8o8 A MANUAL OF PHYSIOLOGY 



plete section of the cord in the cervical or upper dorsal region, and 

 conclude that the spinal reflexes in the higher animals are far more 

 dependent on the upper portions of the central nervous system than 

 in the frog. 



The phenomena of spinal shock and its varying severity in 

 different animals may be accounted for by the rupture of the paths 

 normally used in the reflexes. The theory that the shock is 

 due to an inhibition set up by the mechanical injury is untenable. 

 For shock affects only the portion of the central nervous system 

 distal (or aboral) to the lesion. When a dog is allowed to live after 

 transection of the cord in the lower cervical region till shock has 

 been recovered from, a second transection distal to the first is 

 followed by only slight and very transient depression of the reflex 

 power, although the direct effect of the second injury ought, of 

 course, to be as great as that of the first. Finally, according to 

 Sherrington, the condition of the spinal reflex arcs in shock differs 

 from th^ condition caused by inhibition, and resembles rather a 

 general spinal fatigue in which conduction along the arc and especially 

 across the synapses is difficult and uncertain. This condition is 

 supposed to be due to the loss of a ' tonic ' influence of higher centres, 

 assumed to be necessary for the maintenance of the normal con- 

 ductivity of the arc. These cranial centres, if they exist, or, at 

 least, the most efficient of them, must be assumed to be situated 

 distal to the cerebral cortex, probably in the pons or midbrain. 

 For section just behind the pons causes much more severe shock than 

 removal of the cerebral hemispheres. 



Peripheral Reflex Centres. The question whether any reflex 

 centres exist outside of the spinal cord and brain, and especially in 

 the sympathetic ganglia, has been the subject of a lengthy contro- 

 versy. That the spinal ganglia cannot act as reflex centres is 

 generally acknowledged, and it is not difficult to see that, for ana- 

 tomical reasons, this must be so. A reflex arc must, so far as we 

 know, in all highly-organized animals, include at least two neurons. 

 There is no proof that an afferent impulse can ascend an axon to a 

 cell-body and there excite an efferent impulse, which, descending 

 the same axon in a separate set of fibrils, gives rise to a reflex con- 

 traction, or a reflex secretion. Now, the cells of a spinal ganglion 

 represent the original neuroblasts from which the posterior root- 

 fibres grew out as processes towards the cord on the one side and 

 the periphery on the other. A sensory fibre passing into the 

 ganglion makes connection with a cell by a T-shaped junction and 

 passes on its course again. No afferent fibres run from the nerve- 

 trunk into the ganglion, to end in arborizations around the ganglion 

 cells, and no efferent fibres arise from nerve-cells in the ganglion to 

 pass out into the trunk. For although a slightly greater number of 

 medullated fibres of small calibre is found in a spinal nerve-trunk 

 immediately distal to the junction of the roots than in both roots 

 taken together, this appears to be due to the passage into the nerve 

 (from the grey ramus communicans) of raedullated fibres which end 

 in the bloodvessels or other tissue of the ganglion (Dale). Here 

 it is evident that there is no possibility of a complete reflex arc. 

 Indeed, it is not certain that the normal afferent impulses pass 

 through the bodies of the spinal ganglion cells at all. For (i) a 

 negative variation can be observed in the posterior roots above 

 the ganglia on stimulation of the trunk of a frog's sciatic nerve 

 more than two days after the death of the animal, when the ganglion 



