256 GALIACEA. 
In one instance, one of the lateral flowers had pedicels, on 
its inner side with a flower developed, and as this was much 
smaller than that in the axilla of the leaf, we may conclude 
that the flowers are all terminal, or rather that each lateral 
flower-bud may become a simple raceme. 
In Galium, the carpella are both developed, and both lobes 
of the disk are equal, in the more petaloid species of this 
genus, the perianth separates all round the base, of the albu- 
en. . 
From casual examination I think that the inner mem- 
brane is distinctly visible in the fruit. 
"Throughout all the stages, I have examined of the deve- 
lopment of the ovulum, and these include very minute buds: 
it would appear that the ovulum consists of a nucleus alone ; 
the same organisation exists in Galium, and probably 
some? Rubiacez. Within is developed an ——— ary § 
of that form which I call albuminary, and the sir 
becomes at last reduced to a. thin coat, cohering most inti- 
mately with the inner layer alluded to, and also slightly with 
the external or calycine? layer, so that the albumen i. e. 
the seed, has no demonstrable covering. 
The membranous nature of the embryonary sac, which 
I should observe is of late formation, i. e. it does not exist 
before impregnation, is well seen in the nearly mature albu- 
men, which presents a hyaline membrano-mucous-looking 
hr this sae, opaque grumous grains of albumen are copi- 
ously developed. 
It follows that the usual account of Galium are altogether 
wrong ; on this head it will be necessary to examine Decaisnes 
Rubia. 
The nucleary ovule, and the reduction of the perianth 
forcibly call to mind Santalacez, from which it is 
in the first place by the relation of the stamina to the 
perianth, and in a less "degree by the bilocularity of the 
ovary, 
