110 BicKNELL: FERNS AND FLOWERING PLANTS OF NANTUCKET 
rounded or cuneate at base; the lateral leaflets broadly rhombic 
and notched or cleft on the lower margin to form a prominent 
rounded lobe; flowering branchlets slender, angled, thinly pubes- 
cent, sparingly armed or unarmed, the lower becoming 2 dm 
long, the upper much reduced; inflorescence, when fully developed, 
a loose corymbose, sometimes compound cluster of mostly 2-5 
slender, erect or ascending pedicels and several, usually conspicu- 
ous, unifoliate, often deeply lobed, glabrate leaves sometimes 4—6 
cm. wide; pedicels mostly 3-6 cm. long, thinly loose-pubescent to 
glabrate, unarmed or sparingly setulose, but destitute of glandular 
hairs, often bearing one or more green bracts, and occasionally 
branched; calyx lobes broad, minutely pubescent on the outer sur- 
face, white-tomentulose within, often with foliaceous termination; 
flowers not usually large, spreading 2-3.5 cm.; petals oblong or 
broader, narrowed towards the base; fruit stibplobose to ovoid, 
often small but becoming of good size; drupelets large, large- 
seeded, thinly arachnoid-pilose when young and even at matu- 
rity, father watery and insipid or acid. 
Frequent or common in open sandy places and in pure sand. 
In fresh flower June 15, 1910; fruit ripe August 4, 1906. Frequent 
on Long Island. 
The trailing blackberry, Rubus flagellaris, well defined by 
Willdenow over a century ago, has ever since been lost to the rec- 
ognition of botanists. Professor L. H. Bailey has told us* that 
the type specimens are still in existence in the Berlin herbarium. 
Until actual comparison has been made of these specimens with 
the plant here brought to notice, a doubt may fairly be entertained 
whether the two are actually the same, although the evidence 
that they are would seem to be reasonably conclusive. Willde- 
now's species has been a continuing mystery. Bailey, who studied 
the type specimens, seems to have doubted whether it was an 
American plant at all. A particular interest therefore attaches 
to this Nantucket and Long Island blackberry, if it shall now 
finally dispel the mystery and resolve the doubt. That the original 
description of Willdenow applies to this plant with really close 
precision might not of itself be enough to justify a conclusion. 
But in addition it will be found that the illustration of the type 
specimens, presumably from a photograph, which Bailey has 
supplied, tT leaves little to be desired in the | way of convincing 
*Evolution of our ative fruits, 368. 
TLoc, Git:, 369. 
