He goes on to show that the transition from plaitwork to knotwork took place _in Italy_ between 563 and 774. But is that not a proof of introduction into Italy, and not of its discovery there? I am rather disposed to think that partly through the adoption of the osier wattle in domestic architecture, partly through the employment of the tartan, the plait in all its intricacy was a much earlier product of the genius of the Celtic race. There is a pretty story in the life of an early Irish saint. He had been put at school, but could not learn. At last, sick of books, he ran away. He found a man at work with willow rods, weaving them to form the walls of a house he was building. He dipped them in water, and laced them in and out with wonderful neatness, patience, and dexterity. And the boy, looking on, marvelled at it all, took it to heart, and said to himself, “These osiers flip out; but when there are patience and skill combined, they can be made into the most exquisite patterns, and plaited together into a most solid screen. Why may not I be thus shaped, if I allow myself to be bent, and am docile in my master’s hands?” So he went back to school. CHAPTER IV. CORNISH CASTLES The ancient camps--Their kinds--1. Rectangular, Roman--2. The Saxon burh--3. The Celtic circular or oval camp--The _lis_ and the _dun_--4. Stone fortresses--Heroic legends in Ireland--The Firbolgs--5. The stone castle with mortar, Norman--No good examples. Anyone with a very little experience can at once “spot” a camp or castle by the appearance from a distance of a hill or headland; and the traveller in Devon and Cornwall will pass scores of them, as he will see by his Ordnance Survey Map, without giving much attention to them, without supposing that they can be of great interest, unless his attention has been previously directed to the subject. It is a pity that anyone should go through a country which may really be said to make ancient camps and castles its speciality and not know something about them. [Illustration: LAUNCESTON] Of hill castles or camps there are several kinds:-- 1. Those that are rectangular or approximately so, and which have been attributed to the Romans. Of these in Cornwall there are but few. Tregear, near Bodmin, and Bossens, in S. Erth, have yielded Roman coins and relics of pottery; but whether actually Roman or Romano-British remains undecided. 2. There are those which consist of a tump or mound, sometimes wholly artificial, usually natural, and adapted by art, and in connection with this is a bass-court, usually, but not universally, quadrilateral. This was the Saxon type of _burh_; it was also that of the Merovingian. The classic passage descriptive of these is in the _Life of S. John of Terouanne_, written in the eleventh century:--