No. 22: Jul-Aug 1982
October 27, 1780. The Maine Coast. An expedition from Harvard and the American Academy of Arts and Sciences, led by Samuel Williams, had set up equipment to observe the predicted total eclipse of the sun. A solar eclipse occurred all right, but the expedition was shocked to find itself outside the path of totality. They saw a thin arc of the sun instead of its complete obscuration by the moon. Modern analyses of this embarrassing incident for embryonic American science blame Williams for miscalculating the path of totality.
Actually, recent computations compound the mystery. The expedition measured the time of the eclipse as 40 seconds later than it should have been for their (erroneously selected) site. Modern analysts insist further that the expedition should have seen an arc 10 arcsec wide subtending an angle of 89°; instead, Williams and his colleagues measured an arc of less than 24°. Finally, the measured duration of the eclipse was far different from what it should have been.
The members of the expedition were skilled and their instruments excellent. What happened? Was it human error? Are modern eclipse-calculating methods in error? Did something astronomical happen, such as a temporary, slight glitch in the earth's period of rotation? The mystery persists.
(Rothschild, Robert F.; "Where Did the 1780 Eclipse Go?" Sky and Telescope, 63:558, 1982.)