NavList:

   

Reply

I had commented, about Jeremy Spencer's paper on Cook's surveying
methods, at the recent Greenwich symposium,

>Well, it isn't so simple to make a map of an unknown coastline as you
>"meander" along it. Land surveyors can measure a baseline and take
>bearings from its ends and triangulate from that. From sea, any
>baseline is somewhat fluid. For a coast that tends North-South, it
>can
>be done by creating a baseline between two sea-positions with spacing
>determined by precise celestial latitude measurement. Where it tends
>East-West, it's much harder, when the surveyor lacks a chronometer,
>as
>Cook did on that first circumnavigation, and had only lunars, and
>dead
>reckoning, to rely on for longitude differences.

and Nicolas de Hilster has responded-
>
"I discussed this bit the second evening in Bar du Mus�e with Jeremy.
As
you all know I am a hydrographic surveyor [NavList 1734] and with my
background I do see an option of using a plane table on board of a
vessel. As the plane table needs a fixed orientation this could be
obtained by using a far direction (mountain, cliff). Any movement of
the
ship would be small in comparison to this far direction. If they added
two visors to the plane table and operated it with two men (one for
orienting the plane table using the visors on this far direction and
one
for drawing the lines using an alidade) it would have been possible to
do a plane table survey.  The survey itself would only take a minute
or
so as the main goal of the plane table would be getting the reference
lines on paper. Filling in the details could be done on sight."

===================

Comment from George.

If anyone is interested to see a copy of those manuscript pencil
charts made by Cook, of parts of the East coast of New Zealand, that
were referred to by Jeremy, they were published by the Hakluyt Society
in 1988, in " Charts and Views of Captain Cook's Voyages, vol 1, ed.
Andrew David. However, Jeremy had the advantage of consulting the
originals in the British Library.

There is meat for some interesting discussion here. Remember, Cook's
Endeavour was a short, fat, round-bottomed, square rigged barque. Not
a good platform for a plane-table survey, except in particularly
smooth conditions. But I accept that a plane table would be possible,
if there were hands given the job of keeping it levelled on the
horizon, and aimed in a constant direction on some landmark, against
the motion of the vessel.

Certainly Cook was familiar with plane-table surveying from on land,
which he had learned in wartime activities in Canada.

A main argument for the plane-table seemed to be the immediacy of the
plotting, with erasures and pasted-in insertions, as though Cook was
making it up as he travelled, not sitting down afterwards with a list
of measured bearings. I accept that the chart looks as though it was
assembled at that very moment of observation, but it seems to me
possible to do that down in the great cabin on the chart table,
plotting-in magnetic bearings of landmarks that were called down from
an observer with azimuth compass on deck, or even from the main-top.
That job, after all, was exactly what an azimuth compass was intended
for, and Cook's officers would surely be familiar with its use.

====================
Nicolas added later-

"It is known that longitude posed a problem as Cook cut some of his
field
sketches in two (in north-south direction) to correct for longitude
errors."

====================

Yes, he pasted in a paper fillet in one place because he needed to
stretch the chart, where his longitudes had been set down wrong.

The problem of longitude shows up particularly in Cook's charting of
the North-pointing Coromandel peninsula (a lovely patch of NZ, that I
recommend to any visitor). His charts show well the East coast and the
West coast of that peninsula, but when you compare with a modern chart
you can see that he got the peninsula far too slim; that is, the East
and West coastlines were too close together. Why should that be? Well,
each coast was linked to its own lunar longitude, and each lunar
longitude, even when measured by Cook, could easily be 30 miles out.
Not only that; his ship's track shows that he had to beat around the
North point at Cape Colvill, and that would have made it very
difficult to link up the East and West coasts of the Coromandel by any
sort of dead reckoning. It's easy to see, then, how 15 miles or so
could have got lost from the width of the Coromandel.

==================

Nicolas concluded-

"As someone after Jeremy's paper suggested it would be a good idea to
check the lunar distance tables Cook used for errors, in order to find
out the influence of them (if there are any) on the survey."

=================

Yes, that was me. There's an interesting paper by Nicholas A Doe,
"Captain Vancouver's Longitudes 1792", in Journal of Navigation, Vol
48, No 3, Sept 95, pages 374 to 388.

This compared Moon ecliptic longitude predictions, from which
Maskelyne's lunar distance predictions were derived, with modern
calculations from the Jet Propulsion Laboratory, for the period late
March to end April 1792. Those discrepancies were all in the same
direction,  by amounts that varied between about 3 arc-seconds and 50
arc-seconds in a way that was roughly cyclic with a period of about
one month. Any navigator using those predictions, anywhere in the
World, in that month would have found that those errors would
inevitably cause his deduced positions to be set too far to the East,
by between 2 and 24 minutes of longitude. Of course, there would be
additional variations due to observing errors, on top of those almanac
errors.

That analysis was done for a period that was two decades later than
Cook's voyages, but I imagine that in Cook's day the Almanac errors
were no less, and it would be valuable if someone were to analyse
those errors, using modern data, covering a longer period than Doe has
done. They would have affected, not just Cook and Vancouver, but any
navigator using the lunar distance tables in the Nautical Almanac, or
any other Almanac pirated from it.

I have advocated such a systematic analysis more than once on this
list, and perhaps old hands have heard more about that hobby-horse
than they wish.

George.

contact George Huxtable at george@huxtable.u-net.com
or at +44 1865 820222 (from UK, 01865 820222)
or at 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to NavList@fer3.com

To , send email to NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---

   

Reply