Each file contains a sequence of pairs of numbers being latitude (negative meaning south)
and longitude (east) in degrees and decimal fractions (not minutes and seconds), and so
should be readable without difficulty. I used to have a file in which these segments were
identified and combined into one file, but it has been lost...
   The values came from an odd process. An actual outline map of New Zealand was attached to
a drum scanner and the positions of the black lines on each east-weat scan line were determined.
These positions were collated into line sequences corresponding for example into the east and west
coast of the North Island, etc. Subsequently, I combined some sequences (such as the east and west coast)
into one sequence.
   The drum scan positions (in odd units) have been transformed by an un-projection calculation
(involving arctangents and logarithms as I recall) that was supplied (and also lost) into latitude
and longitude positions.
   Unfortunately, plotting of sample positions on an actual map showed that there appeared to be
systematic errors in positioning with a maximum in the south of the South Island), amounting to
a mile or so. A process for correcting this has been considered, but has run into the sand.
I have however corrected some obvious hiccoughs, such as islands not having a closed loop,
or badly misplaced points. Such corrections have conformed to whatever appeared to be the
local offset.

                                       Good luck,
                                       Nicky McLean.