to incorporate the intermediate parts, it was necessary not only to carry
his scale of longitude 20½' more west, but also to reduce the extent of
the coast. The chart was compared with the narrative and chart in
Hawkesworth, and the log book of the Endeavour with them all; when it was
found that reductions might be made in various places upon one or more of
the above authorities, for differences between them were frequent and
sometimes considerable, and in one instance alone a reduction of 12' in
the chart was obtained. It is said in Hawkesworth (III, 202), "As soon as
we got within side the reef (through Providential Channel) we anchored in
nineteen fathom;" and afterwards (p. 204), that the channel, "bore E. N.
E. distant ten or twelve miles." In the first chart the distance is 14½
miles, and nearly the same in that which accompanies the narrative; but
in the log book it is said to be 2½ miles only, which corresponds with
having anchored as soon as they got within the reef, and has been
adopted. In some cases it was not easy to make a choice between these
different authorities; but I have commonly followed the narrative and log
book when they were found to specify with precision, and they generally
produced such corrections to the chart as brought the longitudes of
places nearer to my positions. Captain Cook's track in Plates XI. XII.
and XIII. is laid down afresh from the log book; and many soundings, with
some other useful particulars not to be found in the original chart, are
introduced, for the benefit of any navigator who may follow the same
route.

The reconstruction of the charts in the Atlas was done upon various
scales, but that no error might escape unseen, the least was of _ten_
inches to a degree of longitude; they were then reduced by Mr. Thomas
Arrowsmith to four inches, this being thought sufficiently large for a
general sailing scale; and each reduced sheet was scrupulously compared
by me with the original before it went into the engravers hands, and the