NavList:

I was browsing through the first edition of the "American Ephemeris and Nautical Almanac" (how about AmE&NA, for short) which contains data for the year 1855. Comparing with the British nautical almanac (formally the "Nautical Almanac and Astronomical Ephemeris" --I use NA&AE, for short) for the same year, it was  interesting to see how much they disagreed. It was easy to find cases where the values for predicted lunar distances in the British and American almanacs differed by 30 arcseconds or more. So if you cleared a lunar observation in the year 1855 using the British NA&AE you could get a longitude 15 miles away from the longitude you would find using the American AmE&NA. Modern ephemeris data indicates that the data in the American almanac were usually better but not by a huge margin. It's interesting to note that a navigator who might have discovered this and adopted the arbitrary, even absurd, policy of averaging the American and British almanac data would have done pretty well and better than using either almanac alone.

 

By the way, by 1866, the British almanac calculaors have begun using new tables for their predictions and their data now match the American data to within 2 or 3 seconds of arc in the handful of cases I checked.