STUDIES ON THE VEGETATION OF ICELAND 139 
species belongs, it applies to all species that there is 
one class of moisture in which the species attains its 
highest F.-percentage and shortest distance between the 
individuals, and outside which the F.-percentage de- 
creases and the distance between the individuals in- 
creases whether we go up or down the scale of moisture. 
The distribution of the species in the scale may afford ground 
for the setting up of a series of types characterised by the magnitude 
of the F.-percentage, the position of the maximum in the scale, the 
number of classes in which the species occurs etc. etc., and in time 
it will be necessary to introduce a terminology in order to charac- 
terise briefly the relations of a species within an area. At the present 
time, while such investigations are still in their inception, there is 
no reason to set up such a system, especially since a good deal of 
material would be requisite for such a purpose. This part of the 
investigation must therefore be left until a later period. In this 
connection it will suffice, as was the main object of our investiga- 
tion, to establish the fact that a species is closely identified 
with a definite degree of moisture of the soil. If there is 
any change in the degree of moisture, no matter in what direction, 
the F.-percentage of the species will change simultaneously, and the 
greater the change in the degree of moisture, the greater, too, will be 
the change in the F.-percentage, until such conditions of moisture 
are reached as entirely exclude the species. The species reacts 
identically to changes in moisture wherever it occurs. 
The table shows how markedly this is the case in the three 
localities Bjork, Lyngdalur, and Arnarvatnsheidi therein indicated. 
These three localities have been selected at random from the areas 
of distribution of the species discussed, and there is no reason to 
suppose that an investigation in other localities under the same 
external conditions would give a picture of the relation of the species 
concerned to the degree of moisture essentially different from that 
shown in the table. Greater certainty might of course be gained by 
an increased number of investigations, in that the influence on the 
magnitude of the F.-percentage of accidental factors, i. e. factors 
not determined by the degree of moisture, would be precluded or 
diminished. 
The distribution of a species in a scale of external 
factors is just as constant and »good« a character ina 
species as any morphological or anatomical character. 
