THE SCANDINAVIAN MOUNTAINS. 91 
decomposition, been obliterated. An example of the latter case 
we have near Kararfvet, where the gneiss is decomposed, and 
the kernels of quartz, which are beautifully furrowed, remain 
from three to six lines above the surface of the gneiss. 
Where no evident furrows are apparent, I have still, with the 
assistance of long practice, been enabled, in most places, as 
well as on such rocks as rise above the surface of the earth, to 
decide on the direction of the furrows to a certainty within from 
three to ten degrees. 
For such purpose I have accurately observed,— 
Ist. The opposing side, to find out which part directly re- 
sisted the boulder-stream, which is generally apparent from the 
form of the rock, as well as by the large hollows which have 
been formed in the places wherein the pieces have been loosened. 
2nd. The lee-side; there the place is easily distinguished 
where the points still remain, and show that they formerly had 
been, as one says, to leeward ; as well as the direction in which 
pieces have been separated ; and 
3rd. The hollowness of the surface after the pieces have been 
broken off, by which it has often been possible to perceive plainly 
the direction in which these pieces have been broken loose. 
Experience and care could in this way produce very good 
observations, as well as give the direction of the normal furrows, 
by which can mostly be seen, if possibly they are side furrows, 
belonging to a height in the neighbourhood. 
On some occasions it has appeared to me, that, under such 
circumstances, the most certain indication is given by the oppo- 
sing side, which, for example, was the case with my observations 
around Norrtelje; on the other hand, in the tract of Medevi, 
the lee-side has shown more plainly where the boulder-flood 
has passed. In the latter place, however, one is not certain to 
within eight degrees, partly from this circumstance, that the 
rocks to a great degree are coarse-grained and brittle ; partly also 
that the whole tract is full of smaller steep rocky hills, which 
make the direction of the furrows in general nearly undetermi- 
nate. The same is the case from Sdderkoping to near Kalmar, 
and it appears to be uncertain if it be not the texture of the 
rocks which is the principal cause of the unevenness of the high 
land, as well as the consequent irregularity in the direction of 
the furrows. Directly to the south of the church of Déderhult, 
a well-furrowed slope, however, appears on the left of the road. 
