any
timber roof.

In the course of the explorations, a probable conjecture was changed into
complete certainty. The very vaults for which inductive reasoning had shown
the necessity were found, if not in place, at least in a fragmentary
condition, and in the very rooms to which they had afforded a cover--and
here we must quote the words of the explorers themselves.

In the most deeply buried quarters of the building, the excavations were
carried on by means of horizontal tunnels or shafts. "I was often obliged,"
says M. Place, "to drive trenches from one side of the rooms to another in
order to get a clear idea of their shape and arrangement. On these
occasions we often met with certain hard facts, for which, at the time, we
could give no explanation. These facts were blocks of clay whose under
sides were hollowed segmentally and covered with a coat of stucco. These
fragments were found sometimes a few feet from the walls, sometimes near
the middle of the rooms. At first I was thoroughly perplexed to account for
them. Our trenches followed scrupulously the inner surfaces of the walls,
which were easily recognizable by their stucco when they had no lining of
carved slabs. What then were we to make of these arched blocks, also coated
with stucco, but found in the centre of the rooms and far away from the
walls? Such signs were not to be disregarded in an exploration where
everything was new and might lead to unforeseen results. Wherever a trace
of stucco appeared I followed it up carefully. Little by little the earth
under and about the stuccoed blocks was cleared away, and then we found
ourselves confronted by what looked like the entrance to an arched cellar.
Here and there these portions of vaulting were many feet in length, from
four to six in span, and three or four from the crown of the arch to the
level upon which it rested. At the first glance the appearance of a vault
was complete, and I thought I was about to penetrate into a cellar where
some interesting find might await me. But on farther examination this
pleasant delusion was dispelled. The pretended cellar came to an abrupt
end, and declared itself to be no more than a section of vaulting that had
quitted its proper place.... The evidence thus obtained was rendered still
more conclusive by the discovery on the under side of several fragments of
paintings which had evidently been intended for the decoration of a
ceiling."[200]

It is clear that these curvil