Classical Topography of the Roman Campagna.—I. 165
chamber, 6'45 m. in diameter, and lined with cement, is reached: beyond
this the tunnel continues and water soon begins to appear, while another
tunnel branches off from it to the S.E. The water in the main tunnel was,
we were told, one metre deep, and there was a footway on each side. The
water was clear and pure, and apparently fed by strong springs; for it
was said that it was never exhausted, though its level sometimes changed.
These underground cisterns provided the water supply for the villa
above. Between the entrance to them and the road a bathroom, with
marble wall lining and mosaic pavement, had been discovered and de-
stroyed. Various tombs had also been found near the road, and had met
with a similar fate.
The farmhouse itself rests upon ancient foundations of red tufa concrete,
orientated N.E. by S.W. A few yards N. of it is a mound, under which
are the remains of another part of the villa. The mound has not been by
any means completely excavated, and the chambers accessible belong
entirely to the basement. The first which is entered is constructed of
small blocks of tufa about 20 centimetres long by 6 to 8 high, with thick
layers of mortar between. This mode of construction, if met with in
Rome, would be assigned to the fourth century A.D. at the earliest, though
it occurs at Hadrian's Villa,1 and in other villas of the Campagna, where
its close conjunction with .opus reticulafum places it of necessity in the
second century. It has, further, recently been found in some passages
under the area of the Forum Romanum, which probably belong to
Republican times (see Classical Review, 1902, 94). Below this is a hypo-
caust, the floor of which is composed of concrete resting upon tegulae
bipedales. I found below the floor the rare stamp C.I.L. xv. 402, belong-
ing to the time of Commodus : other stamped bricks were said to have
been found and destroyed. The floor is supported by hollow terra-cotta
socket pipes, each, not counting the socket, 32 cm. long ; at least three were
placed one above the other (the space below the floor is not completely
cleared out): sometimes the top one is reversed, so that the socket does
not fit into the pipe below, and there is often a small piece of brick
between the pipe and the floor of tiles. The chamber measures 2-92 m.
by 3'53 m.,and there are six rows of supports one way and seven the
other. In the course of the excavations a dolium and several rough
household pots were found.
1 See Winnefeld, JafoWi des Instituts, Erganzungsheft iii. p. 26 sqq.
chamber, 6'45 m. in diameter, and lined with cement, is reached: beyond
this the tunnel continues and water soon begins to appear, while another
tunnel branches off from it to the S.E. The water in the main tunnel was,
we were told, one metre deep, and there was a footway on each side. The
water was clear and pure, and apparently fed by strong springs; for it
was said that it was never exhausted, though its level sometimes changed.
These underground cisterns provided the water supply for the villa
above. Between the entrance to them and the road a bathroom, with
marble wall lining and mosaic pavement, had been discovered and de-
stroyed. Various tombs had also been found near the road, and had met
with a similar fate.
The farmhouse itself rests upon ancient foundations of red tufa concrete,
orientated N.E. by S.W. A few yards N. of it is a mound, under which
are the remains of another part of the villa. The mound has not been by
any means completely excavated, and the chambers accessible belong
entirely to the basement. The first which is entered is constructed of
small blocks of tufa about 20 centimetres long by 6 to 8 high, with thick
layers of mortar between. This mode of construction, if met with in
Rome, would be assigned to the fourth century A.D. at the earliest, though
it occurs at Hadrian's Villa,1 and in other villas of the Campagna, where
its close conjunction with .opus reticulafum places it of necessity in the
second century. It has, further, recently been found in some passages
under the area of the Forum Romanum, which probably belong to
Republican times (see Classical Review, 1902, 94). Below this is a hypo-
caust, the floor of which is composed of concrete resting upon tegulae
bipedales. I found below the floor the rare stamp C.I.L. xv. 402, belong-
ing to the time of Commodus : other stamped bricks were said to have
been found and destroyed. The floor is supported by hollow terra-cotta
socket pipes, each, not counting the socket, 32 cm. long ; at least three were
placed one above the other (the space below the floor is not completely
cleared out): sometimes the top one is reversed, so that the socket does
not fit into the pipe below, and there is often a small piece of brick
between the pipe and the floor of tiles. The chamber measures 2-92 m.
by 3'53 m.,and there are six rows of supports one way and seven the
other. In the course of the excavations a dolium and several rough
household pots were found.
1 See Winnefeld, JafoWi des Instituts, Erganzungsheft iii. p. 26 sqq.