NavList:

   

Reply

Hi all,

I apologize to the List for my delay in responding to Frank’s reply to my 
original postings under the above subject; unfortunately, I simply do not 
have the time to compose properly reasoned replies to much of what is posted 
these days and spend much of my time lurking. However, as Lunars have again 
been featured, it may be appropriate to now respond, as follows ...

In NavList 3269 I opined that Lunars generally had received a bad name as far 
as accuracy is concerned primarily for two reasons.

I stated in 3269
1.  The relative inaccuracy of early instruments. Let's remember that the 
octant, as originally produced was not fitted with a tangent screw and gear 
for easy and positive measurement of small arc increments. Although the 
vernier appeared rather early on, it was the incorporation of the tangent 
screw that allowed more positive fine adjustment. I also certainly cannot 
believe that some of the earlier instruments made of wood with ivory arcs, 
although works of art, could be relied on as respects accuracy over a wide 
range of condition to be met at sea.

Frank responded in 3272, in part 
Certainly, instrument accuracy was always a big issue and still is for lunars. 
I agree with that. Regarding your specific comments on octants, I know of no 
one who shot lunars with a wooden octant in the era. A metal sextant --with 
tangent screw, telescope, and assorted shades-- were all recommended right 
from the very beginning of the lunars era. Even in 1763, Maskelyne spells out 
these requirements in detail. There's clear evidence 
that navigators carried two types of instruments: octants, like you've 
described above, for daily latitude observations (almost always Noon Sun) and 
sextants for lunars. Of course, if there WERE any navigators who declined the 
advice to get a proper sextant, and this advice was in practically every 
navigation manual and guide to shooting lunars, they would definitely have 
been disappointed with the results. Wooden octants arecomparable in accuracy 
to modern plastic sextants (e.g. Davis).

Then he went on to post in 3789, in part ...

Here's an article published in 1847 discussing the uses of lunars and  
chronometers. The principal lunar story actually refers to events in 1825.  
Apparently, this navigator had good success taking lunar observations with a  
"quadrant" which in that era would have meant an ordinary wooden octant.

I now respond further ...

It was not my intent to imply that octants were employed in Lunar observations 
to the exclusion of any other instrument type, and merely cited their 
construction as an example of the potential for inaccuracies. Frank knows as 
well as I do that there were also wooden sextants manufactured – how many I 
don’t know, but enough have wound up in collections and museums to give us a 
good sample. Even if a metal sextant was the 18/19th century navigator’s 
instrument, what was the arc made of – early instrument makers were known to 
sometimes use common bone, ivory, or brass for the arc before silver became 
the commonly accepted material for this purpose; Spencer, Browning & Co., 
London (1849 – 1870), and others,  marketed  brass sextants with an ivory 
arc, and I have personally handled a metal 19th century sextant possessing an 
ivory arc which had become so contracted as to render the instrument useless 
for any purpose. Certainly, it is possible to
 reference log books containing worked up Lunars, but unless that logbook also 
specifies and describes the instrument used in some detail you really don’t 
know what was used, do you now. Assumptions not based on hard evidence do not 
make historical fact. Also, what degree of accuracy can be ascribed to such 
records of Lunar calculations at sea as have come down to us by way logbooks; 
unless associated with a landfall, or comparable navigational event, you 
again simply don’t know.

John Bird, London astronomer and scientific instrument maker, manufactured 
both brass and wood instruments; in 1759, he produced a mahogany framed 
sextant with an ivory arc – an example is in the Scheepvaart Museum, 
Amsterdam; he is reputed, but not so proven, to have made a large wood framed 
sextant in 1771 for Captain Cook to use on his 2nd voyage – this instrument, 
or an example thereof, is presently in the Science Museum, London. H. 
Limbach, of Hull, England, produced an ebony framed Sextant with brass index, 
ivory arc, and ivory vernier, and there are on display other signed wood 
framed sextants "equipped with, tangent screw, telescope, and assorted 
shades”, at least into the 1830s. I cannot say what these wood framed 
sextants were used for or how accurate they may have been – neither can I say 
that they were not used for Lunars. Scientific expeditions, supported by 
Governments appear to have been well funded and supplied with the best
 instruments available; not so the average Merchant Vessel, aboard which 
officers were generally required to provide their own instruments, and I 
seriously doubt that they universally provided themselves with the assortment 
recommended by some navigational texts, regardless of how valid such 
recommendations may be. As respects the United States, the best of 
instruments were not always available; witness the US Exploring Expedition, 
of 1838 – 1842, which found it necessary to send an emissary to Europe in 
search of suitable instruments to outfit its vessels. I am pleased to find 
that Frank, in 3789, has found an 1825 navigator who apparently ... “had good 
success taking lunar observations with a  "quadrant" which in that era would 
have meant an ordinary wooden octant.”   ... I think there may have been many 
more.

I also stated in 3269

2.  The mathematical methods of clearing the distance, especially the so 
called approximate methods were a wonderment of complication, in some 
instances requiring the application of numerous corrections which in 
themselves sometimes defied reason. Most of the epitomes I have read did not 
apply to reason, but simply listed rules to follow. In my view, at least, a 
relatively simple explanation of the problem related to a trigonometrical 
format would have better served the mariner. The entire subject of spherical 
trigonometry does not appear as complicated as some of the approximate 
methods.

Frank responded  in 3272........

Here, I DO disagree. The great majority of the lunars that you will find 
worked out in 19th century logbooks are series methods (erroneously known 
sometimes as "approximate" methods). The steps involved in working them were 
no more complicated than working an ordinary time sight --something which 
nearly every navigator could do with ease. The amount of work was about three 
times greater, but the steps and mathematical skills were not greater. I've 
seen many examples of lunar observations worked out on spare pages in old 
logbooks, and not once have I seen any of the straight triangle methods, ike 
Borda's, in use. They used the series methods, like Bowditch's Principal 
Method, Thompson's method (adopted into Bowditch in 1837 just before his 
death), Witchell's method, variants of Lyons' method, Bowditch's Original 
Method (known in Europe as the method of Mendoza Rios, although never claimed 
by MR himself) and so on. There's no doubt that other direct
 methods, including Dunthorne's and Borda's and Krafft's were in fact used, 
but they were less popular. The choice of method was not considered a "big 
deal" in the day. Many navigators seem to have learned a couple of different 
methods if for no other reason than for intellectual amusement. 

I now respond further ...

My dictionary defines approximate: 1) as an adjective = “ nearly exact, not 
perfectly accurate or correct”, and 2) as a verb = “to come near to; approach 
closely to”. So really what is the discussion about. Is it now claimed that 
the methods mentioned in 3272, including the “straight triangle methods” are 
perfectly accurate or correct – if not, they are and have been properly 
called "approximate". I have long felt that Lunars should be perhaps 
categorized as “Scientific” or “Commercial”, as well as by period. Certainly, 
some excellent observations were returned by early Government expeditions; 
these were,however, well funded, supplied with the best instruments and 
astronomical tables available, and frequently included astronomers in their 
complement of scientific personnel.

Accuracy in the clearing of a Lunar Distance requires, in addition to other 
factors normally considered, attention to the Moon’s parallax in azimuth, the 
spheroidal shape of the earth, and the true shape of the Sun/Moon disks as 
affected by refraction. These factors were apparently early on not thought 
necessary of consideration in practical usage and of relatively small 
magnitude – given Maskelyne’s own expectations of instrumental accuracy to 
only 1-minute of arc, as well as Longitude accuracy of from 15 to 30-minutes 
of arc, and including log entries posted on this List wherein Lunar 
observations are recorded to only the nearest minute of arc, an effort to 
improve accuracy by a few seconds of arc here an there may well have been 
extraneous. Regardless, as instruments improved, tables became more accurate, 
and scientific minds became tuned to the problem, more attention was 
obviously invited to these lesser corrections - unfortunately, by this 
 time Lunars were already partially on the way out.

While acknowledging the method of Friedrich Wilhelm Bessel (1784 – 1846), a 
German astronomer and mathematician of the highest repute, as being then the 
only truly rigorous method of clearing a Lunar Distance, William Chauvenet, 
in his 1868 paper entitled “Navy Scientific Papers No. 1 – Astronomy: 
Comprising Suggestions to US Naval  Officers”, details a more modern and 
balanced view of Lunars. This paper, once discussed on this List extensively, 
details and quantifies the errors to which Lunars were subject previously by 
reason of tabular inaccuracies prior to 1855, neglect of certain corrections, 
instrumental error, and observational error. I will not extend this writing 
by further quoting from this paper, but will strongly recommend it to budding 
Lunarians and for review by the pundits of this List for it lays bare the 
inadequacies, at least in the authors opinion, of the older methods. He also 
published a set of tables to facilitate the use of
  an “approximate” (his words) method, devised by him, to clear the distance 
with a mean error of not more than 2-seconds. Chauvenet’s method seems to 
have impressed the US Hydrographic Office, as the 1888 edition of HO 9 
(Bowditch) includes its use, to the exclusion of others. 

In a paper presented before the Royal Society of Victoria in 1889, E. V. 
White, FRAS, reported a mean of 42 Sun/Moon Lunars, observed by him at a 
place of known Longitude between August 27 and October 2, 1887 and cleared by 
Chauvenet’s method, to have exceeded the true Longitude by 4.8 seconds in 
time; the probable error of a single observation is said to have been +/- 21 
seconds in time. Mr. White also has a number of less than favorable remarks 
to make regarding the accuracy of sextants encountered over his some 40 years 
prior experience in their use.

I certainly agree with George's philosophy regarding Lunars, as set forth in 
5647, and perhaps carry it a step further. Lunars have evolved, from a rather 
crude beginning to an ever more scientific approach. The solution to the 
Lunar Problem is based on sound principle, and perhaps still awaits the more 
perfect means of measurement - experiments were at one time in progress 
utilizing the photographic plate and apparently with some success. Who knows; 
someday, somebody may come up with a startling breakthrough. In the meantime 
....

Regards,

Henry


      

--~--~---------~--~----~------------~-------~--~----~
Navigation List archive: www.fer3.com/arc
To post, email NavList@fer3.com
To , email NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---

   

Reply