OF ARTS AND SCIENCES. 
VII. 
CONTRIBUTIONS FROM THE CRYPTOGAMIC LABORATORY 
OF THE MUSEUM OF HARVARD UNIVERSITY* 
V. —ON THE ANATOMY AND DEVELOPMENT OF 
AGARUM TURNERI, Post. & Rupr. 
James Ellis Humphrey. 
Presented June 16, 1886. 
Tiie family of Laminariaceae includes several genera of large, coarse 
marine algae of an olive-brown color and leathery substance. In 
most of these, the thallus shows a distinct differentiation into stipe 
and lamina, and is attached to the substratum by branched, root-like 
processes, or rhizoids. This family contains forms which are the 
largest of Thallophytes, and indeed among the largest of plants. 
Not to mention several contributions by earlier writers on the 
Algae, three papers have recently been published which deal with the 
anatomy of Laminariaceae, all by German writers and including ob- 
servations on four genera. Reinke has described the anatomy of 
Laminaria saccharina, L. digitata, and Alaria esculenta ; f Will, that 
of Macrocystis luxurians ; J and Grabendorfer, that of Lessonia ovata.§ 
* The studies whose results are here recorded were carried on in the Cryp- 
togamio Laboratory of Harvard University, under the direction of Dr. W. G. 
Farlow. 
The following are the titles of the preceding papers of this series : — 
1. Notes on some Species in the Third and Eleventh Centuries of Ellis’s North 
American Fungi. Proc. Am. Acad , XVIII. 65-85. 
2. The Cryptogamic Flora of the White Mountains. Appalachia, III. 232-251. 
3. Notes on some Species of Gymnosporangium and Chrysomyxa of the 
United States. Proc. Am. Acad., XX. 311-323. 
4. Notes on Arctic Algae; based principally on Collections made at Ungava 
Bay by Mr. L. M. Turner. Proc. Am. Acad., XXI. 4G9-477. 
t Beitrage zur Kentniss der Tange,—Pringsheim’s Jahrbiicher fur wiss. 
Botanik, Bd. X. p. 371, 1876. 
f Zur Anatomie von Macrocystis luxurians, — Botanische Zeitung, 42 Jahr- 
gang, p. 801, 1884. 
§ Beitrage zur Kentniss der Tange, — Bot. Zeit., 43 Jahrgang, p. 641,1885. 196 
PROCEEDINGS OP THE AMERICAN ACADEMY 
These papers show that there is a great uniformity in the character 
and arrangement of the tissues composing the thallus in the different 
genera. 
The adult stipe of the Laminariaeeae may be said, in general, to be 
composed of four forms of tissue. At the centre is a bundle of twisted 
and intertangled hypha-like filaments, among which are scattered long 
trumpet-shaped or funnel-formed cells, placed by twos with their large 
ends together. Will describes in detail a zone of tissue, composed of 
sieve-tubes with very abundant sieve-plates, which surrounds this core 
in the adult stipe of Macrocystis. Outside of the central filamentous 
tissue, or medulla, is a layer of elongated cylindrical cells, whose 
side walls are very much thickened and abundantly marked with 
pits, their intercellular spaces being filled by a slimy substance. In 
Macrocystis and Lessonia, this layer, which may he termed the inner 
cortex, is sharply defined against the medulla; but in Laminaria the 
medullary filaments penetrate its inner cell-rows, and cause a blending 
of the two tissues, so that no clear line of demarcation can be fixed 
between them. Next without the inner cortex are rows of thin-walled 
cells, more angular than the last in form. These compose a tissue called 
the outer cortex. They become smaller toward the surface of the stipe, 
and pass gradually into the single superficial layer of cells which forms 
the epidermis. This latter layer consists of small cells whose exposed 
outer walls are rounded and much thickened, and which contain the 
olive-brown pigment characteristic of this group of Alga;. The cells 
of both outer cortex and epidermis are traversed by very delicate 
partitions, both parallel and oblique to their walls. In Alaria, the 
cortical tissue seems not to be distinguishable into outer and inner 
layers. 
The lamina is, anatomically as well as morphologically, a terminal 
expansion of the stipe, and shows the same tissues as the latter, but 
in quite different proportions. The medulla resembles that of the 
stipe, but the trumpet-shaped cells seen in the latter are not present. 
The cell-walls of the inner cortex are thinner than in the stipe, and 
contain no pits. The outer cortex forms but a small part of the 
thickness of the hlade, and consists of small, thin-walled cells pass- 
ing into an epidermal layer, as in the stipe. In Alaria, the stipe is 
prolonged into the lamina as a midrib, whose anatomy is that of the 
stipe itself. 
The rhizoids consist of a mass of cells somewhat thick-walled at the 
centre, and decreasing in size and thickness of wall toward the sur- 
face, the outer layer forming an epidermis. No trace of medulla is OP ARTS AND SCIENCES. 
197 
present. At their points of contact with the substratum, the exposed 
cell-walls become very much, aud often irregularly, thickened. 
Each of the papers above quoted contributes something in regard to 
the growth or development of the species treated, and reference may 
be had to them for full details. It seems well established that the 
seat of growth in length is the point of union of stipe and lamina, 
where growth takes place either periodically or constantly. How the 
stipe grows in thickness seems as yet hardly clear, lleinke describes 
a “ meristem cylinder ” composed of the inner cells of the outer cor- 
tex in Laminaria, to which he ascribes a function similar to that of the 
cambium of Dicotyledons. And Will describes a peculiar layer im- 
mediately outside of the medullary tissue in the young stipe of Macro- 
cystis, which serves as the source of increase of the medulla; this 
ring disappears in the adult stipe. The cells of the epidermis in- 
crease by the formation of walls at right angles to the surface, thus 
keeping pace with growth in thickness, as well as in length. 
The differentiation of the frond into stipe and lamina is evident 
from a very early stage. When quite young, the stipe consists of a 
cortex and a few thick-walled axial cells, from which the medul- 
lary filaments develop later, as outgrowths. At a corresponding 
stage the lamina consists of two large-celled layers and epidermis, 
although an earlier condition is known in which it is but a single cell 
thick. The lamina is annually renewed from the region of growth, 
and Eeinke considers it probable that not much change occurs in 
the structure of a given lamina, but that the successive steps in the 
development of the complex adult frond from the comparatively 
simple young condition are accomplished by an increase in complexity 
of structure in successively formed fronds, until the adult type is 
reached. 
Agarum Turneri. 
Agarum Turneri, the so-called Sea Colander, is one of the Lami- 
nariacem, having a cylindrical stipe 10 to 30 cm. in length, which 
becomes flat toward its upper end and finally expands into a broad 
lamina of an ovate or oblong shape, with a strongly cordate base and 
crisped margin. Through the middle of the lamina passes a broad, 
flat continuation of the stipe, the midrib. The lamina is abundantly 
perforated by holes of considerable size, which will be treated in detail 
later. The stipe splits up, at its base, into numerous branching rhi- 
zoids, by which the plant is attached to the substratum. The whole 198 
PROCEEDINGS OP THE AMERICAN ACADEMY 
plant often reaches a meter in length. This alga is an inhabitant of 
the Arctic Ocean, and extends southward on the western shore of the 
Atlantic to Boston, and in the Pacific to California and Japan; being, 
like other Laminariacem, a lover of cold waters. At East port, Maine, 
it occurs at low-water mark, but south of that point it grows in deep 
water, from which it is frequently washed ashore. 
At the suggestion of Dr. W. G. Farlow, these investigations have 
been undertaken with the purpose of learning the anatomy of the 
adult and of the young frond, and the mode of formation of the per- 
forations of the lamina, in Agarum Turneri. The material used has 
been alcoholic, and has consisted of young fronds collected at Eastport 
in September, 1877, by Dr. Farlow, and of adult plants obtained at 
Marblehead, Mass., in .November, 1885, by myself. 
1. Adult Anatomy. 
The stipe of the adult frond does not differ in its structure from 
the general type of the Laminariaceie. The medulla and inner 
cortex together compose about one third of the whole diameter of 
the stipe at its base, but gradually enlarge so that, at the upper 
end, they form one half of its diameter. The component threads of 
the medulla are rather loosely matted together, and have many 
rounded cells scattered among them. These threads penetrate the 
innermost layers of the inner cortex quite freely, causing a gradual 
passage from one tissue to another, as in Laminaria. The cylindri- 
cal cells of the inner cortex average from three to four times as long 
as broad. Their side walls are much thickened and pitted, while 
their end walls are thin. The outer layers belonging to this tissue 
consist of smaller cells than the rest, and are clearly distinguishable 
from the inner rows of the outer cortex, which lie next to them. 
These latter form a darker and more refractive layer of cells, since 
they are very rich in protoplasm. The cells of the outer cortex 
become smaller and thinner-walled toward the surface, and their super- 
ficial layer is modified to form an epidermis. The epidermal cells 
have their exposed ends rounded and much thickened, and contain 
the characteristic olive-brown pigment. As in Laminaria, the cells of 
the whole outer cortex are crossed by very delicate parallel and oblique 
walls. The epidermis and cells immediately below present all the 
appearances of growing tissues. 
The midrib continues the tissues of the stipe, but the ratios between 
the different tissues vary markedly from those in the stipe. The outer OP ARTS AND SCIENCES, 
199 
cortex, inner cortex, and medulla occupy about one, five, and two 
eighths, respectively, of the radius of the midrib. That is, the in- 
crease in proportion of the inner cortex, observed in passing upward 
from the base of the stipe, continues rapidly into the midrib, until the 
outer cortex is reduced to a very thin tissue. The outer cortex con- 
sists of a few layers of thin-walled cells arranged in rows perpendicular 
to the surface, the exposed outer layer being modified to form an epi- 
dermis. The inner cortex forms an abrupt change from the last, con- 
sisting of rounded cells with walls not so thick as in the corresponding 
cells of the stipe, and without pits, but much thicker than those of the 
outer cortex. The outer cells of this tissue are somewhat smaller 
than the rest. The medulla is identical in character with that of the 
stipe, and its filaments penetrate the inner cortex. 
The passage from midrib to lamina is rather gradual, and the latter 
becomes thinner towards its edge. In the lamina, the distinction 
between outer and inner cortex is lost, and the whole cortical tissue 
consists of thin-walled cells. The inner layers are rather large-celled, 
but the outer cells are smaller, while those of the epidermal layer 
are quite small and have their outer walls thickened. (Fig. 7.) 
At the margin, where the frond is quite thin, the cortex is reduced 
to a very few cell-layers. The medulla becomes gradually thinner 
toward the margin of the lamina, where it consists of a few fine fila- 
ments fused into an indistinct mass. The rhizoids originate from the 
end and lower portion of the stipe, branching freely and irregularly, 
and often are considerably expanded at the ends which come in con- 
tact with and attach the plant to the substratum. Their tissue is a 
direct continuation of the outer cortex of the stipe, and is similar to it 
except in being rather smaller-celled. The outer cell-layer forms an 
epidermis, and the layer which comes into immediate contact with 
the substratum has its exposed walls thickened very strongly, and 
often irregularly, so that it not infrequently happens that the cav- 
ity of a cell is nearly or even wholly obliterated by the thickening 
of its wall. 
2. Growth. 
The cambium-like layer described by Reinke for Laminaria has 
also been mentioned as occurring in Agarum, but there appears to be 
another region of growth in the latter genus, which includes the epi- 
dermis and cell-layers immediately underlying it to the number of 
two or three. I believe that growth in thickness takes place in these 
two regions, each meristem adding, by its activity, to the tissue lying 
within it. 200 
PROCEEDINGS OF THE AMERICAN ACADEMY 
As in Laminaria, the epidermis follows growth in thickness by the 
increase of its cells through the formation of radially perpendicular 
walls. 
The lamina of Agarum has never been observed to present any 
evidences of periodic renewal so plainly shown among the species of 
Laminaria, especially the digitate forms, during the spring months, 
although it has been collected in great quantities at all seasons on the 
coast of Massachusetts. This fact and the presence at all seasons of 
abundantly forming perforations, and only these in the basal part of 
the lamina, make it quite certain that the growth is continuous. 
Even in very young fronds, the exact tip is not known. Many 
fronds, when very young, are markedly attenuated at the upper end, 
but none has been found with the actual tip still present. In any 
but a very early stage, the attenuate character of the upper end is lost 
and the lamina is ragged and water-worn at its apex. 
3. Anatomy of Young Frond. 
The specimens of the young frond of Agarum Tarneri which have 
been at my disposal are as small as any known, but none is sufficiently 
young to show the one-layered condition of the blade known in Lami- 
naria, if it exists in this species. A young frond of 3 cm. long has a 
stipe 4 or 5 mm. in length and not over .5 mm. in diameter, with 
several branched rhizoids at its base. The lamina is oblong in shape, 
about a centimeter broad and of the thickness of very thin writing- 
paper, with the margin slightly crisped or wholly Hat. Through the 
middle of the lamina the midrib appears as a rather broad flat band, 
of about twice the thickness of the lamina and of a firmer texture. 
(Fig. 1.) Such a frond, while much simpler than in the adult condi- 
tion, includes all the tissues of the latter except the medulla. 
The stipe has an outer cortex composed of radial rows of rather 
thin-walled cells, the whole of which is deeply colored by the olive- 
brown pigment. This tissue is covered on its surface by a homo- 
geneous cuticula, in consequence of which the exposed ends of the 
outer cells, while markedly rounded, are not thickened. The rest 
of the stipe, composing about two thirds of its diameter, consists 
of elongated cells closely resembling those of the inner cortex of 
the adult, but with somewhat less thickened walls. (Fig. 2.) No 
medullary threads are present at this stage, but a little later, when 
the perforations of the lamina begin to appear, they may be seen 
arising as outgrowths from the axial cells of the stipe. The struc- OF ARTS AND SCIENCES. 
201 
ture of the midrib agrees completely with that of the stipe. A com- 
parison of the two with the adult shows that the midrib undergoes 
the less change of structure, remaining always more primitive in con- 
dition. The lamina of such a frond as we are considering consists 
of two or three rows of large, squarish, thin-walled cells, covered by 
epidermal layers of much smaller cubical cells, containing the brown 
pigment. (Fig. 3.) 
About simultaneously with the appearance of medullary threads in 
the stipe, they are seen between the large-celled layers at the base 
of the lamina, and gradually push out towards its edge as the plant 
grows older. The lamina does not further increase in complication of 
structure for a considerable time. 
4. Perforations of the Lamina. 
The very young lamina of Agarum Turneri is quite imperforate for 
a time, but when a length of 3 to 4 cm. is attained, the perforations, 
afterward so characteristic, begin to be formed in all parts of the 
frond. They continue to be developed during the life of the plant, 
especially at the base of the lamina and near its middle along the 
midrib, but also wherever growth has produced a considerable un- 
perforated space. Any large frond will furnish an abundance of 
perforations in all stages of development. (Figs. 4 and 5.) The 
rarity of the occurrence of such holes among plants and the evident 
definiteness of their development give to their study much interest 
and importance. 
The first indication of the formation of a hole is the appearance, on 
either of the faces of the lamina, of a small, conical, papilla-like ele- 
vation, with a corresponding depression on the opposite surface. 
(Fig. 5, a.) This papilla continues to push up until a rupture occurs 
at its tip (Fig. 5, i), and a small opening, like a pin-hole, is made 
through the lamina (Fig. 5, c). If a piece of a lamina with papillae 
in various stages of development be held to the light, the tips of the 
papillae will be seen to be translucent, and those most so which are 
nearest the rupturing point. This shows that there is a gradual de- 
crease in the thickness of the tissue at the tip of a papilla, from the 
time of its beginning until the rupture takes place. The hole now 
increases in size with the growth of the frond (Fig. 5, rf), and may 
reach 2 cm. in greatest diameter and become very irregular in shape, 
though usually of a broadly elliptical form. It retains for a long time 
the character of a perforation of the lamina with a raised margin 202 
PROCEEDINGS OF THE AMERICAN ACADEMY 
when viewed from one surface, or a depressed one when viewed from 
the other face of the lamina. 
The microscopic study of the details of the development of the per- 
forations lias required only sulhciently thin sections of their various 
stages. Free-hand cutting wholly failed to give these. A piece taken 
from the base of a lamina, and containing several early stages in the 
development of perforations, was then imbedded in paraffine and cut 
into very thin sections by means of the microtome. The sections 
were then mounted in balsam, according to the usual method for thin 
sections of animal tissues. Although this seems a hard treatment for 
an algal tissue, the tough, leathery character of the frond enabled it to 
withstand well, and very satisfactory results were obtained. 
Sections prepared in this way show that a perforation originates as 
a depression of the epidermis of either of the surfaces of the lamina. 
(Fig. 6, a.) This depression is formed by a rapid multiplication of 
the cells of the epidermis in that region and the incurving of the layer 
in consequence of its increased length. (Fig. 7.) After the inward 
bending has begun, the continued growth of the epidermis causes it 
to press deeper and deeper into the underlying tissue, which is com- 
pressed and pushed aside, until the medulla is reached. (Fig. 6, b.) 
The depression has, at this stage, the form of a basin with sloping 
sides and a nearly flat bottom, lined by the unbroken epidermal 
layer. The bottom of this depression lies against the medullary tis- 
sue, while its sides are in contact with the compressed cortex and form 
angles with the medulla. 
Further growth appears to take place only in that portion of the 
epidermis which forms the sides of the depression, and has two effects. 
Since the bottom of the depression remains passive while the sides 
are tending to push on into the medulla, the former is stretched with 
increasing force until a rupture occurs near its middle; this whole 
bottom layer of epidermis now dies and falls away, leaving the sides 
unrestrained. Under the influence of their rapid growth, the latter 
press into the medulla like a circular punch, cutting out that part of 
the medulla formerly in contact with the bottom of the epidermal 
depression. (Figs. 6, c, and 8.) Meanwhile, that part of the cortex 
of the other surface of the lamina which corresponds in position to 
these lost tissues has begun to show a changed appearance. Although 
not directly affected by the growth of the depression just described, 
it becomes dark and shrivelled in appearance (Fig. 8), and soon 
breaks away from the living cortical tissue surrounding it, escaping into 
the water as a small block of dead and collapsed cells. (Fig. G, d.) OP ARTS AND SCIENCES. 
203 
In this way the perforation of the lamina is completed, and we have 
a small hole, not usually exceeding .5 mm. in diameter, formed 
partly by the pushing aside, and partly by the death and breaking 
away of the tissues which formerly occupied its place. 
Where the tissues have been pushed aside by the ingrowing epi- 
dermis, they remain covered by it, but where a part of the tissue has 
died and fallen away, as in case of the cortex of one surface, that 
part which becomes thus exposed is not at first protected. But very 
soon the exposed living cells of the cortex begin to thicken their outer 
walls and to acquire the characteristic brown color of epidermal cells. 
Finally they become completely indistinguishable from and continuous 
with the cells of the original epidermal layers of the two surfaces of 
the lamina. (Figs. 6, e, and 9.) In this we have distinct evidence 
that the epidermis consists simply of modified cortical cells, and that 
any cortical cells may become epidermal in character, if exposed to 
external influences. 
Further changes in the perforations affect their size only, which 
increases with the growth of the lamina. 
The results of the investigations detailed above may be summarized 
as follows: — 
1. The structure of the adult frond of Agarum Turneri agrees 
closely with that of other Laminariacese, and especially with that of 
Laminaria. 
2. The frond of Agarum grows in length at the union of stipe and 
lamina, as in other Laminariacese; growth in thickness probably takes 
place in two regions. 
3. The structure of the young frond is much simpler than that of 
the adult, but its relations to the latter are clearly traceable. 
4. The perforations of the lamina are formed by the simultane- 
ous occurrence of an indentation of one surface and the death of the 
corresponding part of the opposite surface. 
These results will serve at least to call up still unanswered ques- 
tions, many of which require for their satisfactory solution access to 
abundant growing material in all stages. For a student who can 
command these desiderata, the work to be done on the Laminariacese 
is as interesting and important as the many unanswered questions 
connected with the study of other groups of marine algse. 204 
PROCEEDINGS OF THE AMERICAN ACADEMY 
EXPLANATION OF PLATES. 
Fig. 1. Young frond of Agarum Turneri about 3 cm. in length. X $. 
Fig. 2. Transverse section of stipe of a frond like Figure 1. cut. Cuticula. 
o. c. Outer cortex, i. c. Inner cortex. X 176. 
Fig. 3. Transverse section of lamina of a similar frond, e. Epidermis, c. Cor- 
tex. X 176. 
Fig. 4. Part of adult lamina, showing perforations in various stages. X $. 
Fig. 6. Section across basal part of lamina, showing various stages in forma- 
tion of perforations, a, b, c, d. Successive stages, f. Plane of 
lamina. X 2. 
Fig. 6. Diagrammatic view of five successive stages, a, b, c, d, e, in the develop- 
ment of a perforation, c'. Cortex, m. Medulla. 
Fig. 7. Detailed figure of an early stage (Fig. 6, a) of a perforation, e. Epider- 
mis. c. Cortex, m. Medulla, i. Apex of depression. X 176. 
Fig. 8. Detailed figure of a later stage (Fig. 6, c). e, c, in, as in Figure 7. 
x. Medullary tissue cut out by growth of epidermis, y. Fragment of 
cortical tissue cut out. z. Cortex of upper surface beginning to 
shrivel. X 176. 
Fig. 9. Detailed figure of section through a completed perforation (Fig. 6, e). 
e, c, m, as in Figure 7, the right-hand portion being represented in 
outline only. v. Remnants of dead cortical cells. X 175. P 'Ll. 
Fig. 2. 
Fif/.l. 
Fig. 3. 
Fig. 6. 
Fig. 4. 
Fig. 3. 
AD NAT DEL.HUMPHREY. 
DEVELOPMENT OF AGARUM. 
Fortes Co Boston. PL.II. 
Tig 7. 
Fig. 8. 
Fig. 9. 
AD NAT DEL.HUMPHREY 
DEVELOPMENT OF AGARUM. 
Forbes Co, Boston.