THE SPECIES AND THE VARIETY AS ECOLOGICAL UNITS 105 
praecox has been found to give a progeny which does not show any 
decrease in fertility . In the further discussion of this complex other 
well defined species of the genus must be taken into consideration. 
Hybrids between the above-discussed »species» and these latter ones 
are occasionally found on the east coast and along the Sound, although 
they are usually very difficult to trace. On the west coast, north of the 
Sound, these hybrids however become so frequent that it is sometimes 
difficult to find representatives of the above-mentioned western type 
except in certain places, where other species of the genus, such as 
Atriplex Babingtoni and A. latifolium, grow but sparsely. Here, then, 
the typicalness of the type depends on the number of other species 
present on the same spot. When these latter are present in sufficient 
number the »swamping» of the variety in question is very likely to 
occur. As A. Babingtoni and A. latifolium are more abundant on the 
west coast than on the east coast there are fewer chances of hybridi- 
zation between A. praecox and these species, but the distinctness of 
the latter variety cannot be due to this circumstance alone, since 
places are known where all the species occur indiscriminately without 
inducing any swamping of A. praecox. It has already been stated that 
the types under consideration flower very early compared with other 
species of the genus, and that this is particularly true of A. praecox 
and A. longipes, while the flowering of the western variety falls some- 
what later. That this circumstance accounts for the differences in the 
degree of hybridization between the eastern and western varieties is 
more than likely. The western variety flowers late enough to find 
particularly early individuals of other species flowering at the same 
time. Hybrids are thus formed yearly, and back-crossing might en- 
sure the persistence of hybridogeneous forms even if the primary 
bastards are less vital than the parents. 
A natural circumscription of the ecospecies is thus seen to exist 
to the east and south, while it is often swamped on the west coast. 
Similar peculiarities in the behaviour of a great number of plants 
and animals are well known. The species may be well defined within 
a certain area and the systematist has no difficulty in tracing it; then 
suddenly a nearly related species (or several of them) is found to 
intrude upon the area at certain points, causing great variation in 
type and unsurmountable difficulties to the systematist. These inter- 
1 Further information with regard to the experimental study will be given 
in publications to appear later. 
