THE SCANDINAVIAN MOUNTAINS. 123 
northern coast of the Gulf of Finland should in the same man- 
ner be the lee-side; the southern coast, the opposing side. 
A good map of North America would serve to show the direc- 
tion of the furrows in that country. 
g- In conclusion, we must moreover, as a proof of the often- 
mentioned assumption, adduce the considerable difference which 
is found in the direction of the furrows on the tops of Omberg, 
Taberg, Billingen, Kinnekulle, and Stora Tiveden (east of Finn- 
rédja church), as well as the furrows at the eastern foot of these 
mountains. On the summits the flood has followed the general 
direction; whilst, on the contrary, near the foot of the moun- 
tain, they follow the irregularities of the country in order to flow 
forward between them. 
The true directions can, however, only be discovered at some 
future time. 
§ 12. 
Of Sand Hills. 
The question of the formation of sand-banks is apparently 
connected with the subject of the foregoing section, and will re- 
ceive its explanation when the main question is answered. In 
the mean time I have directed my attention to them, as their ex- 
istence had previously excited so much inquiry. 
In the next place I must remark, that they are not correctly 
laid down on the maps of Baron Hermelius and M. af Forsell. 
They are seldom found of such great extent as shown on the 
maps; they are composed of several small ridges, which appear 
to have no other connexion between them than that the public 
road goes over the whole of them, for this reason only, that our 
forefathers, in road-making, appear in preference to have chosen 
the elevated lands. 
The lesser hills, on the other hand, are situated oftenest on the 
lee-side of some rock; but they are sometimes found without 
any such protection. 
They have mostly the same direction as the furrows in the 
same tract, and seem readily to have fixed themselves where 
these have diverged or converged; but, in general, it appears 
equally difficult to discover any rule for their formation as to 
explain the cause of the snow-drifts after a storm even when the 
wind has been unvarying. The least unevenness on an other- 
