236 
PROFESSOR W. C. WILLIAMSON ON THE ORGANIZATION 
In fig. 60 definite relationships become assignable to each of these tissues ; d clearly 
represents the position of the micropile, immediately below which we have the flask- 
shaped cavity, d, which I have designated the lagenostome, the cavite pollinique of M. 
Brongniart. This cavity is bounded by a thin but very well-defined membrane, c. Its 
length, from the surface of the testa at the micropile d to the base where its walls 
appear to converge, is about ’04. In its interior is a mass of very delicate and regular 
parenchyma, h, the densest portion of which extends upwards into the neck of the lage- 
nostome towards the micropile, but which neither here nor yet lower down entirely fills 
the cavity of the lagenostome. The single nucular membrane f subdivides at its upper 
portion into three parts ; one of these, f', is deflected on each side so as to become 
united with the upper part of the endospermic membrane at (j . The second one, f", 
proceeds upwards and inwards, on each side, to unite at an acute angle with the lower 
part of the wall of the lagenostome. A third part passes upwards, on each side, to 
form the canopy e. This latter structure appears in this specimen as if it was a mem- 
brane of considerable thickness ; but such is not the case. This structure is thrown into 
remarkable longitudinal folds, and the section has here cut through the membrane in 
the superficial plane of some of these folds. Its thin sharp edge can readily be discerned 
in the original specimen. The endospermic membrane, g, is here seen to be detached 
some distance from the base of the lagenostome ; but this will be shown not to have been 
the normal condition of the seed. A faint line connects the two sharp basal angles formed 
by the walls of this lagenostome ; but this line has evidently not been a continuous tissue, 
but merely some minute and detached organic atoms which have been torn from some 
other tissue upon which the base of the lagenostome rested. So far as this section is 
concerned the centre of that base has been an open one. 
Fig. 61 is a tangential section which has passed through the oblate sides of the 
lagenostome but has not intersected its narrow neck. In its leading features this 
section corresponds with fig. 59. Thus we have the testa, a, connected with the nucular 
membrane, f ’, by numerous shreds of tissue. That membrane divides at its upper part, 
on each side, into three layers, as in fig. 60, whilst at its upper part it further exhibits 
the crenulated contour of fig. 59 ; but here the crenulated curves are but three in number. 
The thin wall (f") of the lagenostome d is now seen to exhibit a clear wavy outline, whilst 
within the interior of the cavity we have a well-defined mass of parenchyma ( h ). This 
being an important section I have, in fig. 62, enlarged the upper part of it to 80 diameters. 
The testa (a) exhibits the three crenulations a 1 , a! which are just detaching themselves 
to form three of the folds of the canopy e. The wavy outline of the wall of the lageno- 
stome appears at c. The nucular membrane has united itself to the endospermic 
membrane at f; but at f traces of the separation of these two membranes present 
themselves. At g the nucular membrane seems to be firmly adherent to the delicate 
parenchyma ( h ) of the lagenostome, whilst that parenchyma has shrunk away from the 
lateral wall ( c ) of the cavity in a very regular manner, leaving an open space ( d ) all round 
the cellular mass. This last point is, as we shall shortly perceive, an important one. 
