1094 CUTANEOUS SENSATIONS. [BOOK in. 



scions sensations, the events in the cortex furnish an indispensable 

 link. And the phenomena of the dog in question on the one hand 

 illustrate how complex the chain is, and on the other hand suggest 

 that the completeness of the loss of sensation in the hemiplegic 

 man is not a pure "deficiency" phenomenon, but is due to the 

 lesion affecting the chain of events in some way or other besides 

 merely removing the link furnished by means of the cortex. For 

 as we previously urged, the dog in question, however curtailed its 

 psychical life may have been, seemed to a casual observer to feel 

 and move much as usual. Neglecting visual and auditory sensa- 

 tions with which we are not now dealing, it needed careful 

 observation to ascertain that some of the animal's movements fell 

 short, the failure being apparently due to the lack of adequately 

 energetic coordinating sensory impulses ; a stronger stimulus than 

 usual had to be applied to the skin in order to call forth the usual 

 movements and other tokens that the stimulus was "felt." As 

 we have before urged, it is impossible to suppose that the mere 

 stump of cerebrum left in this case could have taken on all the 

 functions of the lost hemispheres; and making as we have pre- 

 viously done full allowance for the differentiation between man 

 and dog, we must conclude that in the more general sensations 

 with which we are now dealing, as with the more special visual 

 sensations, the full development of a complete sensation is a 

 complex act of more stages than one between the afferent impulse 

 along the afferent nene and the affection of consciousness which 

 we subjectively recognise as ' the sensation : ' the cortical events 

 are only some among several. It follows that any analogy 

 between the cortical events which play their part in a sensation 

 and the cortical events which immediately precede the issue of 

 impulses from the motor region along the fibres of the pyramidal 

 load is misleading ; the highly differentiated motor localisation 

 does not justify us in concluding that there exists a similar 

 topographical distribution of sensations. 



680. We may now attack the problem in a different way, 

 and instead of beginning with the cortex begin with afferent 

 impulses started along afferent nerves from their peripheral 

 endings, and attempt to trace them centralwards. And first we 

 may call to mind what anatomical guidance we possess. ( 569.) 



We have seen that the fibres of posterior roots, the channels of 

 afferent impulses, end in the spinal cord in at least two main ways. 

 One set are continued on, not broken by any relays, as the median 

 posterior tract, and by this tract representatives of all the spinal 

 nerves are connected with the gracile nucleus in which, G10, the 

 median posterior column ends. The other fibres of a posterior 

 root appear to end in the grey matter not far from their entrance ; 

 but from the grey matter there starts the cerebellar tract, which, 

 though not conclusively proved to be, may be assumed to be an 

 afferent tract. We may therefore probably suppose that afferent 



