210 
Figs. 17, 18.—Cnicus altissimus, var. discolor, (Muhl ) Gray ; 17, style x 30; 
18, brush hairs x 500. : : 
Figs. 19-21.— Centaurea Americana, Nutt. 19, 20, style x 30; 21, brush 
hairs x 500. 
Figs. 22, 23.— Zaraxacum officinale, Weber. 22, style x 30; 23, brush hairs 
X 500. 
On the Evolution of Parasitic Plants. 
When we observe the complete parasitism of Phoradendron, 
Arceuthobium and Conopholis, it is difficult to conceive that time 
was when they grew wholly in the ground. It is not necessary 
to use the expression, now common, that their ancestors were 
terrestrial. There seems no reason why these plants, as now ex- 
isting, may not in the past have grown wholly in the ground. 
When we consider the immense number of parasitic Loranths, 
and the close relation of that order to Santalaceous plants, we 
might assume some considerable evolution in order to bring the 
former to the parasitic condition, but for the fact that we find 
members of the latter family partially parasitic In 1853 Mr. 
Jacob Stauffer of Lancaster, Pa., published a paper with drawings 
of a parasitic Comandra umbellata. Carefully washing out roots I 
could find no such attachment, but have since been rewarded by 
a beautiful case. The Comandra was obtaining most of its support 
from an attachment to a large root of Vaccinium stamineum. A 
case with an English ally of Comandra, Thesium linifolium, is 
noted in Bromfield’s Flora Vectensis, and approvingly referred to 
in the third edition of Gray’s Text Book. What is there to 
prevent that which is occasional from becoming habitual and 
permanent ? 
Partial parasitism is observed in many families. Mr. Stauffer, 
in the paper cited, illustrates Gerardia quercifolia and Gerardia 
flava in a similar condition, but I have never been able to find a _ 
case. This has also been reported of Castil/eia. 1 exhibited at 
the Boston meeting of the American Association for the Advance- 
ment of Science carefully prepared colored drawings by Lunzer 
of C. coccinea showing its partial parasitism on the roots of grasses. 
So far as these instances are concerned it is clear that a plant 
may easily assume the habit of attachment without any serious 
_ modification of structure. We want only some illustrations of 
