1883. ] The Achenial Hairs of Townsendia. 1105 
and to expel the bubbles at the fracture. On placing unbroken 
hairs in fluid, one may observe the fluid penetrating the wall and 
very quickly displacing the enclosed air. 
The quasi-glandular appearance of the diverging and recurving 
extremities of these hairs in some of the species, may be a case 
of sphacelation, the delicate apex. contracting as it withers, and 
its contents becoming discolored. It is not impossible, however, 
that the basal cell of the upper semi-hair (the one next the apex 
of the achene) is a glandular sac; as it appears to become gorged 
with fluid, and so to erect the hair which swings on the basal cell 
of the lower semi-tube as on a hinge. 
Many of the hairs of T. spathulata, and sometimes those of other 
glochidiate species, have one of the semi-hairs ending sooner than 
its comrade, sometimes just at the place for diverging. In this 
way we get a uni-glochidiate condition (like an anchor with only 
one fluke). 
The examination of T. Watsoni started a difficulty which I have 
not been able to solve. Gray describes its achenial hairs as 
“short, obtuse, or having 2-3 obtuse apical teeth ;” I find them in 
the specimens under examination to be-.32 to .34™ long (with a 
few short ones intermingled), the semi-tubes not diverging, some- 
what swollen at the extremity, occasionally one of the semi-tubes 
outgrowing its comrade and then slightly bending outwards. I 
do not find any of them lobed or toothed or triple (an extra tooth 
may grow on one of the semi-hairs). But it ought to be remem- 
bered that the entire condition may be in order of growth the 
forerunner of the divided condition; this I have seen to be the 
-case in other genera. The same explanation may apply to 7. 
condensata, whose achenial hairs in the specimen before me are 
obtuse, one of the semi-hairs slightly. exceeding the other ; per- 
haps they divide at a later stage. 
_ With these reservations and explanations the following analysis 
IS Submitted as an attempt to bring together all the glochidiate- 
Capitellate Species, and to marshal the other species around them. 
4. wilcoxiana may get a sub-section for itself, being well charac- 
terized by its very long pappus (11™™ the average of the other 
Species being 6mm) and its long and. delicate duplex-hairs with 
their Circinately rolled tips. 7. Watsoni seems, by its achenial 
airs, to come near T. condensata, but other characters prev Se 
itir approximation. 7. condensata has a number of slender 
