vi BioTIAN ASTERS 
The genus Aster has long been reputed one of most pronounced 
variability ; and of all Asters, perhaps the most variable are those 
of the Biotian section ; some other groups, as Gray's G/azdulosi, 
being quite stable in comparison. It is to their variant forms that 
the Asters owe their supreme interest to the student of the evo- 
lution of species; exhibiting a genus evidently engaged at this 
present time in the development of new lines of descent. Many of 
the latter show apparent stability and therefore demand recognition 
as species ; a few seem unstable and rank merely as subspecies ; 
and a much larger number are still too uncertain for absolute 
rank, and I leave them under the indefinite term form. 
Since these descriptions left my hand for the press, the world 
has been enriched by the appearance of the monumental work of 
de Vries on the origin of species by mutation. I have no doubt 
that many if not most of the new species of Aster here described 
are similarly mutants proceeding from dissimilar parents. Explicit 
proof is not so readily attainable in Aster, however, owing to the 
perennial habit and the number of years before flowering. Work- 
ing without the opportunities of garden-sowing but entirely from 
plant-colonies as established in nature, I have been forced to draw 
my conclusions as to specific difference rather from collateral com- 
parisons than from a series of lineal descendants. The length of 
preliminary growth before blossoming, often seemingly seven 
years, interferes decidedly with observations on aster-seedlings of 
successive sowings. When the working life of a botanist passes 
the century mark, then perhaps such a botanist can see with his 
own eyes the successive generations of an Aster like the Upsalan 
A. tardiflorus of 18 years before flowering. 
Asters are urther complicated by remarkable responsiveness 
to environment, producing an unusually wide area of variation- 
range on each side of a norm or center of species-character, a norm 
or species-type which may have been, in very many cases, of mutant 
origin. With closely allied and parallel types, each with wide 
variation-range, it is natural that many outliers belonging to allied 
species should overlap and should present confusing characters. 
If these individuals were regarded as forming such connecting links 
as to invalidate species, close distinction of species would in many 
groups of Aster become hopeless ; and itis in this sense that Aster 
