SEEDLINGS AND SURCULI 19 
made difficult by the length of life of the individual aster, each 
plant living without flowering it may be from three to seven years, 
or in some cases even twenty or more. By each plant I mean not 
only each seedling, but also each rootstock branching from each 
seedling, many at least of which remain some years without 
further growth than the development of radical leaves. My main 
source of evidence as to variability of aster seedlings, as would be 
true in case of most other perennials, is the comparison of simul- 
taneously-existing colonies of the same species. Such colonies 
can include but a slight range of generations, and many, some- 
times all, belong to one. Careful comparisons of this kind may 
give rise, however, to inferences as correct and as convincing as 
any known to logic. I have made these comparisons, with this 
inquiry in view, during the last seventeen years, and through a 
region extending from the Androscoggin to the Potomac. My 
conclusion is that while offspring is like parent in most cases, there 
is relatively a very large percentage of sports in Aster ; relatively 
large as compared with most other plants; though not abso- 
lutely a large number, for it may be only one individual among 
thousands. 
VARIATION OF ROOT-PROPAGATED PLANTS 
My most careful direct efforts have been devoted to determin- 
ing how far and in what ways the members of a single colony vary 
from each other. By a single colony I mean the plants descended 
from a seedling and still bound together by a continuous system 
of rootstocks or still indicating that origin by their position, 
although many of the connecting rootstocks have been lost by 
decay. Such a colony in the MJacrophylii may be very large, per- 
haps a rod or more in diameter, and in case of similar compound 
colonies of A. guiescens I have known the ground for 100 feet 
radius to be practically occupied by the radicals to the exclusion of 
almost all other herbage. Such a colony in the Curvescentes may 
be of large size also, 2c feet or more; and among the Divaricati 
sometimes the same, though more often not over three or four 
feet. Among the other broad-leaved species colonies are smaller. 
Among the narrow-leaved they are with great difficulty traceable, 
because of the early decay of connecting rootstocks, making it 
