ISTHMIA NERVOSA: 
A Study of its Modes of Growth and Reproduction. 
J. D. COX, A. M., LL. D. 
(Reprinted from the American Journal of Microscopy). ISTHMIA NERVOSA. 
A Study of its Modes of Growth and Reproduction. 
J. D. COX, A. M., LL. D. 
[ From the American Journal of Microscopy. ] 
concentric tubes, which slide over each 
other during the multiplication of the dia- 
tom by division; but in regard to the mode 
of growth of the hoop itself, the causes of 
variation in the size of the frustules, and 
especially as to the formation and function 
of the conjugate (or sporangial) frustule, 
the facts observed seem to point to con- 
clusions so different from his as to warrant 
a detailed presentation. 
T the time Dr. 
Wallich’s article 
upon the tele- 
scopic hoops of 
Biddulphia, Isfh- 
mia, and some 
other diatoms, 
appeared in the 
Monthly Micro- 
scopical Journal, 
of February of 
last year, it hap- 
pened that I had 
in hand a considerable quantity of the Isth- 
mia nervosa, in situ upon seaweed, from the 
vicinity of Santa Barbara, California. lat 
once determined to make a somewhat full 
examination of this material, and to learn 
what facts in the life history of this diatom 
could be drawn from the study of the dead 
specimens. 
At the risk of repeating some things too 
well known, I shall endeavor to give a con- 
nected description of the appearance of the 
diatom at the several stages of its growth, 
up to the appearance of the conjugate frus- 
tule, and through such subsequent steps as 
I find sufficiently supported by the evidence 
before me. 
1. Forms of the Diatom.—The Isthmia 
nervosa has both a rhomboid and trapezoid 
form. The heavy costae, or ribs, which 
give the name to the species, are found on 
the valves only, but not on the hoop. These 
nervures or costae are very similarly ar- 
ranged in all the frustules. Those of one 
valve are nearly parallel to each other, 
pretty regular in direction, running from 
the margin next the hoop to near the sum- 
mit of the dome-like valve, where they 
reticulate with each other. Those of the 
other valve start from the margin in a sim- 
ilar way, but branch much sooner, and run 
together in a coarse network, of which the 
Part of the material I mounted dry in 
situ, so that it could be examined as either 
an opaque object or a transparent one; the 
rest was cleaned in acids, part mounted in 
balsam, and part kept in distilled water for 
examination in that condition. 
The examination fully sustained Dr. 
Wallich’s statement, that the hoop of Isth- 
mia is often found to be of two or more 2 
ISTHMIA NERVOSA. 
most marked lines or ribs partially sur- 
round and isolate the nose-like projection 
of the valve, from which the gelatinous 
stipes exudes through the fine seive-like 
perforations, which here take the place of 
the much larger areolse found on other 
parts of the valve. The costae project 
boldly into the interior of the shell to a 
depth of about one-twelfth of its shortest 
diameter, and their interior arrangement 
about the point by which the diatom is at- 
tached, is such as to indicate, or at least 
strongly to suggest, a separate cell, having 
the office of secreting the gelatinous sub- 
stance of the stipes or stalk. (Fig. 1.) 
be seen that the stipital ends are in the 
same position, thus maintaining a general 
analogy to the positions of buds and stems 
upon any larger plant. (Fig. 1, a, b.) 
4. Modes of Connection in the Filament of 
Diatoms.—The apparently capricious way 
in which the several diatoms are attached to 
each other is easily accounted for, since a 
single slide of the plant in situ will gener- 
ally give specimens which mark nearly 
every conceivable step in their growth and 
separation. Bringing these steps together, 
we find that the simplest form of connec- 
tion is that which occurs when the stipes of 
No. 2 attaches itself to the top of No. 1, 
their growth crowds them beyond the hoop, 
and the two remain as shown in Fig. 1, a, b. 
No. 2 may attach itself to the edge of the 
hoop of No. 1. In this case No. 2is grad- 
ually crowded out of the hoop by the 
growth of its twin brother below', until its 
stipes can only hold on the very edge of the 
hoop, and the still further growth of No. 1 
makes it turn outward until it takes the 
position shown in Fig. 1 at c. If, now, 
No. 1 completes a second division, No. 3 
may remain attached to the others as shown 
above (Fig. 1, d), or may become loosened 
! and fall off entirely, leaving the first two in 
a position in which the hoop of No. 1 pro- 
jects beyond the frustule, and No. 2 remains 
attached to its side. Besides the various 
combinations and permutations of these 
modes of attachment, the loosened frus- 
tulea may attach themselves wherever they 
may fall, either to the alga on which the 
colony lives, or to other diatoms amongst 
which it may lodge. 
Fig. 1. 
2. Relative Numbers of the two Forms.— 
The fission of the diatom is always trans- 
verse diagonally, the line of division run- 
ning in such a way as always to divide a 
rhomboidal form into two trapezoids, and a 
trapezoid into one rhomboid and one 
trapezoid. From this it results that the 
number of trapezoids is largely in excess of 
the rhomboids, for, from the binary divis- 
ion of one of each kind, must result three 
trapezoids to one rhomboid. (Fig. 1, a, b.) 
4. The Hoop.—l have already stated my 
accord with Dr. Wallich in finding the hoop 
composed of concentric elliptical tubes, 
sliding over each other like the draw 
tubes of a telescope. Indeed this is so 
obvious a fact that the most casual observer 
could hardly be excused for overlooking it. 
Not only is it shown by the thicker band 
about the middle of the diatom, when the 
areolse of the different lamellae of the hoop 
are confused by lack of coincidence with 
each other, and where, by focussing upon 
3. Position of the Stipes and Point of At- 
tachment.—Assuming the point of attach- 
ment of the frond or filament to be its lower 
end, the valve with irregular costse, which I 
will call the stipital end of the diatom, is 
always down. Thus, in a specimen in 
which division is complete, except that the 
two new frustules are still held within the 
elongated hoop of the original one, it will ISTHMIA NERVOSA. 
3 
either, the areolse of that lamella may be 
brought out sharply, the other becoming 
dim; but at the sides of the diatom, where 
the curve of the hoop becomes nearly per- 
pendicular to the line of vision, the double, 
and sometimes triple thickness, is seen al- 
most as clearly as if looking at the edge of 
a longitudinal section. 
ing of the suture is very small, the areolae 
of the inner film have evidently slid from 
under the exterior ones, by the same dis- 
tance as the opening of the suture, thus 
showing that they were before concentric, 
and suggesting (what further observation 
satisfactorily proves) that the same process 
of doubling the membrane or film has gone 
on upon the other side of the suture in the 
other half of the diatom, but is for the 
nonce invisible, because the adhesion of 
But Dr. Wallich draws the conclusion 
that the hoop grows by addition of silex at 
the free margin, “much in the same way 
that the epiderm of the molluscous shell is 
secreted by the margin of the animal’s man- 
tle.” My observations tend to establish a 
different conclusion, viz., that the growth 
of the hoop is in accordance with the laws 
the films is undisturbed, and their areolae 
remain concentric. 
If the slide is rich in broken specimens 
(so often the most instructive), a little 
search will probably be rewarded by the 
discovery of a cast hoop in which, beyond 
the suture, the thin film above described is 
still adherent, not as a gradual tapering or 
diminution of the substance of the hoop, 
but starting from the edge of the suture by 
a shoulder, or sudden diminution of the 
thickness, which the light marks by a dis- 
tinct line, beyond which the silica grows 
evanescent and the areolae gradually fainter, 
toward the irregular and scarcely visible 
edge. (Fig. 2, b.) 
of vegetable growth, rather than animal, 
and harmonizing with the theory of a mem- 
brane in which the silex is deposited, which 
membrane may be divided and split by the 
process of intussusception, or a new mem- 
brane may be formed within the old, from 
the protoplasm or live cell contents. Let 
us enumerate some of the facts bearing upon 
the question. 
Upon almost any slide of Isthmia will be 
found numerous specimens, in which will 
be noticed not only the arrangement of the 
areolse in the form of the quincunx, made 
by pretty regular diagonal rows, but the 
face of the hoop is divided by one or more 
lines of clear silex, running straight around 
the diatom, and breaking up the dotted 
field into two or more parts, as shown in 
Fig. 2, a. Bringing different individuals 
under the eye, one will soon be found in 
which this clear line has divided so as to 
show two smooth edges a little separated, as 
if the two halves of the diatom had broken 
asunder through the middle of the hoop. 
This suture, however, will be found not to 
cut clear through the shell, but the areolae 
of an inner lamella will be seen below the 
opening, and upon one side or the other of 
it careful focusing will trace successively 
the faint outlines of other areolae in an ex- 
ceedingly thin film of silica, which fades 
away to invisibility as it is traced toward 
the junction of the hoop with the valve. 
Pig. 2. 
I shall presently give reasons for think- 
ing that the separation of the parts of the 
hoop at the suture does not take place, 
when the diatom is growing normally, 
until the inner hoop has acquired its growth 
in thickness, and extends from valve to 
valve, or at least to the next suture above 
or below the one at which the division is 
supposed to take place; but the next point 
to be immediately noticed is that a similar. 
It may also be seen that when the open- 4 
ISTHMIA NERVOSA. 
suture occurs near the edge of the hoop 
where it connects with the valve, and when, 
by the growth of the diatom, the inner or 
outer hoop separates at that place, the line 
of the division can be plainly seen running 
across the large horseshoe-shaped areolse 
which there form the characteristic marking 
of the shell. We thus have lines of suture 
always at the two extremities of the hoop, 
and, frequently, one, two, or even three, 
crossing it at intervals. The intervening 
sutures are not invariably present, but are 
found oftenest in the largest diatoms, and 
especially in those whose width is greatest 
in proportion to their length. The suture 
may appear on the inner hoop at a different 
place from that on the outer, or when none 
appears on the outer hoop except at its 
ends, and vice versa. This mode of “break- 
ing joints ” will at once be seen to be use- 
ful in strengthening the structure, and pre- 
venting the slightest accident from break- 
ing the diatom quite in two in the middle. 
(Fig. 3, c.) 
found to be about one-sixth of the shorter 
diameter of the valve. The coste of the 
valve terminate in this flange or diaphragm, 
with a tapering curve reaching to its inner 
edge, thus supporting and strengthening it. 
Upon the valve, this flange makes a sharp 
right angle with the outer edge, or is even 
a little concave. The hoop has a similar 
diaphragm, buttressed and supported in like 
manner by the nervures of the large areolse 
which mark the end of the hoop; the ap- 
pearance of these so-called areolae being, in 
fact, wholly caused by the nervures or 
braces which support the diaphragm, and 
having no real similarity to the areolae 
which mark the general surface of the 
diatom, but are analogous to the costae of 
the valves. The hoop diaphragm has one 
very important peculiarity, viz., that its 
inner edge is recurved or bent down over 
the edge of the valve diaphragm, with a lip 
of about the thickness of the latter plate, 
thus giving great firmness of connection, 
and effectually preventing lateral displace- 
ment of the parts. The hoop and its dia- 
phragm do not meet at a sharp angle, as in 
the case of the valve; but the side of the 
hoop rounds into the diaphragm with an 
easy curve. The connection of the parts is 
shown diagrammatically, in longitudinal 
section, in Fig. 3, a, h. 
I have occasionally found the diaphragm 
of both valve and hoop greatly widened, 
projecting into the interior of the diatom 
much further than common, so as to make 
a partition more than half closing the con- 
nection between the two parts. Its inner 
outline in these cases, is of the form of a 
broader ellipse, but it seems to have grown 
as a thinner film of silica, back of the ordi- 
nary diaphragm, or rather so extending 
from it as to leave the box-like fitting of 
parts, as shown in Fig. 3, so that the thinner 
portions are not in immediate contact with 
each other, as the thicker and narrower 
rims or flanges are. Upon these widened 
diaphragms the lines of the supporting 
nervures extend, as upon the narrower 
ones, but becoming fainter and wider as 
they go, giving a beautiful appearance of 
Fig. 3. 
5. The Connection of Hoop and Valve.— 
Before proceeding further in the descrip- 
tion of the manner in which the sliding of 
the telescopic hoops upon each other ac- 
commodates the growth of the diatom, it 
will be well to examine the structure of the 
connecting parts of hoop and valve, since 
these are very important items in the list 
of conditions which control the phenomena 
in question, and they do not seem to have 
been carefully examined or understood. 
In the fully formed valve of the diatom, 
the edge next the hoop has a broad flange 
projecting inward, whose width is usually ISTHMIA NERVOSA. 
radial lines pointing to the open centre. 
(Fig. 4.) 
ever, be traced by examples of almost every 
stage of its progress. It is found with the 
film so thin that it can only be discovered 
at all by its outline, where a portion of it is 
seen on edge, as it were. In these cases, 
no membrane could be conceived more deli- 
cate, appearing slightly crumpled by the 
pressure which the new valves, being back 
to back, seem to apply to each other by the 
growing force within. At this stage, no 
trace of costse or areolae can be seen. Soon, 
however, as a somewhat more advanced 
specimen is found, faint, hair-like traces of 
the costas are seen. These are found soon 
to be more and more developed; areolae, 
between them, begin to appear, and, finally, 
the fully formed valve, with its oval edges 
resting on the hoop diaphragm. The living 
force within each pair of valves would now 
seem to have gathered its energies for the 
next step, for we find the new valves separ- 
ating from the old ones, and the pair of 
new diatoms severally growing as the parent 
did before them. In this separation the 
new valves leave the hoop diaphragm be- 
hind them, with its braces looking like large 
areolse, and they each seem to be simply 
forced apart from the old valve by the grow- 
ing force of the cell contents, as two pistons 
in one cylinder would be by steam injected 
between them. 
It will thus be seen that the provision for 
a firm connection between valve and hoop 
indicates permanence, and the fact will be 
found in accordance with this. The valve 
and hoop never sever their connection at 
this place during the life of the diatom, and 
in the examination of a great number of 
individuals in which the separation has 
occurred, I have scarcely found half a dozen 
in which the recurved lip of the hoop dia- 
phragm has not been fractured in places, 
thus indicating that the separation was by 
violence and not by natural causes. I re- 
gard it therefore as a settled point that the 
Fig. 4. 
end of the hoop (including the large areolae 
and the diaphragm) after it is once grown, 
never separates from its adjacent valve dur- 
ing the life of the diatom, but that all 
divisions by multiplication, and losses or 
separations of intermediate parts of the 
hoop, leave the connection between this and 
the valve undisturbed. 
7. The Sliding of the Telescopic Hoops, and 
Formation of New Hoop Diaphragms.—The 
growth of the diatom to a size sufficient to 
contain two full sized valves within the 
hoop, may or may not have occurred with- 
out the sliding of the tubes of the hoop 
upon each other. According to my obser- 
vation, it depends upon the fact of the hoop 
of the parent diatom being broad enough 
at the time of its total separation from its 
parent, to equal the depth of two new 
valves. 
6. The Formation of New Valves -in the 
Growth of the Diatom.—The diatom may 
properly be considered as having attained 
its full growth when its valves are so far 
separated as to leave room for the division 
of the cell contents into two, and the forma- 
tion of two new and full-sized valves within 
the hoop. When this work is completed, 
we have no longer one diatom, but two, and 
their union from that time may be regarded 
as mechanical merely. The study of the 
dead specimens can tell nothing directly, 
about the formation of the nuclei and 
division of the endochrome. The growth 
of the siliceous film of the valve may, how- 
In the stages of its history immediately 
following the appearance of the conjugate 
(or sporangial) frustule, as I shall notice pre- 
sently, the hoop of the parent is narrow, 
the whole diatom being very broad in pro- 
portion to its length. Consequently the 
splitting of the hoop by intussusception, or 6 
ISTHMIA NERVOSA. 
the growth of new layers within, and the 
sliding of the tubes upon each other, ac- 
company, in such cases, the growth of the 
cell contents, and may have to be repeated 
several times before room enough for the 
new valves is found. In other cases the 
parent diatom, having room enough within 
its girdle, the telescopic sliding does not 
begin till the new valves are formed, and the 
pair of young diatoms begin to crowd each 
other by their own growth. 
be seen of the growth of a new hoop dia- 
phragm, with its strong braces making the 
large horse-shoe areolse. These also may 
be found in all stages, from the faintest 
trace and scarce visible outline, to the 
strongly marked and complete form. In 
one instance, not only were the faint out- 
lines of the large areolse seen, but through 
the whole interior of the hoop were trace- 
able oval outlines partly compressed into 
hexagonal form, like a transverse section of 
a vegetable stem with closely packed oval 
cells. This is the only instance I have 
found of such indications of the beginning 
of the areolse in a mere outline, but every 
slide I have is abundant in specimens, 
showing that when the inner film of the 
hoop is new and very thin, the areolse are 
very much larger than on the exterior, often 
so much so as to reduce the connecting sil- 
ica to quite narrow bands. Whilst the soli- 
tary example, therefore, will hardly warrant 
an induction from it, the statement as to 
the much larger size of the areolse on new 
films is given as one fully proven. 
What, then, is the process and method of 
this extension of minute draw tubes? I 
have already described the sutures which 
girdle the hoop at different places. The 
full-grown diatom seems to have its hoop 
also ripened into preparation for its new 
office. The separation of the lamellse takes 
place, like the ripening of the shuck of a 
nut, which not only loosens it from the nut 
within, but prepares the seams between the 
segments to open at the slightest applica- 
tion of force. So the concentric films of 
the hoop, and the sutures crossing it, are 
prepared for separation when the growth of 
the cell contents demands more room. The 
weakest joint first yields to the strain, and 
the division goes on until the young plants 
are full grown and drop apart. 
Here, however, I must notice a fact which 
stumbled me. In the cases which I have 
noticed in paragraph 4, where the thick 
hoop has been found with a thinner part 
attached, and this last diminishing to a 
knife edge, the areolse on the thin part have 
not been larger, but smaller if anything, as 
if split from the bottom of a depression 
corresponding to the outer areolae, (Fig. 2, 
b). After careful examination of numerous 
instances, including those in which the 
loosening at the suture is only partial, the 
best solution I can suggest is the following. 
These cases would seem to exist only when 
the separation has been forced, and prema- 
ture, i. e. before the splitting of the hoop 
by intussusception is fully ripened. For in 
most cases of apparently natural growth, 
where the separation of the hoop takes 
place at one of these sutures, the doubling 
of the lamellse on the loosened side extends 
all the way from the suture to the dia- 
phragm, or to the next suture, and the 
sliding of the parts follows the usual appear- 
ance, the areolse of the two films being sim- 
The simplest case is that of a diatom with 
a hoop wide enough for the formation of 
the new valves. Here the ripened hoop is 
at least double in thickness; the young dia- 
toms press upon each other, and one thick- 
ness of the hoop separates from the other, 
at one end by a suture of the inside film, 
and at the other end by a suture of the out- 
side film, and the two slide over each other 
till they slide apart. Many cases of this 
kind have been observed, in which the parts 
of the old hoop are persistent, and form 
the hoop of the new diatom. When the 
halves fall apart, the growth of the cell con- 
tents continues to force the valve out to the 
end of the hoop, or until the diatom is 
again ripened and has its growth. Then is 
witnessed a singular phenomenon. When 
the two valves are separated to the distance 
required for the division of the contents, 
and not till then, the first indications may ISTHMIA NEEYOSA. 
7 
ilar in size. The weight of evidence seems 
to be that the doubling or tripling of the 
lamellae occurs in both of the two methods 
indicated, viz., by the splitting of the ex- 
terior lamella, and by the rapid growth of a 
new one within, the latter case being the 
one in which the areolae appear of very con- 
siderably greater size. 
frustule its descendants decrease regularly 
in width, but gain in length, until the pro- 
portions are exactly reversed, the width of 
the conjugate frustule being equal to the 
length of that from which it sprung, at the 
moment of its full growth, and the conse- 
quent division of its cell contents, while its 
length, when leaving its shell, is only equal 
to the width of its attenuated parent. 
The telescoping of the hoops exhibits all 
the combinations which could be expected 
from the arrangement of sutures which 
have been described. The thick band com- 
monly visible in the growing diatom may 
be caused by the simple lapping of two 
tubes, but it may also be caused by a real 
band, the division taking place at its two 
edges simultaneously, and the divided inner 
hoop sliding both ways out of the band, 
which is loosened, and falls when the separ- 
ation is complete, giving one of the most 
common forms in which a perfect cast hoop 
is found, both margins being smooth and 
unbroken, (Fig. 3, a). 
The facts illustrating and supporting 
these positions are as follows: In carefully 
going over my slides of the dry diatoms on 
algse in situ, I happened upon a specimen 
in which the end of a very wide but short 
frustule was still inserted in a small hoop 
and valve. The line of division between 
the valves of the large diatom was at right 
angles (or nearly so) to that of the small. 
Here was evidently a case of a chick with 
part of its shell still on its tail. Had the 
same been found in a cleaned specimen, it 
might have been accounted for by sup- 
posing the two accidentally to have fallen 
into such juxtaposition; but they were here 
as they had grown, and were the terminus 
of a filament which was plainly before my 
eyes. (Fig. 5, b). A hunt stimulated by 
Before leaving this part of the subject, I 
will mention the fact that with high powers 
the areolae on both valve and hoop can be 
seen to have a thin film of silica over or 
under them, the margin of each areola hav- 
ing an indented appearance, which I inter- 
pret to be little thickened projections from 
the thick margin upon the surface of this 
thin film, analogous to the braces or pro- 
jections of the costae upon the diaphragm 
of valve or hoop above described. 
8. The Conjugate or so-called Sporangial 
Frustule.—For manifest reasons, it would 
appear to be improbable that much evi- 
dence on the subject of conjugation should 
be found in the dead remains of diatoms. 
But in this respect, as in many others, the 
indestructible character of the material has 
been the means of preserving testimony 
which presents itself unexpectedly, and is 
of decisive weight on one or two important 
questions. 
Fig. 5. 
I regard it proven by my observations 
that the so-called sporangial frustule in 
Isthmia propagates by division in the same 
manner as all the other frustules. It seems 
almost equally clear that from this large 
this discovery, resulted in the finding of 
over a dozen more instances, all attached to 
the small parent frustule in a similar man- 
ner, but varying in their own length as they 
had progressed in growth, by the widening 
of the hoop. These being arranged in the 8 
ISTHMIA NERVOSA. 
order of their growth toward maturity and 
division, made a regular series, ending with 
two, one of which showed the faint outline 
of the nascent pair of new valves within the 
hoop, (Fig. 6, a), and the last was a com- 
pleted example of division into two new 
frustules (Fig. 6, b). Later, among the 
cleaned but unmounted material was found 
an aborted frustule, in which conjugation 
had taken place, but the perfect growth of 
the valves had been stopped apparently by 
the breaking of the shell of the parent dia- 
tom too soon. As in the other cases, the 
end of the conjugate frustule was inserted 
into the parent shell, of which nearly half 
remained, and from which it projected 
straight out. Its coat was, of course, sili- 
cified, having withstood the boiling in acids, 
but its areolae were confused and its shape 
imperfect, and no demarcation of valves 
distinguishable, (Fig. 5, a). 
Whoever will take pains to compare the 
phenomena of the growth and division of 
Isthmia with the results of modern botani- 
cal investigations into the origin of the cell 
wall, its growth and division by intussus- 
ception, and the division of the cell con- 
tents, so as to form two or more “ daughter 
cells,” will, I am sure, be impressed with 
the evidence he will find that the general 
laws of 'vegetable physiology, as seen in 
larger plants, are as imjili fitly obeyed in 
these minute structures as an,, . here in the 
vegetable kingdom. 
We have only to bear in mind the fact, 
discovered early by Prof. Bailey, that cellu- 
lar membrane may be demonstrated to co- 
exist with the silica in the diatom walls, 
and the following general statement of 
Sach’s in regard to the morphology of the 
cell, will be found to harmonize very per- 
fectly with the phenomena I have attempted 
to describe: “The systems of striation and 
stratification of a cell wall, intersect one 
another, like the cleavage planes of a crys- 
tal splitting in three directions. But since 
the striations and laminae consist of lam- 
ellae of a measurable thickness, composed 
of alternately denser and less dense sub- 
stance, the cell wall appears to be composed 
of parallelolipedal pieces, distinguished by 
their contents of water. If we for a mo- 
ment disregard the stratification, and as- 
sume that we have two intersecting systems 
of striation, then where two denser striae 
intersect, the densest or least watery places 
are always to be found; where two less dense 
ones intersect, the least dense or most 
watery; and where places of greater and 
less density intersect, areolae of intermedi- 
ate density are formed.” (“Text-Book of 
Botany,” Eng. Ed., p. 30.) The whole con- 
text of the passage quoted, the succeeding 
section on intussusception, and the treat- 
ment of the origin of the partition walls in 
tissue cells, which increase by bi-partition, 
in the same work (p. 72), describe the steps 
of growth in this diatom as perfectly as 
general statements can do. The specific 
peculiarity of the diatom is that the de- 
position of the silica in the membrane of 
The relation of other filaments to these, 
among the specimens in situ, was such as 
to force the mind very powerfully toward 
the conclusion that the strings of large dia- 
toms were those nearer in the degrees of 
Fig. 6. 
descent to the conjugate frustule, and the 
strings of smaller ones were those further 
removed from the ancestor, but nearer to 
the advent of a new conjugation. In the 
presence of these facts it must require direct 
and positive evidence to support the theory 
that the conjugate frustule is a sporangium, 
giving rise to a numerous brood of new dia- 
toms at once. ISTHMIA NERVOSA. 
9 
the cell wall, so stiffens the parts as to make 
the separation of the lamellae of the cell 
wall appear more like mineral cleavage, and 
has brought about the adaptation of the 
parts and of the natural forces to their pe- 
culiar work, so that the singular feature 
of the sliding of the hoops upon each 
other in mechanical fashion, has taken the 
place of the sort of accommodation which 
the softer parts, p* .the common vegetable 
structure f°r the growth of the new 
cells. 
society of the Pacific coast, where these 
forms abound. 
It is a little singular that the observed 
facts which have been detailed in the fore- 
going paper, tend strongly to support the 
earlier view of the character and office of 
the conjugate or sporangial frustule, as 
against the later one. As early as 1818, 
Mr. Thwaites, in the paper in the “Annals 
of Natural History,” in which he made 
known his discovery of conjugation in cer- 
tain species of diatoms, spoke of the occur- 
rence of this phenomenon as “a mixture 
of Endochromes; after which process fissi- 
parous division proceeds as before,” and 
this view prevailed until arguments from 
analogy, but, so far as I know, without 
definite support from actual observation, 
led to the opinion that the conjugate frus- 
tule was in fact a sporangium. Dr. Carpen- 
ter has expressed this opinion in the last 
edition (1875) of his work on the “Micro- 
scope and its Revelations,” saying (p. 317): 
“It has been already shown that the spor- 
angial frustule, even when it precisely re- 
sembles its parent in form and marking, 
greatly exceeds it in size; and this excess 
seems to render it improbable that it should 
reproduce the race by ordinary self-divis- 
ion. Appearances have been observed 
which make it probable that the contents of 
each sporangial frustule break up into a 
brood of gonidia, and that it is from these 
that the new generation originates.” 
Many of the appearances of the diatom, 
at different stages of its growth, indicate a 
degree of plasticity and elasticity of the 
walls, especially in the case of the new form- 
ing valves, which very evidently yield to 
each other, crumpling the membrane more 
or less, before the silication has become 
strong, and presenting quite a different ap- 
pearance from that of the strong and mature 
shell, which is finally exposed to the action 
of the outer water. 
I have spoken of the gelatinous stipes as 
exuded from the sieve-like perforations of 
the lower end of the shell. This is, of 
course, an induction from the appearance 
of different examples of this stipes, and 
from the manner in which it is attached to 
other shells, or to the algae on which the 
plant is found. It would be difficult to de- 
scribe all these appearances, but a summary 
of them in one’s own mind gives great 
strength to the conviction that the mode of 
growth is as stated. It is quite easy to un- 
derstand how a similar exudation may give 
rise to the longer stipes of Gomphonema, 
Achnanthes, Licmophora, etc. It is only 
necessary to imagine the exudation more 
rapid, and rapidly solidifying on exposure 
to the surrounding water, to see how a long 
column may thus be built up, I, however, 
fully recognize the fact that scientific 
methods forbid that this should be regarded 
as anything more than a suggestion to be 
verified or disproved by actual observation 
of the life history of these plants, an inves- 
tigation quite within reach, and which we 
may hope will stimulate the zeal of the 
members of the vigorous microscopical 
In like manner, Dr. Wallich, in the arti- 
cle to which reference has already been 
made, says, “ The sporangial frustule, in- 
stead of being, as heretofore assumed, the 
primary or parent frustule of a new and 
vigorous generation, constitutes in reality 
the expiring phase in the life cycle of a 
generation that is passing away,”—that it 
is a “sporangial cell,” “doomed, on the 
liberation of its living contents, to death 
and immediate decay.” 
The observations I have detailed appear 
to me a demonstration of the fact that the 
conjugate frustule is the “parent of a new 
and vigorous generation,” and that the ear- 
lier belief is the better one. But whilst 10 
ISTHMIA NERVOSA. 
thus enforcing the danger of making nega- 
tive assertions in matters of science, I will 
not fall into the same error by denying 
that there are among the diatomacese any 
sporangia] cells producing broods of goni- 
dia. The cases I have observed in Isthmia 
nervosa show that this diatom is propagated 
from the so-called sporangial frustule by 
ordinary self-division. I have looked with 
some care for evidence of distinct observa- 
tion by others of the origin of any diatom 
from gonidia and have not lighted upon it. 
In regard to the Isthmiece, at least, Dr. 
Carpenter says the phenomena of conjuga- 
tion “have not been clearly made out in 
this group.” (“The Microscope and its 
Revelations,” p. 314.) 
near the connection between hoop and 
valve, no cast hoops would be found, except 
when accident had broken the frustule, for 
the two lamellae would adhere, one to the 
old and the other to the new valve, and this 
would satisfactorily answer the queries pro- 
pounded by Dr. Carpenter in another pas- 
sage, where he says, “In some other cases 
all trace of the hoop is lost, so that it may 
be questioned whether it has ever been pro- 
perly silieified, and whether it does not be- 
come fused (as it were) into the gelatinous 
envelope. ”—ld., p. 314. 
The field is a most fascinating and an 
open one, and let us hope that careful ob- 
servations will be pushed until the life his- 
tory of all the diatomacese shall be thor- 
oughly known, a work which the venerable 
authority just quoted well says, “may ad- 
vantageously occupy the attention of many 
a microscopist who is at present devoting 
himself to the mere detection of differences, 
and to the multiplication of reputed spe- 
cie It is also to be hoped that the valu- 
able work already done in this direction by 
such able naturalists as Prof. H. L. Smith, 
and perhaps others, may be soon brought 
within the reach of all who are interested 
in these studies, so that the task of investi- 
gation may be lightened by avoiding un- 
necessary repetition of steps already taken, 
or by the encouragement of finding verifi- 
cation for observations which might other- 
wise be involved in doubt. 
Upon the topic of the mode of division 
of the hoop and the sliding of the laminae 
upon each other, the same caution should 
be observed lest too large or too hasty con- 
clusions are drawn from the facts. In re- 
gard to Isthmia, the cases observed seem 
conclusively to prove that the hoop proper 
is a permanent part of the diatom, except 
when the separation at the sutures is such 
as to leave a portion of the hoop free at 
both edges, in which event the separation 
of the young frustules drops this segment 
of the hoop. In other cases it remains at- 
tached to one or the other valve, and forms 
the outer coat of the wall of the new grow- 
ing diatom. If we suppose that in any 
species no sutures occur in the hoop, except THE 
AMERICAN JOURNAL OF MICROSCOPY. 
Devoted to the Diffusion of a Knowledge of the Best Methods of Using the 
Microscope; and of all Valuable Improvements in the Instrument, its 
Accessories, and all New Methods of Microscopical Investigation. 
$l.OO per Year. Specimens Sent Free. 
HOW TO USB THE MICROSCOPE; 
Practical Hints on the Selection and Use 
OF THE MICROSCOPE. 
INTENDED FOR BEGINNERS. 
By JOHN PHIN, Editor of the American Journal op Microscopy. 
Price, free by mail, 75 cents. 
IN PRESS. 
HOW TO SEE WITH THE MICROSCOPE. 
Being Useful Hints Connected with the Selection and Use of the Instrument; also 
Some Discussion of the Claims and Capacity of Modern High-Angled Objectives, 
as Compared with those of Medium Aperture. With Instructions as to the 
Selection and Use of American Object-Glasses of Wide Apertures. 
By. J. EDWARDS SMITH, M. D., 
Professor of Histology and Microscopy; Corresponding Member of the 
San Francisco, Dunkirk, and other Microscopical Societies, etc., etc. 
INDUSTRIAL PUBLICATION COMPANY, 
176 Broadway, New York.