128 
Sargant . — The Reconstruction of a Race of 
member. So far, however, this hope has grown fainter with every advance 
in knowledge. 
Starting from the other side of the gulf, much has been done of late 
years to investigate the origin and development of the male and female 
gametophyte among Gymnosperms. The variety of structure in allied 
genera is astonishing, particularly when contrasted with the uniformity 
of Angiosperms. But while the identity of the food-tissue in the embryo- 
sac with the prothallus of Pteridophytes is clear in every form, no com- 
parison with the endosperm of Angiosperms is even suggested by the 
development of such food-tissue, except perhaps among the Gnetaceae. 
Accordingly, I propose to discuss first the anomalous embryo-sacs 
hitherto described among Angiosperms, and then to refer very shortly to 
the isolated cases of Gnetum and Ephedra. 
The most complete account of the Angiospermous embryo-sac before 
fertilization is given in Coulter and Chamberlain’s Morphology of Angio- 
sperms (19), published in 1903. Guerin’s more critical account (30) of 
1904 is substantially the same. Both authorities remark on the extra- 
ordinary uniformity of this structure. The number of exceptions which they 
record is certainly very small compared with the number of species investi- 
gated. Even this small number should, I think, be considerably reduced, 
but on the other hand two species of Cypripedium must be added (Pace, 63). 
Species belonging to five genera are described as showing considerable 
deviation from the normal type. They are : — 
Peperomia pellucida and P. hispidida. 
Gunnera — several species, but chiefly G. Hamiltoni . 
Tulip a sylvestris and T. Celsiana. 
Juglans nigra and J. regia. 
Trillium grandiflorum. 
The anomalies in Peperomia pellucida were first described by Campbell 
(15) in 1899. R fuller and rather different account was published in the 
following year by Johnson (47), including observations on allied forms. 
Johnson has quite recently described a second species of Peperomia (49). 
Its endosperm formation shows anomalies similar to those of P . pellucida^ 
but with certain interesting departures from that type. 
In Peperomia , as in Lilium and other genera, the embryo-sac mother-cell 
becomes the embryo-sac without division. In P . pellucida sixteen nuclei 
are formed within the embryo-sac by repeated division of the original 
nucleus and its descendants. These nuclei are distributed pretty evenly in 
the peripheral layer of cytoplasm. One near the micropyle becomes in- 
vested with a definite mass of cytoplasm and can afterwards be recognized 
as the nucleus of the ovum. A neighbouring nucleus, which generally lies 
immediately under the micropyle, is likewise invested by cytoplasm, and 
