typical form, and sheaths enclosing nerve-tubes. This neuroglia is,
however, not so well developed as we have seen it in Annelids and
Crustaceans. It took me some time, therefore, to come to a full
understanding of the real structure of the nerves and the nerve-tubes.
Fresh isolated nerves, as well as nerves macerated in different ways,
I could, usually, though always with some difficulty, split up into
fibrillae, which, however, varied somewhat in size (fig. 17). In
sections, longitudinal as well as transverse, I came to similar results,
and could really distinguish longitudinally running darker stained
fibrillae, of somewhat varying size; in transverse sections they
95
appeared as dark points; in longitudinal srctions as dark lines,
and in oblique sections I could see them as short semi-erect rods.
There could be no doubt that real fibrillae of some kind were pre-
sent, and I was almost disposed to consider HALLER'S description
of the nerves, as consisting of bundles of Primitivfibrillen , to be
right. Besides these fibriUae I could, however, in transverse sections
see a reticulation, as if produced by transsected tubes; in this re-
ticulation the transsected fibrillae were always situated in the walls
of the meshes, and especially in the junctions of the walls of several
meshes (i. e. nerve-tubes) the whole had, consequently, somewhat
of a resemblance to transverse sections of nerves of Annelids and
Crustaceans. I could, however, in preparations obtained in my usual
ways come to no certain results. There was, besides, also another
circumstance which made me certain that those fibrillae could not
be real nervous fibrillae as HALLER and RAWITZ suppose, this was
their staining, which was quite that specially pertaining to neuroglia.
Finally, I succeeded in finding a method of preparing the nerves
which gave a quite clear idea of their structure. This method con-
sists in treating pieces of Patella (the pedal muscular disc in which the
pedal nervercords are imbedded) for about 48 hours in osmic acid
(i %) then washing them, afterwards cutting them and staining the
sections as described above (p. 77).
Examination of sections obtained in this way left no doubt as
to the structure of the nerves ; they contain nerve-tubes having di-
stinct neuroglia-sheaths, in which thicker longitudinally running fibres
are situated (cmfr. fig. 19). l ) These neuroglia-fibres are the Primitiv-
fibrillen of HALLER and RAWITZ. Besides this, there occur stouter
neuroglia septa orriginating in the outer neurilem -sheath. They
ramify and divide the nerve-tubes into larger or smaller bundles
(fig. 19, s, s 7 ). Fig. 19, Jc, k are nuclei occurring in these neu-
roglia-septa.
The nerve-tubes are of somewhat varying size, and are usually
slender and seem to contain but few primitive tubes. I measured
nerve-tubes having a diameter of .002 .006 Mm. In some nerves
the nerve-tubes are, however, much smaller. Fig. 20 is, for instance,
] ) A glance at the origin of the small lateral nerves a' and b' (fig. 19) gives
a very clear idea of the structure of the nerve-tubes, s', s' are the neuroglia-septa
which separate the bundles of nerve-tubes, forming these small nerves, from the rest
of the great nerve. In a and b some of these nerve-tubes are transversally trans-
sected in a' and b' they are longitudinally transsected. It is very distinctly see
that the fibrillge in fact belong to the tube-sheaths.
- 9 6-
an illustration of a longitudinally transsected nerve, drawn under
higher powers than fig. 19. The nerve-tubes are, however, narrower
in fig. 20, than they are in fig. 19 (cmfr. a' and 6'). 1 )
The reason why the longitudinal neuroglia fibres in the nerves and
pedal nerve-cords are isolated in macerated preparations is, evidently,
that they are stouter and stronger than the rest of the tube-sheaths,
and thus are separated from them. In carefully treated macerated
preparations it is, however, possible to see the nerve-tubes. Fig. 17
and 1 8 represent, for instance, such preparations, where nerve-tubes
with a longitudinal striation are distinctly visible, and even to some
extent isolated ; an isolation of them is, however, as a rule, extremely
difficult, for the reason just indicated above. In the extremities o
these nerve-tubes we can see, in some of them, isolated fibrillse,
which, partly, are neuroglia fibrillae, partly, spongioplasmic fibrillae.
Pearls of hyaloplasm are adherent to the sides of them (fig. 17, a, &, d).
The Ascidians.
Finally, I will here mention a group of invertebrates which some
time ago were the subject of my study, the results of which I have
hitherto only given in a preliminary report (1. c. 1886). This
group is the Ascidians. The peripheral nerves of the Ascidians
have a structure very similar to what is described of the peripheral
nerves of Homarus (cmfr. fig. 7), exept as regards the large tubes;
those do not occur in the Ascidian-nerves.
The whole nerve is, in the Ascidians, divided by the neuroglia,
or inner neurilem, into large bundles, these are again divided into
smaller and these, again, into still smaller bundles, the subdivision being
repeated until we at last arrive at the nerve-tubes, which are very
slender, and contain but few primitive tubes; a great many nerve-
tubes seem to consist, even, of only one primitive tube, something
similar to what we have described in Homarus. Very often, it is
even very difficult to decide what are only primitive tubes, and what
are nerve-tubes, and again what are bundles of nerve-tubes, as there
is often seen, in transverse sections, a subdivision into smaller and
smaller tubes, and, the higher and higher we employ the powers of
the microscope the more and more do we trace out minute tubes.
At last, however, under very high powers we are able to observe
a ) Sometimes large vacuoles are seen in transverse section of nerves (cmfr.
fig- 19, c i d)] whether these vacuoles really are transsected nerve-tubes I am not
in a position to decide at present.
- 97
some very slender tubes inside which no subdivision is visible, and
these we must supose to be the primitive tubes, a small bundle of
which generally forms a nerve-tube.
Fig. 21, which represents the transverse section of an anterior
nerve of Phallusia venosa, will, I hope, give easily, at a glance, a
correct idea of the structure of the peripheral nerves of the Ascidians. l )
Fig. 22 represents a part of a longitudinal section through a nerve,
seen under high powers of the microscope; the primitive tubes are,
here, visible, being transsected in their curved course, the curvature
in which arises from the contraction of the nerve, ts, is are the
sheaths of the nerve-tubes. Some of them (a and 6) are stouter
and more prominent than the others , those are probably transsected
neuroglia-septa enclosing bundles of nerve-tubes. In one of these
septa is seen a neuroglia-nucleus (n).
The division of the nerve-tubes into different smaller and larger
bundles is, in the Ascidians, as well as in the Molluscs, Crustaceans
etc. less prominent near the origin of the nerves; close to the nerve-
roots there is, in the Ascidians, scarcely any division visible, and a
great many of the nerve-tubes appear, even, to be broken up into
nerve-tubes of a much smaller diameter ; the reason of this appearance
we will in a later chapter have an opportunity of discussing.
The structure of the nerves of the Ascidians reminds much of
what we have found in the Molluscs. The nerve-tubes are separated
by only very thin sheaths (cmfr fig. 22, ts) in which thickish fibres
are situated, especially along the concreting longitudinal edges; in
transverse sections these fibres, when transsected, appear as darker
dots, situated in the corners of the meshes, and are produced by
the transsected sheaths of the nerve-tubes (vide fig. 21).
Summary.
The results of, these, our researches on the minute structure of
the nerve-tubes of various invertebrates, we may, I think, assume to
be applicable to the nerve-tubes of all invertebrates of such a high
organisation as to posess a well developed nervous system, i. e. all
invertebrated bilaterates. From LANG'S memoirs, we may gather
that the Polyclades and Trematodes have nerve-tubes of the same
type described in this chapter (cmfr. p. 52 54); from FRAIPONT's
') The primitive tubes are not seen in this illustration.
98
memoir on the Archiannelids we may gather that this is the case
also in those animals (cmfr. p. 57), etc. etc. 1 )
Of the results of our researches we may therefore give the
following summary:
i) The nerve-tubes of the invertebrated bilaterates consist of an
external consistent sheath with a viscous contents.
2] The sheaths are formed by, or belong to, the connective sub-
stance extending through the whole nervous system, and which I
call neuroglia. 2 ) In these sheaths nuclei (neuroglia-nuclei), occur more
or less sparingly.
3) The contents of the nerve-tubes consist of primitive tubes,
which are extremely slender tubes or cylinders, separated from
each other or rather formed by membranes or sheaths of a firm
supporting substance, spongioplasm, very much resembling the neu-
roglia-substance ; these slender cylinders of spongioplasm contain a
hyaline, viscous substance, hyaloplasm, which is the real nervous
substance, and which very often is exuded from fresh isolated nerve-
tubes in form of small hyaline pearls.
Fibrillae and fibres, as most writers describe them, do not, con-
sequently, in my opinion, exist.
4) A concentration towards a kind of axis is visible in a great
many largish nerve-tubes of Homarus and Nephrops; this axis may
be more or less narrow, and consists of a bundle of central primitive
tubes which have stouter spongioplasmic sheaths and smaller diame-
ters than the other primitive tubes. A similar concentration in the
centre of the nerve-tubes can, as a rule, not be observed in the
other animals examined ; only a slight indication of it, I believe to
have observed in some nerve-tubes of Nereis.
2. The structure of the ganglion cells, and their processes,
in invertebrates.
Homarus vulgaris.
The contents of the ganglion cells of the lobster consist of
nucleus and protoplasm enclosed in a thicker or thinner membrane
or sheath.
a ) In Myzostoma I have previously indicated a similar structure.
2 ) In my memoir on Myzostoma (1885) I have called it the inner neuri-
lem, as mentioned before.
99
These membranes or sheaths of the ganglion cells have
the same structure as the sheaths of the nerve-tubes, and are formed
of the same substance, viz. the neuroglia; nuclei, having the appear-
ance of common neuroglia-nuclei, occur frequently in them, and are
generally adherent to their outer side; I have seldom observed nuclei
situated on the inner side as described of the nerve-tubes. The
sheaths, enveloping the cells, very often consist of several layers or
membranes, similar to what is described of the sheaths of the
nerve-tubes.
I can not decide, whether inside these layers of neuroglia-sub-
stance there also occurs a thin membrane, spongioplasmic membrane
(different from the neuroglia-sheath), belonging to and arising from
the protoplasm of the ganglion cells. It would, consequently, be a
cell-membrane as many writers describe it to be, but the existence
of which other writers deny. 1 ) The importance of such a membrane
existing, or not existing, is, however, in my opinion, not great as
may be seen from the subsequent description of the structure of
the cell-protoplasm.
When several writers deny the existence of a cell-membrane,
as well as a membrane of connective substance, I think that must
arise from their examining macerated preparations, and imperfectly
stained sections. In macerated preparations, it is often very difficult
to distinguish the enveloping membrane from the cell-protoplasm, it
having a concentric striation which is very similar to the outer
layers of the protoplasm ; neuroglia-nuclei are, however, usually ad-
herent to its outer side and indicate its real neuroglia nature
(cmfr. fig. 23, ri).
In successfully stained sections, e. g., sections stained with
HEIDENHAIN'S haematoxylin-method (vide p. 76), the neuroglia mem-
branes are, even under the lower powers of the microscope, distinctly
visible (fig. 24 29). In such preparations, these membranes also
appear to be intimately connected with the neuroglia-reticulation
extending between the ganglion cells (cmfr. fig. 24), and they look as
if they belonged to that substance, which, in my opinion, is also
*) If such a special spongioplasmic membrane occurs, which is, in my
opinion, not improbable, it would, I think, be so intimately connected with the
surrounding neuroglia-membranes (cmfr. my subsequent description of the cell-
protoplasm and its reticulation) and have an appearance so perfectly similar to
them that, I cannot understand how authors, working with macerated preparations,
can describe special cell-membranes as being different from the surrounding
connective-tissue* membranes.
7*
100
the case, as will also be subsequently referred to. In some parts of
the ganglia, where small ganglion cells are situated closely together,
the neuroglia-membranes form cavities in a similar way as the
waxen-walls in a honeycomb, and one ganglion cell is situated in
each cavity (vide fig. 40).
The membranes also extend into, and envelope, the processes
of the ganglion cells (vide fig. 26). Round the processes of a great
many cells, sheaths of a quite peculiar structure occur, these sheaths
will be mentioned in the description of the processes.
The protoplasm of the ganglion cells of the lobster has
been described in different ways by the many authors who have
treated the subject. In one of the later and more important me-
moirs on this subject, viz. that by FREUD (1882), it is, as we have
seen (p, 34), described as consisting of two substances, a reticulated
substance, and a hyaline viscous substance.
FREUD has specially examined ganglion cells in the fresh state.
I have also examined them in the fresh state, but could not arrive
at a lucid conclusion as to the structure of their protoplasm, neither
could cells isolated by maceration afford much better information
in this respect.
I saw a kind of reticulation, as described by FREUD. This re-
ticular appearance was especially distinct in macerated preparations
(cmfr. fig. 23) the meshes having a circular shape (they have be-
come too distinct in fig. 23; the illustration not being a successfull
one) ; I could, however, not convince myself whether this appearance
was really produced by a substance having a reticular structure or
not. Besides this reticulation I could generally to some extent trace
out a concentric striation (cmfr. fig. 23, it was, however, never so
distinct as it appears in the illustration).
On examination of carefully prepared sections l ) somewhat more
light was, however, thrown on the subject, though I must confess,
that there are several points in which I could still wish to come
to more clearness.
Fig. 24 is a part of a section through two large ganglion cells
in an abdominal ganglion. We can there see that the membranes
(cm) enclosing the cells are stout, and are very distinctly marked,
J ) The best preparation-method for this purpose I found to be, fixing in
chromo-aceto-osrnic acid and staining with Heidenhain's haematoxylin (cmfr. p. j6\
or fixing in Lang's sublimate-solution and staining in Heidenhain's haematoxylin
(cmfr. p. 77). To give the preparations, afterwards, a nuclear staining by Delafield's
hsematoxylin is a very good practice (cmfr. p. 76).
101
In connection with these membranes we find, however, a very
strange structure. In B it has the appearance as if a series of deeply
stained oblong corpuscles were situated along the inner side of the
membrane. Upon closer examination we will see that these cor-
puscles are, as a rule, connected with the membrane by very
slender filaments. In some cases these filaments are even very thick
and the dark corpuscles appear thus to be transsected fibres or septa
issuing from the membrane (cmfr. fig. 24, b). On the other side
we will, on close examination, find slender filaments issuing from
the inner extremities of those dark corpuscles or fibres and pene-
trating into the protoplasm of the cells ; indeed, we will see that
they anastomose and form a reticulation, extending through the
cell-protoplasm and distinctly visible in its outer layers. The
meshes of this reticulation are also largest in the outer layers near
the surrounding membrane (vide fig. 24). They have, here, very
often, an oblong form, going in a radiate direction towards the centre
of the cells, and having, sometimes, the appearance as if the meshes
were formed by filaments radiating from this centre to the surround-
ing cell-sheaths; in fig. 24 B such an arrangement is very distinct.
Within these large meshes smaller meshes are generally seen (fig. 24).
These smaller meshes have, however, a somewhat different appear-
ance, they being not so distinctly visible; I am not sure whether
they are formed in quite the same way.
Towards the inner part of the cell-protoplasm (which is more
deeply stained) the meshes of the first mentioned reticulation be-
come much smaller and become similar to the small meshes just
mentioned, indeed, they cannot be distinguished from each other.
The meshes are so narrow that they are visible only under the
higher powers of the microscope, and, even then, not very distinctly ;
in small cells they are especially very difficult to observe.
In the filaments of this reticulation, granules and thickenings
occur, giving the cell-protoplasm, seen under lower powers of the
microscope, a granular appearance. These thickenings are, especially
in the large ganglion cells, very prominent in the outer layers of
the protoplasm.
In the small meshes of this reticulation, a hyaline substance is
suspended, very similar to the hyaloplasm of the nerve-tubes.
The question which now, as regards the reticulation, very na-
turally forces itself upon us is, whether is it a real spongy reticulation
extending through the protoplasm of the cells, or is it a reticulation
produced by a transsection of tubes in the same way as we have
102
seen in the nerve-tubes. My opinion upon this point is, that this reti-
culation is, to some extent, a real one having partly a quite spongy
structure without the formation of tubes; to some extent the latter
must, however, also be the case, as we shall soon see. The substance
forming the reticulation has, as mentioned, a great resemblance to
the sheaths of the cells as regards its staining etc., and it is, in my
opinion, the same substance which forms the sheaths of the primi-
tive tubes in the nerve-tubes, viz. spongioplasm.
Before we subject the significance of this substance to a further
examination I think it will be well to look somewhat at the struc-
ture of the processes of the ganglion cells and their origin in the
cell-protoplasm.
The contents of the processes, within the cell, its
structure and origin in the cell-protoplasm. - - The ganglion
cells of the lobster are, as a rule, unipolar; if they have several
processes, .there is always one of the process which has a special
appearance, and which is the real nervous process; if other processes
are present they are, as I call them, protoplasmic processes, and
they have, in my opinion, a nutritive function, as will be subse-
quently mentioned. At present we will only consider the structure
and origin of the nervous process. ! )
In well prepared sections, a distinct longitudinal striation of these
processes is easily observed. On observation of transversally trans-
sected processes, it is seen that this striation is produced by primi-
tive tubes having the same structure, and constituting the nervous
processes in the same way, as described of the nerve-tubes.
As regards the staining, the contents of the nervous processes
differs from the protoplasm of the ganglion cells; the latter having
generally a much darker staining. This difference is, for instance,
very prominent in fig. 37 & 38, where the contents of the longitu-
dinally transsected processes have a very light staining and a distinct
striation.
The origin of the nervous processes differs somewhat in the
various cells. In fig. 26 it has the appearance, as if the contents of
the nervous process arises from a convergence of primitive tubes from
the whole body of cell-protoplasm, in which they have a uniform
extension, and to which they generally give, to some extent, a con-
a ) I have in reality, never observed a multipolar ganglion cell of quite in-
dubitable multipolar shape, in the nervous system of the lobster. I am afraid that
it has very often been neuroglia-fibres issuing from neuroglia-sheaths which have
given rise to descriptions of multipolar ganglion cells.
103
centric striation. Some of the primitive tubes we can in many pre-
parations even trace, for some distance, in their course through the
cell-protoplasm, of which they are not, as we have previously seen
(cmfr. reticulation described p. 101), the only constituent, but of
which they are, however, a principal part. l ) These primitive tubes
have the same structure and diameter as those we have described
in the nerve-tubes; they consist of hyaline contents, hyaloplasm,
enveloped in sheaths of spongioplasm, which has the same staining
and appearance, and is evidently the same substance, as that pre-
viously described (vide p. 101), as forming the reticulation in the
cell-protoplasm. In fig. 27 and 28 it is very distinctly seen that
these spongioplasmic sheaths ot the primitive tubes are intimately
connected with the peculiar peripherically situated fibres (fig. 27, & ;
fig. 28, 6) issuing from the surrounding neuroglia-membrane.
In a great many ganglion cells these primitive tubes have not,
however, such an uniform extension through the protoplasm as
illustrated in fig. 26. In the large cells they are generally united
to bundles, distinctly distinguished from the rest of the protoplasm.
In successfully stained sections, 2 ) where they are transversally trans-
sected, they are distinctly visible as larger or smaller light areas
situated in the deeply stained protoplasm. Very often, when the
bundles of primitive tubes circulating in the cell-protoplasm are
numerous and small, we get sections having the appearance illustra-
ted in fig. 25 and 27.
In ganglion cells containing such bundles of primitive tubes circulat-
ing in the protoplasm, the nervous process arises in such manner that
the bundles unite to form the process, as illustrated in fig. 27 and 25.
This union takes place generally within the cell-protoplasm, and the
contents of the nervous process has then, for some distance, an un-
divided course through the cell- protoplasm, and can be traced as a
large light area through a series of transverse sections (cmfr. fig.
28 and 29). It is surrounded by thicker or thinner deeply stained
fibres (fig. 28, s and &; fig. 29, s and s') of the same substance as forms
the reticulation of the protoplasm (cmfr. p. 101). Some of these
*) That it really is primitive tubes, with slightly stained hyaline contents, and
not fibrillae which circulate in the protoplasm of the ganglion cells, is easily
seen in preparations deeply stained by haematoxylin (e.g. fig. 37); they are distinctly
visible as light concentric lines where they are longitudinally transsected (cmfr.
also fig. 38).
2 ) The fixing method and staining described p. 76 and p. 85 can especially
be recommended for this purpose.
104
fibres are, in some sections, seen issuing from the sheath of the
ganglion cell. Similar fibres can also be traced into the prosesses
(fig. 27, 6) where they have a longitudinal course (fig. 30, sf, sf"). Some-
times, in the cells, they are so closely situated that it looks as if they
almost formed a sheath (cmfr. fig. 28) surrounding the process-contents
for some distance into the cell-protoplasm, where they gradually
disappear and obtain the same appearance as the reticulated spongio-
plasm, which spongioplasm often apparently gives, even the smaller
bundles of primitive tubes in the protoplasm, a kind of relatively
firmer surrounding layer resembling a thin membrane, which layer
is generally more deeply stained by haematoxylin than the surround-
ing protoplasm (vide fig. 25, 27, 28, v and 29, v, v 7 ).
Sometimes the primitive tubes are united to a few and very
large masses situated peripherically in the cell. 1 ) Fig. 38, A represents