147. At this stage of the discussion it is needful to consider a point which will spontaneously occur to every instructed reader, I mean the interesting fact discovered by Dr. Waller, that when a sensory root was divided, the portion which was still in connection with the ganglion remained unaltered, whereas the portion which was only in connection with the spinal cord degenerated; and _vice versa_, when a motor root was divided, the portion connected with the cord remained unaltered, the portion severed from the cord degenerated. The observation has been frequently confirmed, and the conclusion drawn has been that the cells in the ganglion of the posterior root are the nutritive centres of posterior nerves, the cells in the anterior horn of the cord being the nutritive centres of the anterior nerves. Another interpretation is however needed, the more so because the fact is not constant.[178] True of some nerves, it is not true of others. Vulpian found that when he cut out a portion of the lingual nerve, and transplanted it by grafting under the skin of the groin, where of course it was entirely removed from all ganglionic influence, it degenerated, but it also regenerated. Pathological observations convinced Meissner that the ganglia are wholly destitute of an influence on the nutrition of the vagus; and Schiff proved experimentally that other ganglia were equally inoperative, since motor nerves could be separated from the spinal cord without degeneration.[179] Not however to insist on this, nor on the other facts of regeneration, in the absence of ganglionic influence, let us remark that Dr. Waller’s examples would not be conclusive unless the teaching of Embryology could be disproved. That nerves degenerate when separated from ganglia is a fact; but it is also a fact that muscles degenerate when separated from a nerve-centre; yet we do not suppose the nerve-centre to nourish the muscles. And against the fact that the sensory nerve remains unaltered only in that portion which is connected with the ganglion, we must oppose the observations of Kölliker and Schwalbe,[180] who affirm that none of the fibres which enter the posterior columns of the spinal cord have any direct connection with the cells of the ganglion on the posterior root. The cells of this ganglion they declare to be unipolar (in the higher vertebrates), and the fibres in connection with these cells are not those which pass to the cord, but all of them pass to the periphery. According to Ranvier, the fibres from the cells join the fibres of the posterior root. Schwalbe found that if the spinal nerve be firmly grasped and _steadily_ drawn, it will often be pulled from its sheath, and the ganglion laid bare;[181] in this ganglion all the cells are found undisturbed, which could not be the case had fibres from _those_ cells entered the cord, since the traction would necessarily have disturbed them. RECAPITULATION.