Of these fifty-one monuments only one is a dolmen, in the sense in which we usually understand it, meaning thereby a single cap stone, supported by three, or, as in this instance, by four uprights. This one is near Exlo, and is one of the few that formed a chamber in a tumulus. A few have three cap stones, and from that number they range up to ten or twelve, with at least double that number of supports. They are all, in fact, of the class which the French call "allées couvertes," or "grottes des fées;" Calliagh Birra's house (woodcut No. 80) and the dolmens at Glen Columbkill are of the same class. But the Drenthe dolmens have one peculiarity not found either in France or Ireland: that they are all closed at both ends, and the entrance, where there is one, is always on the longer side. In this respect they more resemble the Scandinavian examples, such as the tomb at Axevalla (woodcut No. 116), or that at Uby (woodcut No. 114).
The annexed attempted restoration of one near Emmen will give a fair idea of their general arrangements. It is 49 feet long over all, and internally from 4 to 6 feet in width. It is roofed with nine or ten stones, some of considerable dimensions. Some of these Hunebeds have a range of stones round them, not arranged in a circle or oval form, but, as in this instance, following the lines of the central chamber. This is the case with another near the same place, which is 125 feet in length over all. When closely examined, however, it does not seem to be one Hunebed, but three ranging in a straight line, with a small space between each. Two have five and one six cap stones. As a rule, each cap stone stands on two uprights, and though frequently they touch one another, as often they form really independent trilithons. It was no doubt this fact that induced Keysler to compare these monuments with Stonehenge, though in fact no two sets of rude-stone monuments could well be more dis-
