THE SHRINES OF ATTICA 259
than Pentclic — from the very hills on which it was
reared, so that it must have seemed, as does the
temple of Bassae in the Peloponnesus, to grow right
out of the landscape. It is easy to see that it was
never completed. Only the fluting of the upper and
lower drums of the columns had been cut in, the rest
being left, as was customary, to be worked off from
these guide-marks when the columns were set up.
The same incomplete tooling is seen on the surface of
the steps.
The old temple and the new are set so close to
each other that they are only a few inches apart— at
one end about eighteen inches, at the other but five
or six. The visitor with a straight eye asks why they
were not built perfectly parallel, when it would have
been so easy to do it. The same divergence in the
foundation lines is seen in other cases, where new
temples were erected close beside those of much ear-
lier date. The explanation of Penrose is that this
difference in orientation comes from the difference in
the Greek calendar. Greek temples, as already shown,
were so built that the rising sun would shine directly
into the front door of the temple on the day of the
year devoted to the god. If the day were changed,
the position of the sun would be changed also. But,
assuming that the same day of the year was nom-
inally retained as the festal day, in the lapse of two
°i" three centuries the uncorrected Greek calendar
would bring about sufficient variation between real and
apparent time so that the sun would not rise on that
day in precisely the same place on the apparent hori-
zon that it did when the first building was erected.
Phe new building was adjusted, therefore, according
than Pentclic — from the very hills on which it was
reared, so that it must have seemed, as does the
temple of Bassae in the Peloponnesus, to grow right
out of the landscape. It is easy to see that it was
never completed. Only the fluting of the upper and
lower drums of the columns had been cut in, the rest
being left, as was customary, to be worked off from
these guide-marks when the columns were set up.
The same incomplete tooling is seen on the surface of
the steps.
The old temple and the new are set so close to
each other that they are only a few inches apart— at
one end about eighteen inches, at the other but five
or six. The visitor with a straight eye asks why they
were not built perfectly parallel, when it would have
been so easy to do it. The same divergence in the
foundation lines is seen in other cases, where new
temples were erected close beside those of much ear-
lier date. The explanation of Penrose is that this
difference in orientation comes from the difference in
the Greek calendar. Greek temples, as already shown,
were so built that the rising sun would shine directly
into the front door of the temple on the day of the
year devoted to the god. If the day were changed,
the position of the sun would be changed also. But,
assuming that the same day of the year was nom-
inally retained as the festal day, in the lapse of two
°i" three centuries the uncorrected Greek calendar
would bring about sufficient variation between real and
apparent time so that the sun would not rise on that
day in precisely the same place on the apparent hori-
zon that it did when the first building was erected.
Phe new building was adjusted, therefore, according