648 
PHANEROGAMS. 
facts concerning them. The perigynous structure of the flower is peculiar to 
Dicotyledons, as is also the occurrence of hollowed axes of the inflorescence, like 
the fig and similar structures, and the cupule, which occur in some families, and are 
dependent on similar processes of growth. 
The Ovules exhibit, in the different divisions of Dicotyledons, all those varieties 
of structure which have already been mentioned in the introduction. Very commonly, 
especially among the Gamopetalae, the nucellus is covered by only one integument, 
which is then often very thick before fertilisation. But on the other hand the 
third integument or aril is much more common than among Monocotyledons. 
When there are two integuments, the outer one — differing again in this respect 
fromi most Monocotyledons — takes part in the formation of the micropyle, enveloping 
the exostome or entrance to it. In some parasites the ovules are rudimentary, 
and in many Balanophoraceas are reduced to a naked few-celled nucellus ; while 
in Loranthaceae they are coherent with the tissue of the floral axis in the inferior 
ovary. 
The behaviour of the Embryo-sac^ before and after fertilisation is similar 
in most Dicotyledons to that which occurs in Monocotyledons. The endosperm 
usually originates by free cell-formation, and is transformed by repeated divisions 
of the first cells which are formed in this manner into a more or less dense tissue, 
which fills up the embryo-sac either before or after the formation of the multi- 
cellular rudiment of the embryo. But in a very considerable number of families 
belonging to altogether different groups the embryo-sac exhibits on the one hand 
striking phenomena of growth, elongating considerably before, impregnation into 
a long tube, and emitting after impregnation one or more vermiform protrusions 
which penetrate into and destroy the tissue of the nucellus and of the integu- 
ments, or "even protrude altogether out of the ovule (as in Pediculans, Lathrcßa^ 
and Thesiimi). On the other hand, in those plants in which the endosperm 
originates by cell-division we learn from Hofmeister that the following variations 
occur : — ' The whole of the cavity of the embryo-sac behaves like the first cell 
of the endosperm in Asarineae, Aristolochiaceae, Balanophoracese, Pyrolese, and 
Monotropeae ; the first division of the sac is the result of a partition-wall which 
divides it into two nearly equal halves, each of which encloses a cell-nucleus 
and again divides at least once into daughter-cells. In other cases the first 
cell of the endosperm includes the upper end of the embryo-sac ; the embryo-sac, 
immediately after fertilisation, appears to be divided by a transverse septum into 
two halves, the upper one of which developes into the endosperm by a series of 
bipartitions ; while no such bipartition of the lower one occurs in Viscum, Thesiu?fi, 
Lathrcca, Rhinanthus, Mazus, Melampyrum^ or Glohularia. The first cell of the 
endosperm fills up the middle part of the embryo-sac in Veronica^ Nemophila, 
Pcdicularis, Plantago^ Campanula, Loasa, and Labiatse ; its lower end in Loranlhus, 
Acanthus, Calalpa, Hebenstreiiia, Verbena, and Vaccinium^ In Nymphcßa, Nuphar, 
and Ceratophyllum, the upper end of the embryo-sac is cut off from the rest 
of the space by a septum soon after fertilisation, and the further development of 
^ Hofmeister, Jahrb. für wiss. Bot. vol. I. p. 185 ; and Abhandl. der kön. Sachs. Ges. der Wiss. 
vol. VI. p. ^iee supra, p. 579). 
