NavList:

Frank,
Perhaps my ability to explain my result isn't that good. What I tried to explain was that the different constants (53° and 62°) were used as a way to provide an approximate "third correction" that always is additive, while the moon's parallax correction could be either additive or subtractive. As the graphical method only provides the moon's parallax correction per se, the only simple way to add a "third correction including everything else", without an additional table, is to take a percentage of the parallax correction. This I think is a quite clever idea, but the result is mediocre. Enclosed is an Excel sheet where one additive case and one substractive case is cleared by three different methods. The graphical method of the 1883 Hints is by far the worst one. To avoid measuring errors I have calculated the graphical method; with measuring errors added it would be more or less useless as a way to rate chronometers, which I guess was it's purpose in Hints. So I do agree with youh "that this proposed graphical method of clearing lunars was just a bad idea".
Lars