516 
outer margin, on stout plants sometimes 2’ long by 1’ wide. In- 
terposed leaflets often only a small entire pair in the distal inter- 
space, sometimes larger, obovate-oblong and dentate above with a 
single minute one or a pair on either side. Racemes slender, 
mostly 5’-10’ long, rarely 16’, rather loosely flowered. Flower- 
buds rounded, scarcely or not at all glandulose. Flowers 3/’-5” 
broad, rather deep yellow; anthers with broad connective. Bracts 
and bracteoles pilose-ciliate, very small, at least the bracts narrowly 
3-cleft. Fruit before maturity oblong and ascending, later reflexed 
or subspreading on short appressed pedicels, minutely subglandu- 
lose and slightly strigose, small, 2-214” long, 1 1%4’’—2” wide, nar- 
rowly turbinate to subhemispheric, often with a somewhat nar- 
rowed nonsulcate basal portion, or the lowest fruit of the racemes 
sometimes depressed-turbinate, marginless or obscurely margined, 
the disk flat or slightly convex with obtuse sepaline process, 
bristles loosely ascending or closer and nearly erect, equalling the 
body of the fruit or only half its length. Roots tuberous-thick- 
ened as in A. sé¢ria¢a, but the swellings often thicker and less tap- 
ering at either end, often club-shaped. (Plate 282, fig. 3.) 
Dry open woods and copses and weedy banks and hillsides. 
Of scattered or solitary habit of growth. 
Begins to flower at New York from the middle to the end of 
July and blooms later than any other species; belated flowers are 
sometimes to be found at the middle of October. . 
Reduced plants are sometimes trifoliate nearly throughout and 
simulate A. pumila. It will be usually evident to the collector of 
such forms, however, that they represent a state of imperfect de- 
velopment under unfavorable conditions of growth. 
This is the most variable of our species and runs into several 
unstable forms. It should not be overlooked, however, that it 
shows a well-defined tendency to separate into two particular 
forms or varieties. Extreme examples of both are common and i 
exhibit so considerable a degree of divergence that the eye always 
gives them instant recognition. Although both varieties are to be 
found holding their characters perfectly under identical conditions 
of soil and situation, intermediate forms, or what appear so to be, ae 
are of such frequent occurrence that I have not been able to sat- 
isfy myself of the expediency of giving a distinctive varietal name. — : 
For the sake of definiteness, however, the foregoing description of xg 
A. mollis has been made to cover only the form represented OF 
: the type, a specimen of which is preserved in the Herbarium of — 
