Scott and Hill.—Structure of Isoetes Hystrix . 433 
species 1 . As Kruch pointed out, the wall lining the cavity 
is distinctly suberized. As Janczewski first showed (1882) 
and Kruch demonstrated in detail (1890, pp. 61, 70), the 
canals in the foliar bundle of Isoetes represent disorganized 
xylem-elements; the large central canal takes the place of 
the actual protoxylem, the first-formed row of tracheides 
(Fig. 20 ,px). These are of rather large size, with transverse 
walls, which persist for a long time, and are pulled widely 
apart by the growth in length of the leaf, so that in a 
microtome-series from a partly-developed leaf, one wall was 
seen about every hundred sections. These transverse walls or 
diaphragms are lignified and reticulated, and are perforated, 
at least when old (Fig. 21; cf. Fig. 22 ,px). Strasburger has 
shown that these canals in the bundle of Isoetes contain water, 
so it appears that they continue to perform tracheal functions. 
The tracheides corresponding to the lateral canals develop 
rather later, and so become less disorganized. In the upper 
part of the leaf they usually persist as tracheides, so that here 
there is only the one median canal, and it is then only 
around the latter that the endodermis is differentiated. The 
first-formed tracheide, as we have seen, is the outermost, so 
that here the wood is developed centripetally (with reference 
to the centre of the stem), as Russow states. 
The protophloem, which is differentiated earlier than the 
protoxylem, forms a short band of apparently empty elements, 
with deeply-staining cellulose walls (sieve-tubes according to 
Kruch), lying immediately opposite the first tracheide (Fig. 20, 
pph). This median phloem becomes obliterated, and it is 
only the later-developed sieve-tubes of the lateral phloem- 
bands which remain functional. Kruch (1890, p. 70) and 
Strasburger (1891, p. 463) state that new sieve-tubes are 
added to the lateral bands, towards the dorsal (outer) side, 
and this appears to be the case. 
The sieve-tubes, though small and difficult to investigate, 
1 The endodermal markings are not shown in our figures, as the microtome 
preparations were not favourable for demonstrating them. They were quite clear 
in sections mounted in glycerine. 
G g 
