GLANDS AND VIOLET COLOR 21 
danthinus, A. nobilis, A. quiescens and kindred. Sometimes deep 
violet-blue and older white heads are to be found over the plant 
at once, as in A. zostemma and in A, versicolor ; sometimes all the 
flowers on a plant are of nearly even development and all change 
together from violet to pale, often deceiving the collector as to 
their identity. In general, presence of violet characterizes the 
glandular Biotian Asters, and white rays (regular among non- 
glandular species) are among them the result of fading. As an 
exception, very little color is developed, even at the beginning, in 
A. violaris and in A. macrophyllus pinguifolius. 
Many of the other cases which we have described as pale lav- 
ender might by some observers be simply termed whitish. In such 
cases, however, the trace of blue is still present to my eye, and is 
noteworthy as an indication of affinity with A. macrophyllus and as 
an accompaniment of the glandular character. 
Other groups in which violet or purple rays accompany glands 
include A. grandiflorus, A. oblongifolius, A. patens, A. claviger, A. 
spectabilis, etc., and so with A. Novae-Angliae, which is royal 
purple. On the other hand, white rays are seldom found in the 
above glandular species ; and the distinctively white-rayed species, 
as A. ericoides, A. paniculatus, etc., are non-glandular. 
Some other non-glandular groups, however, as A. cordifolius, 
A. laevis, A. salicifolius and congeners, are blue or violet. 
Pure blue rays are rare in Aster and do not occur in the Biotian 
group; forms of A. azureus, A. cordifolius and kindred approach 
close, however, to a pure blue. 
Of the relation of glands and color, we can simply say that 
glands never or very rarely occur among the white-rayed asters, 
but do occur widely among the violet. It is yet to be proved 
what chemical connection, if any, exists between the development 
of these glandular secretions and the presence of violet color. 
Yellow rays. — Their absence in Aster and presence in nu- 
merous related genera was made the ground of generic distinc- 
tion for Aster by Vaillant as far back as 1720. Asa Gray and 
many other writers have since emphasized this as a remarkable 
and valid character. Popular language has grown to reflect this 
fact, unconsciously observed, perhaps, at first, until “a yellow 
aster" has become recognized as a /usus naturae. 
