HEADWARD EROSION OF HANGING VALLEYS 423 



and Lodalskaupe are good examples in the Loen region of fragments still 

 surviving this form of attack. Headward erosion by this process seems 

 only possible in a valley if there be an axial wealaiess to work along. 

 Had the process gone a little farther the head wall must have broken 

 down and a high level cross-canyon resulted between the valleys of Loen 

 and Stryn. The interest of this possibility is in the abundance of cross- 

 fiords and valleys in west Norway and the difficulty of explaining them 

 on a scheme of pure ice erosion. In the Bergen region, where these are 

 very systematically developed at right angles to the line of continental 

 glaciation, the geologic map shows the cross-valleys are typically the out- 

 crops of sedimentaries and other apparently weaker rocks among the 

 gneisses. 



The easiest botner to visit in Norway lies in the outskirts of Bergen, 

 overhanging the Svartediget precisely on the line of an outcrop of fine- 

 grained quartzose gneiss mapped by Doctor Kolderup (1902) as far as 

 the opposite wall of the valley, and declared by him certainly a sediment. 

 The layers stand vertical at the lip of the botner and are axial to it. 

 Bergen itself lies in a ten-mile-long valley a thousand feet deep in the 

 upland, but curving around in a crescent along an outcrop of Silurian 

 rocks starting from and returning to the sea.. Such valleys, when their 

 floors are under water, are as much fiords as those that run from the land 

 seaward and as characteristic. That these have sometime ended in a 

 botner amphitheater of rock is an assumption; but the valleys described 

 do, and a continuation of activities that have not entirely ceased in these 

 valleys would certainly make of Praestedal and Tjugedal cross-valleys 

 such as abound. 



It is reasonable to suppose that Tjugedal may once have served as a 

 partial outlet valley for the ice of the Skaala outlier of the Jostedal 

 glacier, much as the valleys over Brengsnaes and Hogrending seem to do 

 today, but the valley of that day must have lacked the botner head that 

 characterizes it now, and its length Avas less by all that headward erosion 

 has given it since. 



The corrosive attitude of the ice as it lies on these uplands is nowhere 

 better exemplified than in the moimtain on the border of Stryn lake, a 

 little north of the Tjugedal. As seen from Hjelle, across the lake, the 

 snow lies sunken into the mountain top as if in the crater of a volcano— 

 an appearance often seen in the arctic regions. The details of this sort 

 of erosion have been cited by Eussell (1887) from Lorange. A similar 

 aspect is present in the Hellsaeterbrae (HI, figure 2) on Loen lake 

 (figure 1, plate 41), which falls in huge masses from a cliff summit 

 4,000 feet above sealevel into two snow cones. The upper one is sinking 



