PHYSICAL GEOGRAPHY 263 



land 1 and in places where the water, for some reason or other, 

 cannot drain off. Here the soil is usually much thinner than 

 in the home-fields hut nevertheless it lias a very great power of 

 ahsorbtion during the process of surface-evaporation ; the greater 

 part of these knolls consists of niohel la-clay. On an entirely hare 

 clay-Hat no knolls are found, and the separate cakes in the ,,rude- 

 mark" rise but slightly in the middle during spring, hut not until 

 they are plant-covered and clad with greensward do they bulge and 

 retain their convexity. On a closer investigation it is probable that 

 several transitional forms will be found between "rudemarks" and 

 knolly flats, but investigators have not as yet made this matter a 

 subject for study. Where similar plant- formations (heather and 

 "moar") occur upon slopes or mountain sides upon which the water 

 easily finds outlets, knolly flats are absent, and k 'rudemarks" are 

 never found in such localities. Upon mountain-sides small eminences 

 of various kinds may sometimes be observed which are due to under- 

 lying stones or ridged mud-flows, etc., but never "thui'ur" proper. 

 Nowhere have I seen any indications of "rudemarks" or u thufur- 

 marks" being at all connected with mud-flow phenomena. 



On the plateau peculiar knolls of usually large size are found 

 which the inhabitants of the different districts call "riistir," "haugar" 

 or u dys.'' These knolls are, as a rule, of irregular, oblong form, 

 and are bare of vegetation on the top, where they consist only of 

 humus and clay. It can be seen that the ground has bulged and the 

 knolls are cracked at the top. In some places there are evident 

 proofs of this being the case: bands of the original soil are seen to 

 lie upon the top of the knoll while the clay and gravel within have 

 poured out of the cracks between the bands. In the neighbourhood 

 of Ullsvatn on Tvidaegra, at a height of 460 metres above sea 

 level, I investigated such knolls in 1898; they were 1--1 1 /* metres 

 in height, 15 20 metres in length and 8 10 metres in breadth. 

 The sides were covered with heather, but Cyperacese grew in the 

 intervening spaces. Similar knolls occur in several other places in 

 the interior of Iceland, but only in one more place did I observe 



1 Dr. H.Jo us son describes a heather-covered "rudemark ' where the cracks 

 between the cakes were covered with Grimmici hypnoides, but the cakes themselves 

 with Calluna and Empctrum. This peculiar vegetation the brownish heather- 

 vegetation divided into numerous polygons by the greyish moss-bands - - occurred 

 only upon level ground (Botanisk Tidsskrift. XXVII, 1905, pp. 43 and 44). 



