EMBRYOLOGY OF THE AMENTIFER.E, 413 
in the lower regions of the strands to be more or less oblique instead of exactly trans- 
verse—as are the walls which give the characteristic appearance to the upper part. 
The Castanea material I have worked upon was gathered in the East of England and 
gave very unsatisfactory results in the later stages. I hope this year to obtain material 
from a district in which this species is better able to carry on the development of its 
fruit. Judging, however, from flowers gathered from August 6, when I first find the 
tracheides appearing, until August 22, I should be inclined to say they are not constantly 
present. A complete series of sections of an ovule cut with the microtome often reveals 
no trace of a tracheide, and this inconstant appearance supports the view that they are 
vestiges of some former structure. 
Quercus. (Pl. LXIX. figs. 28, 29.) 
This genus has not afforded me anything distinctive to relate. It closely resembles 
Castanea in the structure of the nucellus, although this never attains the great length 
that is so characteristic in Castanea. The axial strand, the position of the embryo-sac in 
this, all the processes in the embryo-sac—including the extrusion of the ceecum (fig. 29, 
c@.)—and the arrangement of the antipodals in a pointed base, can be readily demonstrated 
to be similar. I have been able to verify Hofmeister’s observations on the course of the 
pollen-tubes, several of which enter and descend a single micropyle. The pollen-tubes 
form short blind branches as shown in fig. 28, c., c. ‘Lhe embryo, being attached to the 
apex of the extruded embryo-sac, is to be seen lying almost in the micropyle up which 
the embryo-sac pressed before fertilization. 
CASTANOPSIS. 
This genus still awaits investigation, but I hope soon to have the material in hand. 
The fruit takes two years to develop, as is well known to be the case with Q. rubra, 
Q. Cerris, and many other species of Quercus. 
On leaving the Cupuliferæ we enter upon the examination of two groups of the 
Amentiferæ which, so far as yet investigated, reveal a striking phenomenon in the 
abnormal course of the pollen-tube, which does not enter the micropyle, but penetrates 
the nucellus from the chalazal region. This exceedingly interesting fact was first noted 
by Treub for the case of Casuarina. I hope now to show that it also occurs in Betula 
and Alnus, Corylus and Carpinus, and I will refer to this group of genera by the dis- 
tinguishing name of * Chalazogamie Amentiferæ.” Тһе whole account of the course 
taken by the pollen-tube, and other associated topics, will be found further on, but L 
will first give a short description of the embryology of representative species cf each 
genus. 
BETULA ALBA, Linn. (РІ. LXIX. figs. 30, 31.) 
The structure of the Young nucellus can be seen from the longitudinal section repre- 
sented in fig. 30. We see again the coaxial system, the archesporium abutting on the 
epidermis, and this epidermis two-layered as in an early stage of Fagus. It is the more 
