108 COSMOS. 



approximated very remarkably to the truth when he gave 

 it at 7' 30". Delambre,* who did not take into account 

 any of the observations made in his own time, with the 



looked by Delambre (Hist, de I 'Astronomie Moderne, torn. ii. 

 p. 653,) that Newton (probably basing his calculations upon 

 more recent English observations of the first satellite) should 

 have approximated within 47" to the true result, (namely, that 

 of Struve, which is now generally adopted,) while the time 

 assigned for the passage of light over the semi-diameter of 

 the earth's orbit continued to vacillate between the very high 

 amounts of IT and 14" 10", from the period of Romer's dis- 

 covery, in 1675, to the beginning of the 18th century. The 

 first treatise in which Homer, the pupil of Picard, com- 

 municated his discovery to the Academy, boars the date 

 of November, 22, 1675. He found, from observations of 

 forty emersions and immersions of Jupiter's satellites, " a 

 retardation of light amounting to 22 minutes for an inter* 

 val of space, double that of the sun's distance from the 

 earth." (Memoires de VAcad. de 1666-1699, torn. x. 1730, 

 p. 400.) Cassini does not deny the retardation, but he does 

 not concur in the amount of time given, because, as he 

 erroneously argues, different satellites presented different 

 results. Du Hamel, secretary to the Paris Academy, (Regies 

 Scientiarum Academice Historia, 1698, p. 143,) gave from 10 

 to 1 1 minutes, seventeen years after Romer had left Paris, 

 although he refers to him; yet we know, through Peter 

 Horrebow (Basis Astronomies sive Triduum Roemerianum, 

 1735, pp. 122-129), that Romer adhered to the result of 11', 

 when in 1704, six years before his death, he purposed bringing 

 out a work on the velocity of light ; the same was the case 

 with Huygens (Tract, de Lumine, cap. i. p. 7). Cassini's 

 method was very different ; he found 7' 5" for the first satellite, 

 and 14' 12" for the second, having taken 14' 10" for the basis 

 of his tables for Jupiter pro peragrando diametri semissi. The 

 error was therefore on the increase. (Compare Horrebow, 

 Triduum, p. 129 ; Cassirl, Hypotheses et Satellites de Jupiter 

 in the Mem. de V Acad., 1666-1699, torn. viii. pp. 435, 475; 

 Delambre, Hist, de I Astr. mod., torn. ii. pp. 751, 782; Du 

 Hamel, Physica, p. 435.) 

 10 Delambre, Hist, de V Astr. mod., torn. ii. p. 653. 



