Primitive Angiosperms. 129 
the naked cell thus formed may represent a synergid. Eight nuclei mass 
together and after fertilization fuse to form the primary endosperm nucleus. 
The six nuclei which remain retain their parietal position and are cut off by 
walls from the cytoplasm of the embryo-sac. In the end they are crushed 
and absorbed by the developing endosperm. 
No male nucleus has been identified in the group which becomes the 
primary nucleus of the endosperm. But this may be due to the absence 
of the stages which immediately succeed the act of fertilization in the pre- 
parations examined. For in all the cases described the male nucleus was 
about the same size as the nucleus of the ovum with which it was fusing. 
This indicates that it had entered the ovum some time. A similar nucleus 
within a group of eight would, of course, be indistinguishable from them. 
The history of P . hispidida is similar, but not identical (Johnson, 49). 
In the eight-nucleate embryo-sac, two nuclei are near the micropyle and 
six at the chalazal end. The succeeding mitosis forms twelve chalazal and 
four micropylar nuclei. One of the latter becomes a well-defined ovum : the 
other a synergid. Two micropylar nuclei travel into the centre of the 
embryo-sac, where they fuse with the twelve chalazal nuclei. This fusion of 
fourteen nuclei becomes the primary nucleus of the endosperm. 
The allied genera, Piper , Heckeria , and Saurnrus , were found by Pro- 
fessor Johnson (48) to form perfectly normal embryo-sacs, and he concludes 
that the peculiarities of Peperomia must be considered as derived from the 
usual type rather than primitive. The facts just described support this 
conclusion. In primitive species we should expect the endosperm to be 
better developed than in the normal embryo-sac — nearer, that is, to the 
original structure, whether gametophyte or sporophyte, from which the 
tissue familiar to us has been reduced. But in both species of Peperomia 
the endosperm in the mature seed consists of forty or fifty cells only, and is 
of very small bulk compared to the perisperm, which is the effective food- 
tissue. Indeed all analogy goes to show that a fusion of many nuclei — 
eight in P. pellucida and fourteen in P. hispidtda — -is far more likely to pro- 
duce a small mass of short-lived tissue than a vigorous body capable of 
development into a permanent organ. 
Peperomia stands alone as a normal deviation from type. For though 
Schnegg (79) has found similar features in the embryo-sac of Gunner a, he 
gives very strong reasons for thinking that in the species examined the 
embryo is formed without fertilization. If so, the example is irrelevant. 
For when the generative nuclei never enter the embryo-sac the origin of 
the endosperm must be as irregular as that of the embryo, and its formation 
is likely to be abnormal in other respects. 
Of the three genera which remain in the list of exceptions on p. 128, 
the only one which need concern us is Ttdipa. In the two species, T. sylves- 
tris and T. Clusiana , Guignard (33) found that the embryo-sac before fer- 
