248 
the stele. Before, however, entering on this subject, 
I ought to say something on the .question of fact. 
In my opinion M. Chauveaud’s figures and descrip- 
tions represent the vascular development in the hypo- 
cotyl and cotyledons of Angiosperms more accurately 
than any others with which I am acquainted. He 
has dealt more fully with Dicotyledons than Mono- 
cotyledons, but I have been able to verify his account 
of the latter to some extent by reference to my own 
preparations, which include a number of species 
closely allied to those which he has cut. Among 
Dicotyledons I have had the great advantage of con- 
sulting the preparations of Miss Thomas and of Mr. 
R. H. Compton. Neither of these botanists made 
their preparations to illustrate M. Chauveaud’s 
theory; indeed, they attacked the subject, as I did, 
with aims distinct from his. Yhere has _ therefore, 
been nothing like complete verification in any single 
species, but so remarkable a correspondence with his 
figures in similar stages of the material, that | am 
satisfied of M. Chauveaud’s fidelity. 
Assuming, then, that his account of the vascular 
development in a young seedling is substantially cor- 
rect, what are we to conclude as to the homology of 
the central cylinder of the stem with that of the root? 
M. Chauveaud himself believes the stem cylinder 
in the upper hypocotyl of a fairly old seedling to be a 
true stele, but one belonging to a later phase of 
evolution than that of the root, and not, therefore, 
strictly homologous with it in the sense in which the 
earliest vascular formations in cotyledon and hypo- 
cotyl respectively were homologous with each other. 
He considers the successive vascular formations which 
we have just followed in these regions —formations 
marked by the appearance of alternate, intermediate, 
and superposed xylem in turn—to represent three 
successive phases of stelar development. The root- 
stele corresponds to the first of these phases only. 
But questions of phylogeny are strictly historical, 
and the only precise meaning that can be attached to 
the expression ‘“‘successive phases of stelar develop- 
ment’ in the seedling of an Angiosperm is that at 
some past period a group of plants in the direct line 
of descent of Angiosperms possessed a stele re- 
sembling that which is now a mere stage in the life 
of the individual. Thus the alternate formation found 
throughout the very young seedling implies an 
ancestral group with an exarch steie in stem as well 
as root, and a leaf-trace of corresponding structure. 
There is nothing at all improbable in this hypo- 
thesis, since groups with exarch steles in stem as 
well as root are found among living and extinct 
plants. But if adopted several important consequences 
would follow. The seedling while it consists of 
cotyledons, hypocotyl, and primary root only—the 
plumule present as a mere bud—must represent a past 
period in race-history when its ancestors possessed 
an exarch stele in stem and root alike; when the 
stem-stele belonged to the stem only, and the insertion 
of leaf-traces hardly modified its structure; when it 
entered the root without change, and therefore no 
transitional region occupied and puzzled the anatomist 
of the period. 
This early stage in the development of the seedling 
is succeeded by that in which the epicotyl begins to 
grow, and as a rule the epicotyl is quite undoubtedly 
modern.® Its vascular skeleton is built up of leaf- 
traces, which are endarch from the first. At the 
cotyledonary node they are inserted on the vascular 
6 The epicotyl in the /icfea has been described as containing centripetal 
xylem. The facts are given by Mlle. Goldschmid (1876), Prof. Gérard 
(1881), M. Chauveaud (1911). and Mr. Compton (1912). The theoretical in- 
terpretation is discussed by M. Chauveaud (¢:c., p. 348) and Mr. Compton 
(Zc., pp. 93-95). There seems every reason to think, as Mr. Compton 
suggests, that the character is adaptive, depending on the twining habit. 
He points out that Victa Kaba, which does not twine, has a normal epicotyl. 
NO. 2295, VOL. 92] 
NATURE 
[OcToBER 23, 1913 
cylinder of the hypocotyl, which has become endarch 
at the top. This transition has been effected lower 
down in the hypocotyl, as described already, by 
the formation first of intermediat@, and then of super- 
, posed xylem, together with the gradual disappear- 
ance of the original alternate xylem. 
Thus the cotyledonary node may be considered to 
mark the interval between two acts in the drama of 
evolution—an interval the length of which cannot yet 
be estimated, but is clearly to be reckoned in geo- 
logical epochs. 
The race-history of the Phanerogamic stem-cylinder 
is at present unknown. How did the ancestral stele 
lose its exarch character, and what intermediate stages 
led up to its present construction from endarch leaf- 
traces? Possibly the development of the hypocotyl 
may give a clue as suggested by M. Chauveaud, 
and the change have been effected by the develop- 
ment of intermediate xylem. Or Prof. Jeffrey may be 
right in deriving the leaf-traces from a siphonostele 
which has been gradually more and more broken up 
by the appearance of foliar gaps. This process is 
said to be exhibited in the young epicotyl.’ Until this 
point is cleared up the exact relationship of the 
vascular cylinder of the stem to that of the root will 
remain obscure. As a matter of convenience the stem- 
cylinder will, no doubt, be called a stele, even though 
anatomists should acknowledge that it cannot be con- 
sidered as strictly homologous with the stele of the 
root. Much confusion of thought would, however, be 
avoided if the two structures were not treated as 
strictly comparable. 
There can be very little doubt that the insertion of 
leaves has brought about the change, and I might 
suggest here that the insertion of leaves on an exarch 
stem-stele would be an interesting subject for re- 
search. The literature of the subject is scattered, 
and its treatment seems to me very incomplete. An 
exarch axis bearing leaves is, of course, exceptional, 
but more common among extinct plants than among 
recent species. So far as my very cursory examina- 
tion of the literature has gone, it seems a general 
rule that the leaf-traces are inserted on the xylem 
poles of the stele.§ 
Hitherto I have considered modern embryology in 
relation to a single problem of internal anatomy— 
namely, the comparison of the vascular system of the 
stem to that of the root. But the evidence of em- 
bryology is also of great weight in questions of 
internal morphology and phylogeny. : 
Several questions of this kind are discussed by 
Hanstein, from whose classical paper I continue to 
date. For example, his account of the embryo of 
Monocotyledons suggests two distinct problems. One 
belongs to formal morphology—namely, the question 
whether a terminal member can be considered as a 
leaf. The other is a question of phylogeny: whether 
Dicotyledons are derived from a monocotylous ancestor 
or Monocotyledons from a dicotylus form. Both these 
questions I have discussed elsewhere,* and only refer 
to them now as examples of the way in which 
seedling anatomy has proved complementary to that 
of the older embryologists. 
The most obvious interpretation of Hanstein’s ob- 
servations is that the single cotyledon of Mono- 
cotyledons is equivalent to the pair found in Dicotyle- 
dons. This would imply that Dicotvledons were 
7 Jeffrey, ‘‘ The Morphology of the Central Cylinder in the Angiosperms.” 
Trans. Canadian Inst.. vi., 1900 
8D. H. Scott, ** Studies in Fossil Botany,” 1908. p 97 (Sphenophylium) 5 
C. E. Bertrand, ‘‘Rema-ques sur le Lesidodendron Harcourtii,’" 1801, 
p. 109; M. Hovelacaue, ‘‘ Recherches sur le Lepidodendron selaginoides,” 
1892, p. 150; F. O. Bower, ‘‘ Origin of a Land Flora,” p. 334 (Selaginella), 
1908 ; C. E. Jones, Trans. Linn. Soc., ser. 2, Vii., 1905, p. 19 (Lycopodium). 
9 E, Sargant, dun. of Bot. xvii., PY, 1903. and 7d xxii., pp. 150-2, 1908. 
