15 
leaves, 2d, the floating leaves with filiform petioles, 3d, the scales on 
the rootstock, and 4th, the floating leaves with short petioles. 
But in order to obtain entire satisfaction as to the difference 
between the filiform petioles and stems we must examine the internal 
structure of the same. The fibro-vascular centre of the stem (Fig. 2) 
IS distinctly divided by a zone of parenchymatic ground-tissue into 
two bundles, each of which contains both vascular {v) and sieve tis- 
sue {sv)^ the latter situated toward the endodermis {en). In the 
fibro-vascular bundle of the filiform petiole (Fig, 3) there is but one 
set of xylem {v) and phloem {$%)) with the woody tissue on the 
inner side, A line bisecting the two bundles of the stem forms an 
angle of about 45^ with the median line, while the bundle of the 
petiole is symmetrically arranged, /. <?., it is bisected by the median 
line. The petioles of the other two kinds of leaves are, in this 
respect, exactly like the filiform ones. It is instructive to observe 
how the double bundle of the stem separates into two branches on 
approaching the node, one of them leading to the inflorescence and 
apex, the other continuing its course into the leaf. 
Otherwise the stem and the petiole do not present any material 
differences. Both have two additional, much more reduced fibro- 
vascular bundles near the circumference, on opposite sides from the 
central one, with which they anastomize, at certain intervals, by 
nieans of obliquely ascending, tender branches. The bulk of the 
stems and petioles consists of large parenchyma cells which have 
very large intercellular channels between them, in the manner of so 
n^auy aquatic plants. We find, stretched obliquely across these air- 
canals, some peculiar diaphragms formed of rather flat cells and 
pierced with numerous triangular intercellular spaces. Such dia- 
phragms are considered as a mechanical contrivance to stiffen the 
^alls of the air-canals in which they grow. Their occurrence in Litn- 
^(tnthenmm is particularly interesting, because together with them are 
^oiind those star-shaped cells so frequently observed in Nuphar 
and Nymphcetv; for it has heretofore been thought that plants "Raving 
fliese stellate ^'interior hairs" are not provided with any diaphragms.* 
'I'he manner in which the former grow, leaves, with me, no doubt as 
to their office. 
The walls of the intercellular channels (Fig. 8, p) are formed of 
single layers of cells among which the star-shaped ones are conspicu- 
ous by their great number, their form and their very thick, lignified 
V^ls. There are, however, not any of the granular crystals of cal- 
c>^im oxalate imbedded in the substance of the wall and projectmg 
fj*om it as with Nuphar and Nymphaa,\ The several arms mto 
which such a cell branches project into the intercellular passages 
l^'^^gs. 8 and 9), and spread in front of the partitions (Fig. 10) into 
which the main body of the cell is inserted. Considering the length 
G. Haberlandt, Physiologische Pflanzenanatomie (1S84), p. 300 
*uh such ce!U as those oi N. alba described by Strasburger, /. c. 
