eclipses of the Sun on the 23d of February, 1477, 
and the 7th of June, 1481, which are indicated 
in the hieroglyphical annals ; on several me- 
morable periods of the conquest ; and on the 
days, when, according to the Mexican records, 
the Sun passes the zenith of Tenochtitlan ; seem 
to prove, that this error of three days did not 
take place ; and that at the beginning of the 
sixteenth century, as we have before observed, 
the dates of the Azteck calendar were more ac- 
cordant with the days of the solstices and equi- 
noxes, than those of the Spanish calendar. 
Without knowing the exact length of the year^ 
the Mexicans would have been enabled from 
time to time to rectify their calendar, as they 
learned from gnomonic observations, that, in 
the first year of the cycle, the equinoxes of spring 
and autumn were some days distant from 7 ma- 
linalli, and from 9 cozcaquauhtli. The Peruvi- 
ans of Cuzco, whose year was lunar, regulated 
their intercalation, not by the shadow of gno- 
‘ mons, which they however very assiduously mea- 
sured, but by marks placed in the horizon, to 
denote where the Sun rose and set on the days 
of the solstices and equinoxes. A periodical and 
exact intercalation, such as that which has been 
known by the Persians since the eleventh cen- 
tury, is no doubt preferable to those sudden 
changes, which are denoted by the title of re- 
forms of the calendar ; but a nation, which for 
