PHYSICAL AND MATHEMATICAL SCIENCES. 1 1
the moon's revolutions under the same conditions as com-pared with those of the sun come round again, so thatvery nearly the same series of eclipses will recur. By thesemeans they were able to predict eclipses with tolerableaccuracy. They observed, with great care, the motionsof the planets *; and according to some testimonies(though contradicted by others), they would appear tohave formed just notions of the nature of comets. He-rodotus t ascribes to them the duodecimal division ofthe day, and the use of the gnomon.
Our knowledge of the Chinese astronomy is entirelyderived from the laborious researches of the Jesuit mis-sionaries stationed in that country, and in particular fromthe works of father Gaubil.^ The Chinese recordsrefer to the date of 2461 b. c., at which they asserta remarkable conjunction of five planets took place.This point has been much discussed ; and it seems, onthe whole, most probable, that it was a purely fictitiousepoch obtained by reckoning backwards. They alsomention a solar eclipse in the reign of the emperorTchong Kang, for neglecting to predict which the twochief astronomers weie put to death. To this, datesare assigned, varying between b. c. 2159 and 2128.By calculating backwards, on modern data, Gaubil andothers find an eclipse in the last-mentioned year; butthe data are so completely uncertain that little weightcan be attached to the statement. No eclipses are men-tioned for many centuries after this.
More importance may fairly belong to the recordedobservation of the length of the shadow of a gnomon,compared at the sutnmer and winter solstices, aboutb.c. 1100. The obliquity of the ecliptic deduced isfound to agree very closely with what would resultfrom the modern calculation of the diminution of theobliquity founded on the theory of gravitation: this
* Diod. Sic. ii. S. + >i. 109.
t See History of Astronomy, Lib. Useful Knowledge, p. 2. el seq .