Somerset part of the Book, the 
Norman scribes are noted by 4, the Saxon by 2. 
Reginald provided himself with 47 books of parchment of 
uniform size, corresponding to those of the Geld Roll, but varying 
in number of sheets. Each of these is to be filled with the 
holding in the successive hundreds of the King, a baron, ora 
thegn respectively, following a definite order, that of the index, 
with the readjustment of making Sutpetret No. g follow Nortpetret 
No. 1, and making Chori Mallet follow Chori Rivell. A Saxon 
scribe takes the first roll, Nortpetret, and enters the King’s manor, 
and two others (see No. 28 Geld Roll), and then passes it on to a 
Norman scribe. He next takes Sutpetret (No. 31 Geld Roll) 
and copies throughout by himself, except one entry (490b). 
Seuenametona. Very likely he was then entering No. 2 Vice- 
comitis Brunetona (15 Geld Roll) containing only 2 manors. 
Next comes No. 3 Chori Rivell (38 Geld Roll), all entries except 
three in Saxon writing. Afterwards, we think, was entered Chori 
Mallet No. 22 (32 Geld Roll), From this stage the index was 
followed without interruption. Thus three or four scribes were 
entering simultaneously from as many hundred rolls, and their 
work crossed, and so ina measure interrupted the index order 
in each fief. 
To test the above theory, take the analysis of the Domesday 
Exon Book, examine the order of hundreds in each fief, do the 
same in terrae occupate, and no doubt will remain that entries 
are made in the order of the rectified index. 
One objection made against the Exon Book is the number of 
corrections and marginal entries, which indeed shows that the 
