310 SEED-STRUCTURE OF RAFFLESIACER AND HYDNORACER, 
at first sight one might readily mistake them for anatropous ovules. 
Through the kindness of Mr. Carruthers, the present director of the 
Botanical Department. at the British Museum, I have had an oppor- 
Brown with the materials for his paper. The seed is attached to the 
cell means of the soft, fleshy funicle, from which it is easily 
detached at the point where the denser tissue of the chalazal swelling 
For an idea of the form this protuberance assumes we may 
refer ye the beautiful plates accompanying Brown’s paper pee 
above, and to fig. 1 of our plate. e outer coating of the seed, a 
well as the whole of the chalazal aeeliing, consists of dark meddish 
cy cells, provi vided with a firm woody membrane, whose outer 
thickened. The side walls of the wedge-shaped outer layer of cells 
er Wi ts 0 : 
d there is an interruption of the hard testa, the aperture being 
closed by a tissue of square, thin-walled, very small cells, each of 
which contains a drop-like mass of a dark brown colour. (Fig. 1 a.) 
This testa encloses the inner hollow space, which, in comparison to 
the size of the entire seed, is relatively small. This space is imme- 
diately bounded by an outer brown, opaque, rie bes. though 
hin, skin, the lateral limits of whose cells are, how distinctly 
visible, Within this—close upon the erupts peat colourless 
stance. A perfect idea of its real structure can a be a 
from a central longitudinal section, as shown at fig. 4. It will 
seen from a study of this figure that it consists of tw wo parts—a 
the two parts is all the mor apren distinguishable because the mem- 
The endosperm layer is sc eg more than one cell in legit oe 
usually six cells. Their number is pretty uniformly the same in er 
four rows of one and the same embryo, and as the par i 
the cells of all the rows in a cross sec ction fall in the same plane, 
follows that we have a kind of storied structure of the whole ae 
echsietine of about be buries, one above the other, each story of 
f the 
broken ty the petits of an extra partition of one or more ©. 
ells. Other very co: i gularties occur from slight hae sem 
of the cells of the embryo (s 4), as must naturally apP 
the rows of cells do not fall in exactly the same e vertical set 
lowermost story, or stratum of cells next the micropyle end, } 
