OCTOBER 23, 1913] 
the real homology between the vascular systems of 
stem and root throughout all vascular plants. This 
was pointed out to me more than twenty years ago by 
Dr. D. H. Scott, and it has been the sheet anchor to 
which I have since clung through much stress of 
morphological weather. No difficulty arises so long 
as we are dealing with roots only, or with the stems 
of those vascular Cryptogams in which the vascular 
system is a closed cylinder, without gaps at the inser- 
tion of the leaf-traces. In such stems the vascular 
cylinder is as well-defined as in all roots, and can be 
described in the same terms. But the case is quite 
different in the stems of Phanerogams, where to all 
appearance the primary vascular cylinder is a system 
built up of leaf-traces, embedded in a parenchymatous 
matrix. And the early anatomists were faced at once 
by this problem in its crudest form. Beginning with 
the anatomy of Phanerogams, they first became 
acquainted with the primary structure of the Dicotyle- 
donous stem. That of the root was not clearly under- 
stood until many years later; perhaps because 
anatomists attempted to interpret it by reference to 
the skeleton of the stem, and in the same terms. 
But there is nothing in the vascular anatomy of the 
root to correspond with the leaf-trace, and the leaf- 
trace is the vascular unit of stem-structure in all 
Phanerogams. Here, as elsewhere, confusion of 
nomenclature went hand in hand with confusion of 
thought, and it is difficult to say which was cause and 
which effect. 
Even when the facts of root-structure were 
accurately known, the conception of the leaf-trace 
bundle as the structural unit continued to be a 
stumbling-block. In 1877 De Bary published his 
monumental work on plant anatomy, and though it 
still keeps its place as the great book of reference 
on that subject, his descriptions of root anatomy 
appear to the modern botanist to be written in a dead 
language. When he calls the vascular axis of the 
root a “radial bundle” it is quite clear that he 
regards this as a purely formal term, not implying 
any true homology between the leaf-trace bundle of 
the stem and the axial core of the root. He does not, 
indeed, consider a bundle as a unit: he defines it 
as a compound structure ‘‘formed of tracheids and 
sieve-tubes definitely grouped.”? But the word 
‘““bundle’’ was already impressed with another super- 
scription. However defined originally, it had con- 
noted the unit of stem-structure to a generation of 
botanists. With that connotation, De Bary’s use of 
the term is in hopeless conflict. Moreover, the con- 
ception underlying that use was already out of date 
in 1877. Modern anatomy dates from 1871, when 
Prof. Van Tieghem published the first of his great 
series of memoirs on the subject. In these the axial 
core of the root was treated as equivalent to the whole 
system of leaf-trace bundles in the stem, though the 
word “stele”’ was not yet invented. This conception 
gained ground from the first; it was popularised by 
the happy choice of a name in 1886. From that date 
the stelar hypothesis has replaced all other schemes 
of vascular anatomy. The advance then made on all 
previous generalisations has been shown by the new 
impulse given to research, and the comparative sim- 
plicity introduced into text-book anatomy. 
We cannot claim equal simplicity, I fear, for the 
technical language of research in this subject, and 
this alone should inspire caution, for obscurity of 
language rarely persists where there is no correspond- 
ine obscurity of thought. 
No one now doubts that the central cylinder of the 
root in Phanerogams is far more closely comparable 
2 ‘Comparative Anatomy of Phanerogams and Ferns.” ist Eng. ed., 
1884, Pp. 400. 
NO. 2295, VOL. 92] 
NATURE 
245 
to the leaf-trace cylinder of the stem than to any one 
of the traces within it. Yet when the comparison 
becomes detailed, difficulties are constantly arising. 
Where, for example, there is a medulla in the root it 
certainly forms part of the stele, which is a solid 
cylinder sharply defined by the specialised endodermis 
surrounding it. But the leaf-traces in the young stem 
surround a massive cylinder of parenchyma, precisely 
resembling the parenchyma of the cortex, with which 
it is in apparent connection through the gaps between 
the leaf-traces. Even the secondary formations do 
not completely divide one system from the other. 
When a specialised endodermis is present it is not so 
clearly defined as in the root: in many cases it is 
not present—in other words, there is no cell-layer out- 
side the leaf-trace cylinder which is differentiated in 
any way from the surrounding tissues. In a few 
instances—most baffling of all—an endodermis sur- 
rounds each leaf-trace. 
The stele in the stem of Phanerogams is not of 
necessity a morphological fiction, because in many 
stems its precise limits cannot be determined. If, 
indeed, the word be used as a descriptive term, its 
value is seriously impaired by every instance in which 
it fails to describe stem-structure with precision. But 
morphology is not merely descriptive. If we suppose 
that the stem-stele in remote ancestors of the Phanero- 
gams was as well defined as that of the root and 
clearly comparable to it, we may attach a real morpho- 
logical meaning to the term when applied to modern 
Phanerogams, provided we can show cause to believe 
that what we call the stele in their stems represents 
the ancestral stele. Its tissues will then have a his- 
tory distinct from those of the cortex, though not 
clearly separated from them. The burden of proof, 
however, certainly lies with those who assert that an 
apparently continuous and uniform tissue can be 
separated into two parts of distinct origin. 
The evidence advanced is of two kinds—one founded 
on the comparative anatomy of stems, and the other 
on the history of the tissues in the individual plant. 
Dr. Schoute has argued the case with great skill from 
the first point of view in his “ Stelartheorie.’’ Depend- 
ing to a large extent on his own researches, he has 
collected a great body of evidence to show that in the 
stems of Angiosperms a specialised layer is commonly 
distinguished from adjacent tissues either by the 
peculiar thickening characteristic of the endodermis 
in the root, or by the presence of starch in its cells. 
He shows that such a sheath surrounds the vascular 
cylinder in a very large proportion of the Dicotyle- 
dons examined, and in a majority of the Mono- 
cotyledons. Among Gymnosperms it occurs but 
rarely. Observing that the Angiosperms in which this 
bundle-sheath is obscure or wanting are commonly 
closely related to species in which it is perfectly well 
defined, Dr. Schoute concludes that its absence in 
such cases must be attributed to reduction. 
Allowing that such a layer is as general among 
Angiosperms as Dr. Schoute believes, grave doubts 
may still exist as to its homology with the endodermis 
of the root. The latter is defined not only by its 
thickened walls, but also by the position of its cells. 
They form the inmost rank of the series of radial 
files which distinguish the inner cortex, and the 
morphological endodermis—the phlaeoterma, as Stras- 
burger calls it—can usually be distinguished by this 
purely morphological character, even when its walls 
are unthickened. In the stem, however, the cells of 
the inner cortex are not radially arranged, except in 
rare cases, such as Hippuris. Thus there is no 
morphological criterion to distinguish the phlaoterma, 
or inmost cortical layer of the stem, from adjacent 
tissues. The bundle-sheaths distinguished by their 
