178 BOTANICAL GAZETTE [Serre 
one considers the adaptations which these plants show; but sey 1 - 
plants of both families show characters which, to say the least, are 
significant. es 
Concerning VANTIEGHEM’s types, the polystele need not be con- 
sidered, for no monocotyledon has yet been found with internal 
phloem; the medullated monostele may be present in such forms as 
Acorus and Smilacina, but the condi.ion may be equally well explained 
by assuming the degeneration of an internal phloeoterma, deriving 
this condition from that shown in Clintonia; what may be called an 
astele or schizostele is probably present in the mature stem of most 
members of the two families, but in none of the cases examined does 
t arise by the breaking up of the stele followed by the uniting of the 
broken ends of the external endodermis on the inner side of the — 
meristeles; on the contrary, wherever the endodermis is discernible 
in the region of splitting up of the stele, there is an internal as well as 
external endodermis which communicate at the leaf gaps (¢ 5 
Clintonia) ; in Symplocarpus each strand which turns into the medulla 
is surrounded by a portion of the internal endodermis. 
Turning to the theory of JEFFREY, a consideration of the figures : 
We 
which accompany this paper shows that there is evidence in the case 
of the two families studied to support his fundamental statement that ag] 
the siphonostelic type of central cylinder “is primitively a fibro 
vascular tube with foliar lacunae opposite the points of exit of the — 
leaf traces” (6, p- 38). That the simple tubular condition is found 
for only a few internodes in most cases is due to the monocotyledons 
having acquired a new mode of insertion of the leaf traces, which has 
replaced the mode characteristic of ferns. However, in rhizomes, 
whose subterranean position has shielded them from the disturbing 
effects of aerial life,a more primitive type of stele is frequently found; : 
seedlings almost universally show a gap in the central cylinder above — 
the point of exit of the cotyledonary trace, unless indeed they are 
Protostelic at this level, as in Trillium. The siphonostelic nature of 
the central cylinder is often retained for several internodes, ae 
sooner or later the medullary strands appear, or the gaps. persist ft 
through an entire internode, in either case resulting in the masking © 8 | 
the essentially tubular nature of the stele. That the young © pe 
cylinder of so highly organized a group as the monocotyledons $ co 
