have given interesting notices and illustrations of the subject
for modern builders.

The task of collecting and examining these details is not easy. It
needs much local knowledge and many local books, all of which are hard
to come by. Here, as in most branches of Roman history, we want a
series of special inquiries into the fortunes of individual Roman
towns in Italy and the provinces, carried out by men who combine two
things which seldom go together, scientific and parochial knowledge.
But a body of evidence already waits to be used, and though its
discussion may lead--as it has led me--into topographical minutiae,
where completeness and certainty are too often unattainable and errors
are fatally easy, my results may nevertheless contain some new
suggestions and may help some future workers.

I have avoided technical terms as far as I could, and that not merely
in the interests of the general reader. Such terms are too often both
ugly and unnecessary. When a foreign scholar writes of a Roman town as
'scamnirt' or 'strigirt', it is hard to avoid the feeling that this
is neither pleasant nor needful. Perhaps it is not even accurate, as I
shall point out below. I have accordingly tried to make my text as
plain as possible and to confine technicalities to the footnotes.

F.H.




CONTENTS