in Scandinavia, but this conjoined arrangement is rare and remarkable, and its similarity to the Livonian example is so great that it can hardly be accidental. The Aschenrade graves, it will be recollected, contained coins of the Caliphs extending down to A.D. 999, and German coins down to 1040. There would, therefore, be no à priori improbability in these graves in Algeria being as late, if the similarity of two monuments so far apart can be considered as proving identity of age. Without unduly pressing the argument, the points of resemblance which exist everywhere between the Northern Europe and North African monuments appear to prove that the latter may be of any age down to the tenth or eleventh century, but any decision as to their real date must depend on the local circumstances attending each individual example.
The preceding woodcuts are perhaps sufficient to explain the more general and more typical forms of Algerian dolmens, but they are so numerous and so varied that ten times that number of illustrations would hardly suffice to exhibit all their peculiarities. Their study, however, is comparatively uninteresting, till we know more of their contents, and till something definite is accepted as to their age. When, however, we turn to examine that, we find the data from which our conclusions must be drawn both meagre and unsatisfactory. Such as they are, however, they certainly all tend one way. In the first place, the negative evidence is as complete here as elsewhere. The Greeks, the Romans, and the early Christians were all familiar with northern Africa, and there is not one whisper as to any such monuments having been seen by any of them. When we consider our own ignorance of their existence till some ten years ago, it may be said that such evidence does not go for much; but it is worth alluding to, as a hint in the opposite direction would be considered final, and as its absence, at all events, leaves the question open. On the other hand, all the traditions of the country as reported by M. Féraud, and others, and repeated by M. Bertrand and Mr. Flower, ascribe these monuments to the pagan inhabitants who occupied the country at the time of the Mahommedan conquest. Thus (page 127): "At the epoch the Mussulman invasion these countries were inhabited by a
