Vegetation of the Island of Gothland. 263 
of marl in the underlying rock. Again, many lakes are partially 
surrounded by sandy drift, and swamp may have been due to 
the formation of an impervious layer in the sand by the 
precipitation of iron compounds (moor pan). It is quite probable 
that more than one, or even that all of these causes have operated 
to give rise to lakes and fens. 
The largest lake existing at the present time is Tingstade trask 
(see PI. IV) which is about 3 km. long and 2 km. across. Near the 
side where there is least vegetation its bottom seems to be formed 
of a very calcareous mud. The water is quite alkaline and appears 
to be shallow in all parts. A considerable portion is colonised by 
Phragmites, while islands of Cladium Mariscus are also seen. The 
Phragmites forms thick beds on the eastern side and we doubtless 
have here the first stage in fen formation. This lake adjoins a 
considerable tract of reclaimed fen on one side, known locally as 
“ myr,” while a few miles away in the neighbourhood of Martebo 
a very extensive marsh formerly existed which was probably part 
of the same series of fens. The Martebo marsh has now been 
drained and most of it is under cultivation, though here and there 
we find the remains of many small lakes, some of which are 
surrounded with interesting vegetation. 
I was able to examine the remains of one of the best of these 
small lakes, Binge trask. In July the site of the lake was marked 
by an extensive tract of dry powdery peat. In places there were 
patches of Phragmites, which seemed to be still growing vigorously, 
some plants of Cladium, almost dead, and a number of bushes of 
Salix repens. The relative vigour of the Phragmites was doubtless 
due to its deeper root-system. To one side of this tract was a 
thicket with dense vegetation apparently little affected by the 
drainage and probably still retaining most of its original features. 
Much of it was birch and willow scrub, but among it were patches 
of very wet spongy bog with the water table less than six inches 
below the surface of the ground. Since the dry bare peat almost 
certainly represents the last vestiges of the former lake and was in 
fact probably recently covered with water, it seemed curious that it 
should be quite dry and powdery, at the same time that the ground 
a few dozen yards away, which had been long above the surface of 
the lake and was thickly covered with vegetation, was absolutely 
saturated with water. 
The most probable explanation of this fact rests on the relative 
depths of the peat. Where the peat is shallow the water gets 
