THE GENOTYPICAL RESPONSE OF THE PLANT 335 
shifting dunes). At Vitemölle the type abounds on the stationary dunes 
some hundred metres from the beach, while it occurs but rarely on 
the outer dune bank of the beach, where the substratum is moving 
and not stable. The same phenomenon may be seen on the coast 
just north of Cimbrishamn. The inference that it is only the spe- 
cialized type of the shifting dune which is able to inhabit the moving 
dune banks of the beach is further strengthened by the examination of 
the Hieracium umbellatum population on certain points on the south- 
coast of Scania, viz. at Falsterbo and Ystad (list nos. 16 and 17). 
Although some of these individuals are procumbent, the structure of 
the leaves of these populations is that of the inland type. The type 
growing in the woods in close proximity to these beach localities also 
shows most resemblance to the inland type. The plant is extremely 
rare on the outer beach dune at Falsterbo. The case is the same at 
Ystad (Nybroan), where only 7 individuals were found on a 5-metre 
broad stretch of 300 metres’ length. For the sake of comparison it 
may be mentioned that a similar stretch at Löderup (in the Sandham- 
mar region), where the type of the shifting dune begins, an average of 
two individuals to the square metre were found growing on the outer 
beach-dune. It seems clear from this that the inland form is unable 
to populate the dune banks. It may also be suggested that the reason 
why a dune type has not become differentiated from the species-po- 
pulation in these localities lies in the fact that the woods harbouring 
the inland type run down to the beach. It is probable that an inces- 
sant intercrossing between the individuals which have succeeded in 
establishing themselves on the dune and the rest of the population is 
maintained, and thus the isolation of a specialized dune form is 
prevented. . 
The same inability of the inland type to populate the beach is 
found at Sofiero (on the Sound), where the type covers the Calluna 
bluffs, about 50 metres distant from the beach proper, and extends 
inwards through the woodland. In this locality not a single individual 
is found to have established itself on the sandy beach below the bluffs. 
The intimate relation between type and habitat is also apparent 
in the case of the so-called climatic varieties (BATESON 1913; NiLssoN- 
EHtLE,1914). The zonation of a species-population into coast types and 
an inland type, repeatedly discussed in the first part of this paper, 
furnishes an example of the same order, as does also the occurrence 
of different types on the east-coast and on the west-coast of Sweden 
(Atriplex sarcophyllum, Matricaria, Solanum etc.). It would seem to 
