l62 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 59 



saga and of Landnamabok that happens to survive. In merely Ice 

 landic records and stories we have no right to expect such informa 

 tion unless from Islendingabok, which is lost. The situation is a 

 natural one. If Hauksbook had happened to be destroyed the date 

 must have been carried along further still, and that would yet prove 

 nothing, except that our evidence would be less in volume and force. 



(5) The second and later principal narrative of voyages to Wineland (the 

 Flateyjarbook s &quot; Grcenlendinga-pattr &quot;) gives a very different account of the 

 discovery, by another, and likewise of the later voyages thither. 



That is true. The natural course of development is for a later 

 version to elaborate hints and w^eave stories about names, filling in any 

 floating legendary data which may come to hand. This is especially 

 true in a decadent artificial period, even at its beginning. The Flatey- 

 book narrative is not unique in its method and qualities, but is a very 

 bad example. 



(6) The first of the two sagas, and the one which is regarded as more to be 

 relied on, contains scarcely a single feature that is not wholly or in part 

 mythical or borrowed from elsewhere ; both sagas have an air of romance. 



This is far from the case, for Helluland, Markland, Kiallarness, are 

 all admitted by Nansen to exist. Straumey, Straumfiord, the moun 

 tains, Hop, the seal headland are veritable. The courses around the 

 great ness into and out of the Gulf are accurately and carefully given, 

 Biarney is true to fact. The Wonderstrands are the typical American 

 coast line found on no other Atlantic shore of which any Icelander 

 short of the fifteenth century would be likely even to hear. The 

 Indians, products, climate, and breeding places are authentic. The 

 Uniped was probably an Eskimo in his kayak. The Greenland part 

 of the tale has many embroideries of fancy. The/re are divers ballads 

 turned to prose attached to the exploring narrative ; but they do not 

 invalidate or obscure it. The saga-man might have chosen ad 

 libitum magical cats and dog-footed monsters, the roc-phoenix and 

 the island of unending laughter, holy white-furred hermits and angels 

 who waited on the table, Judas and his hounding devils, the sea- 

 monster that took the saint a-traveling on its back, the isle of women, 

 the pool of youth, and the river of death. His Celtic sources (as 

 supposed) would have done this. Why did he stick to the facts in 

 stead? Surely because he was not following Celtic models, but 

 relating facts. 



(7) Even among the Greeks of antiquity we find myths of fortunate isles far 

 in the western ocean, with the two characteristic features of Wineland, the 

 wine and the wheat. 



