NavList:

Mark Coady, you wrote:
"My home port is in West Cove Noank."

Oh sure. I've heard of Noank. :) Hey, have you seen the sign over on the west side of the peninsula (about 2000 feet due east from the head of West Cover) commemorating the marriage of Amelia Earhart in Noank? And do you know the connection to NavList? We also had a NavList dinner at the Seahorse right there on West Cove with visitors from all over the US (Ken Gebhart, owner of Celestaire and distributor of the Astra sextants, flew in from Kansas to join us).

You wrote:
"The last errors were in calculated GMT. with 2 -3 minutes TIME (GMT)"

Ok. Assuming that all other things were done correctly, this would typically imply an error of 1-1.5 minutes in the lunar arc.

You added:
"I know my index correction was in error."

Well, maybe that's all it was. Do you really think your IC was wrong by a minute of arc or more? That would normally be considered an unlikely error.

You added:
"I expect better. Doing straight altitude sights with that same sextant last summer I was getting positional errors of <2 miles often enough to make me happy."

For altitude sights that would be reasonable, but I agree for lunars you can expect better. So there's a puzzle! And puzzles are fun.

You also wrote:
"So far I was using straight calculator spherical trig (variations on the spherical law of cosines) to resolve relative bearing angle with semidiameter corrected altitudes, followed by calculating true lunar distance using RBA, and computed altitudes."

Those expressions "RBA" and "relative bearing angle" come from John Karl's book (or some other source directly based on it). The name "azimuth difference" has been around a lot longer. I don't know why John felt he needed to muddy the waters with new terminology and an acronym, too. The technique you're describing should, of course, work fine, but you have to be careful about details. I would not be surprised if you discover that some of your several minutes of error in computed GMT is coming, not from your observations, but from your particular style of clearing the sights. Try the online tool I described earlier. It has been in wide use by lunarians all around the world since 2004. It has many advantages including the direct calculation of the ephemeris data. All you have to enter is your known location and the measured lunar arc (plus the usual details like IC). It does everything else and immediately returns the error in your lunar observation. Simple. There are also slick nineteenth century methods for clearing lunars and very fast modern calculator methods for clearing lunars. Here's a little essay on Easy Lunars (a calculator method, also easily adapted fro slide rule use) which I wrote way back in April 2004: http://reednavigation.com/lunars/easylun.html. I plan to revisit this soon.

You added:
"I just went through last month and worked all the sample problems in the Stark Tables. Have not used them for my own sights yet."

Bruce Stark's tables are interesting to ponder, and his publication helped launch a renewed interest in lunars among navigation enthusiasts back in the 1990s. But their time has come and gone. They're not traditional lunar tables. They're a realization of a unique, idiosyncratic modern technique invented by Bruce Stark circa 1990.

Frank Reed