Structure of the nerve-cell.-Every nerve-cell consists, as has been already mentioned, of a nucleated enlargement-the nerve-cell-body (nerve-cell proper of some authors)—and of one or more processes. If there is only one process, this is an axon and becomes the axis-cylinder of a medullated nervefibre, or forms the bulk of a non-medullated fibre. If two processes, both may be axons, as with the cells of the spinal ganglia. Generally, however, there is one axon and one or more dendrons, the latter ramifying almost immediately and usually ending not far from the cell-body. The axon may come off either directly from the cell-body or from a dendron.

While it is universally conceded that the axon of the cell is a conductor of nerve-impulses, the view has been taken that the dendrons or protoplasmic processes are merely for purposes of nutrition, serving the office of roots to the cell-body. This view was upheld by no less an

B

A

FIG. 26.-TO ILLUSTRATE THE SO-CALLED LAW OF CON-
DUCTION IN THE CEREBELLAR NEURONES. (Cajal.)

A, an afferent (moss-) fibre conveying impulses to the dendrons of a, the granule-cells. The axons of these pass them on to the dendrons of b, a basket-cell. The axon of the basket-cell again conveys impulses from its cell-body to the cell-bodies of the cells of Purkinje, c, and by the axons of these, B, impulses are conveyed away towards the white substance. C, another afferent fibre passing at d to the dendrons of the Purkinje cells, and carrying impulses from some other source. From the dendrons they may be supposed to pass to the cell-bodies and then away by the axons. The arrows indicate the course taken by the nerve-impulses, which is seen in every case to be towards the cell-body in the dendrons and away from the cell-body in the axons.

authority than Golgi, the eminent anatomist of Pavia, who regarded all connexion between the nerve-cells as taking place between axons and their collaterals, the ultimate branches of which he supposed to terminate in a fine nervous network in the grey substance.1 The existence of such a network had been previously conjectured by Gerlach. But the conducting function of the dendrons was put beyond a doubt by the discovery -made originally by Cajal,' and since abundantly confirmed (1) that the branchings of the axons do not end in a network, but have free terminations, and (2) that the axons of one nerve-cell form close contact-connexions (synapses) with the dendrons of another, of such a nature that conduction can only take place through the dendrons towards the cellbody and axon, and away from the dendrons and cell-body by way of the axon and its collaterals. This constitutes the so-called law of conduction' or 'of dynamic polarisation' within the neurone (fig. 26).* The spinal ganglion-cells offer, however, an important exception to this general law, for in them the peripheral process, which is unquestionably an axon, habitually conducts impressions towards the cell-body.

Many cells have long axons which terminate by arborisations at a distance from the cell-body

(fig. 27 A). Others are provided with short axons which ramify in the grey matter near the cell-body (fig. 27 B). These form distinct types (types i. and ii. of Golgi) which may be termed the long-axoned and short-axoned cells respectively. The cells of type i. effect communication between the central nervous system

Golgi's first observations, made by the well-known method discovered by him, were published in the Gazetta medica lombarda in 1873. A further important contribution to this subject appeared in 1886 under the title 'Sulla fina anatomia degli organi centrali del sistema nervoso.' These and other articles on the same subject were collected and republished in German in 1894 under the title 'Untersuchungen ü. d. fein. Bau d. centr. u. periph. Nervensystems.'

Stricker's Handbook of Histology, 1871. It is represented in fig. 22a.

3 Riv. trimestr. micrograph. 1888 and 1889.

4 Cajal, Riv. trimestr. 1897. This hypothesis applies, of course, only to nervous impulses set up naturally, and not as the result of artificial excitation of the fibres.

and the periphery, and between parts of the nervous system which are distant from one another. Those of type ii. effect communications between adjoining

grades of transition exist
between the two types,
although in most cases
the differences are easily
recognisable. As a general
rule, the longer and larger
the axon the larger the
cell-body from which it
arises: hence a vast dif-
ference is found in the
size of nerve-cells, in
general correspondence
with the very varying
length of the nerve-fibres
which spring from them.
This relation between
size of cell-body and size.
and length of axon is,
however, not
exceptions.

without

FIG. 27A.-A LONG-AXONED NERVE-CELL FROM THE
CEREBRAL CORTEX. (Golgi.)

a, axis-cylinder process with collaterals; d, d, dendrons;
b, body of cell.

The body of a nerve-cell always contains a characteristically large spherical nucleus with a very distinct nucleolus and membrane, but the chromatin network

of the nucleus is delicate or may not be discernible. Like other cells, the nerve-cell contains a centrosome (Lenhossék, Lewis), but it is not always easy of demonstration.

In the protoplasm of some nerve-cells an irregular network of fine canaliculi (fig. 28 A and B) has been described by Golgi1 and others; according to Holmgren this

B

b

a A

EC

is occupied by processes of extraneous cells (trophospongium). By other methods a fine superficial network (fig. 29) may be demonstrated on the surface of many nerve-cells; whether this is related to the neuro-fibrils or whether it is of neuroglial origin is uncertain.

The cytoplasm of many nerve-cells contains small angular patches of a proteid material (neuro-chromatin) which stains deeply with basic dyes (figs. 30, 31), apparently owing to the presence of nucleoprotein. These also contain a small amount of iron (Macallum). The angular patches were first fully described and their importance recognised by Nissl 2-hence the name Nissl granules applied to them. They undergo changes with alterations in activity of the nervecell, becoming built up during rest and broken down during activity (fatigue-effect; fig. 31). The breakingdown process is known as chromatolysis, and is brought about by various agencies besides fatigue, such as the action of certain drugs and the effect of diseased

FIG. 29.-SUPERFICIAL NETWORK OF GOLGI SURROUNDING
TWO CELLS FROM THE CEREBRAL CORTEX OF THE CAT.

Ehrlich's method. (Cajal.)

A, large cell; B, small cell. a, a, folds in the network; b, a ring-like condensation of the network at the poles of the larger cell; c, spinous projections from the surface.

1 First in spinal ganglion cells; subsequently in the cortical cells and elsewhere. (Verh. d. anat. Gesellsch. 1901).

2 Tagebl. d. Naturforscherversamml. zu Strassburg, 1884; Centr. f. Nervenh. u. Psych. v. 1894. The granules had been previously seen by other observers, but their functional significance had not been recognised.

conditions of the nervous system, or of the body generally. Most strikingly is this change of the neuro-chromatin effected if the axon of the nerve-cell be cut in any part of its course (axonal reaction). A few days after such a lesion the Nissl granules begin to alter and break down, and may ultimately disappear; the cytoplasm becomes swollen, and the nucleus assumes a peripheral position (fig. 32). In most cases the Nissl granules are built up again after a time. But sometimes this is not the case, and in these instances the cytoplasm undergoes ultimate atrophy. Some nerve-cells are altogether destitute of Nissl bodies, but they are rarely absent from the larger type of cell. In these they also occur in

FIG. 33.-NERVE-CELLS OF KITTEN (FROM THE ANTERIOR CORPORA QUADRIGEMINA)
SHOWING NEURO-FIBRILS. (Cajal.)

a, axon; b, c, d, various parts of the intracellular plexus of fibrils.

the dendrons, but never in the axon nor in the part of the cytoplasm from which the axon issues.

Nerve-cells occasionally contain clumps of brown or black pigment lying near the nucleus.2 Where many cells in any part of the nervous system are thus pigmented, a characteristically dark appearance is presented by the grey matter, as in the so-called locus niger.

1 For the literature of this subject, see Barker, Nervous System, 1899; Cajal, Textura del sistema nervioso, 1899; Schäfer, article 'Nerve-cell in Text-book of Physiology, 1900; and Van Gehuchten, Bull. de l'Acad. roy. de Belgique, 1897, and Anatomie du système nerveux, 1907. Chromatolysis may also occur in motor nerve-cells as the result of cutting corresponding sensory neurones (Warrington, Journ. Physiol. xxiii. xxiv. xxv. and xxx., 1897 to 1904; and Bräunig, Arch. f. Anat. u. Phys. 1903) ? On the pigment in nerve-cells, see Obersteiner, Arbeiten a. d. neurol. Instit. Wien, x. 1903.