entire side-slabs, the ends being filled up either with entire slabs or
with masonry of small stones. In rare cases the side-slabs are replaced
by masonry walls. The average size of the cover-slab is 6-1/2 by 5 feet.
The dolmen itself is, of course, built directly on to the platform, and
the space between it and the circle is filled up with rough stones. The
orientation of the dolmens varied considerably, but the cover-slab was
never placed in such a way that its length ran up the hill-slope,
probably because in moving the slab into place this would have been an
awkward position.

Another equally fine site is that of Bou Merzoug, near Oulad Rahmoun,
about an hour's railway journey from Constantine. The place is naturally
adapted for a settlement as there is a spring of water there. This
spring was later utilized by the Romans to provide water for the city of
Cirta. The dolmen-graves lie in great numbers on the hill at the foot of
which the spring rises, and extend down into the valley. Each dolmen
lies in the centre of a stone circle. This last is in some cases formed
by very large slabs set on edge, but more often by two or three courses
of rough oblong blocks. Many of the graves are badly damaged. One of the
finest had an outer circle about 27 feet in diameter, and an inner
circle 14 feet in diameter. Between these two a third circle, much more
irregular and of small stones, could just be distinguished. But in most
cases it was impossible to make out clearly more than the one outer
circle and the dolmen within it. The dolmen itself consisted of a large
slab resting on walls formed of several large blocks, the spaces between
which were filled up with smaller stones. None of the stones used were
worked. The dolmens were not oriented according to any fixed system. M.
Féraud states that the separate graves were united together by open
corridors formed by double or triple rows of large stones, but no traces
of such a system could be found by the later visitors to the site,