STUDIES ON THE VEGETATION OF ICELAND 155 
ciable change, are indifferent. Passing upwards from the level of 
the sea to the snow-line in a mountain district, we find the same 
changes. 
In Iceland the same holds good for the variation according to 
altitude: The H percentage remains unchanged throughout the al- 
titudinal zones, the arctic life-form Ch increases rapidly, while the 
southern life-forms, Pt, Ph, G, HH, and Th, decrease in the same 
proportion. 
These changes in the life-form spectra are connected with, or 
caused by, a decrease in the warm temperature of the summer and 
an increase in the cold temperature of the winter. The summer 
temperature grows lower and the summer of shorter duration, while, 
on the other hand, the winter grows longer and more severe the 
further northward we go. 
If we calculate the biological spectra for the Icelandic types of 
vegetation and compare the spectra for the types on snow-bare soil, 
on soil with a normal snow-covering, and on soil with a constant 
snow-covering, we get the same variations as above on passing from 
arctic towards temperate regions or from the snow-line towards 
the level of the sea. 
"The result will be the same whether we base our calculations 
on the number of species or on the number of individuals, i. e. on 
the frequency sum. 
There is no reason to believe but that, in this case loo, it is 
the temperature conditions which determine the occurrence and 
development of the life-forms. The temperature conditions favourable 
to the artic species are due to the fact that the snow is blown away 
so that the cold can act with its full force on the vegetation with 
the result that southern types cannot survive, while arctic types 
thrive. Where the snow-covering is deep and constant, the case is 
reversed: here the environment will be unfavourable to arctic but 
favourable to temperate types because the snow-covering will prevent 
the extreme variations in temperature from reaching the vegetation. 
Hence, under conditions where the external factors 
are not easily observable, the individual life-forms may 
with full justice be employed, as above, as indicators of 
environment. Thus, in Iceland, many Ch will indicate 
severe cold, many Ph, K, and Th relatively favourable 
temperature conditions. 
From a scientific point of view, as a means of checking life- 
