OF SOME SPECIES AT THE EDGES OF THEIR RANGES 45 
and its seedlings are taking possession of many cut-over woods. 
But its terminus is perfectly sharp; not a single individual has 
been found further northward, although a constant lookout has 
been kept for it. The increase in available habitats is giving it 
a continuous range, whereas formerly it occupied only isolated 
stations in the crevices like Silene rotundifolia. This change 
greatly increases the chances of successful reproduction and the 
plant gives every evidence of spreading northward. But like 
Silene it has no special means of seed dispersal and its progress is 
very slow. 
A large proportion of the western plants terminating here 
behave similarly. Afzelia macrophylla, for example, is rare in 
our area having been seen only once by the writer but under cir- 
cumstances which clearly indicate its powers of invasion, for its 
seedlings were more than holding their own against the former 
occupants of the soil. The eastern edge of the range of these 
species consists of a circle of such outliers located where chance 
seeds have dropped in advance of the range. The large propor- 
tion of western plants which are rare with us and at the same 
time are known far beyond our area may indicate that these are 
not merely characteristics of the individual species but that there 
is a somewhat systematic migration of western plants eastward. 
The converse of the same proposition is presented by such a 
plant as the hemlock, Tsuga canadensis. The main body of its 
range ends in the Sugar Grove region, where it is common, especi- 
ally in the deepest canyons. It occurs in numerous outlying 
stations, however, far to the west and south. One of these is 
in the mountains of western Alabama;* another is on the Green 
River in Kentucky; a third,t which is better known, is near Green- 
castle, Indiana, where it occurs together with the yew on a lime- 
stone soil. Since it is usually found on rocks one might suppose 
that the causes of its termination were physiographic—the absence 
of suitable habitats in the intervening country. But although 
most often found on rocks, both sandstone and limestone, it is 
by no means a chasmophyte by preference but is at its best 
* Mohr, C. Plant Life of Alabama. Cont. Nat. Herb. 6:34. Igor. 
+ Coulter, Stanley. Indiana Geol. Surv. Rep. 24: 616. 1899. Since this paper 
was in type I have found that is by no means the only station in Indiana as is stated 
by Coulter. 
