C. ESTABLE 329 



addition, the important cross synapses, with all the neurones whose 

 dendrites are spread within the plexiforni layer. 



In an article, published in 1923, we called iiiossy-cclliilar conicoarcbcUar 

 system, a complex of neurones directly and functionally linked, on the 

 one hand, by means of axo-dendritic cross synapses, on the other hand, 

 through synapses of the mossy fibres with the dendrites of the grains 

 (glomeruli). The afferent synaptic spectrum of each Purkinje neurones is 

 integrated by a small number of axo-somatic, axo-dendritic and axo- 

 axonic synapses of pre-synaptic terminal pars, while the cross axo-dendritic 

 synapses are astoiiisliinoly ntiinerons, existing in the proportion of about one 

 million to each Purkinje neurone (Estable, 1923). 



In the spreading and beautiful dendritic arborescence, the slender 

 axons of the grains, in parallel fibres, establish as many intimate contacts 

 with Purkinje neurones as verticils are formed by the microcluhs or thorns 

 bristling on their surface (Figs. 9, to). 



It may be deemed fantastic to talk about one million cross synapses for 

 each Purkinje neurone. There are, however, many techniques that confirm 

 this figure. Furthermore, the axon of each single grain contacts all the 

 Purkinje neurones in a cerebellar convolution, succeeding one another in 

 the same direction. 



Three more significant facts prove the importance of the axo-dendritic 

 cross synapses: (i) they are not missing in any centre examined; (2) the 

 said synaptic type is the only one in the afferent spectrum of Golgi short 

 axon cerebellar neurones; (3) the cross synapses are remarkably prevalent 

 over all the other synaptic types in the brain cortex. A single axon, 

 among billions in the plexiforni layer of the brain cortex, relates itself by 

 means of cross synapses, with the neurones that, from the different 

 layers, send their dendrites up to the pial surface. 



In a re-examination of the neuro-histological bases of neuro-physio- 

 logy, it is particularly important to study the retina. 



Most heterogeneous ramifications occur in the outer-plexiform layer: 

 [a) thin filaments branch oR the feet of the cones and rods and in them no 

 dendritic penetration has been noticed, as happens in the widest part of 

 the feet; (b) dendrites of variable thickness, of two origins, from bipolar 

 neurones and from horizontal neurones (there is no evidence that all the 

 dendritic ramifications of the bipolar invagmate in the pedal widening of 

 cones and rods); (c) axonic ramifications of the horizontal neurones and 

 perhaps of the stellate neurones of the inner granular layer, besides certain 

 ascending axons similar to the retinopetal fibres of the inner plexiforni 



