116 BOTANICAL GAZETTE [AUGUST 
During wet periods bog plants show a tendency to move shore- 
ward. This is due to local conditions rather than to historic 
factors. Such movement is only temporary. 
In the summer of 1909 the island in the center of Dead Lake 
had begun to reappear. During the late summer the ground 
appeared in the higher parts, showing that the water level was get- - 
ting lower. In less than ten years the island disappeared and in 
part reappeared, which indicates that these wet and dry periods 
may be of short duration. This is contrary to the belief, often 
expressed, that in southern Michigan they last about thirty years 
each. 
The effect of the dry periods upon vegetation is also very 
marked. The relation of such periods to plant succession has 
been emphasized by Davis, who believes that they offer an 
explanation of the xerophytic structures of bog plants (9, p. 160). 
During such periods the surface layers of the peat become exceed- 
ingly dry; this may extend to a depth of several feet. Fires 
which reduce the surface several feet are of common occurrence. 
The effect of fires upon plant succession has been given by 
PENNINGTON (13). : 
During this time there is very little water available in these 
surface layers for plants, and, as it will be shown later that the 
movement of water in peat is slow, it can easily be seen that the 
‘habitat is very dry, even omitting the usual factors of humus 
acid, low temperature, bog toxins, etc. The following table is 
given by Davis taken from WARRINGTON (19). 
TABLE II 
PARTS OF WATER PER 100 PARTS OF DRY SOIL 
Type of soil When plants wilted cared gag 
Coarse Sandy 6.0.2. ce es 1.15 
Sandy panien..; .. 2... sc. 4.6 .00 
ine humus sand... 6.6.55 6.2 3-98 
Sa WO Soe es 7.8 5.74 
RUIN, 5s eS 55 ck cs vs 9.8 5.20 
Bec ek ip akecss 49.7 42.30 
Davis says: “If these results are correctly reported, it appears that peat 
may appear very wet and yet contain no water for plants growing in it, so that 
