Thom does some speculation with posts and the sun's shadow. The north-south line can have been accurately calculated this way certainly. I don't trust Ruggles' stuff at all. He approaches as a sceptic and discounts errors that could only be measured with modern equipment. The equinoxes are floating points between the two solstices, really, and can't have altered position much, consequently. The amount of significance given to the equinox versus the solstice would be, perhaps, Easter as against Christmas. Thom suggested that the year was anciently split into 1/16ths, which he deduced from statistical alignments. Great work, incidentally!