NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Precision of lunars
From: George Huxtable
Date: 2007 Apr 22, 18:10 +0100
From: George Huxtable
Date: 2007 Apr 22, 18:10 +0100
Frank has just ridden out, once again, on one of his favourite hobby-horses. Not that anyone should mind that; we all have our own pet topics. | Although he was a great authority on late 19th century navigation, | Lord Kelvin was absolutely NOT an authority on lunars. Well, how is Frank able to state that, with such certainty? Has he found some error, in Kelvin's text about lunars, that the rest of us have missed? It's likely, being a rich man, with a rich man's yacht, that Kelvin had the very best equipment of the day on board, which in those days would have included multiple chronometers. But he had experience of practical navigation at sea that far outweighs mine, or Frank's, or that of most of us on this list. He was a man with boundless intellectual curiosity, and a thoroughly logical mind, and must have been shipmate with many seamen who had been brought up on lunars. So I reckon that Frank does him wrong to belittle his authority in that way, unless he can offer hard evidence to back it. The lecture | you're quoting from was delivered in 1875. This is forty or fifty | years after lunars ceased being used even as a backup measure aboard | British ships and twenty to thirty years after they ceased that role | aboard American vessels. That's another "sweeping statement" from Frank. I wonder what firm evidence he can provide. I can quote Victor Slocum's biography of his father Joshua, as master of the "Constitution" (not THE "Constitution"!), a passenger vessel plying between Honolulu and San Francisco. On one passage, which cannot have been earlier than 1872, his chronometer broke down but his vessel made port on schedule. "But as it had always been rated by lunars, the mishap made no difference as far as the navigation of the vessel was concerned." Victor Slocum, himself a ship master, adds - "In the hands of skillful observers and patient computers the Lunar Method is reliable to within a quarter of a degree longitude, which would be the distance of a high landfall." That's an estimate that most of us would go along with, I suggest. And he continues- "I have before me the log of the ship Clive which in 1859 made a six months' passage, from the English Channel to Madras, and sighted the landfall within an hour of the expected time. That could not be beaten at this day. Captain Shaw reported that he checked his chronometers, which had altered their rates considerably, by lunars. He said he could by this means keep to the sea four years, the period of time that Lunar Tables in the Nautical Almanacs were published in advance." So, those are two examples to contradict Frank's opinion about the early demise of lunars, as a backup to chronometers. It's clear from the lecture that Kelvin was | trying to warn young navigators not to be seduced by the stuffy old | advocates of lunars who still enforced their teaching in navigation | schools. Kelvin may have thought that; I just don't know. But that isn't what he says in his lecture, as I read it. I ask Frank to recount what Kelvin said, to back up that view. True, Kelvin pointed to the big disadvantage of lunars, their imprecision; and so he should, and so did every other commentator. Frank's comments about the Sumner method, however, are to the point. In fact, his principal astronomical suggestion to navigators | has nothing to do with lunars: his suggestion is that they should all | be using Sumner's method instead of the methods that are usually used | aboard ship in 1875. As we've discussed previously on the list, even | decades after Sumner's method was published, the great majority of | navigators were still shooting separate sights for latitude and | longitude --they just didn't see the merit of Sumner's method over the | common meridian sights for latitude and time sights for longitude. | There's a moment in the lecture where you can almost see him standing | there: Kelvin announces that he is publishing some tables (indeed he | did) to help facilitate Sumner's method and he says "I hold in my hand | copies of these tables which are soon to be published" (or words to | that effect). | | Kelvin is dismissive of lunars for the same reason that Lecky was | dismissive of them at about the same time. Everybody with common sense | knew very well that the best backup for the chronometer was another | chronometer. It was rather silly that all those poor students were | still studying lunar distance calculations so many decades after they | had fallen out of use. Kelvin is simply repeating the "common | perception" of the accuracy of lunars decades after they were commonly | used. He is not describing his own research or saying anything about | the fundamental accuracy of sextants. Certainly, Lecky was dismissive of lunars, but I do not see Kelvin as being dismissive. He simply treated lunar distance as a way of obtaining longitude, with its defects.. What Frank has provided above is the Frank Reed view of lunars, rather than the Kelvin view of lunars. By the way, the "best backup for a chronometer" was two extra chronometers, not just one. If you only had two, and they disagreed, you were hardly any better off, as you didn't know which to trust. | Somewhere along this thread you speculated that Lord Kelvin had access | to better ephemerides for the Moon than those available in the | official almanacs. No way... Exceedingly unlikely. There were very few | people on Earth who dealt with modelling the Moon's motion, and Kelvin | would have had no reason to hunt down that research since it was | irrelevant to his practical advice to navigators. I would add that the | nautical almanacs were improved just a few short years after Kelvin's | lecture and the inaccuracy due to the almanac data went away. Lecky | notes this improvement in his book, but of course, it was too little, | too late. I agree with most of that. But experienced astronomers were still applying their minds to improving the predictions of the Moon's motion, for several reasons beside navigation, around that period (as they are still). Hansen had made great improvements in 1857, which were applied in the Nautical Almanac from 1862, but the great American astronomer, Simon Newcomb, made further corrections, which were applied from 1883 onward. These are what Lecky refers to in a footnote, added after his first edition. The main text had said that there was a small longitude error of 6' to 8' due to uncertainty in the position of the Moon as given in the tables, which must correspond to positional error of the Moon of about 15 arc-seconds. The footnote adds that by 1891, this had been reduced to a "mean error" of just over 1". He doesn't state any scatter, however. Newcomb published two papers in 1878 on the Moon's motion, as follows- "Researches on the motion of the Moon", in Washington Observations for 1875, Appendix 2, page 1 to 280. "Corrections to Hansen's tables of the Moon, prepared and printed for the use of The American Ephemeris and Nautical Almanac." Washington. If anyone has easy access to a Nautical Almanac from 1883 to 1922, it contains an appendix of 2 to 4 pages of Newcomb's corrections to Hansen's tables, which I would greatly like to see. Earlier, there were published (1853, 2nd ed. 1865) by B. Pierce (a name that's appeared recently in these columns) "Tables of the Moon", under "Tables prepared for The American Ephemeris and Nautical Almanac". I haven't seen any of those, and they may be hard to find, but I expect Newcomb's, at least, would be valuable. He wrote very clearly. ================ Kelvin had quite a lot of stuff in his lecture about chronometers, and one example is instructive. Two chronometers were put aboard the ship Tenasserim, which left Liverpool in Jan 1874, and arrived in Calcutta 4 months later. On arrival, they differed by 4 minutes 35 seconds, after correction by "the ordinary method". Kelvin was advocating a new method of correction for temperature variation, based on a daily temperature reading, and a quadratic temperature law, with its two constants determined in advance, for each chronometer. With the corrections made that way, one differed from Greenwich by only 3.5 seconds, the other by 8.5 seconds. It's clear, then, that on long slow sailing-vessel voyages, the chronometer was not necessarily the answer to the maiden's prayer, when adjusting for temperatere by a different technique can have such a large effect on the result. At least, with lunar distances, the error didn't grow with time; with a chronometer, it certainly did. So I suggest that a large influence on the replacement of lunars by chronometers was the introduction of steam. Voyages were then got over in a shorter time, and the long voyages between Europe and the East would be broken by a passage through Suez. As a result, there was less time for big errors to accumulate. 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 -~----------~----~----~----~------~----~------~--~---