linnets, in a world entranced from movement or the passing of time. And then I fell asleep. CHAPTER V LYNTON (_continued_), COUNTISBURY, AND NORTHWARD The word "Lynton," Mr. Chanter tells us in his interesting monograph on the village, means the town on the lyn, and "lyn" is the Celtic word, not for river, but for pool, and occurs in this meaning all over England, in Northumberland, Yorkshire, Kent, Herefordshire. It is strange, perhaps, that this rushing mountain stream should have been named from its very rarely occurring pools, but the authority is indubitable. The Celtic folk who named it, the "early Britons," as our childish history books used to call them, were not, of course, the first inhabitants of this wild and wooded spot; there are neolithic remains--hut circles and burial-places--fairly thickly scattered along this coast, and a certain number of flint implements have been found. The hut circles in the Valley of Rocks, of which traces still remain, though many of them have been destroyed quite recently, within the last two hundred years or so, belong to this period, and it is probable that the earth-camps of Lynton and Countisbury, of Parracombe, Martinhoe, and Ilfracombe, were built by the immense labour of this vanished people. Remains of the early Bronze period show that there was a moderate population in this district before the Roman Conquest. Of Roman remains there are none, save a few coins of doubtful authenticity found at Countisbury, which are supposed to have been scattered and buried by a resident clergyman at the close of the last century, with the avowed intention of "fogging" later antiquarians--surely the strangest "fourberie" ever indulged in by a reverend gentleman. All other evidence points to the fact that the Romans never occupied North Devon, though they may have held in temporary garrison one or other of the existing camps of the district.