Helen M. Armour. 
52 
embryo-sacs were present (Fig. 21). These had evidently arisen by 
the simultaneous development of two adjacent segments of the 
mother-cell. After this stage, the development of the embryo-sac 
proceeds in the usual manner. Three antipodal cells, each with 
definite wall and a deeply staining nucleus (Fig. 22), are formed at 
the narrow basal end of the embryo-sac, and an egg apparatus at 
the micropylar end, while the two polar nuclei lie side by side in the 
centre of the sac. The fixation of the egg apparatus was imperfect, 
hut it was observed to consist of three large cells with no definite 
wall, and with densely granular nuclei (Fig. 23). The largest of 
these cells is presumably the ovum, while the synergidae are in close 
contact with it. A section through the ovary and base of the 
staminal scale of a flower ready for fertilisation shows the position 
of the mature embryo-sac (Fig. 4). The micropylar end of the 
ovule is now directed to the base of the ovary cavity. Fig. 4 also 
shows the position at the insertion of the staminal scale of the 
small outgrowth which is frequently regarded as a perianth. The 
vascular bundle supplying the chalaza is seen to continue in the 
ovary wall for a short distance above the insertion of the ovule. 
None of the specimens examined showed any trace of pollination, 
nor of the subsequent stages up to the formation of the young 
fruit. Apparently, however, after fertilisation does occur, the 
staminal scale breaks off at the place of narrow insertion, for it is 
not found on even the youngest fruits. In the fruit, three layers 
are clearly marked off (Fig. 12). The outermost is a somewhat 
thick succulent coat formed from the carpellary wall. Within it, 
and derived from the outer integument, is a hard fibrous layer 
whose inner margin of elongated cells represents the inner 
epidermis of the integument. The third layer is the inner integu¬ 
ment, which forms a layer of thin walled tissue. In the seed, the 
nucellus has practically disappeared. At the basal end it persists 
longest, but there is nothing to suggest that its function is that of a 
perisperm. Within the seed-coats the space is thus occupied by 
the large embryo-sac filled with endosperm. The embryo, which is 
embedded in the endosperm, is small and shows no differentiation 
of its parts (Fig. 13). 
The morphology of the flower of Chlorantlius now falls to he 
considered in the light of the foregoing investigation. Upon this 
subject various opinions, expressed for the most part in systematic 
works, have been held. The flower has been regarded as composed 
of a male and female flower united together and looking like a single 
flower. 1 Most authors, however, have considered the flower to be 
1 Hooker. Flora of British India, Vol. V., p. 100. 
