80

HISTOKY OF PHYSICAL ASTBQNOMY.

by means of liis own observations and those of Hodierna, but his labourswere attended with very imperfect success.

The earliest tables which enjoyed any confidence among astronomers werethose of Cassini, which first appeared in 1008 at Bologna. Picard, the cele-brated French astronomer, having compared them with a number of observedeclipses, found them to be even more accurate than their author anticipated.Fie, in consequence, recommended him to Colbert as an astronomer, whosetalents would be an ornament to France , and, at the suggestion of thatminister, Louis the Fourteenth invited him to his capital. Cassini, uponhis arrival in Paris , resolved to perfect his previous researches on themotions of the satellites, and during many years he continued to makeobservations on their eclipses. In 1693 he published his second tables,which far exceeded in accuracy any previous efforts of the kind. Thoseof the first satellite especially were found to represent the times of theeclipses with remarkable fidelity, and by means of them the longitudewas determined with a precision hitherto unknown. In these tables theorbits of the four satellites were considered to be circular; they were in-clined to Jupiter ’s orbit at equal angles, and their nodes had all a commonposition. Cassini estimated the inclination of the orbits at 2° 55', andhe fixed the nodes in 10 s 14° 3(1' of longitude; both of these elementswere supposed invariable. He did not employ the equation of light inhis tables, although at one time he was favourably disposed towards thehypothesis of Roerner. He perceived that the successive propagation oflight explained the irregularities in the eclipses of the first satellite whenthe earth was in different positions of her orbit; but, finding that it didnot account in an equally satisfactory manner for the irregularities of theother satellites, he rejected it altogether, and instead of it he used in thetables of the first satellite an empiric equation depending on the relativepositions of the Earth aud Jupiter . Although the error in an eclipse ofthe first satellite seldom exceeded 1' of time, yet it happened occasionallythat it rose to 5' or 6'. The inequality which principally occasioned thiserror was certainly not easy to discover ; but it is surprising that a similarinequality in the second satellite, which rises to a much greater magni-tude, should have escaped the sagacity of Cassini. He also failed tonotice the principal inequality in the fourth satellite, although it causesthe times of eclipses to vary to the extent of an hour. Notwithstandingthese defects, the tables of Cassini mark an important epoch in thehistory of the satellites, and their construction will ever remain a monumentof the ingenuity and patience of their illustrious author.

Maraldi I., the nephew of Cassini, also devoted much attention to thesubject of Jupiter ’s satellites. He admitted, in common with his relative,that the equation of light gave a very satisfactory account of the errors inthe first satellite, when the earth was in different parts of her orbit, buthe maintained that, if this equation was founded upon true physicalprinciples, it should vary from the perihelion to the aphelion of Jupiter ’sorbit, a conclusion which the observations on eclipses did not seem to himto warrant. He also remarked, that if the errors in the times of theeclipses depended upon the successive propagation of light, they shouldbe equal for all the satellites when the earth was in the same parts of herorbit. It did not occur to him that other irregularities might exist in themotions of the satellites, and might cause the errors of eclipses to be verydifferent for each satellite. It is true that the orbits were supposedcircular, and as long as astronomers entertained this belief there could