ENGELMANN—CUSCUTA. 455 
duced almost to a point; it runs in the direction of the inte- 
rior angle of the seed (longitudinal, DesM.), or at right an- 
gles with it (transverse, DesM. , Or it has an intermediate, 
oblique direction. In some sections I find these characters 
sufficiently distinct; in others they seem to be less reliable; 
in the eee Cuscutee I have often found them interme- 
Neatly ripe seeds are smoother and larger, whet soaked, than 
ripe ones, 
The e mbryo has been supposed to offer good characters ; 
but I hare reason to believe that those embryos with one or 
few cireumvolutions (such as the one figured by Webb, Phyt. 
Can. IIL, pl. 142, fig. 14) are taken from unripe seeds. 
Another character which I at one time relied on for Lc 
tic distinction of Cuseutew is found in the c calyx. Usually it 
is gamosepalous, but in some American species it is formed of 
entirely distinct and imbricate sepals, not different from the 
g bracts—a character which prevails in Convotou- 
i proper, where only one genus ( Wilsonia) is peels 
rs welt e stem, but pri stow in nie ares 
wer tend 7 
os the or absence of 
The shape and ai of dalys a oro) a 
char (tube and thes) hen furnish important but not pen perem 
eters. necestens 
