NORSEMEN IN NORTH AMERICA— BR0NDSTED 387 



splayed. This small window 4 must have been the only means of 

 watching the region north of the tower for the inhabitants of the first- 

 floor chamber. 



Next we find traces on the wall of an inner stairway running from 

 the first to the second story : six rather small, rectangular step sockets 

 in an oblique line (see fig. 6 and pi. 7, lower). Then follows window 

 No. 1, not doubly splayed, just left of the fireplace (see pi. 5, upper) . 



The second-floor chamber had three small windows (5-7), only 

 slightly splayed inward. Between windows 6-7 are niches 5 and 6. 



None of the essential architectural details — floors, windows, fire- 

 place or stairs — can reasonably be regarded as later additions to the 

 first building. To me the tower seems to be an original whole. 



Now where shall we look for a similar primitive building fashion ? 

 Not in Iceland, where tlie people built with earth and wood. Nor 

 in Greenland, where stone was used, it is true, but generally without 

 mortar. But we do find a similar technique in medieval buildings in 

 the Scottish isles. And yet, a comparison with Newport Tower shows 

 that its masonry is a trifle more irregular and primitive. 



However, this kind of masonry cannot be used as any chronological 

 criterion, or the consistency and appearance of the mortar either. 

 Both phenomena are, so to say, timeless ; we see how the same teclmical 

 mode of building held out among the poor populations of the Scottish 

 isles right up to tlie present day. Thus neither the masonry nor the 

 mortar of Newport Tower lend themselves for use as chronological 

 evidence. 



In its style Newport Tower undoubtedly contains medieval features. 

 The pillars themselves and the flat arches of the arcades and over 

 the fireplace are typically Eomanesque elements. The double splay 

 in some of the windows is also a common Eomanesque feature. 



Concerning the fireplace, tlie question has been asked whether the 

 lack of a chimney is not an antique element. Perhaps it is, but this 

 chimneyless vent, conditioned primarily as it is by the cylindrical 

 form of the tower, subsidiarily by the secondary use of the tower as a 

 windmill, cannot be taken as a chronological foothold. For even if 

 the builders were ever so familiar with chimneys, practical con- 

 siderations may have prevented the inclusion of one here. 



Thus there remain, as typically Romanesque architectural details, 

 the pillars, the arches, and the double splay. These medievalisms are 

 so conspicuous that, if the tower were in Europe, dating it to the 

 Middle Ages would probably meet with no protest. 



For what purpose was Newport Tower built? Three answers to 

 this question have been to the fore : Windmill, watchtower, and church. 

 First let us take the windmill theory. 



