1903] Fernald,— New Bidens from the Merrimac Valley 9I 
involucre are as long as the disk, but in this they differ strikingly 
from Æ. comosa whose broad flowering disk much exceeds the inner 
involucre. In its achenes the Merrimac plant is somewhat interme- 
diate between B. connata and B. comosa. ‘The achenes of the former 
are rather tetragonous in section, the ribs on the two faces being very 
conspicuously thickened and keel-like; and the inner achenes are 4.5 
to 6 mm. long. In B. comosa the achenes are flat and essentially 
nerveless, the innermost 8 or 10 mm. long. The achenes of the 
cylindric-headed plant of the Merrimac shores are essentially flat, but 
they usually have a well defined though narrow mid-rib on each face, 
and the innermost achenes from 7 to 9 mm. in length. Thus in its 
achene the Newburyport and Salisbury plant stands between P. con- 
nata and B. comosa, though it differs from both in the shape of its 
head in which character it strongly simulates the local and otherwise 
unique Z. didentoides. 
The awns of Bidens bidentoides are upwardly barbed, instead of 
with the retrorse barbs which are ordinarily associated with Bidens. 
On this account the plant of the Delaware flats was long supposed to 
be a Coreopsis. Similarly when in 1866 A. H. Smith found near 
Philadelphia a plant resembling in all other characters Bidens fron- 
dosa, but with the awns upwardly barbed, the plant was supposed to 
be a hybrid between Coreopsis bidentoides and Bidens frondosa, and 
was later referred to by Dr. Gray as “doubtless a hybrid.” ! Subse- 
quently however, a plant quite identical with the Delaware River 
material has been found as far east as Cape Breton Island (Macoun, 
no. 19,168), fully 800 miles from the nearest Bidens (Coreopsis) biden- 
toides, so that the hybrid origin of the plant seems quite out of the 
question. This extreme of B. frondosa with upwardly barbed awns 
may be called var. anomala, Porter, a name under which the plant 
was distributed by the late Thos. C. Porter. 
Dr. N. L. Britton has recorded? the occurrence of downwardly 
barbed awns in Bidens discoidea which commonly has the barbs 
ascending, and Dr. K. M. Wiegand has recorded 3 upwardly barbed 
awns in B. connata, concluding that such variations are rarely or 
never due to hybridization. In view of these exceptional tendencies 
already observed in the related species of Bidens it was interesting to 
yn H8. 7p 092,22007 
? Bull. Torr. Bot. Club, xx (1893) 280. 
3 Bull. Torr. Bot. Club, xxvi (1899) 400. 
