Kepler was the first astronomer to calculate eclipses with some approach to scientific form, as exemplified in his Rudolphine Tables. His method was of course geometrical. But La Grange, who applied the methods of more refined analysis to the problem, was the first to develop a method by which an eclipse and all its circumstances could be accurately predicted for any part of the earth. To many minds, the prediction of an eclipse affords the best illustration of the superior knowledge of the astronomer: it seems little short of the marvelous. But recalling that the motion of the moon follows the law of gravitation, and that its position in the sky is predictable for years in advance with a high degree of precision, it will readily be seen how the arrival of the moon's shadow, and hence the total eclipses of the sun, can be foretold for any place over which the shadow passes. All these data derived by the mathematician are known as the elements of the eclipse, and they are prepared many years in advance and published in the nautical almanacs and astronomical ephemerides issued by the leading nations. Buchanan's "Treatise on Eclipses" will supply all the technical information regarding the prediction of eclipses that anyone desirous of inquiring into this phase of the problem may desire. So important are total eclipses in the scheme of modern solar research, and so necessary are clear skies in order that expeditions may be favored with success, that every effort is now made to ascertain the weather chances at particular stations along the line of eclipse many years in advance. This method of securing preliminary cloud observations for a series of years has proved especially useful for the eclipses of 1893, 1896, 1900, and 1918; and had it been employed in Russia for totality of 1914, many well-equipped expeditions might have been spared disaster. The California and Mexico totality of 1923 does not require this forethought, as the regions visited are quite likely to be free from cloud; but observations are now in process of accumulation for the total eclipse of 1925. The out-look for clear skies on that occasion, the total eclipse nearest New York for more than a century, is not very promising. The path of totality passes over Marquette, Michigan, Rochester and Poughkeepsie, New York, Newport, Rhode Island, and Nantucket about nine in the morning.