irregularly circular, flat-bottomed vessel, from the cave of Nabrigas, on which the finger-marks of the hand that moulded the clay are still clearly distinguishable on the baked earthenware.  That is the great merit of pottery, viewed as an historical document; it retains its shape and peculiarities unaltered through countless centuries, for the future edification of unborn antiquaries. Litera scripta manet, and so does baked pottery.  The hand itself that formed that rude bowl has long since mouldered away, flesh and bone alike, into the soil around it; but the print of its fingers, indelibly fixed by fire into the hardened clay, remains for us still to tell the story of that early triumph of nascent keramics.

The relics of palaeolithic pottery are, however, so very fragmentary, and the circumstances under which they have been discovered so extremely doubtful, that many cautious and sceptical antiquarians will even now have nothing to say to the suspected impostors.  Among the remains of the newer Stone Age, on the other hand, comparatively abundant keramic specimens have been unearthed, without doubt or cavil, from the long barrows—­the burial-places of the early Mongoloid race, now represented by the Finns and Lapps, which occupied the whole of Western Europe before the advent of the Aryan vanguard.  One of the best bits is a curious wide-mouthed, semi-globular bowl from Norton Bavant, in Wiltshire, whose singular shape suggests almost immediately the idea that it must at least have been based, if not actually modelled, upon a human skull.  Its rim is rough and quite irregular, and there is no trace of ornamentation of any sort; a fact quite in accordance with all the other facts we know about the men of the newer Stone Age, who were far less artistic and aesthetic in every way than their ruder predecessors of the interglacial epoch.

Ornamentation, when it does begin to appear, arises at first in a strictly practical and unintentional manner.  Later examples elsewhere show us by analogy how it first came into existence.  The Indians of the Ohio seem to have modelled their pottery in bags or nettings made of coarse thread or twisted bark.  Those of the Mississippi moulded them in baskets of willow or splints.  When the moist clay thus shaped and marked by the indentations of the mould was baked in the kiln, it of course retained the pretty dappling it received from the interlaced and woven thrums, which were burnt off in the process of firing.  Thus a rude sort of natural diaper ornament was set up, to which the eye soon became accustomed, and which it learned to regard as necessary for beauty.  Hence, wherever newer and more improved methods of modelling came into use, there would arise an instinctive tendency on the part of the early potter to imitate the familiar marking by artificial means.  Dr. Klemm long ago pointed out that the oldest German fictile vases have an ornamentation in which plaiting is imitated by incised lines.  ’What was no longer wanted as a necessity,’ he says, ’was kept up as an ornament alone.’