ants
obtained water, it is a reasonable supposition that these were really
cisterns. Similar excavations were discovered all over the area of the
ruins.

Leaving the second terrace, and passing up the ruined stairway, we find
ourselves on the summit platform of the third terrace, and see before us
one of the long, low, richly ornamented buildings of Yucatan. This cut
presents us an end view, but gives us a good idea of the building as a
whole. It does not occupy the entire summit; there is a wide promenade
all around it. Its length is three hundred and twenty-two feet; its
width, thirty-nine feet, and its height twenty-six feet.

Illustration of End View.------------

In order to understand clearly the arrangement of the rooms, we will
here give the ground-plan. The two end portions may have been additions
to the original structure. There are, at any rate, reasons for supposing
the small rooms in the two recesses of later construction. We must
notice that we have here the usual three parallel walls and two rows
of rooms. All the walls are massive, the rear wall especially so. It is
nine feet thick throughout, and so are the transverse walls of the two
recesses. Supposing the rear wall might contain rooms, Mr. Stephens made
an opening through it. He found it to be solid.

Illustration of Ground Plan.------------

Illustration of Cross-section of Uxmal.----------

The stones facing the walls and rooms are smooth and square, and the
mass of the masonry consists of rough, irregular fragments of stone and
mortar. This cross-section makes this meaning plain. We can but notice
what an immense amount of useless labor was bestowed on the walls
and ceilings of this building. We gather more the idea of galleries
excavated in a rocky mass, than of rooms inclosed by walls. The rooms
are very plain; no attempt at decoration was observed. In one or two
instances the remains of a fine coat of plastering was noticed. "The
floors were of cement, in some places hard, but by long exposure broken,
and now crumbling under foot." The arches supporting the roof are of the
same style as those at Palenque--that is, triangular,--though, in this
case, the ends of the projecting stones were beveled off so as to form
a smooth surface. At Palenque, we remember, the inequalities were filled
with cement. Across the arches were still to be observed beams of wood,
the ends buried in the wall at both sides. The supposition is that they
serv