of this southern face of the mountain down to a certain point, one
realizes that even without a wall the place would be practically
impregnable. Add to this the fact that all the stones necessary for a
wall were obtained during the scarping of the arx on the side toward the
Sabines,[31] and needed only to be rolled down, not up, to their places
in the wall, which made the task a very easy one comparatively. Now if a
place can be found which is naturally a suitable place for a lower cross
wall, we shall have what an ancient site demanded; first, safety,
because the site now proposed is just as impregnable as the citadel
itself, and still very high above the plain below; second, pasturage,
for on the slope between the lower town and the arx is the necessary
space which the arx itself hardly supplies; and third, a more reasonable
nearness to the fertile land below. All the conditions necessary are
fulfilled by a cross wall in Praeneste, which up to this time has
remained mostly unknown, often neglected or wrongly described, and
wholly misunderstood. As we shall see, however, this very wall was the
lower boundary of the earliest Praeneste. The establishment of this
important fact will remove one of the many stumbling blocks over which
earlier writers on Praeneste have fallen.

It has been said above that the lowest part of the wall of the arx, and
the two walls from it down the mountain were built at the same time. The
accompanying plate (III) shows very plainly the course of the western
wall as it comes down the hill lining the edge of the slope where it
breaks off most sharply. Porta San Francesco, the modern gate, is above
the second tree from the right in the illustration, just where the wall
seems to turn suddenly. There is no trace of ancient wall after the gate
is passed. The white wall, as one proceeds from the gate to the right,
is the modern wall of the Franciscan monastery. All the writers on
Praeneste say that the ancient wall came on around the town where the