Next message: George Huxtable: "Re: Sumner and the Smalls lighthouse."
A few years back, Jan Kalivoda wrote a couple of posts to this list about
Thomson's Tables for clearing lunar distances (which were adopted as Bowditch's
Second Method in 1837). He noted that the calculation of the "third
correction" table was considered mysterious in the 19th century. For anyone who read
this account back then, I just wanted to note here that the table is not at
all mysterious, and it can be calculated directly. It's a lot of work because
there are thousands of entries, but the steps involved are simple, and the
majority of cases had already been tabulated before Thomson's time. Most similar
works tabulated the linear refraction plus the Moon's quadratic term.
Thomson adds in the quadratic cross-term. This additional calculation rarely
changes the result by even a tenth of a minute of arc (equivalent to three minutes
of longitude in the result) except when the lunar distance is less than 30
degrees and even then only when the Moon's altitude is rather low [Jan
Kalivoda's earlier post noted a difference of a full minute of arc however this was
only correct for methods which ignored the quadratic corrections entirely]. To
a navigator, this was simply a number to be extracted, never mind the
details, and it was a very popular method, involving about 30% less work than other
similar methods.
By the way, I believe it was Baron von Zach who started the urban legend
that they're was something extraordinary in the calculation of Thomson's table,
although Thomson himself may have had a hand in it. There's a paper about the
tables by the Baron briefly described in the Monthly Notices of the Royal
Astronomical Society in 1829 which can be found on the web via adsabs:
http://adsabs.harvard.edu/
(if you've never used this service, when you do a search and it says 'zero
records found', give it about thirty seconds. It's working on it...)
-FER
42.0N 87.7W, or 41.4N 72.1W.
www.HistoricalAtlas.com/lunars
