242 EXCAVATIONS AT THE ARQIVE HERJEVM IN 1892.

course varies considerably from a straight line at many points
and it bears all tlie marks of belonging to a late period of build-
ing. Only the steps are preserved for a considerable portion of
its length, and we could not prove tbe existence of a stylobate on
which columns had stood. We did find fragments of columns,
which might have belonged to the stoa, and a well preserved
angle-triglyph of corresponding dimensions; also, what may have
been a piece of the cornice, with finely worked ornament of mse-
ander-pattern in relief. At a point nearly opposite the eastern
end of the temple, the front line of the stoa projects very consid-
erably toward the south, thus forming a main central portion with
probably two receding wings; though toward the east we could
not follow the excavation far enough to establish this detail. The
main portion was cleared back to the rear wall, which makes are-
taining-wall against the slope above, and is built of unhewn
stones, rather carelessly laid. The whole space was completely
filled with bases of statues and steles of all forms and dimensions.
Some must have borne figures or groups of considerable size; all
were packed together with no attempt at order or arrangement.
Still more stood before the stoa, in the area between the central
hall and the wings. "We were surprised to find nothing more
than the bases themselves, no trace of a statue and no single frag-
ment of an inscription. The marble had of course been burned
for lime during the Middle Ages. We did find in situ, on a stone
projecting before the step next to the uppermost, a relief repre-
senting two doves facing each other. It was not work of the best
period, and, as has been said, the stoa seemed on all accounts
late. Moreover, one of these cross-walls dividing this middle
space passed directly over a statue-base, this fact pointing to a
still later reconstruction. At the west end of the stone was a
curved line of foundation, the special meaning of which we could
not determine.

Several minor discoveries may be briefly reviewed. An irreg-
ularly shaped pavement, made up of large square stones and resting
upon a rough foundation of rubble and earth, is situated near the
northeastern corner of the temple-stereobate, and is nearly of the
same level. It may be explained conjecturally in a variety ot
ways.