106 GOTE TURESSON 
mediate areas harbouring mixed types would seem to be best suited - 
for the study of the differentiation of the species from an ecological 
point of view, as are also the zones of contact between varieties 
belonging to the same ecospecies to be discussed in the following. 
The case so often seen where a species at its boundary-line hybridizes 
and gets swamped by other species in spite of the fact that these same 
species may occur side by side within the »natural» area of that spe- 
cies without any swamping, is most likely due to the different geno- 
typical constitution of the species near the border line. This has been 
shown to be the case in the above-mentioned western Atriplex variety, 
and a certain peculiarity, which distinguishes this variety from all 
other known types of the ecospecies, viz. the late flowering, has been 
pointed out as a reasonable cause of the mixing with other species 
on that coast. Other cases of hybridization in the zone of contact 
between different species will probably also be found to be due to 
changes in the genotypical construction of the species at their limits 
of distribution. 
The factors likely to bring about changes in the genotypical con- 
struction of species at different points of their geographical distribu- 
tion have now to be considered. In order to evade the complications 
arising when parts of a plant population become geographically isola- 
ted from the bulk of the population, only such plants as combine an 
extensive and uninterrupted range with an extraordinary frequency 
have been investigated. Certain Atriplex species, as for instance the 
one discussed in the above and known under the different names of 
A. praecox, A. longipes and A. hastifolium, comply with this demand 
fairly well. The Atriplex ecospecies just discussed occurs and has been 
followed along the coast from Stockholm round southern Sweden up 
to the middle part of Bohuslän, on the west coast. Cultivations have 
shown that the hereditary variation within each of the three above- 
mentioned varieties is by no means slight, and different forms be- 
longing to these varieties have been described and named by systema- 
tists. When well acquainted with the material one has no great diffi- 
culty in stating from which of the three coast strips a certain form 
has been collected. For in spite of the considerable variation within 
each of the three varieties there are certain characteristics which are 
common to all the forms of the particular variety. When therefore 
it is found that the ecospecies is split into three well-defined varieties, 
each occupying different but distinct stretches of a continuous strip 
of coast, the question becomes pressing whether or not climatic and 
