NavList:

   

Reply

Hello, Jan, you wrote:
"But I saw his early  edition in Paris two years ago, where his first method
of lunars stood only in  the appendix and therefore Thomson's method (later
the second) held the first  position in the main text, if I remember it well.
Not very important, of  course."

No, not important, of course. For what it's worth, here are some  details:

1795: Bowditch develops a modification of the standard series  method for
clearing lunars. I call this "Bowditch's Original Method". Word  spreads rapidly
in New England that there is a mathematical "genius" in Salem.  Someone arr
anges an honorary Harvard degree for him to give the self-educated  expert a
little paper respectability.
1799/1800: Bowditch and Blunt insert  Bowditch's new method in their
Americanized edition of Moore's Practical  Navigator, replacing a variant of Lyons's
method. The second method included is  the same one Moore included: Witchell's.
1802: After a more extensive  revision, Blunt and Bowditch retitle the book:
the New American Practical  Navigator. The two lunars methods are Bowditch's
Original Method and Witchell's  Method. There is a small note in the preface,
in this edition only, noting that  his 1795 lunars method is "somewhat similar"
to one published by Mendozo Rios  [sic] in 1797. In fact it is identical to a
method previously published by  Mendoza Rios. Bowditch has been beaten to
press. Since the claim of a new and  unique method of clearing lunars seems to
have been important for sales of the  Navigator, I suspect Bowditch was strongly
encouraged to enhance his  method.
1804: Bowditch publishes an unbound supplement to the Navigator. It  was
known as an "appendix" but we wouldn't call it that today. This contains  tables
and directions for his enhanced version of his original clearing method.  I
call it "Bowditch's Principal Method". It incorporates a number of simple
tabular tricks for shortening the length of the calculation and avoiding
subtraction (which is always more error-prone than addition), but it is  fundamentally a
refinement of the same method published previously.
1807: The  second edition of the New American Practical Navigator is
published. It  incorporates the tables from the 1804 supplement. Bowditch's Principal
Method is  now described first --it is "method 1". The original method is
pushed back to  method 2, and Witchell's method is now method 3.
1811: In the third edition  of the Navigator, Bowditch notes with some pride
that his Principal Method has  been described (reviewed?) in the Connaissance
des Temps for the year  1808.
1826: Several editions have gone by with no significant change to the  lunars
methods or tables. In this edition, there is a "new form" to Table XX.  This
is the double-entry quadratic correction table required in most of the  series
methods in this era. Bowditch has realized by this point that he can add  in
the quadratic cross-term and gain a little extra range of reliability (at
lower altitudes and somewhat shorter distances) without adding any work at all
for the practicing navigator. The trick here is the same as in Thomson's tables
 but the resolution is coarser in Bowditch's Table XX so interpolation would
be  required for high accuracy. It's better than Thomson's though under
non-standard  atmospheric conditions.
1837: Several more editions pass without significant  change to the lunars
material, but in this edition there is a big change. The  major table from
Thomson's method has been added to the book. This is accurate  and easy to use but
it suffers from its dependence on a mean value for the  Moon's HP and mean
values for the state of the atmosphere. Bowditch's modified  version of Thomson's
method is given position 2 in the list of methods. Since  this is a direct
descendant of Lyons's method, which Bowditch deleted from the  Navigator 38
years earlier, it is impressive that this method is bumped up to  second place.
Several editorial errors creep in at this edition. The phrase  "regulation of
the watch", which refers to setting the deck watch to local time  is erroneously
replaced with "regulation of the chronometer". The intent of the  passage is
not really lost, and I doubt it caused any trouble in the era, but  it's
clumsy. Also, there is a "typo" in the directions for Bowditch's Original  Method
(third method now) which might have caused some real problems for  navigators
trying to use it.

The methods for clearing lunars in the  Navigator from 1837 until 1880 are:
1) Bowditch's Principal Method 2) Thomson's  method (modified slightly) 3)
Bowditch's Original Method, otherwise known as the  method of Mendoza Rios, and 4)
Witchell's method.

Nathaniel Bowditch died  in 1838. His son Ingersoll took over the publication
and though he was a major  patron of astronomy, he did little to improve the
Navigator. I don't think it  interested him, and it seems to have sold
reasonably well regardless. The plates  and rights to the Navigator were sold to the
US Navy in 1868. They continued to  publish the book without any real changes
until 1880 while working on a major  overhaul. Right up until 1880 the
examples in the chapter on lunars are based on  observations and almanac data from
the year 1836.

The overhaul of the  Navigator was completed in January of 1881. It's an
entirely new book after this  date in most respects. All of the methods for
clearing lunars are dropped and  replaced by Chauvenet's rather eccentric method. It
is interesting to note that  Chauvenet scoffed at the method of "one Mendoza
Rios", which he had seen  recently in Raper, when he first described his new
method. Ironic considering  that this was almost exactly the same as Bowditch's
Original Method.

Chauvenet's method shifts to an appendix in 1903 and finally disappears
after 1912 (the year the pre-computed lunar distance tables were dropped from  the
American Nautical Almanac). It is remarkable that a few paragraphs from the
explanatory material and advice on shooting lunars have been copied unchanged
from edition to edition of Bowditch all the way from 1802 to  1912.

-FER
42.0N 87.7W, or 41.4N  72.1W.
www.HistoricalAtlas.com/lunars

   

Reply