Daniel Graham, in his latest book, argues that Anaxagoras had measured the size of the sun with the help of a solar eclipse. The idea was proposed earlier, but with other eclipses than that of 17 February 478 BC that went over the Peloponnesus and could be observed in Athens. I will show that Graham's attempt is based on assumptions of modern astronomy which do not hold for a flat earth, and that its execution yields insurmountable problems. An alternative method, proposed by me, Graham calls "excessively complicated". I will show that a rough and simpler version of the same method also leads to the desired result.