GENERAL REMARKS. 215 
Has the style invariably 2 vascular bundles, indicating its 
binary composition ? 
How is fecundation effected ? 
In several the ovula cohere, at least partially with the 
ovaria, this is however not of much importance, since the seeds 
appear to be invariably free. Prenanthes no. (2) 445, p. 251 
however leads me to suppose, that the seed may appear per- 
fectly free, while in fact the testa adheres to the pericarp ; 
this is not the case in the example cited. If this be true, 
the immediate covering or the apparent testa will be due 
either to the remains of the nucleus, or to the embryonary 
Sac; in Prenanthes the latter is the case. 
Allthat I have hitherto examined agree in the single- 
ness of the integument and its intimate cohesion with the 
nucleus. They all agree in the mode of development of 
e embryonary sac which is attached by both its ex- 
tremities, but chiefly by its base, and in the first appear- 
ance of this as a vesicle, at the apex of the nucleus. 
The embryo however appears.to be formed about its 
middle, instead of its apex as in usually the case. With 
respect to the raphe, they seem to differ, in some I cannot 
find it at all, in others it is developed gradually, and in these 
it appears to be generally, if not always complete, terminat- 
ing in the margin of the foramen. The embryonary sac is 
Very large in Prenanthes, no. (2) 445. p. 251. 
Burrumpootur : March 23rd, 1830. i 
am still more inclined to doubt the fruit being an Akeni- 
um; in every instance the testa coheres to the pericarpium. 
is is at once proved by the presence of the raphe on parts 
of it, when detached from its cohesion. The cellular mem- 
branous coat enveloping the embryo, owes its origin to the 
embryonary sac; this is at once evident by its having at 
: ends, remains of its original attachment. The nucleus, 
Which however is never distinct, disappears altogether. 
Although the transition from the perfect flowers of the 
dise, to those of the imperfect radial ones in those cases in 
