66 RAGNARGRANIT 



silence the firing cell, the reason being that by the electrical tetanization of 

 the extensor afferents excitatory drive is well maintained in spite of the 

 lowered depolarizing pressure. Such experiments show that it is necessary to 

 distinguish between excitatory drive and depolarizing pressure. 



In experiments in which the motoneuron has been silenced by recurrent 

 inhibition it is often observed that maintained stretch does not succeed in 

 reactivating the cell although antidromic stimulation is stopped. This is 

 difficult to explain unless one assumes that recurrent inhibition penetrates 

 into the spinal cord beyond the circuit completed with the projections of 

 Renshaw cells upon motoneurons. Frank and Fuortes (1956) showed that 

 neurons located further inside the spinal cord are influenced by antidromic 

 stimulation and this has since been confirmed. It is therefore possible that 

 recurrent inhibition does something to the interneurons which leads to 

 removal of excitatory drive, provided that drive is low. 



A general theory of the physiological role of recurrent inhibition follows 

 from the results obtained. The recurrent control will preferentially be directed 

 towards removal of discharges or states of excitation which are badly sup- 

 ported by excitatory drive, lingering after-discharges, subliminal fringes, 

 near-threshold activity in general, and so, as it were, will hold the reflex to 

 its task. The present author has often wondered why interneurons fire at 

 such high rates and why aff'erent activation often is so much in excess of 

 what is the immediate apparent need (see e.g. in Granit, 1955, p. 247) but it 

 is clear that if excitatory drive is as important functionally as depolarizing 

 pressure, then what superficially looks like excess activity is merely what is 

 required to maintain low-rate operations of neurons provided with recurrent 

 collaterals. As is well known most nervous centres possess recurrent col- 

 laterals. The motoneurons are by no means an exception. The views of 

 ourselves (paper no. 1) and Brooks and Wilson (1959) with regard to special 

 functions of recurrent inhibition fit well into this general theory. Also, 

 whatever organizational features be ascribed to the recurrent system, the 

 inhibitory effect will have to be in accordance with the general rule that has 

 emerged from the work now reviewed. 



Recurrent inhibition on a tonic discharge can, as we have seen (paper 

 nos. 1 and 3), be made cumulative in the sense that it generally silences the 

 discharge, the intervals between the eff'erent impulses increasing from spike 

 to spike. This is done by reducing the amount of excitatory drive by which 

 anyone depolarizing pressure is maintained. However, assuming drive to be 

 sufficient, what is then the relation between (control or) normal frequency of 

 discharge F„ and that during recurrent inhibition F,? 



In order to be able to reply to questions of this type it is necessary to be 

 able to vary the firing frequency Fn of any given cell and try recurrent inhibi- 

 tion on it. Many motoneurons are so heavily stabilized in firing rate that 

 they cannot be used in a study of this particular kind. They simply refuse 



