Scott and Hill.—Structure of Isoetes Hystrix. 427 
tions completely confirm Russow’s statement, but give it 
greater precision. The phloem of the leaf-traces (as to the 
nature of which there is no longer any doubt) passes over 
directly into those latticed elements of the intracambial zone, 
which alone constitute the true phloem of the stem (Fig. 12). 
At the point of junction, transitional elements occur, which 
belong equally to the leaf-trace and to the cauline phloem. 
As we follow the stele downwards, we find that its primary 
wood becomes more and more disorganized, the tracheides 
breaking down so completely that they must become perfectly 
functionless. The parenchymatous elements of the wood 
remain living, and appear even to grow in length, forming 
a trabecular network, in the meshes of which the remains of 
the disorganized tracheides are scattered. Possibly a pull may 
be exerted on the central wood by the leaf-traces, when 
stretched in consequence of secondary growth (Farmer, 1890, 
p. 41), but if so it has little effect in enlarging or distorting the 
cylinder, which is smaller in the lower than in the upper part 
of the stem, and usually shows little change from its normal 
cylindrical form. 
The change to the triangular section of the basal part of 
the stele, as above explained, is due to the development of the 
adventitious roots. In /. Hystrix , the roots are developed 
along three lines, corresponding to the three furrows. The 
root-bearing furrows start from the centre of the base of the 
stem, and run outwards and upwards, dying out below 
the region of the living leaf-bases (see Plate XXIV, Fig. 29). 
The three prominent arms, seen in transverse sections of the 
lower part of the stele, correspond in position to the three 
furrows. 
The structure of the basal part of the stele in Isoetes depends 
so completely on the distribution and development of the 
roots that a few words on this subject are necessary here, 
though the structure of the roots themselves will be deferred 
till later. 
The order of succession of the roots of Isoetes appears to 
have been first correctly made out by Hofmeister. We worked 
