Differentiation and Pattern in Monstera deliciosa. The Idioblastic Development of the
Trichosclereids in the Air Root
Author(s): Robert Bloch
Source: American Journal of Botany , Jun., 1946, Vol. 33, No. 6 (Jun., 1946), pp. 544-
551
Published by: Wiley
Stable URL: https://www.jstor.org/stable/2437589
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Wiley is collaborating with JSTOR to digitize, preserve and extend access to American Journal
of Botany
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
-DIFFERENTIATION AND PATTERN IN MONSTERA DELICIOSA. THE
IDIOBLASTIC DEVELOPMENT OF THE TRICHOSCLEREIDS
IN THE AIR ROOT'
Robert Bloch
To THE student of both normal and abnormal
cell behavior anything which can be learned about
conditions under which cells differentiate (develop
different properties during ontogenesis) is of theoretical
and practical consequence. At the plant
growing point embryonic cells in all likelihood start
out as morphologically very similar and genetically
identical totipotent units, although probably no two
cells in the organism remain exactly alike for long;
very soon cells can be distinguished which exhibit
specialized combinations of various cellular properties
and activities (as to growth, membrane
structure, contents, and physiological character).
Usually such cells differentiate in association with
the formation of specific tissues in typical locations
within the plant organ; sometimes, however, they
are arranged in less regular fashion, namely, as
individuals scattered through the ground tissue, to
which Sachs applied the term "idioblasts." Among
the latter those cells appear of special interest which
occur as solitary, actively proliferating elements
among the more static and regularly shaped elements
of the fundamental tissue. Notable among
these cells, which often combine developmental and
structural features of bast fibers, sclerotic parenchyma
cells, hairs and other forms (for classification
see Tschirch, 1885; Foster, 1944), are certain
actively growing, prosenchymatous and branched
sclereids which ramify between elements of the
ground tissue, such as the well-known astrosclereids
of leaves (de Bary, 1884; Solereder, 1908; Foster,
1944). In a number of plants these cells are known
to develop by proliferation of internal initials abutting
on the intercellular space system as in
the Nymphaeaceae and Menyanthoideae (Giirtler,
1905), the Araceae (van Tieghem, 1866), and Rhizophora
(de Bary, 1884; Gfirtler, 1905); where the
lacunae are large, the development of these cells reminds
one of that of external hairs from an epidermal
surface. The terms "internal hairs" (Meyen,
1837) and "trichoblasts" (Sachs, 1882; Giirtler,
1905) applied to these cells refers to this property;
they are also frequently described, together with
other kinds of idioblastic sclereids, as "spicular
cells" (Solereder, 1908). Both the terms trichoblast
and spicular cell are also used, however, to denote
other kinds of plant cells; the present writer, therefore,
proposes the use of the term trichosclereid for
such forms as are described in this paper. These cells
can be distinguished clearly both morphologically
and developmentally from the brachysclereids which
are differentiated in the hypodermal region of the
cortex of Monstera air roots during later development
(Bloch, 1944).
1 Received for publication February 8, 1946.
Comparatively little is known about the induction
and development of trichosclereids and other
forms of idioblastic sclereids. Gurtler (1905) studied
their development in Nymphaeaceae and other
plants; Foster (1944, 1945) has recently investigated
the development of sclereids in compact tissue
of the petiole of Camellia 'aponica, as well as in the
more highly lacunate tissue of the lamina and petiole
of Trochodendron araloides. Characteristically in all
cases the idioblasts developed from small initials
which, however, did not occur in any definite pattern;
in Trochodendron, according to Foster, all
spongy parenchyma cells retain the capacity to form
sclereid initials. The most outstanding feature of the
sclereids in Camellia and Trochodendron appears to
be their capacity to grow under different intercellular
conditions. Thus in Camellia, cell processes intrude
between densely arranged collenchyma or
parenchyma cells and cease growth when a large
space is entered, while in Trochodendron, where
sclereid formation occurs over a relatively long period
of time, the branches of the sclereids grow
through spaces and intervening tissue alike. The
potency here of considerable portions of the primary
wall to grow out would appear as a property which
these cells share with those of the ordinary parenchymatous
type which retain the capacity for wall
growth for considerable time. It appears significant
that in Camellia, where the growth is almost entirely
intercellular, but in close contact with the parenchyma
walls, pit connections are apparently re-established
between the walls of parenchyma cells and
sclereids, much as in other intrusively growing
fibers, while in Trochodendron pits are restricted
to the central part of the sclereid which remained in
contact with parenchyma cells.
The present writer has worked with the long,
pointed, fibrous and more or less branched cells
which ramify in air spaces of various tissues of Araceae,-
especially the Monsteroideae. These have been
described several times on account of their very peculiar
and variable forms (van Tieghem, 1866;
Wiesner, 1875; Lierau, 1888; Gurtler, 1905; Solereder,
1919; Solereder and Meyer, 1928), but relatively
little is known about their development. Van
Tieghem (1866) mentions their origin from small,
parenchymatous initials which form processes into
the adjacent intercellular space system. These processes
may be seen readily enough in transverse sections
through mature regions of the mesophyll of
leaves or the cortex of roots and air roots (fig. 8);
the early origin of the initials, however, requires
more intensive study. The author noticed previously
(1944) that in the air root of Monstera deliciosa the
initials of these sclereids appear very early near the
544
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
June, 1946] BLOCH-MONSTERA DELICIOSA 545
apex of the root meristem, and that these initials
send out fast growing processes which grow upward
and downward between the parenchyma cells. The
tissue in which they arise is a rib meristem of considerable
symmetry, and such tissue would appear
to have certain natural advantages for the study of
early intercellular relationships in idioblastic de-
velopment.
MATERIAL AND TECHNIQUE.-Tips and mature
parts of air roots of Monstera deliciosa were killed
in C R A F and the development of trichosclereids
studied from the youngest stages onward, both in
longitudinal and transverse sections, 12-20 IL in
thickness, stained with safranin and fast green. Air
roots of various sizes, from 1.5 to 15 mm. diameter
and of varying length, were studied. Young stages
of trichosclereids in hand sections of fresh material
may be profitably studied after staining of the young
cell walls and cytoplasm with ruthenium red. The
form and length of fully grown sclereids were
studied after maceration and mechanical isolation.
THE, DEVE'LOPMENT OF THE MOTHER TISSUE AND
INTERCELLULAR SPACE SYSTEM.-The development of
the sclereids is so closely related to changes in the
cortical parenchyma that it is advisable to describe
these first. Both transverse and longitudinal sections
were studied. At a distance of about 700 K from the
apex of the cone of the meristem the cortical parenchyma
cells have begun to vacuolate, and show in
longitudinal sections the typical- configuration of a
rib meristem. In it numerous files of cells may be
distinguished, each of which is made up of the derivatives
of one meristematic cell of the cortex, now
consisting of 3-5 cells (fig. 3). It will be noted that
these files are not always exactly continuous, but
occasionally stagger or alternate somewhat; as the
cells vacuolate and expand longitudinally; however,
this discontinuity becomes less distinct.
The intercellular space system begins to develop
very early and may be recognized in both longitudinal
and transverse sections. A number of channels
may be seen to run longitudinally between the files to
within a short distance of the apex, and in the longitudinal
sections small intercellular spaces may also
be seen at the corners where the cell files abut on
each other (fig. 3). Both the somewhat tapering
shape of each file and the presence of air spaces
where the files abut on each other, make it easy to
recognize the boundaries of the original cortex cells.
Transverse sections show that the air space systems
undergo considerable changes during ontogenesis.
Already near the apex three kinds of spaces may
be distinguished: triangular, somewhat larger quadrangular,
and still larger pentangular spaces (fig.
1 a-c, 2). The number of air channels with more
than three sides increases as vacuolation and differentiation
of the parenchyma proceed, and is due to
various factors, namely (1) some cell divisions which
meet an air space (fig. 1 a); (2) cell divisions in
opposite position (fig. 1 b); and (3) particularly
early and later loss of contacts between cells (fig.
1 c-d). The air spaces finally become much larger
than the diameter of the trichome-like sclereid processes
which originally entered them (fig. 1 d, 8), and
in the case of later loss of contacts several air spaces
containing sclereid branches may coalesce so that
finally several branches which belong to different
sclereids are seen in the same space (fig. 1 d).
INITIATION OF TRICHOSCLEREIDS AND THEIR
GROWTH IN RELATION TO THE INTERCELLULAR SPACE
SYSTEM.-In the cortex of young air roots, varying
in diameter from 1.5 to 3.0 mm., polar, differential
divisions occur, often as near as 700 y from the apex.
These divisions cut off one small cell, and very frequently
a pair, at the basal (proximal pole) in many,
but not in all cell files (fig. 4, 5). In this manner a
distinct difference in cell size results. These last
formed small cells may be readily recognized, on
account of their dense and more deeply staining
cytoplasmic content and their nuclei which are relatively
large in relation to the cell volume (fig. 4, 5).
In roots with comparatively little meristematic activity
often only one small cell is-formed in one cell
file.
The small cell located at the basal (proximal) end
of the cell file becomes a trichosclereid initial or
-mother cell from which processes develop into the
adjacent parenchyma. Generally this cell is the only
one which develops into a trichosclereid. If there is
a second small cell present, it usually expands somewhat
longitudinally, like the cells of the ordinary
parenchyma; sometimes, however, both cells develop
into trichosclereids (fig. 9). Thus both relative
size, resulting from an unequal, differential division,
and relative position within a cell file characterize
the sclereid initials and make it easy to
locate them at their earliest stage, i.e., when the
cortical mother cells begin to divide. The foliar
sclereids of Camellia and Trochodendron, according
to Foster (1944, 1945), can be recognized only
by the larger size of their nuclei, but in the present
case a more determined type of idioblastic development
occurs, from divisions which are both differential
and bear a constant positional relation to the
cells of the mother tissue.
The sclereid initials abut on several longitudinal
air channels, usually one of them being a larger
four- or five-sided space (fig. 2, 4, 8). After very
little longitudinal expansion of the initial, tubular
processes begin to grow out from it which enter the
four-sided intercellular channel (fig. 2, 6). At this
stage the growth of the invading branch thus appears
free, like a root hair, or a tylosis pushing into a vessel.
The originally small branches grow with great
rapidity and also expand somewhat and thus often
reach the end of an intercellular channel (fig. 7).
Throughout their growth the trichosclereids are thus
characterized by the fact that they closely follow
the developing intercellular system and by their
ability to change their direction of growth, to branch
and to fork at their ends when obstacles of various
kinds are encountered. In this manner forms are
produced which closely resemble certain fibers, such
as those found in Cannabis and Luffa (Sinnott and
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
546 AMERICAN JOURNAL OF BOTANY [Vol. 33,
b
I ~~~~~d
Fig. 1-I1.-Fig. 1. Transverse sections through cortical yarenchyma of air root, showing origin of quadrangular and
pentangular air spaces by early cell divisions (a and b), bv early loss of cell contacts (c), and by later loss of cell contacts
(d).-Fig. 2. Transverse section through cortical pa enchyma, showing two trichosclereid initials, one of whiclh
sends out a process into a quadrangular air space. X 450Q Fig. 3. Longitudinal section through root meristem 500 A
from apex, showing files of cells. X24,0. (In all longitudiral sections the root apex is toward the bottom of the page.)
-Fig. 4. Similar section, showing formation of trichosc!ei-eid initials at basal (proximal) end of two cell files abutting
on longitudinal intercellular space. 700/A from apex. X240.-Fig. 5. Similar section; onle nucleus (upper left)
is in propbase. X240.-Fig. 6. Similar section, 2200 IA frorn apex, sbowing the formation of processes into an intercellular
space (compare fig. 2). X240.-Fig. 7. More advanced stage, in which the basal end of the process has reached
the end of an intercellular space; 2200 1u from apex. X 240. Fig. 8. Transverse section similar to figure 2, showing
both the initial part of the trichosclereid and several processes in the thickened mature stage. The initial part is
pitted. X195.-Fig. 9. Longitudinal section, showing parts of two mature trichosclereids which have developed from
divisions as in figure 4. X240. Fig. 1(0. Four comparatively short trichosclereids from macerated material, showing
diversity of branching and distribution of pits; the walls are somewbat swollen. X105.-Fig. 11. Trichosclereid, the
uppr rocssofwhih as oredtwobrnchs rond prenhya ell X05
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
June, 1946] BLOCH-MONSTERA DELICIOSA 547
Bloch, 1943), although these grow in much closer
morphological and physiological contact with the
adjacent parenchyma cell walls. The type of growth
observed in the Monstera trichosclereids appears to
be that formerly described as "intrusive," in which
growth occurs mainly at the tips of proliferating
portions of the cells, with no appreciable shifting or
sliding of larger portions of cell walls (Sinnott and
Bloch, 1939, 1943). In the inner cortical regions
trichosclereids occur whose processes follow closely
the course of the intercellular canals in zig-zag
fashion (fig. 12). The diameter of these intercellular
channels changes at regular intervals owing to
the curved surface of the adj acent longitudinal
parenchyma walls, and the trichosclereid processes
show corresponding wider portions. It was found
that this relation had not changed after maturation
and sclerification of the processes (fig. 11, 12) which
would indicate that there had been no slipping or
sliding of large portions of the processes within the
intercellular canal, any possible slidi'ng occurring
only at the tip. Furthermore, small lateral branches
and outgrowths, which protrude into the adjacent
parenchyma, are never subsequently torn out of
place.
The growth of the sclereid processes is free within
an intercellular space or in rather close conta'ct with
parenchyma membranes, but wherever an obstacle
in the form of a cell wall is encountered, the processes
are easily deformed and deflected from their
course. In the early stages the advancing arms can
frequently be seen in close contact with the middle
lamella (intercellular substance) of the parenchyma
cells at the end of an intercellular. space (fig. 7), and
in fully differentiated trichosclereids the very narrow
and pointed tips of long processes may occasionally
be seen in close contact with the parenchyma
walls. It is difficult, of course, to say whether and to
what degree the sclereid might actively contribute
to the separation of the primary membranes of the
cortex cells'. In Monstera this process seems to take
place relatively easily, and the intrusive growth of
the processes may be considerably facilitated by the
fact that it proceeds along the course of natural air
space formation. The wall of the processes seems
never to be in very intimate connection with the
parenchyma walls since it separates most readily
under various treatments.
Branching of the trichosclereid may he associated
with the formation of new air spaces adjacent to it,
or may be induced by obstacles encountered by the
processes, such as cell contacts, or by growing parts
of other trichosclereids. Mostly two or four processes
are seen to develop from' one initial (fig. 10);
H-shaped cells, known to occur in many species, are
common (fig. 15). Branching and bending of processes
around intervening cells are frequent (fig. 11),
and very irregular forms were figured by van Tieghem
(1866) from the leaves of Spathiphyllum lanceaefolium
and Tornelia fragrans. The tip of the
growing sclereid process may be induced to branch
by even a slight obstacle such as the young tip of
another process, growing in the same air space, but
in the opposite direction (fig. 14). Branching of
processes within an air space, however, does not
seem to occur as readily as in other types of freely
growing hairs, for example in some of the internal
hair cells described by Giirtler (1905).
Sclereid initials may remain dormant for some
time until suitable conditions for their further development
arise. They then may grow into a space
full of other more mature sclereid processes.
SECONDARY THICKENING OF THE TRICHOSCLEREIDS.
-In the mature stage the walls of the sclereid arms
are considerably thickened and the lumina reduced;
the cells possess pointed, tapering tips, much like
bast fibers, and the thickness of the secondary layers
decreases gradually toward both the tip and the
initial (fig. 13, left). The basal or central initial portion
of the cell is usually less thickened (fig. 13, 16),
a fact already mentioned by Wiesner (1875) for
sclereids of the mesophyll. The same author noted
an irregular thickening of the wall the external contour
of which, similar to the bast fiber of jute, is not
always parallel to the inner one. In the present investigation
the same irregularities were noted. Wiesner
also found that the secondary wall of the arms
shows some spiral striation, but no pits. In the present
study it was found that pits occur in the initial
portion of the cell which contains the nucleus, but
no pits were found in the branches. Obviously the
initial part of the cell retains from the beginning
close morphological and physiological contact with
the neighboring cells, similar to the sclereid of
Trochodendron.
As the air spaces grow wider, the growth in width
of the processes does not keep pace with them. In
Raphidophora decursiva, another species of the
Monsteraceae, however, the branches fill almost the
entire space, even in the mature, sclerotic stage,
showing the imprint of the outlines of the parenchyma
on their walls (fig. 17). Van Tieghem noted
the same in Raphidophora pinnata (1866; p. 153).
The mature walls of the Monstera sclereids are
impregnated with suberin, but not truly lignified.
They stain readily with Sudan III and safranin,
but do not give the lignin reaction with phloroglucin
and hydrochloric acid, unlike the sclereids in other
members of the Araceae.
Final thickening and suberization of the wall do
not take place until a considerable distance from
the root tip. This distance varies with the type of
air root and with its diameter; in a thick air root,
80 cm. long and 15 mm. wide at its base, the sclereids
at 20 mm. from the tip and 7 mm. diameter were still
thin-walled.
Striking differences in the onset of wall thickening
and chemical impregnation occur, not only between
individual sclereids, but also between the various
parts of the same 'cell. At the same root level sclereids
are found which are fully mature, while others
may be still thin-walled or occasionally even beginning
to develop (fig. 13). Furthermore, parts of the
long arms of one cell, often the tip or a portion some
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
548 AMERICAN JOURNAL OF BOTANY [Vol. 33,
13
16
1.5
12 14 1 7
Fig. 12-17.-Fig. 12-16 showing later stages of growth and differentiation of trichosclereids. X290.-Fig. 12.
Trichosclereid, showing relation between regions of lateral expansion in process and width of intercellular canal.Fig.
13. Three trichoselereids which have developed close to each other, the two at left are fully differentiated, the
one at right still thin-walled. Note that the secondary layers of the wall become thinner toward the tip and toward
the initial part of the cell.-Fig. 14. Two trichosclereids in the same intercellular canal, one process grows upward and
has split in two where it encountered the small process growing downward.-Fig. 15. Two H-shaped sclereids.-Fig.
16. Trichosclereid showing beginning of secondary thickening of the wall. The central initial portion is still thin-walled
and shows the nucleus.-Fig. 17. Transverse section. through process of mature trichosclereid is air root of Raphidophora
decursiva, showing collenchymatous thickening of parenchyma cells around original air space, quadrangular
trichosclereid process which fills the air space completely, and small quadrangular lumen of process. X360.
distance from it, may begin to thicken and change
chemically, while other parts are still thin-walled,
and this may account for the subsequent variation
in the thickness of the secondary layers at maturity
which were noted by Wiesner and the writer. In
fibers, such as those of Luffa, the growing tip is
generally the last to lignify; in the present case the
central, initial parts and adjacent areas of the
branches mature last (fig. 16).
The diameter and lumen of the trichosclereid
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
June, 1946] BLOCH-MONSTERA DELICIOSA 549
arms may also vary considerably, not so much because
of differences in the thickness of the secondary
layers, but on account of the previously described
expansion of the processes in regions where the intercellular
channels naturally widen. Especially in
the inner cortical region this occurs frequently, and
a peculiar zig-zag-shaped course of the branches
results. In the regions where these expand laterally
the secondary layers are frequently, but not always,
somewhat thinner (fig. 12).
All these observations seem to indicate that both
growth and differentiation are more variable and
irregular in sclereids than in fibers, bast cells, parenchyma
cells, etc. The process of wall differentiation
seems not to be entirely controlled by a gradient
which affects all cells at the same level evenlv, but
rather to be an individual reaction of the cell in relation
to local conditions. Such conditions must vary
considerably, since the cells are of considerable
length; cells 2-3.5 mm. long are not rare. Environmental
conditions would thus determine to some degree
not only the growth reactions in various parts
of the same cell but also the ultimate differentiation
of its wall.
DIsCUSSION.-The trichosclereids in the air root
of Monstera deliciosa are interesting in several respects.
The most striking feature is their origin from
nmother cells near the apex, formed by late, polar and
differential divisions, analogous to those occurring
in the hypodermis, at the basal ends of cell files of
the cortical parenchyma. The position and origin of
these "idioblasts" are thus not accidental, random
ones, but have a definite relationship to the cell pattern
in the mother tissue from which they arise. A
difference between sclereid initial and ordinary
parenchyma in cell size and in the relation of the
nucleus to the cell volume results, and highly embryonic
initials are formed adjacent to cells somewhat
more advanced in maturation, comparable
to the trichoblasts of the surface layer of grass
roots, which also form a characteristic pattern in
the mother tissue. The root hair cell, however,
later undergoes considerable longitudinal expansion,
but the sclereid initial grows very little, its
expansion being limited almost entirely to the tubular
branches. Developmentally the trichoselereids
of Monstera can best be understood if one remembers
that they are formed as densely cytoplasmic cells at
a time when their neighbors are already in the vacuolating
and expanding stage. Like the trichoblasts of
the epidermis they are located on surfaces abutting
on air, under conditions which apparentlv favor
specific expansion, growth and subsequent differentiation
of the wall.
Compared with the other cells of the cortical parenchyma
the development of the trichosclereids
would seem to indicate a highly specific nature as a
re'sult of differential reactivity. A similar difference
in behavior occurs between the root hair cells (trichoblasts)
and the ordinary cells in the surface
layer of roots, but in this case the distinction is frequently
not very sharp, and ordinary epidermis cells
may be induced to form root hairs also. In the present
case it was never observed that other cortical
cells were capable of developing into a trichosclereid,
while previous regeneration experiments
showed (Bloch, 1944) that, at a later stage of general
differentiation and developmental potency of
the ground tissue, any cortical parenchyma cell in
the air root of Monstera becomes able to differentiate
into a hypodermal brachyselereid, if the conditions
for their differentiation are experimentally
created which in a certain position exist adjacent to
an external or internal surface.
In other cases of idioblastic development of selereids
similar patterns and intercellular relationships
may exist, but may not always be as pronounced and
easily recognizable as in the case here described.
Foster (1945) in his description of sclereid development
in the spongy parenchyma of Trochodendron
has emphasized that all parenchyma cells appear
to retain the capacity to enlarge and to ramifv, and
that "the entire career of development of an idioblastic
sclereid is marked by a 'rugged individualism'
which suggests that the normal correlative
forces operative during the differentiation en masse
of parenchyma elements have been replaced locally
bv new and unique factors." It is significant in this
connection that the definition of the term i"idioblast"
used by Sachs and other authors for these and similar
cells does not refer to any rules or factors which
might control their distribution in the mother tissue.
Their scattered distribution makes them often appear
less typical a part of the general tissue pattern
than other more compact tissue systems. Adding to
this feature the characteristic "invasive" or rather
long-continued character of their growth, which distinguishes
them so sharply from their neighbors,
the impression may be gained that these cells possess
capacities specifically different from the majority
of plant cells in closely correlated cell aggregates.
However, the distribution and developmient
of the idioblastic trichosclereids of Monstera, which
have a close genetic relationship to a pattern of cell
divisions in the mother tissue, would appear to be as
much under organized control as other types of solitarv
cells, such as stoma cells, certain external hair
cells, and other forms which have not been considered
to be of idioblastic nature, since more was known
about their developmental history. In the growth of
the trichosclereids no phenomena were observed
which would be fundamentally different from those
already known to occur generally in the growth of
plant cells, various methods and possibilities of
which have been discussed in previous papers (Sinnott
and Bloch, 1939, 1943). It is known that many
plant cells are able to grow in several ways in response
to changing external conditions. A parenchyma
cell may show very restrained growth in most
types of tissue, under certain conditions its membrane
may exhibit considerable differential growth,
or may grow freely, for example as a tylosis, into an
adjacent lacuna. The trichosclereids of the Monstera
air roots arise from densely cytoplasmic initials
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
550 AMERICAN JOURNAL OF BOTANY [Vol. 33,
formed in the meristem by relatively late divisions
and in a position abutting on lacunae which are already
present; or fast developing. These facts are
obviously related to the subsequent growth of the
selereids, which is not shared by their neighbors,
and should be taken into account in its further analysis.
Such "individualistic" behavior, especially emphasized
by Foster (1945), appears as a natural
concomitant in this extreme type of differentiation
in which cells in different stages of development and
therefore of different reactivity and developmental
potency are exposed to probably very similar, if not
the same, external conditions. It is not, in the writer's
opinion, an instance of morphogenetically uncontrolled,
independent, or pathological growth. The
phenomenon of idioblastic development of this kind
in plants rather appears as part of the more general
problem of differentiation and pattern.
All histological differentiation can be interpreted
as a process in which cells become individually different
from each other. During this process they go
through different stages of reactivity or developmental
potency (Bloch, 1944) upon which varied
local conditions have differential ways of "control",
thus many individual characteristics are developed.
How this is achieved in a specifically and well coordinated
manner within the pattern of a given genotype
is a major problem of organized development.
The trichosclereids in the air root of Monstera
offer still another interesting aspect. It has been
suggested that their function may be a mechanical
one, somewhat comparable to a network of fibers. It
appears of interest to compare the pattern in Monstera
with the mechanical system in the air roots of
some orchids, for example Cattleya, the so-called
"Reseau de Soutien." While in the cortex of air roots
of Monstera and other Araceae a mechanical system
is achieved by a formation of a pattern of fiber-like,
sclerotic hairs, in the "Reseau de Soutien" of the
cortex of orchid air roots, a continuous network is
formed in a whole group of cells by branching bands
of thickened and lignified walls, comparable somewhat
to the Casparian strip in the endodermis or the
thickening in xylem elements (Sinnott and Bloch,
1945). One natural consequence of these differences
is that in the air root of Monstera a branch root
could readily make its way through the cortex, simply
pushing aside the sclerotic processes which form
no continuous mechanical system and therefore offer
little resistance. In the air root of orchids, however,
the continuous network would form a real obstacle,
and it was found (Bloch, 1935) that here sectors of
the root in which branch roots are being formed
from the beginning remain free of the lignified
thickenings.
SUMMARY
In the air root of Monstera deliciosa trichosclereids
are formed early in the cortical rib meristem,
which proliferate between parenchyma cells like
hairs, sending out long processes which are prosenchymatous
like those of astrosclereids or fibers. The
cells are thus clearly distinguished from the brachysclereids
of the same root which differentiate at
a much later stage from non-specialized cortical
ground parenchyma cells of the hypodermal region
or adjacent to internal surfaces.
The trichosclereids originate from late, differen
tial, polar divisions of cortex cells near the root apex,
analogous to those which give rise to passage cells
in the hypodermis. These special divisions occur at
the basal (proximal) ends of cell files in the rib
meristem, and the trichosclereid initials can be recognized
by their position, small size, dense content
and large nucleus. Their subsequent specific activity,
which distinguishes them from their differently reactive
neighbors, consists in the formation of prosenchymatous
processes which grow rapidly into the
intercellular space system as it develops between the
cortical parenchyma cells. This behavior appears to
be correlated with their late appearance as "embryonic"
cells and with the presence of air spaces.
Their subsequently long-continued growth is comparable
to that of the external trichoblasts of the
grass root epidermis and other plants. The distribution
of trichosclereids in the air root of Monstera
can thus be traced back to a pattern of differential
cell divisions in the fundamental tissue which precedes
it, and later idioblastic development appears
as the outcome of specific differentiation rather than
uncontrolled growth of solitary cells.
The growth of the processes of the trichosclereids
resembles that of other intrusively growing fibers
and sclereids and bears a close relationship to
changes in the intercellular space system; later differentiation
of the wall of the sclereids is comparable
to that of ordinary fibers or sclerenchyma cells,
although somewhat more irregular. Pit canals occur
only in the initial, central portion of the sclereid,
and secondary thickening and suberization are not
always initiated uniformly over the entire length of
the long tubular branches of a sclereid, but are apparently
often subject to variable local conditions.
DEPARTMENT OF BOTANY,
YALE UNIVERSITY,
NEW HAVEN, CONNECTICUT
LITERATURE CITED
DE BARY, A. 1884. Comparative anatomy of the vegetative
organs of the phanerogams and ferns. Oxford.
BLOCH, R. 1935. Observations on the relation of adventitious
root formation to the structure of air-roots
of orchids. Proc. Leeds Phil. Soc. Sci. Sect. 3: 92-101.
. 1944. Developmental potency, differentiation
and pattern in meristems of Monstera deliciosa.
Amer. Jour. Bot. 31: 71-77.
FOSTER, A. S. 1944. Structure and development of sclereids
in the petiole of Camellia japonica L. Bull.
Torrey Bot. Club 71:302-326.
1945. Origin and development of sclereids in
the foliage leaf of Trochodendron aralioides Sieb. &
Zucc. Amer. Jour. Bot. 32:456-468.
GiURTLER, F. 1905. Ueber interzellulare Haarbildungen,
insbesondere uiber die sogenannten inneren Haare
This content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms
Junie, 1946] ERICKSON-GROWTH OF TOMATO ROOTS 551
der Nymphaeaceen und Menyanthoideen. Inaug-Diss.
Berlin. 92 pages.
LIERAU, M. 1888. Ueber die Wurzeln der Araceen. Bot.
Jahrb. 9: 1-38.
MEYEN, F. J. F. 1837. Neues System der Pflanzenphysiologie
I, p. 3a1.
SACHS, J. 1882. Textbook of Botany. 2nd ed. Oxford.
pp. 84?-85.
SINNo,rr, E. W., AND R. BLOCH. 1939. Changes in intercellular
relationships during the growth and differentiation
of living plant tissues. Amer. Jour. Bot.
26: 625-634.
AND . 1943. Developimient of the fibrous
net in the fruit of various races of Luffa cylindrica.
Bot. Gaz. 105:90-99.
AND 1945. The cytoplasmic basis of intercellular
patterns in vascular differentiation. Amer.
Jour. Bot. 32:151-156.
SOLEREDER, H. 1908. Systematic anatomy of the dicotyledons.
Vol. 1 and 2. 1183 pages.
1919. Beitrage zur Anatomie der Araceen.
Beih. Bot. Centralblatt 36, I. 60-77.
AND F. J. MEYER. 1928. System. Anat. der
Monokotyledonen. Heft 3.
VAN TIEGHEIM, PH. 1866. Sur la structure des Aroidees.
Ann. Sci. Nat. 5 se'r. Bot. 6: 72-210.
TscHrRcH, A. 1885. Beitrage zur Kenntnis des mechanischen
Gewebesystems der Pflanzen. Jahrb. Wiss.
Bot. 16:303-335.
WIESNER, J. 1875. Ueber das Vorkommen von Haaren
in den Interzellulargangen des Mesophylls von Philodendron
pertusum. Oesterr. Bot. Zeitschr. 25: 4-7.
GROWTH OF TOMATO ROOTS AS INFLUENCED BY OXYGEN IN THE
NUTRIENT SOLUTION'
'Louis C. Erickson
THE GROWTH of many species of plants in water
culture is greatly improved by bubbling compressed
air through the nutrient solution. This response
suggests that the same factors important in soil
aeration, i.e., oxygen and carbon dioxide (cf.
Clements, 1921; Cannon, 1925; Conway, 1940),
are involved in water cultures also.
In spite of the large number of accounts written
on the benefits of aeration in water cultures there
is meager information in the literature concerning
the concentrations of oxygen and carbon dioxide
that occur in aerated and non-aerated solutions.
The result has been that some investigators attribute
the aerational responses of the plants to
oxygen while others attribute them to carbon dioxide.
Also, there is little information on the effect
of oxygen tension on root structure, although data
are available on the influence of various concentrations
of oxygen on- the absorption of ions and the
production of organic acids.
This paper presents the results of an investigation
of the influence of oxygen content in the nutrient
solution on the growth of tomato roots. The
aims were to determine the concentrations of oxygen
and carbon dioxide occurring in aerated and
non-aerated water cultures during a period of several
weeks and to interpret the- growth responses
of the plants on the basis of experiments in which
these factors were controlled. In a later paper the
influence of oxygen concentration on root structure
will be considered.
REVIEW OF LITERATURE.- The recognition of
aeration as a factor in the water culture of plants
was made at the time this method of growing plants
was undergoing its first rapid development. Sachs
(1860) recognized that a certain amount of air,
1 Received for publication February 19, 1946. The
author is indebted to Drs. D. I. Arnon, A. R. Davis, and
H. S. Reed for suggestions.
in addition to the minerals, must be present in the
water for the best growth of the plants.
Arker (1901) made an intensive study of the
effects of aerating water, soil, and mud cultures of
Lupinus albus and Helianthus annuus and found
that the roots of both of these species made more
rapid growth when air was bubbled through the
cultures.
Hall, Brenchley, and Underwood (1914) noted
a superior growth of barley and lupins in cultures
of coarse sand over those in water, fine sand, or silt.
While the root systems of plants in kaolin and
coarse sand developed freely, those of fine sand and
silt were restricted. The authors suspected aeration
as the disturbing factor and tested that assumption
by arranging comparative water cultures in which
one series was not aerated whereas in the other
bottles a continuous current of air was bubbled
through the solutions. Both the barley and lupins
made the greater growth in the aerated solutions, as
indicated by the dry weights.
Free (1917) found that passing oxygen, nitrogen,
or air through the solutions had no appreciable
effect on buckwheat, when compared with open or
sealed controls, but that in cultures treated with
carbon dioxide the plants wilted within a few hours
and died within a few days.
Stiles and Jorgensen (1917) obtained significantly
better growth of barley and balsam by aerating
the culture solution, but no response was obtained
in buckwheat, which was in accord with the finding
of Free. Stiles and Jorgensen recognized several
possible factors as operating in their experiments
and therefore refrained from correlating the beneficial
effects of aeration with oxygen supply, removal
of carbon dioxide, and other causes..
Pember (1917) did not obtain beneficial effects
from daily aerations of barley.
Beals (1917) grew two Zea Mays plants in water
culture and aerated the solution in one of the conThis
content downloaded from
147.251.40.63 on Wed, 26 Apr 2023 12:02:55 UTC
All use subject to https://about.jstor.org/terms