Igto] DACHNOWSKI—ARID HABITATS 327 
more centrally on the island, and that is the relative absence of toxic 
ingredients in the water and in the upper limits of the substratum. 
The transpiration quantity of such hydrophytic plants is correspond- 
ingly less when grown in bog water from the central habitat. The low 
rate of transpiration can be increased again with the addition of adsorb- 
ents, or with the dilution of the bog water to a point no longer fatal in 
its effects. The peat substratum of the border zone permits, on that 
account, a dense luxuriant growth of hydrophytes, and neither a smaller 
percentage of free oxygen in the water, nor the absence of a mineral 
soil, nor the incoherency of the substratum affords precarious condi- 
tions for growth. 
How much of the salts dissolved in the lake water is retained by 
absorption in the humus soils along the margin of the bog has not 
been determined. There is usually a slight difference between the 
total mineral content of bog water and that of the lake water adjoin- 
ing. The figures obtained, however, are not high. The osmotic 
Pressure and the acidity have been found to be the same for both 
habitats. The stress laid by various authors upon the relation of 
these two factors to plant societies in bogs, in so far at least as this 
region is concerned, will not hold. They are not the factors in the 
selection and distribution of species. 
Water culture experiments are here considered as giving a 
relatively normal indication of conditions obtained in the bog, 
since the bog island itself is merely a water culture on a larger scale, 
with the mineral soil at a depth of more than 1o™ from the surface 
vegetation. In the bog habitat of the island, however, the ratio 
between the amounts of water absorbed and transpired is never 
constant. It varies most during the growing season, and yet it must 
be always more than unity, if the plants are to survive the occasional 
periods of extreme physiological drought during the summer and 
autumn months. This is evident in the scattered small trees of 
Acer rubrum, Rhus vernix, and in Cephalanthus occidentalis, Osmunda 
cinnamomea, and other plants from the adjoining relatively forested 
maple-alder zone. The stunted growth of these plants, their numer- 
ous dead branches, and ragged crown of foliage exhibit to a marked 
degree the physiological action of the substratum. The resistance 
offered by the invading mesophytic vegetation is, indeed, but slightly 
