406 Proceedings of Royal Society of Edinburgh. [sess. 
conspicuous “bottom-moraine,” consisting of the shells of 
Anodonta , etc., which may he scattered over the ground in 
thousands ; stones are polished, and the ice, striking against the 
trees, causes the rifts and wounds referred to above. Many trees, 
on the prominent points bordering the lakes, are killed by these 
heaps of ice, which are piled up year after year on the shores of 
our lakes (see fig. 1). We may also mention that the ice-slabs in 
spring often break great apertures in the closed stocks of Phragmites 
and Scirpus , detaching large patches of rhizomes a square metre 
(over a square yard) in extent and throwing them on shore ; the 
ice may in the course of a few hours cover over a peaty shore 
with sand, or cover a sandy beach with peat-forming material. 
With reference to the temperature of the Danish lakes, it is to 
be regretted that the observations are rather deficient. Still, it 
may be stated generally that the temperature of the water follows 
very closely the changes in the temperature of the air. Having 
exemplified this statement in my Plankton paper, I shall here 
only remark that the surface waters of our lakes are generally 
very warm in summer, often attaining a temperature as high as 
23° C. (73° F.), and in hot summers the water may maintain a 
temperature of 20° to 23° C. (68° to 73° F.) for more than a 
month : it is very rarely that the surface temperature in summer 
falls below 16° C. (61° F.). Almost every winter most of the 
lakes are frozen over, though the length of the period during 
which they are ice-bound varies greatly in different years, but 
never exceeds more than about four months. The observations I 
have made show that the lakes are usually frozen for one or two 
months, generally from about 15th January to 15th March, but 
exceptionally they may not be frozen at all. As we have often a 
short spell of frost in November and December, followed by thaw, 
usually followed again by the customary long period of frost in 
January to March, the smaller lakes may have two ice-bound 
periods — a short one in December and a longer one in January to 
March, but in the larger and deeper lakes only the latter period 
prevails. As most of our lakes resemble each other as regards 
height above the sea, latitude, depth, and form of basin, it will be 
understood that they vary little in temperature. It may generally 
be said that the deeper and narrower the lake and the steeper the 
