528 
similar to the one that a species had left, there would be 
no advantage in modification of structure, and none could 
be brought about by Natural Selection. If the locality were 
favourable to the species, its members would increase, but in 
the absence of their normal enemies they would not suffer 
from one of the greatest disabilities in the struggle for 
existence, and the element of selection by death would be 
wanting. If, on the other hand, the conditions were 
unfavourable, they would not at first have their normal 
fertility, and the output of life would not be forthcoming. 
Thus, in all cases, one or other of the conditions of Natural 
Selection would be wanting, and consequently any modi 
fication which might be necessary could not be brought 
about by Natural Selection. 
We have already considered the case in which the im 
migrant species masters or drives out a native population ; 
and we have seen that the phenomenon could not have 
been brought about by Natural Selection. 
What has just been said will apply to those cases in 
which settlement in a new locality takes place under the 
most favourable conditions to the emigrant. Hut the 
difficulties are much increased when we consider those 
“ means of dispersal,” on which apparently much reliance 
is placed. We have to consider the danger accompanying 
the transit ; the probability that the animal or plant will 
not be able to find a footing in the new sphere, and the 
difficulties which will accompany its adaptation to new 
conditions, if it should be fortunate enough to pass suc 
cessfully through these initial difficulties. 
It is said of mammals that they might be dispersed 
by being accidentally carried out to sea, and in this way 
succeed in reaching islands many miles distant. In the 
arctic regions icebergs originate in glaciers, which descend 
into the sea, and often bear masses of gravel, earth, and