in. _Alignements_ are more
difficult to explain, for, from their form, they cannot have served as
temples in the sense of meeting-places for worship. Yet they must surely
have been connected with religion in some way or other. Possibly they
were not constructed once and for all, but the stones were added
gradually, each marking some event or the performance of some periodic
ceremony, or even the death of some great chief. The so-called
"Canaanite High Place" recently found at Gezer consists of a line of
ten menhirs running north and south, together with a large block in
which was a socket for an idol or other object of worship. Several
bodies of children found near it have suggested that the monument was a
place of sacrifice.

Other megalithic structures can be definitely classed as dwellings or
tombs, as we have seen in our separate treatment of them. It is not
improbable that, if we are right in considering the dolmen as the most
primitive form of megalithic monument, megalithic architecture was
funerary in origin. Yet, as we find it in its great diffusion, it
provides homes for the living as well as for the dead. In their original
home, perhaps in Africa, the megalithic race may have lived in huts of
wattle or skins, but after their migration the need of protection in a
hostile country and the exigencies of a colder climate may have forced
them to employ stone for their dwellings. In any case, in megalithic
architecture as seen in Europe the tomb and the dwelling types are
considerably intermixed, and may have reacted on one another. This,
however, does not justify the assertion so often made that the
megalithic tomb was a conscious imitation of the hut. It is true that
some peoples make the home of their dead to resemble that of the living.
Among certain tribes of Greenland it is usual to leave the dead man
seated in his hut by way of burial. But such a conception does not exist
among all peoples, and to say that the dolmen is an imitation in stone
of a hut is the purest conjecture. Still more improbable is Montelius's
idea that the corridor-tomb imitates a dwelling. It is true that the
Eskimos have a type of hut which is entered by a low passage often 30
feet in length, but for one who believes as Montelius does that the
corridor-tomb is southern or eastern in origin such a derivation is
impossible, for this type of house is essentially northern, its aim
being to exclude the icy winds. In the south it would be into