264 THORODDSEN 



them to be as large as those at Ulfsvatn, viz. in East Iceland on 

 the plateau south-east of Sna^fell, at a height of 690 metres above 

 sea-level. This was in 1894, but the inclemency of the weather pre- 

 vented me from making a halt at the place. Here, also, several of 

 the knolls had burst and discharged a large quantity of clay-soil. 

 This form of knoll occurs especially at the edge of the bogs where 

 the foot of the knoll stands in water during the melting of the 

 snow: where the supply of water is too abundant these knolls do 

 not occur. It is a fact peculiar to all knolls, both in the lowlands 

 and on the plateau, that they consist of mohella-soil and humus 

 intermixed with clay; this is true also of the "rudemarks" with the 

 exception that more gravel and stones are found in them than in 

 the plant-covered ground. As mentioned above, there is no direct 

 connection between the mud-flow phenomena and the "rudemarks" 

 and knolls, but it may happen that a flat with either of these sur- 

 face-forms is afterwards affected by mud-flows; this is then a se- 

 condary phenomenon. Where subsoil-ice is absent, as for instance 

 in the neighbourhood of the hot springs which are scattered in 

 hundreds all over the country, I do not remember to have seen 

 "rudemarks" or "thufur" proper. Upon the plateau itself "rude- 

 marks" are rather rare, but, on the other hand, various mud-flow 

 phenomena are common; thus, stones are often seen arranged in 

 bands on the slopes, and in various other ways which, however, I 

 regret to say, have not as yet been more closely investigated. On 

 basaltic plateaus covered with coarse basalt-gravel and blocks of rock, 

 patches of clay are now and then seen which appear, in the 

 middle, to have thrust up from below and spread out slightly on 

 all sides. 



Besides these knolls there are many other similar eminences 

 in the gravel-covered ground which are called knolls, but are due 

 to quite different causes and have a different origin. Sometimes 

 nodules on the greensward are due to the form of the substratum, 

 as where a thin layer of soil occurs upon lava, or upon a stony 

 and rugged bottom. In blown sand, knolls are formed around roots 

 of Ely mus arenarins, Halianthus peploides and other plants. Where 

 the ground- water stands very high, as in many extensive boggy 

 tracts, so that the entire surface freezes into an ice-cake, only small 

 knolls of organic material are seen, formed of moss and Carices. 

 On hard grassland, on gravelly flats and similar places, small 



