Page 6 - gt1607

GSA TODAY | JULY 2016 England and Wales), which gave the best fit on the Smith County
maps. One additional piece of indirect evidence is that projection
tests on the small-scale General Index to Cary’s 1794 New Map of
England and Wales, which does contain internal coordinate lines,
give RMS errors of 1448 m (4750 ft) for Cassini against 1637 m
(5370 ft) for Bonne. The difference in error is too small for any
definitive statement to be made as to which projection was used,
but on balance it is more likely that a Cassini projection (on a
sphere) was used for both the Smith 1815 and County maps.
TRANSFORMATIONS
Having decided that Cassini was overall the best-fit projection, 410 subscribers for 414 copies of his 1815 map. Not all subscribers
the maps were transformed to that projection. Graticules on the
1815 map provided some limited information, which was supple-
mented by John Cary’s locations of towns and villages. For the
county maps, some 1,800 graticule longitude and latitude points
were transformed using town and village locations and RMS
errors obtained. In general, transformation errors were lower in
southeast England and tended to increase to the north and west.
THE 1ST PRINCIPAL TRIANGULATION
The 1st Principal Triangulation of the UK and Ireland was
conducted over six decades and was a remarkable achievement
that significantly improved the accuracy of nineteenth-century
maps of the country. The Principal Triangulation commenced in
1783 under the direction of General William Roy with the objec-
tive of connecting the observatories of Greenwich and Paris and
determining the difference in longitude between them. After
Roy’s death in 1790, the survey was extended across the UK and
Ireland under the direction of Colonel Williams, General Mudge,
General Colby, Colonel Hall, and finally Colonel Henry James.
Apart from improving map accuracy, the triangulation allowed
Alexander Clarke to compute a spheroid for Earth and make
density calculations. The first stage in the triangulation was the
measurement of an accurate baseline between Kings Arbour (now
the site of the Heathrow Airport car park) and the Poor House at
Hampton. This was first achieved by iron bars and deal rods and
then remeasured using glass tubes. Corrections were made for
temperature and humidity and the line accurately leveled; the
final accuracy is estimated at an amazing one inch in 27,400 ft.
After completion of the line, a number of triangulations were
made in a southeasterly direction toward the Kent coast. The
triangulations were made using a specially commissioned theo­
dolite constructed by Jesse Ramsden. As the triangulation
proceeded, it become obvious that there were large inaccuracies in
existing maps of the country. In 1799, Mudge and Dalby made a
number of severe criticisms of the “erroneous state of our maps.”
Cary’s maps were not specifically mentioned by Mudge and Dalby
and while at the local level Cary’s maps are accurate, errors
become amplified in country-wide maps.
EVOLUTION OF SMITH’S GEOLOGICAL MAPPING
Smith’s 1815 map exists in at least six main variants in five
series as recognized by Eyles and Eyles (1938), based on changes to
the geology, coloring, and topographic detail. These include an
early unnumbered issue (Series I); an early issue second series
numbered 1–100 (Series II); another, mostly second issue third
series, numbered a1 to a100 (Series III); a third issue, numbered
6
b1–b75 or possibly to b100 (Series IV); and a late unnumbered
series (Series V), which is divided into Va maps, similar to late
Series IV and Vb, possibly part of a second 1830s series (Sharpe,
2007, and personal commun.). The early and late unnumbered
maps are not signed by Smith, while the numbered issues are
usually signed. The map is dated 1st August 1815; however, two
early Series I unnumbered maps were issued before this date, with
a further 20 in September–October 1815. However, most maps
were not issued until after 2 November 1815, when Smith began
signing and numbering his maps. Countermarks on the very late
Z map (which hangs in The Geological Society of London’s
Burlington House) date it as 1836 or later. Smith’s memoir lists
took their copies, and some maps went to non-subscribers; the
total production run was probably around 350, of which 120–130
may still survive (Tom Sharpe, 2014, personal commun.).
In order to demonstrate the evolution of Smith’s geological
mapping, extracts from various maps around the city of Oxford
have been produced from the GIS and are shown in Figure 2. All
three Smith maps use the same basic scheme of colors and tinting
technique, although there are some significant changes in geolog-
ical content. All the maps show a succession of strata, oldest
(Great Oolite) in the northwest and getting progressively younger
to the southeast (Chalk). The Smith A map at the top of the panel
is a very early unsigned copy and was probably the Geological
Society subscription copy. Tinting on the A map tends to be light,
almost tentative, often grading to no color. Smith himself says, “In
many cases the edge of one stratum terminating on another is so
gradual, as not to admit of any distinction or definite line”
(Smith, 1815, p. 8). Where there is no color, the formation is
assumed to be intermediate between the strata above and below.
However, on early issue maps (like the A map) the Greensand is
purposely shown colorless. Also, in early maps the Coral Rag is
not shown as a separate formation, only sandstone between the
Clunch Clay and the Oaktree Clay. The later issue P map (Series
Va, possibly 1816, Tom Sharpe, 2014, personal commun.) is more
fully and confidently colored. On this map, the Coral Rag appears
and the Oaktree Clay is extended to the base of the Chalk. Smith’s
Oaktree Clay is somewhat problematic; on the 1815 map legend,
he shows “Blue Marl or Oak Tree Soil” beneath the Greensand and
above the Purbeck stone. However, on his later cross sections,
geological tables, and the county maps, the Oaktree Clay is clearly
below the Portland stone and above the Clunch/Coral Rag,
making it the equivalent of the modern Kimmeridge Clay. The
changes in geological coloring are accompanied by new engraved
geological lines. The third map is a composite of parts of Smith’s
geological maps of Oxfordshire and Berkshire published between
1819 and 1820. The topographic and cultural information shown
is more detailed because the county sheets are at a larger scale; the
geological boundaries are also more detailed. For general compar-
ison, the last map illustrates a simplified version of the modern
geology published by the British Geological Survey.
GEOLOGICAL SIGNIFICANCE OF SMITH’S MAPS
William Smith did not make the first geological map. Jean-
Étienne Guettard (1715–1786) and Philippe Buache (1700–1773)
published a map in 1746 containing a large amount of regional
lithological information that significantly shows an attempt at
the distribution of the chalk in France and England. Later in the