March 15, 1894] 



NATURE 



46; 



motor and sensory cells, since morphologicaHy motor cells are 

 found in the olfactory bulb and the retina, and Goigi's sensory 

 cells are sparsely found in the same regions, and, consequently, 

 it is impossible to deduce the function of a cell from its shape 

 and mode of branching. 



The connection of the axis cylinder with the sensory cells of 

 the grey matter is not by the mediation of a network, but by free 

 arborisations around cells. 



In birds and mammals the cells in the root ganglia have an 

 axis cylinder which extends from the periphery, and the internal 

 branch, entering the cord by the posterior root, bifurcates in 

 the white matter. An ascending branch can be traced for 

 several centimetres along the posterior column, and is found to 

 end by arborisations around ceils in the grey matter. The 

 descending branch has a similar distribution. All branches, 

 however, do not bifurcate. Collateral branches, long and short, 

 pass off in bundles at right angles from the main branch audits 

 bifurcations ; the destiny of the short collaterals is the grey 

 matter where their varicose arborisations surround the cells in 

 the head of the posterior horn and the cells of Clarke's column. 

 The long collaterals pass in a bundle from the ascending 

 or descending oranches and ramify in the substance 

 of the anterior horn, where they come in contact with 

 the bodies or the protoplasmic prolongations of motor cells. 

 From this distribution it is obvious that the extremity of the 

 long collateral is in contact with the body or the protoplasmic 

 processes of the motor cell. For this reason Prof. Cajal speaks 

 of the long collaterals as " sensitivo-fnotor," though KoUiker's 

 term " reflexo-viotor" enables the physiology of these to be 

 the more easily grasped. 



The grey matter of the cord contains at least four types of 

 cells — the commissural, where the axis cylinder of the cell is in 

 connection with the opposite antero-lateral column by way of 

 the anterior commissure, cells in connection with the antero- 

 lateral and pristerior columns of. the same side, motor cells in 

 connection with the anterior root and '■^ pluricordonaV^ cells, 

 where a complex axis cylinder furnishes two, three, or more 

 m^dullated fibres in connection with the columns of one side or 

 of both. 



Cajal holds that, according to the strength of the excitation, 

 impulses entering by the posterior root may pass by the long 

 collateral to the motor cells, and the expression of this is a 

 reflex, or where the excitation is stronger, besides this route, 

 the short collaterals as well as the ascending and descending 

 branches of the bifurcated sensory fibres may conduct, in con- 

 sequence of which other cells are thrown into activity. 



Cajal considers the retina as a nerve ganglion formed of 

 three tiers of neurones, the first of which includes the rods 

 and cones, together with their processes as far as the external 

 granular layer ; the second composed of the bipolar cells, and 

 the third of the ganglia nic cells. 



The internal and ex'ernal molecular layers are the regions 

 where the connections of the neurones are established. The 

 excitatory process, started in the rods and cones, passes along 

 the bipolar cells, the ganglion cells, the fibres of the optic 

 nerve, into the fusiform and pyramidal cells of the geniculate 

 body and the corpus quadrigeminum. 



The optic nerve contains also centrifugal fibres which ter- 

 minate by varicose arborisa'ions around the spongioblasts of the 

 retina, to which they carry impulses started by nerv.jus excita- 

 tions of central origin, the significance of which is obscure. 



In the cerebellum a tiansverse section shows three concentric 

 layers of neurones ; the first, or molecular layer, consists of 

 small stellate cells, the second of the cells of Purkinje, and the 

 third of the granular layer. All these elements have con- 

 nections of two kinds — intrinsic, which place the cells of the 

 three layers in connection with each other, and extrinsic 

 between the cerebellar neurones and the neurones of other 

 nervous organ«. 



The connections of the granules, which are nervous organs, 

 with the cells of Purkinje, are of great interest. The former 

 possess three or four very short protoplasmic processes, each 

 of which breaks up into an arborisation. An axis cylinder of 

 exceeding fineness passes up to the molecular zone, bifurcating 

 at various level-. During their course they come into intimate 

 contact with the protoplasmic processes of the cells of Purkinje. 

 Since each of these parallel fibres traverses the total thickness 

 of the grey matter of a cerebellar convolution and ends by free 

 extremities at the surface, it follows that a single granule is able 



NO. 1272, VOL. 49I 



to act on a multitude of cells of Purkinje. Each of these last is 

 under the influence of a considerable number of granules. 



The extrinsic relations (those between the cells of the cere- 

 bellum and those of other nervous centres) are very difficult to 

 establish. 



As Golgi first showed, the cells of Purkinje give rise to nervous 

 prolongations of the long type of which the termination is 

 unknown, and, on the other hand, there end in the grey matter 

 of the cerebellum axis cylinders coming from other organs, of 

 which the situation is very uncertain. These are the fibres 

 fnoHssztes and the fibres grhnpantes. The fibres mottssiies ter- 

 minate in the molecular layer by collateral processes which are 

 in contact with the protoplasmic expansion of the granules. The 

 ultimate twigs terminate in a varicosity, or in a small ramifica- 

 tion. T^vQ fibres grimpaiites traverse the granular layer, cours- 

 ing along the cells of Purkinje, and .surrounding the ascending 

 stem and the proplasmic branches with an elongated terminal 

 arborisation quite comparable with that of a motor fibre in 

 muscle. 



It appears therefore that the cells of Purkinje may receive 

 nervous impulses from other centres, either by means of the 

 fibres inoiissues, or by means of the fibres grimpantes ; whilst 

 the small stellate cells of the molecular layer, as « ell as the large 

 stellate elements of the granular layer belong to the second 

 type of Golgi's cells, appearing to have no relations with the 

 extrinsic fibres. These last cells are therefore styled "associa- 

 tion corptucles," as they appear to have for their exclusive lole 

 the association of the cells of Purkinje, or the granules, into a 

 dynamic whole of which the significance is unknown. 



In the cerebral cortex, f )r the sake of clearness, three main 

 layers may be distinguished, a molecular layer, a layer of large 

 and small pyramidal cells, and a layer of cells of various shapes. 

 The molecular layer, which is always found in the brains of 

 vertebrates, is formed of a very complicated plexus, the principal 

 factors of which are the peripheral ramifications of the pyramidal 

 cells, the terminal nervous arborisations of certain cells of the 

 pyramidal layer of which the axis cylinders are ascending, and 

 the ramifications of certain cells of fusiform or triangular shape, 

 the greater part of whose expansions become horizontal, and 

 resolve themselves into a large number of twigs. One may 

 compare these elements with the spongioblasts of the retina and 

 with the granules of the olfactory bulbs, as they also are 

 without a differentiation into protoplasmic and nervous ex- 

 pansions. 



The layer of pyramidal cells, the thickest layer of the cortex, 

 consists of many elongated cellsof pyramidal form, the principal 

 characteristic of which is the possession of a protoplasmic stem, 

 terminating in the molecular layer as a more or less horizontal 

 arborisation of fibres, covered with spiny processes, and giving 

 off many lateral and descending protoplasmic branches, and 

 finally giving rise tc a descendirg axis-cylinder c mtinued to the 

 white substance. The last layer consists of cells of variable 

 form, usually elongated, one of the prolongations very often 

 going towards the surface. The axis cylinder penetrates the 

 white substance, and resembles that of the pyramidal cell. 



In their passage through the grey matter all the axis cylinders 

 of the pyramidal cells and the cells of variable shape give off a 

 large number of ramifying collaterals, which terminate freely 

 aroand the nerve cells. The whole of the ramifying collaterals 

 form in the grey substance, and around the cells, a plexus of 

 extreme complexity, in ^\hich are also present ramifying col- 

 lateral twigs from the white substance and terminal arborisa- 

 tions of fibres of association. 



The connections of the pyramidal cells of the cortex may be 

 distinguished as j///t'r/?i/rt/ (belonging to the molecular layer) 

 and Ji-ep (belonging to the subjacent layers). 



In the molecular layer each protoplasmic "plume" of the 

 pyramidal cells is in contact with an almost infinite number of 

 terminal nervous fibrillar derived from the terminal arborisations 

 of fibres of association originating in cells in the hemisphere of 

 the same or of the opposite side ; from special cells in the 

 subjacent layers ; from special cells in the molecular layer 

 itself; from collateral fibres from the white substance, or fron: 

 the deep layers of the grey substance, and from other situations. 



In the molecular layer, then, each pyramidal cell may be 

 influenced not only by the cells of the same region of the cortex, 

 but also by others which lie in other lobes, it may be of the 

 same side or of the opposite side of the brain. It is also probable 

 that the molecular layer receives the ultimate ramifications o' 



