261 
of Edinhurgli, Session 1883-81. 
attention of the editor was given (very properly in such a work) to 
the Scotch topography, and the far more valuable part of the map 
— as a map — was treated as an annex of little importance to the 
subject in hand, attempt to transcribe or examine the mile 
distances was made either by Innes or by Gough. 
When I saw this map at Oxford, and transcribed as well as I 
could in an hour or two the lists of distances, it seemed very evi- 
dent that all the distances in the south and east of England had 
been rewritten by some hand well accustomed to mediaeval script ; 
and this accounts for the ends of many words going off into a mere 
series of strokes, since the re writer was not certain of part of the 
name, and just inked over what he could see. This rewriting 
deserves careful study in considering the map ; and as it was ap- 
parently done before the sixteenth century, and the old writing 
must have been fading then, and yet is not illegible in parts even 
now, it is some evidence as to the great age of the map. What is 
now much needed is a critical examination of this map, identifying 
all the places by comparison with modern maps, and recovering 
any traces of the first writing of names and distances where it has 
become all but illegible. This work needs a good palaeographer ; 
but for the question of the miles, which is what we have at present 
to consider, such care is not required ; for, if a few doubtful dis- 
tances are omitted, it will not perceptibly affect our results. From 
a comparison of the readings of Basine, Gough, Innes, and my own 
notes, the distances may be pretty safely settled in all legible cases ; 
and after omitting those in Wales and Cheshire, which show a 
different mile, there are 130 distances available for examination. 
The mean value of the old mile from these is 1265 ± 9. 
Having now described the various data available for fixing the 
old mile, we will place the results all together and compare them, in 
terms of thousandths of the statute mile ; stating the mean result 
from each source, the probable error ( ± , i.e., what amount of varia- 
tion the truth is as likely to exceed as to lie Vv^ithin), and the 
average error of a single length, which shows the relative accuracy 
of the different sources. 
for Leeming, and the placing of Abergavenny on the west coast of AVales for a 
town which must be Aberdovy. 
* The map is framed and glazed, and screwed high up against a pillar in a 
poor light ; hence it is not easy to study. 
VOL. XII. 
s 
