306 ROMAN FRONTIER DEFENCES AND ROADS. [CHAP.
Berenice on the Red Sea. Finally, the provinces of northern
Africa were connected with one another by a continuous road
which skirted the shore of the Mediterranean ; and in the western
portion of that region, where there is a wider belt of cultivated
land, this was supplemented by other highways, which approached
nearer to the southern frontier of the empire.
Let us now notice the specimens which have come down to us
of the documents by which information was fur-
nished witn regard to the roads in the Roman
empire. We have already seen1 that these docu-
ments were of two kinds, (i) Itineraria adnotata^ which contained
lists of the principal stations on the roads, accompanied by
computations of the distances between them, but without any
geographical remarks or explanations, so that they somewhat
resembled the Railway Guides of the present day; and (2)
Itineraria picta, where the same details were given in a form more
nearly approaching that of a map, with the addition of various
geographical features, especially the courses of the rivers. The
wall-map of Agrippa would seem to have been the original source
from which the main facts in both of these were drawn; but we
cannot doubt that much additional material was from time to time
embodied in them, which was furnished by the Roman archives,
for these could not fail to possess a catalogue of the Roman roads,
with measurements of their length according to the milestones,
and this catalogue would be gradually enlarged. The former of
the two classes is represented at the present day by
'nine iti&eruy. *e Antonine Itinerary, or, to give its full title,
Itinerarium Provinciarum Antonini Augusti. The
emperor to whom the first publication of this work is here referred,
is commonly supposed to have been either Antoninus Pius, or
Marcus Aurelius Antoninus, but some writers are disposed
to ascribe it to Caracalla, who also bore the name of Antoninus.
It is clear, however, that the edition of it which we
possess is not earlier than the time of Diocletian,
since the name of the city of Diocletianopolis, which
was so called after him, occurs in it, and Perinthus is here called
Heradea—a name which it did not receive until shortly before
1 ?. 5ufra> p. 236.