of the basilica, but when we consider the vast number of
basilicas that have perished compared to the few that have survived, and
the fact that the origins and traditions of the building show it to have
been, as Vitruvius describes it, essentially a columned structure, there is
ample justification for the view expressed earlier in this article. There
can be little doubt that the earlier basilicas, and the majority of
basilicas taken as a whole, had a central space with galleries, generally
in two stories, round it, and some arrangement for clerestory lighting.
Later basilicas might vary in architectural scheme, while affording the
same sort of accommodation as the older ones.

The relation of the civil basilica of the Romans to the Christian church
has been extensively discussed, and the reader will find the controversy
ably summarized in Kraus's _Geschichte der christlichen Kunst_, bk. 5.
There is nothing remarkable in the fact that a large church was called a
basilica, for the term was applied, as we have seen, to structures of many
kinds, and we even find "basilica" used for the meeting-place of a pagan
religious association (_Roem. Mitt._ 1891, p. 109). The similarity in some
respects of the early Christian churches to the normal form of the columned
basilica is so striking, that we can understand how the theory was once
held that Christian churches were the actual civil basilicas turned over
from secular to religious uses. There is no evidence for this in the case
of public basilicas, and it stands to reason that the demands on these for
secular purposes would remain the same whether Christianity were the
religion of the empire or not. Moreover, though there are one or two civil
basilicas that resemble churches, the latter differ in some most important
respects from the form of the basilica that we have recognized as normal.
The early Christian basilicas, at any rate in the west, had very seldom, if
ever, galleries over the side aisles, and their interior is always
dominated by the semi-dome of an apse that terminates the central nave,
whereas, with the doubtful exception of Silchester (_Archaeologia_, liii.
549), there is no instance known of a vaulted apse in a columned civil
basilica of the normal kind.

[Illustration: FIG. 6.--Plan of Basilica adjoining the Forum of the Roman
city of Timgad, in North Africa.

(From Gsell's _Monuments antiques de l'Algerie_, by permission of A.
Fontemoing.) ]

When buildings