618 Yendo. — On the Mucilage Glands of Undaria. 
the same degree as the other glandular cells, and the epidermal one hardly 
at all. I can only explain this arrangement of cells as an aberrant mode 
of formation of the gland. The two septa would probably decompose 
subsequently, resulting in an oblong glandular cell. 
After the epidermal cell has completely atrophied and the supra- 
glandular space has been first formed, the glandular cell ceases to grow 
towards the surface of the lamina. The addition to the size of the gland 
takes place mainly at the inner end of the cell (Fig. n). As the glandular 
cell increases in size, the cells around it are evidently pressed aside. But 
the epidermal cells seem to have been disturbed in a very little degree by 
the pressure, and their division and growth continue as the lamina extends. 
This will be treated further in connexion with the growth of the supra- 
glandular spaces. 
The glands add to their size at their inner ends, displacing the loosely 
arranged surrounding cells. Hence their shape, which was first ovate or 
pyriform, with the sharper end towards the surface of the lamina, becomes 
roundish-pyramidal or compressed-conical. In the next stage, most of 
them, if not all, grow horizontally, elongating in the direction of the longer 
axis of the pinnule. In the cross-section of the lamina such glands appear 
as flattened cones with the apices protruding towards the supraglandular 
spaces. But in the surface view it is clearly to be seen that the glands have 
their largest diameter at the point below the supraglandular space, and that 
the horizontal elongations are outgrowths from such principal part (Fig. i). 
It was not ascertained whether any neighbouring cortical or hyphal 
cell has or has not amalgamated with the glandular cell during its growth. 
Judging from the arrangement of the cortical and hyphal cells around a 
completed glandular cell, I am inclined to believe that none has fused to 
add to its size. The direction of the outgrowth is due to the fact that the 
hyphal cells are largely running parallel to the margin of the pinnule — the 
direction of the growth of the pinnule — and thus it has fewest obstacles to 
its elongation in that direction. 
As the primary epidermal cells in the transitional regions are cubical, 
the supraglandular spaces, when they are first formed by the normal 
process above referred to, are naturally square in surface view. They are 
hardly larger than the size of an epidermal cell. As the pinnule grows 
larger the spaces are also extended in area, retaining their former shape or 
elongating into a rectangular form. This is due to the fact that an epidermal 
cell divides into four successively by partition walls perpendicular to one 
another. In the fresh specimens, as well as in those fixed in picric acid, 
this mode of cell multiplication is not clearly seen, for the cells are closely 
compressed together. But when the material is fixed in the sublimate 
solution, the boundaries between the groups of the cells are vividly shown 
by the hyaline middle lamella. Each group consists of four cells cruciately 
