Structure of the Nerve Cell

  • Santiago Ramón y Cajal


Two classes of membranes appear in nerve cells: 1st., the fundamental membrane, an extremely delicate cuticle constantly present in all cells of the gray matter; and 2nd., the connective-endothelial capsule that is characteristic of sympathetic and dorsal root ganglion cells.


Methylene Blue Purkinje Cell Nerve Cell Silver Nitrate Cerebellar Granule 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    [All authors attribute the discovery of the superficial net to Golgi, and Bethe designated it as the Golgi net; only Athias, in his recent book, credits us with part of the priority of this discovery. Indeed, if we just recognized that Golgi was the first to have given a precise description of the pericellular reticule, we should be allowed to recall that the existence of this reticulum was indicated by us one year earlier (Cajal, 1896f), in our article on the application of methylene blue impregnation to cerebral cells. And we quote: “Occasionally, methylene blue deposits only on cell surfaces, when tissue is fixed by the method of Bethe; it delineates then a membrane dotted with clear vacuoles. This reticular aspect often show very beautifully on the bodies of pyramidal neurons, appearing in perpendicular sections, fresh and impregnated with the lubrification method. However, we shall not risk to conclude from this aspect and without further investigations the fibrillar and reticulated nature of the fine envelope surrounding nerve cells.”]Google Scholar
  2. 2.
    [According to the recent investigations of Veratti (1902), Sánchez (1907) and Holmgren (1908), this tubular apparatus is simply the homologue of the very fine terminal tubular system which we discovered (Cajal, 1890g) in the muscles of insects with the Golgi method. It would be therefore a general cytological arrangement presenting modifications and adaptations corresponding to the structure and functional activity of each cell type.]Google Scholar
  3. 3.
    For example, Lenhossék (1895b), who in his first article advocated for an extremely delicate spongy texture of the axon hillock, is inclined today to consider it as a fibrillar net on the basis of his studies on dorsal root ganglia. The fibrils would form a kind of spiral at the level of the hillock, the center of which would represent the beginning of the axon (Lenhossék, 1896b, 1897).Google Scholar
  4. 4.
    [Poisoning the animal with iodides and bromides, followed by the postmortem change of these salts into silver iodides and bromides, and finally reduction of the latter by light and photographic developers.]Google Scholar
  5. 5.
    [Pyridine fixative, ammonium molybdate mordant, thionin stain, etc.]Google Scholar
  6. 6.
    [Immersion of pieces in silver nitrate, then initial reduction of the various formed salts by heating in an incubator, and finally complete reduction with pyrogallic acid or neutral hydroquinone.]Google Scholar
  7. 7.
    However, my observations in the cat with the methylene blue method, proved that there is only one nucleus in each cell.Google Scholar
  8. 8.
    The main details of this description are taken from my article: Structure of the neural protoplasm (Cajal, 1896b).Google Scholar
  9. 9.
    We could not confirm the findings of Rohde (1895), who described two networks of spongioplasm in the cells of Helix, one situated peripherally and formed by fine threads; the other located more centrally and made of thick trabecules; the fine net would continue in the axon. Neither could we recognize the penetration of neuroglial filaments into the neuronal protoplasm, or the existence of nuclei of neuroglial cells within the neuronal body. Rohde (1893) described the presence of vessels traversing the protoplasm in nerve cells of the electric lobe of torpedo and in other animals. This is true but not new. We have already reported it (Cajal, 1888c), but remained almost unknown to scholars, even to Lenhossék who does not refer to it in his recent work on the texture of nerve cells of Torpedo. We shall discuss later the investigations of Apathy on the structure of nerve cells of worms.Google Scholar
  10. 10.
    It is necessary to place the rheofores longitudinally, and not across the nerve, to elicit axonal discharges by electric stimulation. This experiment has been invoked by Pflüger and others in favor of the fibrillar structure of axons and the lack of conductivity across them.Google Scholar
  11. 11.
    Of the authors who have worked with the Golgi method, only Berkley (1896) accepts the exclusively longitudinal conduction; but because he can not overcome the difficulties arising from the mentioned facts of pericellular connections, Berkley assumes the existence of certain sites of current emision along nerve fiber terminals represented by the varicosities, and certain others along the dendrites, represented by the collateral spines. Terminal fine nerve fibers would be protected, between varicosities, by a fine pellicle of myelin. This conjecture fails just when considering that spines are lacking precisely in the somata and dendritic trunks which are covered with pericellular axonal arborizations (recall the Purkinje cells and their axonal baskets). It should be added that it is impossible to confirm the existence of myelin between varicosities, at the level of which no structural change can be observed. The said varicosities are simply accumulations of cyanophilic material, which are normal in axonal branches, but probably become much more noticeable some time after death.Google Scholar
  12. 12.
    Lugaro accepts the transmission by contact and the possibility of the passage of impulses from the axonal fibrils to the soma and dendrites, but not in all cases. According to this author, the axonal fibrils could also come in contact with other fibrils of the same nature, resulting in sometimes cellulipetal and other times cellulifugal axonal conduction. Such assertions by Lugaro have no bases in positive observational facts; they are mere suppositions born from theoretical preconceptions. Neither does his theory of double conductive mechanism appear better founded, since it raises insurmountable difficulties even in the mere theoretical domain. It is one or the other: either the cytoplasm or cellular juice has a transmitting property or it does not have it. In the latter case, the in toto conduction of the impulses arrived through the pericellular arborizations is impossible; if it has such a property, the individual conduction (the second mode of transmission) can not be understood, because it is clear that the impulse arrived to the soma through an elemental fibril of the axon would communicate to other neighboring filaments, and the transmission would result just as diffuse in one or the other reception mechanism. And in the case of restricting the conductive role of the cellular juice to only short distances, the in toto transmission (by the arrival of impulses through pericellular arborizations) would be partial and well defined.Google Scholar
  13. 13.
    The degree of separation of the arms of the compass would be 1.1 mm for tactile impressions in the tongue, 2-2.3 at the tips of fingers, 6 in the nose, etc. Physiologists report similar distances for thermal sensations. When the arms of the compass become closer to produce a single impression, we must admit that either both points touch branchlets of the same arborization (and this assumption appears to be the most probable) or while one of them touches the branchlets, the other corresponds to a gap devoid of pressure nerves.Google Scholar
  14. 14.
    We may still add that there are sensory processes having axonal properties, or at least those of very thick branches, as it occurs with the central fiber in the corpuscle of Pacini, Krause, etc. Are we also going to suppose in these cases that such fibers consist of a bundle of independent conductors?Google Scholar
  15. 15.
    According to Retzius, just the outer hair cells reach to 12,000 in man, and as each of these cells has 16 or 20 hairs, if one supposes that each of the hairs continues with a conducting filament and receives a particular impression, the number of differentiated sounds would be close to 300,000, which is obviously excesive.Google Scholar
  16. 16.
    Wolkmann and Aubert have shown that for the image of a luminous point to be visible at 60 degrees from the fovea, it must have a 150 times greater diameter than that of the image distinctly perceptible in the said pit. The perception of two points requires an even greater separation.Google Scholar
  17. 17.
    At first sight, it appears that excitations reaching the soma by way of a long dendrite must travel through the axon later than those arrived through a shorter dendrite. However, the hypothesis that all excitations arriving at the same receptor apparatus fuse in a single impulse in the soma or on reaching the axon, appears credible if we consider the velocity of neural transmission (28 meters per second), the small differences in the length of dendrites, and the known phenomenon of persistence or storage of impressions received by the neuron. In fact, calculating the time it takes the neural excitation to travel 0.5mm (the maximum difference in length among branches of a single dendrite), results in about 1/60,000 sec, i.e. an interval very much shorter than that necessary to fuse two succesive visual, auditory or tactile exitations, into one. This interval is, in fact, 1/37 sec, 1/133 sec, and 1/640 sec, respectively.Google Scholar


  1. a.
    Fig. 40. — A, B, two neurons of the ventral cochlear nucleusGoogle Scholar
  2. b.
    To avoid confusion, we have changed throughout Cajal’s terms protoplasmic chromatin for protoplasmic chromatic substance.Google Scholar
  3. c.
    Fig. 41. — A, B, two cortical neurons; b, unidentified.Google Scholar
  4. d.
    Histologie omits mentioning Fig. 43 in the text.Google Scholar
  5. e.
    Fig. 44. — B, nucleus; a, single stem.Google Scholar
  6. f.
    Fig. 45. — a, Golgi apparatus in gland cell; b, gland cell; c, d, Golgi apparatus in epithelial cell; e, surface of mucous membrane.Google Scholar
  7. g.
    Cajal uses the term myeloplasias for mast cells.Google Scholar
  8. h.
    Fig. 52. — b, same as a, i.e. bundles of neurofibrils; f, obliquely cut superficial neurofibrils.Google Scholar
  9. i.
    It is interesting that in spite of the wrong concept of transmission of the nerve impulse through intrasomatic channels, Cajal already considers the axon hillock as a region where conduction velocity may change.Google Scholar
  10. J.
    It took a few more years to discover the process of nerve impulse conduction as a membrane phenomenon (Bernstein, 1902, Pfiüger’s Arch, 92: 521-562; Adrian, 1913-1914, J Physiol, 47: 460-474).Google Scholar
  11. k.
    Fig. 54. — a, axon; b, bundles of neurofibrils converging toward the axon.Google Scholar
  12. l*.
    Fig. 55. — A, B, two neurons.Google Scholar
  13. m.
    Fig. 57. — Histologie reads in error f instead of g; f, neurofibrils entering a dendrite; A, B, C, D, four cortical pyramidal neurons.Google Scholar
  14. n.
    Fig. 59. — A, cell soma.Google Scholar
  15. o.
    Fig. 63. — I, II, III, IV, four neuronal nuclei.Google Scholar
  16. p.
    These early descriptions of nucleolar fine structure were confirmed by electron microscopy (Bernhard, Granboulan, 1968, in Dalton, Haguenau (eds) Ultrastructure in Biological Systems, Academic Press, New York, 3:81-149).Google Scholar
  17. q.
    Histologie reads in error nucleus instead of nucleolus.Google Scholar
  18. r.
    Fig. 68. — A, B, large cells; C, small cell.Google Scholar
  19. s.
    Fig. 69. — G, f, unidentified.Google Scholar
  20. t.
    This section is in great part superseded by the description and discussion of neurofibrils on p. 141 et subseq.Google Scholar
  21. u.
    It is hard to find the meaning of myeloplaxies, since apparently Cajal differentiates them from mast cells at this point. See annotation g.Google Scholar
  22. v.
    It is difficult to interpret the meaning of “Golgi corpuscles” as used here by Cajal. They cannot refer to Golgi type II cells, or to the Golgi cell of the cerebellar cortex since both are short axon neurons.Google Scholar
  23. w.
    The concept of summation is embedded in this thought.Google Scholar

Copyright information

© Springer-Verlag Wien 1999

Authors and Affiliations

  • Santiago Ramón y Cajal

There are no affiliations available

Personalised recommendations