(248) 
certaining how many rows of secondary cells, if any, lie between 
two successive primaries. There is yet another aid. The central 
one of the three superficial cortical node-cells sometimes is small, 
sometimes large, but very often grows out into a papilla or by 
further elongation into a spine-cell which may be very long and 
conspicuous. Spine-cells are never borne except at a cortical node- 
cell, hence they are often of assistance in locating the latter. Spine- 
cells usually occur singly, more rarely in pairs or threes, and become 
more scattered as the cells become mature through the elongation of 
the intervening cortical internodes. They can never occur except 
upon corticated stems. 
It is unusual to find such perfect or regular development, and 
deviations from this type are of high diagnostic value in separating 
species or at least groups of species. In some cases secondary cells 
develop on both sides of a primary but grow very much more in 
one longitudinal direction than the other ; sometimes they are formed 
on one side only ; sometimes no secondary cells occur, in which case 
the spine-cells are usually very well developed. The first of these 
is undoubtedly a case of abnormal triple cortication, but as the num- 
ber of cortical cells seen in a cross-section would be about twice that 
of the leaves it has sometimes been described as double; the second 
case is true double cortication; the last single. All these would 
cause a reduction in the number of cells seen in a transverse section; 
at times, however, the ends of two secondary cells slip past one 
another, thus increasing the number between the primaries for a 
part of the distance. In most species when they differ the primary 
cells are larger or more prominent than the secondary; sometimes 
the reverse is the case. 
The cortex of the leaves is much simpler. The number of rows 
is either once, twice, or thrice that of the leaflets; but all the 
cortical cells of the leaves are derived from the leaf-nodes, for the 
cells sent upwards and downwards do not form nodes and inter- 
nodes but remain undivided, meeting similar cells from the next 
underlying or overlying nodes near the middle of the internodes. 
No cortical cell, however, grows upwards on the leaf from its basal 
node, so that the lowest leaf-internode is corticated only by the cells 
descending from the leaf-node above it; in some cases these too are 
wanting and the lowest node of the leaf remains uncorticated. The 
terminal internodes of the leaves are often uncorticated. 
The permanent plants are anchored by long hyaline rhizoids 
1 ch grow downwards from the lower nodes of the stem. 
w 
