General Instructions for Surveying', possibly written byMajor James Rennell, ca February 1791 (Series 60.05) - No. 0023

[Page 23]

Hummock on any part of it, it should be preferred to any part of the line of the Coast and the Views of the Land will explain the position effectually.
So much discrimination  being expected a Scale sufficiently wide must be adopted for the particular Charts; of which you will prepare a regular series from the commencement to the Conclusion of your Survey, in numerical order, each Chart containing a proper Portion of Matter for Plain Chart Projection: that is 3 to 4 Degrees of Latitude & the Scale of these Charts we direct to be five Nautic Miles to an Inch.
This is to be accompanied by a General Chart on a scale of one Inch to a degree of Latitude; & the Projection such, as that the degrees of Longitude may diminish nearly in the same Proportion as on the Globe. It is to be constructed from the Observations of Latitude & Longitude, which are to furnish the primary Points: & the Matter of the Plain Charts, is to be reduced into the Intervals afterwards.
It appears from the result of former Operations, that where the differences of Longitude between two observations by Time Keepers, are very small, such particular Observations are sometimes found to differ from the Longitude deduced from the Bearings of the Coast; altho the Agregate of several Observations (or rather one considerable portion of difference of Longitude may agree with the Chain of Triangles) as well as some of the intermediate Observations & Triangles. This may be owing in part to the minute Accuracy required in taking the Time of which a few seconds only are known to be often equivalent to a Nautic Mile, & partly to the uncertainty of the distance between the