EMBRYOLOGY ОЕ THE АМЕХТТЕЕВЖ, 419 
Amentifere the recurved branches found are simply cæca and probably contain no 
portion of the fertilizing element. 
There are several cases of branching pollen-tubes recorded by Hofmeister among the 
Monocotyledons in the ‘ Neue Beitrage zur Kenntniss der Embryobildung der Phanero- 
сатеп, 1861. Тһе branched tubes of Pothos longifolia are represented in Taf. x. figs. 1 
and 2, and those of Hippeastrum aulicum in Taf. xiii. figs. 4and 6. But more interesting 
examples of the branching of pollen-tubes occur among the Gymnosperms, where we find 
in the Сиргеззтея one tube applying its broad extremity to the necks of a whole group 
of archegonia at the same time, and then forming short narrow protuberances which grow 
down into the separate archegonia. | 
Comparative account of the Archesporium in the foregoing Genera. 
We find in reviewing the various examples described in the preceding pages certain 
features in their archesporium which they have in common, not only with one another, 
but with the normal type which has been shown by Strasburger and others to be 
almost universal among the Angiosperms. 
Thus we observe in all eight genera the subepidermal origin of the archesporium and also 
the presence of tapetal cells. Again, the archesporial cells are arranged in strands, formed 
at first of definite tiers of cells cut off from their respective initial subepidermal cells. 
But whereas in the normal Angiospermic type we find but one of these strands, which is 
in the genera now under consideration we find a 
large number of these rows derived from a subepidermal layer, and it is difficultin many 
cases to point out one row as more distinctly characterized than some others (РІ. LXVII. 
fig. 8, Pl. LX XI. fig. 48). This system of strands, forming what I have previously referred 
to as a sporogenous tissue, is connected with the base of the nucellus by a regular 
- series of cells which I have referred to as the pedicel of the sporogenous tissue. 
Again, we find oblique divisions occurring in the cells of this coaxial system of strands 
he sporogenous tissue. The spindle- 
which lie between the tapetum and the pedicel of t | 
shaped cells thus arising, in the case of Castanea, develop into large and conspicuous 
tracheides, but it is chiefly in the Corylaceæ that they become ы marked feature, thongh 
they do not form tracheides. Without trespassing too far into the region of theory, it 
may be of interest to note that the sporogenous tissue of Fagus seems to belong to a 
more primitive type than we find in the other genera. This A has d 
Е к + | 
support from Eichler’s opinion, who deduced from the (оға титр E m 
4 1 
that the Cupuliferz are а less specialized group than the Cory 
Fagus also there are two integuments to the ovule and a very well-developed epidermal 
cap. In the latter character, among others, there is considerable — "— 
nucellus of Fagus and those of Gnetum, Taxus, and other E ~~ ae - 
type we find, on the one hand, à reduetion in the extent of the archesport 
of the other Cupulifere, until there is very little that is not normal in перон ie ea 
the other hand, a reduction in bulk in the case of Corylaces, Aper poto = ~ 
although in Corylacee, especially, we can still claim for a large part о 
qme 
nucellus a sporogenous nature. 8х2 
therefore named “the axile row,” 
