1883.] Achental Hairs and Fibers of Composite. 35 
The achenial surface of the sub-order, Cichoriex, is de- 
void of hairs, and is covered by imbricating 
flat denticulate cells. The inner cells of the peri- 
carp develop fibers, enclosing crystalloids, which 
aid in the dehiscence of the fruit, much after the 
manner of the fibrous layer of pollen-sacs. Thus A 
I found the so-called indehiscent fruit of dande- inai 
i ; ea : ‘ Fic. 8.—Epiderm- 
lion in the act of dehiscing, by the aid of its of peticarp of A, 
fibers, when moistened, pressing out the seed, and Eten cotula and 
of its crystalloids serving as props and wedges, vx wiper 
the tapering form of the seed being well fitted for ri 3 Is enclosing spi- 
the process. ine 
Krigia virginica has simple red-brown spines over its fruit, and 
chicory has elegant multicellular hairs, corresponding with the 
perianth surface. The thistle group of the Cynaroidez agree as 
Ye 
Fic, jii FiG.10. 
Fic, 9.—Denticulate epidermal cells of pericarp of Lactuca scariola (Cichoriez). 
Fic. 10.—Endocarpal fibrils and crystalloids of Cirsium lanceolatum (Cynaroi- 
dez). 
to absence of ‘double hairs, and as to the fibers and crystalloids 
with the Cichariacez. 
Professor Asa Gray suggests that the mucilaginous filaments 
of Senecio are probably of service by gluing the achene to the 
soil, its pappus being thrown off. There is much mucilage in 
and about the filaments of the Cichoriez, and it will be an inter- 
esting question to determine what are the functions of these and 
the crystalloids, — 
The consideration of the facts stated above suggests a some- 
what different line of affinities from that usually adopted, and a 
. reéxamination of the tribal unity of Cynaroidez. We give the 
orders in the subjoined table, according to the arrangement of 
Bentham and Hooker. But it is manifest that Anthemidez and 
Arctotidez and Mutisiez are misplaced, and that other readjust- 
ments are to be made, if we are to marshal the groups according 
to the character here discussed. Yet the parallelism between the 
structure of the hairs and the affinities of the groupe, as founded 
on other characters, is singularly complete. 
# 
