Primitive A ngio sperms. 1 3 1 
in L ilium. The only variation in the details of fertilization is that the 
nucleus of one synergid takes the place of the lower polar nucleus. 
This explanation seems likely, for the Orchids have always been con- 
sidered as highly specialized forms, probably derived from ancestors with 
flowers of liliaceous symmetry. But if the four-nucleate embryo-sac of 
Cypripedium really represents the eight-nucleate sac of Lilium the exact 
origin of each nucleus becomes important. Is the chalazal nucleus sister 
to the ovum nucleus, or to one of those in the synergidal cells ? 
These are the major irregularities hitherto described in normally fer- 
tilized embryo- sacs. The minor irregularities chiefly concern the antipodal 
cells. One of these variations may be of some phylogenetic importance. 
The antipodals occasionally multiply — in some cases forming quite a tissue 
at the base of the embryo-sac — either before or after fertilization. This 
character is common in certain groups, as the Gramineae and Compositae. 
It has been recorded lately among the Gentians (Guerin, 29), as well as in 
several isolated genera (Guerin, SO, pp. 39 - 41 ) ; and the tissue thus formed 
may possibly represent the primitive prothallus or female gametophyte. 
Considering the vast number of embryo-sacs described, these anomalies 
are exceedingly few. Indeed, if we exclude all apogamous and partheno- 
genetic forms, we find the formation of the ovum and of the endosperm to 
be alike in every genus examined, with the single exception of Peperomia. 
This is the more surprising, as forms likely to show primitive features have 
been picked out for examination, and, again excepting Peperomia , they have 
all proved perfectly normal in essentials. Such are Piper and Heckeria 
(Johnson, 48), Casuarina (Frye, 26), Carpinus (Benson, Sanday, and Ber- 
ridge, 10 ), and Drimys (Strasburger, 86 ). 
In short, the early history of the embryo-sac is wonderfully uniform 
throughout the Angiosperms. This cannot be attributed merely to the 
great reduction of the female gametophyte. For the nuclear divisions 
within the embryo-sac are extremely characteristic in the final orientation 
of their products. Nor can the uniformity observed be put down to in- 
sufficient evidence, for the early history of the embryo-sac has been followed 
in a great number of species. 
The formation of the endosperm is likewise very uniform throughout 
the group. In every normal case 1 it is undoubtedly formed after fertiliza- 
tion. Its origin from a nuclear fusion has been observed in numberless 
instances ; indeed there is no normal case to the contrary. The regular 
occurrence of a triple fusion in which one element is a male nucleus has 
also been recorded in many cases. The only real exception on record 
is Peperomia , where, as just stated, more than three nuclei take part in 
the fusion, and the presence of a male nucleus within it has not been verified. 
1 By normal case I mean an embryo-sac in which the ovum is normally fertilized. 
