who were no doubt ig norant of the position of the 
stars of which the divisions of the ecliptic were 
composed. It is possible, that nations relapsed 
into barbarism had preserved but a confused re-^ 
membrance of the names of the nacshatras ; an4 
that in reforming their calender, they might 
have chosen among the names those of the signs 
of the solar zodiac, without following the order 
anciently adopted. It is possible also, and I am 
inclined to give the preference to this latter 
opinion, that the zodiac composed of twelve signs 
may have had its origin from an ancient lunar 
zodiac, in which the nacshatras were arranged in 
an order more analogous to that which we ob- 
serve at present in the dodecatemoria of the 
people of Thibet and Tartary. In fact, the divi- 
sions of the ecliptic, which Sir William Jones, 
Colebrooke, and Sonnerat, have published, differ 
essentially from each other. The arrow, which 
according to one Indian writer is the eighth 
nacshatra, is only the twenty-third according to 
another. We shall see presently, in speaking of 
a Roman bas relief described by Bianchini, that 
in the East solar zodiacs formerly existed, which 
had the same signs, though placed in a different 
order. Moreover, the return of the Sun from the 
tropics toward the equator, and the phenomenon 
of the equal duration of the days and the nights, 
must have led to great changes in the figures 
