the Wends—the Wendish, of course, only to a very limited extent. The oldest examples of the Old Northern language are not, however, to be found in the Icelandic, but in the names and words graven on stones in runic characters in Scandinavia, Denmark, and Britain. This method of attempting to read some of our disguised or altered place-names appears to be reasonable to the archæologist, who looks not merely to the historical statements of the old chroniclers and the names for his evidence, but also to the surviving customs, to anthropological and archæological discoveries, to folk-lore, and all other sources from which information bearing on the settlement may be gleaned. The value of the information that may be gathered from these sources to the historian or philologist is great. We can see on the Ordnance map of England many names whose origin goes back only to recent centuries, but we find also in every county many others of extreme antiquity. If we could fully understand them we should know much relating to the Anglo Saxon period of our history of which we are now ignorant. Even the different ways in which the homesteads in different parishes or townships are arranged, whether they are scattered or clustered in groups, give information by which the archæologist is able to assist the historian. The scattered homesteads may in some districts be as old as the British period, or in others may have been formed first by emigrants who came from some old Continental areas where the Celtic arrangement survived. There are many other and more numerous areas where nucleated villages exist, in which the homesteads are collected, some arranged on the plan of having roads radiating from them—i.e., the star-like way, similar to the German type common between the Elbe and the Weser. In other instances we find collected homesteads of an elongated, rounded, or fan-shaped form enclosing a small space, around which the original houses were built. These resemble the village types east of the Elbe, in the
