234 THE FORMATION 
in indentity. As in the case of polygenesis, the relatively small number of 
typically distinct habitats makes it clear that two different species of wide 
distribution, bearing to each other the relations of xerophyte to mesophyte, 
of hydrophyte to mesophyte, or of poophyte to hylophyte, might often find 
themselves in reciprocal situations, with the result that they would give rise 
to the same new form. The final proof of the polyphylesis of species is 
afforded by the experiments of DeVries in mutation. DeVries found that 
Oenothera nanella arose from O. Lamarckiana, O. laevifolia, and O. scintil- 
lans; Oenothera scintillans arose from O. lata and O. Lamarckiana; Oeno- 
thera rubrinervis from O. Lamarckiana, O. laevifolia, O. lata, O. oblonga, 
O. nanella, and O. scintillans, etc. Whatever may be the rank assigned to 
these mutations, whether form, variety, or species, there can be no question 
of their polyphyletic origin, nor, in consequence of the connection of 
mutations with variations through such inconstant forms as O. scintillans, 
O. elliptica, and O. sublinearis, of the possibility of polyphylesis in any 
two distinct though related species or genera. 
KINDS OF INVASION 
282. Continuous and intermittent invasion. With respect to the fre- 
quency of migration, we may distinguish invasion as continuous, or 
internuttent. Continuous invasion, which is indeed usually mutual, occurs 
between contiguous formations of more or less similar character, in’ which 
there is an annual movement from one into the other, and at the same time 
a forward movement through each, resulting from the invaders established 
the preceding year. By far the greater amount of invasion is of this sort, 
as many readily be seen from the fact that migration varies inversely as the 
distance, and ecesis may decrease even more rapidly than the distance 
increases. The significant feature of continuous invasion is that an outpost 
may be reinforced every year, thus making probable the establishment of new 
outposts from this as a center, and the ultimate extension of the species 
over a wide area. The comparatively short distance and the regular alter- 
nation of migration and ecesis render invasion of this sort very effective. 
An excellent illustration of this is seen in transition areas and regions, which 
are due directly to continuous and usually to mutual invasion. Intermittent 
invasion results commonly from distant carriage, though it may occur very 
rarely between dissimilar adiacent formations, when a temporary swing in 
the physical factors makes ecesis possible for a time. It is characterized by 
the fact that the succession of factors which have brought about the in- 
vasion is more or less accidental and may never recur. Intermittent invasion 
is relatively rare, and from the small number of disseminules affected, it is 
of little importance in modifying vegetation quantitatively. On the other 
