312 BOTANICAL GAZETTE [OCTOBER 
In the internodes of the seedling, this structure, as described 
above, occurs in the apical internode, and with no indication, so 
far, of any development of secondary hadrome inside the leptome. 
To: trace this, the origin of the secondary hadrome, we must 
examine the next internode, the one below the apical, shown in 
fig. 9. Here are seen the innermost strata of the cortex (C) and 
the endodermis (End); also, on the left side of the section, a part 
of one of the primary mestome strands (M) has been drawn, con- 
sisting of leptome (Z) and a short ray of vessels. The broad inter- 
fascicular tissue consists, on the other hand, only of several well 
defined leptomatic strands, and of a tangential cell-division within 
the large, thin-walled parenchyma inside the leptome. There is 
actually no indication of any distinct pericycle in this section either, 
but there is certainly a very pronounced indication of the presence 
of a meristem in the periphery of the primary parenchymatic 
(medullary) ray. The result of these tangential cell-divisions 
within the leptome may be seen in fig. 10, which has been drawn 
from the epicotyl of the seedling. It consists of the development 
of thick-walled libriform in such a way that the innermost tangential 
cell wall becomes lignified, and very rapidly so, while the outermost 
is still meristematic, so as to give rise to another libriform cell in 
the same radius as the first developed. There is no distinct ring 
of cambium, therefore, since the innermost cell wall of each cell, 
which divides tangentially, becomes lignified almost at once, while 
the outermost remains active so. as to yield another cell to the 
secondary hadrome, and always in the same radius. 
In comparing the structure of these internodes of the seedling 
with that of an old internode of a mature specimen, which is drawn 
in fig. 15, we notice the continued growth of the interfasciculat 
hadrome (#) as many radial layers of thick-walled libriform, while 
the leptomatic strands are about the same, and located in 
walled parenchyma; no pericyclic sheath seems to be differen- 
tiated in this section either. This internode (fig. 15) was from the 
aerial portion of the stem, and examining the structure of the 
pseudo-rhizome, we notice a marked difference (fig. 16) consisting 
in the development of a partly stereomatic pericycle outside the 
primary mestome strands (s in fig. 16). As shown in this figure, 


