In reading and comparing these two series of sacrificial hymns, some points of contact present themselves, and, observing this, it occurred to me that some Sanscrit word ending in _Medha_--_i.e._, sacrifice, and conveying the meaning of _human sacrifice_, might by ancient Indian astronomers have been attached to the constellation, which for us represents the hapless Andromeda: for if we suppose that the constellations Cassiopeia and Cepheus were imagined in India, but adopted with an appropriate legend into the Grecian sphere--the names of the personages in the legend at the same time suffering a Grecian change--it would be easy further to suppose that the Indian name of the constellation near to them, transformed and misunderstood, came to represent in Grecian story not merely a _human_ sacrifice, but that of the much-to-be-pitied daughter of the proud Cassiopeia. Whether these fanciful speculations concerning the names of the actors in the ancient legend be adopted or not need not affect our judgment as to the reasonableness, or otherwise, of the date, 3500 B.C., and of Lat. 23° N. for the origin of the constellational group here discussed. PLATE XXIII.[124] [124] The figures in this plate have been drawn from the globe adjusted to the following dates and latitudes. Figs. 1 and 2, 3589 B.C., Lat. 35° N. Fig. 3, 3050 B.C., Lat. 35° N. Fig. 4, 1443 B.C., Lat. 40° N. The probable dates for the first imagining of four constellations are here given--namely, for the Centaur, Ophiuchus, Auriga, and Perseus. For the Centaur the date in round numbers of 3500 B.C. (fig. 1) is suggested: at that date his huge figure would have well marked, in opposition, the beginning of the calendrical Aries-year; or, in conjunction with the sun, the beginning of the seventh month of the same year. It is not necessary, at that date, to attribute a low latitude to the astronomers who designed this figure: in that of 35° N., as shown in the diagram, the whole constellation would then have been well above the horizon. The much earlier epoch of 6000 B.C. might perhaps be claimed for the Centaur. At that date, as I have assumed, the calendrical and the solstitial year coincided. (Compare Plate XVII. and Plate IX.) As between 6000 and 3500 B.C. I have often hesitated, but on the whole I have come to think the later date, as here given, the more probable.