H.—ANTHROPOLOGY. 147 
which are here visible can be assigned respectively to the Aqua Claudia 
and the Anio Novus, the line going south belonging to the construction 
of the tunnel under the Mons Aeflanus by Paquedius Festus in a.p. 88, 
mentioned in an inscription,!® careful investigation shows that they 
branch off from one another at the north edge of the valley, and that the 
south branch falls slightly more rapidly than the other. 
The south branch is undoubtedly still attributable to Paquedius 
Festus, and the western to the main aqueduct; but the problem of the 
existence of one specus only (which confronts us again at Ponte degli 
Arci, though not after we have passed Tivoli) remains at present 
insoluble. 
The level of the bottom of the specus, at the beginning of the existing 
arches going southward (the northern extremity of the aqueduct near 
the road has disappeared), is 248°57, and at the end of the bridge it has 
fallen to 248°17, or 40 cm. in 349 m., which represents a fall of 1 in 872°5, 
| or 1:15 per 1,000. On the western branch the levels are 249°91 at the 
_ beginning of the bridge, and 249°83 at the end, or 8 cm. in 156 m., 7.e. 1 in 
1,950, or 0°51 per 1,000. Both of these falls are below the average fall in 
the long stretch of arches between Capannelle (where the aqueducts emerge 
from their long underground course) and Rome, which varies from 3-22 
to 0:96 per 1,000. The general average is 2 per 1,000, but there is much 
variation." 
The brickwork of the western branch, which is singularly well preserved, 
and shows no traces either of any earlier construction (except for a few 
loose opus reteculatum cubes of the original period) or of later restoration, 
is of the type which must be attributed to a period considerably later 
_ than Septimius Severus. 
The bridge across the main valley, on the other hand, has a con- 
siderable amount of the original opus quadratum preserved, concrete 
faced with opus reticulatum with stone quoins being used at the south 
end, as it probably was at the north. The whole of the central part is 
encased in reinforcements of concrete in which three different periods may 
be traced. At the extreme south end nothing is visible but post-Severan 
brickwork. 
+ In the next valley to the south is the only instance known to me of 
the existence of an alternative channel on an aqueduct bridge. Both 
jpecus appear to have been maintained to the last and there is no sign of 
either having been put out of use. 
The original structure was the straight (western) channel, in ashlar 
masonry of volcanic tufa, quarried on the spot, with the bridge-heads in 
concrete faced with opus reticulatum, which was perhaps the material of 
the channel. The whole structure was then encased in concrete, and the 
channel also restored. The brickwork is good. 
The alternative specus, on the other hand, is faced with greatly inferior 
_ brickwork, of a later period. 
L The channels are of the usual width, 1:14 and 1:17 m., respectively, 
but of exceptional height, the main specus being 2°94 high, and the branch 
no less than 3:34 m. at the point of departure. There is no trace of 
deposit now in either channel, and it may be that the alternative channel 
10 0.1.L. xiv. 3530, rivom aquae Claudiae sub monte Aeflano consummavit. 
U Livellazione, 75 sqq. 
L2 
