Mr. GnirrrrH on the Ambrosinia ciliata of Roxburgh. 267 
peduncles when present being interposed. The rudimentary stolones always 
correspond to the axils of every alternate sheath. In the young spadices the 
antheriferous portion is as it were sessile on the top of the ovaria, the filiform 
portion being a subsequent development. The inner surface of the convolute 
limb is at the same period smooth and shining. The septum is perfectly 
developed at a very early period, and then closes up the tube completely ; it 
already covers, but does not adhere to the apex of the spadix, and hence the 
subsequent slight obliquity of the upper portion of this body. At an inter- 
mediate period the apex of the spadix adheres strongly to the septum, but 
subsequently regains its original free state. The septum appears to originate 
in a production downwards from the commencement of the division of the 
limb, to which part it always corresponds. The antheriferous portion is well 
supplied with vessels, fascicles of which pass off from the central bundles to 
the anthers, corresponding to the central cellular part or septum that exists 
between the thecæ. The anthers may from a very early period be compared 
to two cups joined together by their contiguous margins; the wide and open 
mouth which they present in their mature state being closed up originally by 
an extremely fine membrane, lining the cavity of the cup and forming a 
convexity where it closes in the opening. Within the cavity thus formed the 
pollen is developed. As the anther increases in size this membrane gradually 
assumes the form of a cone, which projects in proportion as it increases beyond 
the margins of the cups or thecze. At the same time it assumes a yellow tint, 
by which, chiefly, I am led to think that it lines the entire cavities of the 
theca. The cellular tissue of the thecze consists of a cutis, which is papillose 
on the margins of the cup, and an inner series of ovate cells arranged with 
their long diameters pointing from the axis. On the membrane of these cells 
very distinct fibres are developed, which almost always have the same direction 
with the cells. These fibres cross each other at very acute angles, and appear 
to be incomplete at either end of the cell, in which they are developed. The 
cone soon becomes more subulate, it remains closed, and is of a yellowish 
tint. The anthers appear to be fully formed at a time when the spadix is only 
half-developed. At a later period the apex of the cone is open, and through 
this opening the contents of the thecze may be Squeezed, assuming, from the 
comparatively small diameter of the apex of the cone, a more or less 
VOL. XX. 2N 
elongated 
