30 Transactions of the Boyal Society of South Africa. 
fertilisation, and 3 antipodals which are not persistent. The raost 
striking feature of the embryo-sac in Brunia is that it is completely 
filled with large starch grains (Figs. 6-8). I know of no other instance 
where an embryo-sac is packed with starch grains as it is in this plant. 
Starch grains are present also in Berzelia, but much smaller and not 
filling the sac (Fig. 4). They are apparently absent from the embryo- 
sac of Staavia (Fig. 5)."^ 
Summary and Conclusions. 
The main points of the above account may be briefly summarised as 
follows : — 
The ovule of the Bruniacece is pendulous and anatropous, with a dorsal 
raphe. There is a single massive integument with a long slender 
micropyle. In Brunia the embryo-sac is packed with starch and is all 
that remains of the nucellus. In Berzelia and Staavia a little of the basal 
nucellar tissue persists. 
In development the megaspore mother cell is about the third cell from 
the apex of the small nucellus. The usual row of four megaspores 
develops, of which the lowest becomes functional. 
The developing embryo-sac very soon absorbs the tissues above it, 
and so reaches the apex of the nucellus. 
The only conclusion to be drawn from this study is the same to which 
ScHONLAND {loc. cit.) was led — namely that the Bruniacece are very isolated 
in the cohort to which they are usually attached. Whether the charac- 
ters of the ovule also point to its antiquity may perhaps be questioned. 
The single integument might be regarded as a primitive character, if it 
were not associated with the ovule of the Composites, while in some of 
those orders generally regarded as primitive {e.g., Proteacece) two integu- 
ments are found. Possibly the word " antiquity" was not used by Schon- 
LAND in a philogenetic sense, but antiquity in the South African Flora must 
(if we accept the author's conclusions), imply also phylogenetic antiquity. 
The other characters of the ovule rather tend to agree with essentially 
modern types {e.g., Conipositce), than with primitive {e.g., Banunculacece) . 
In conclusion, I am glad to take this opportunity of thanking Mr. 
E. P. Phillips, who accompanied me on a collecting trip, and whose inti- 
mate acquaintance with the local flora was of great assistance in obtaining 
the requisite material for this study. 
* Since the above was written, Miss E. L. Stephens has kindly shown me some 
slides of the embryo-sac of Brunia, in which she has demonstrated that the starch 
almost completely disappears before fertilisation. She has also drawn my attention to 
a paper by D'Hubert (D'Hubert, E., Recherches sur le sac embryonnaire des plantes 
grasses, Ann. Sci. Nat. Bot., viii. 2: 37-128, pis. 1-3, fig. 66, 1896), in which a similar 
formation and subsequent absorption of starch is noted in various fleshy plants. 
