272 



FKIDTJOF NANSEN. 



M.-N. Kl. 



italics), give the most perfect agreement with the computed values, an(i the 

 differences are on the whole no greater in the cases of the greatest heights 

 than at the lower ones, which indicates that the formula is fairly correct. 



The observed heiglits of the Tapes-line are in most cases somewhat 

 lower, often between i and 2 metres lower, than the computed ones. This 

 mav be due to the fact that the terraces have not, as a rule, been built 

 quite up to mean water-level, but ma\' have stood one or two metres 

 below it, even in cases where shore pebbles have been found on their 

 surface. It is also obvious that when a terrace is raised above sea-level, 

 its surface ma}' sink more or less, owing to compression, and to the loss 

 of water contained in its layers, Szc. Our formula is, however, based 01 

 the measurements of shore-ledges cut in solid rock, where this has not 

 been the case, and which may have been formed slightly above mean 

 sea-level. 



There are some observations of the upper limit of submergence and 

 the Tapes-line by Kaldhol in Romsdal which are not given in our table 

 because they showed too great disagreements. Introduced in a map, 

 these observations exhibit great irregularities in their distribution, and 

 this is also to some extent the case with Kaldhol's observations in Xord- 

 mor. Lower values of the heights, both of the upper limit of sub- 

 mergence and of the Tapes-line, very often occur a good deal farther 

 inland than higher values, and it seems very difficult to draw the probable 

 isobases according to these observations. It seems to me to be probable 



