NavList:

Alex E wrote:

"I am saying (answering the question why there was no "Lunar era") that there was no "Lunars era" in history of navigation because"...

 

I think you may be jumping too quickly to an answer. We've got to establish the facts on the ground first...

But I would like to address your points one by one.

You said:

"a) accuracy is too low"

 

Thirty to sixty minutes in longitude is an enormous improvement over pure dead reckoning on a long ocean voyage. But did average navigators understand the concept of error bars? I don't think so. I don't think practicing scientists in this era even had a good handle on the issue.

 

"b) reduction is too hard."

 

This wasn't true. The math was tedious; it wasn't hard. But there WAS an associated expense. You had to study this stuff, or hire someone who had studied it.

 

"c) a better method (chronometer) became available at almost the same time."

 

In the logbooks, longitude by chronometer takes over starting c.1835 (to reiterate, that's for American commercial vessels). Lunars were available starting in 1769. That's over SIXTY YEARS --a very long lag. Lunars definitely were used in this earlier period but they were not used the way that some modern navigation enthusiasts imagine. They were not ever a primary method of determining longitude at sea except for a very few, like Bowditch. Instead, just as they were an occasional check on the "longitude by chronometer" in 1850, lunars in 1800 were an occasional check on the "longitude by DR".

 

Something to consider: Longitude by dead reckoning does work. A good 19th century navigator could probably count on it up to perhaps five miles a day average error. Since this is a random error most of the time, the cumulative error would be about 5miles*sqrt(days). After 14 days, this would amount to about 19 miles expected error in longitude. That's just barely enough to be detectable by a lunar observation. So perhaps the "once every two weeks" pattern for taking lunars was all that was required and a completely practical solution to the problem. Of course, if the dead reckoning is going wrong because of a steady current instead of random error, the difference in longitude could be much greater in the same period of time...