230 RECHERCHES SUR LE TRADESCANTIA VIRGINICA. 
M. Gravis’s description is extremely lucid, and his pea Indg- 
most cases excellent. A detailed discussion of his results 
interesting of M. Gravis’s results is in the matter of the anatomical 
structure of the seedling, a topic of great and, as we think, of 
rar hima importance in morphological anatomy. In the Ete 
mbryo there is already a central cylinder sharply marked off fro 
the surrounding tissue in the region of radicle and hypocotyl. The 
seedling possesses a hypocotyl which, though very short when 
germination takes place in full li ght t, has a well-marked structure 
of its own. There is, as usual, a well-defined central cylinder, 
surrounded by a typical en Se a “aipeaily continuous with that 
of =~ root ; but the stele is distinctly bilateral about the plane of 
metry of the cotyledon, and exhibits a mixture of, rather than 
a kramaition between, the characters of root and stem. It contains 
undles which descend from the first epicotyledonary leaf, 
the remaining two, with precally developed xylems, placed right 
and left of the plane of symmetry, and continuous wit <i 
bundles of the sonnets petiole. All five xylems, though nore 
distinct in origin, eventually come into contact, forming an irregular 
xylem mass in the centre of the stele. The three former (first leaf- 
hese 
It is clear that here the general statement of Gérard as to 
connection of the vascular strands of root and stem, embodied 
ieghem’s Traité and usuall wae as of universal coals 
cation, does not fairly represent the facts. The symmetry of the 
root is not completely “determined” a the cotyledon-traces, as in 
the typical case among dicotyledons, but incompletely by first leaf 
and cotyledon-traces conjointly. Further, Gérard’s account of the 
passage of the xylem-strands from the centripetal (exarch) arrange- 
ment found in the root to the centrifugal ae characteristic of 
the stem as a “torsion” through 180° is not in this, as indeed it is 
not in many other cases, a happy or pit ts e one. Sometimes it 
applies well enough. For instance, where a xylem-strand of the 
root-cylinder doubles as it is traced upwards, each half may fairly 
be said to rotate through 180°, till the protoxylem which was at first 
directed away from, finally becomes directed towards the centre of 
the stele; but in others, where no such division takes place, the 
protoxylem plunges, so to speak, through the pg near valsoaue and 
comes out the other side wiekase others again | Trad: 
one or more of the root-stele protoxylems are not a up pink 
