^3- 



HA.NDBOOK OF PHYSIOLOGY ^ NEUROPHYSIOLOGY II 



among the cortical cells firinc; earliest, discharge of a 

 far greater number is delayed until the rising limb of 

 the C wave. This pattern is consistent with the sugges- 

 tion of Lorente de No (75) that corticofugal cells are 

 mainly excited through cortical interncurons inter- 

 posed between them and the primary afferent plexus 

 in layers III and I\'. 



The data presented in the left part of figure 15 give 

 only a partial picture of the contriijutions of cortical 

 cellular spikes to the total cortical activity resulting 

 from an incoming afferent volley because only initia- 

 tion of firing is represented. Many cortical units 

 (especially Betz cells) fire repetitively in response to 

 a single shock to the foot pad or a peripheral nerve. 

 The discharge train of a Betz cell may occupy 10 to 

 15 msec. To estimate total contributions for each of 

 140 recorded units, probabilities of spike discharge 

 (irrespective of the spike position in the train) were 

 computed for each millisecond interval up to 44 

 msec, after a shock to the contralateral forepaw. Five 

 discharges were measured to compute the probabili- 

 ties.' Summation of the proijabilities of individual 

 units gives a measure of the probable number of 

 spikes during each interxal. Data from 74 Betz cells 

 and 230 other cortical elements are illustrated in the 

 right part of figure 15, along with the reconstructed 

 surface recording. The greatest spike activity occurred 

 during the intennediate time range between peak 

 positivity and peak negativity of C, but compared to 

 that in figure 15 (/c/0, the distribution is skewed to 

 the right, with 1 9. i per cent of the total spike activity 

 coinciding in time with the descending limb of the 

 negative portion of C. The firing pattern of the Betz 

 cells considered alone was similar to that of the whole 

 population: 18.4 per cent during the descent of C, 

 57.0 per cent from peak positivity to peak negativity 

 of C and 24.6 per cent later than mean peak nega- 

 tivity of C. 



A more instructi\e analysis of the spread of cortical 

 excitation involves comparing the activity of units 

 isolated at different depths in the cortex. Unfortu- 

 nately, vk'ith present methods, estimating the depth of 

 units recorded with microelectrodes is subject to 

 several errors which may be serious since the cat 

 somatosensory cortex is only 1800 to 2000 n thick. 

 For example, dimpling of tissue by the advancing 

 electrode probably makes apparent depth greater 

 than actual depth. This error can presumably be 

 minimized by using fine electrodes or by opening the 



' Usually, the first five discharges recorded after isolation of 

 the unit were selected for measurement to minimize cflfects of 

 injury and deterioration. 



pia, although the latter procedure certainly entails 

 injury at least to the superficial cortical layers. Also, 

 there is no certainty that recording depth is a valid 

 measure of depth of cell body because volume pickup 

 from distant cells and recording from parts of cells 

 (dendrites) remote from the perikaryon are possible. 

 (Despite \igorous and often apparently judicious ar- 

 gument to the contrary, it must be admitted that the 

 suggested contributions of individual cell parts — 

 dendrites, axons, perikarya — to unitary recordings, 

 whether they be intra- or extracellular, are no better 

 than educated guesses.) In the present account, 

 micromanipulator readings of depth have been as- 

 sumed to approximate the depth of the cell body. 



Figure 16 (left) shows the initial latency of units 

 orthoclromically fired by shocks to the contralateral 

 forepaw, plotted as a function of depth. Units not 

 satisfying the criteria for Betz cells are shown as dots; 

 Betz units as crosses. The ordinate also shows the 

 average limits of the cytoarchitectural layers as de- 

 termined from measurements on frozen sections (cf 

 67). Within each layer there was considerable spread 

 in latency values as might be expected from horizontal 

 synaptic spread within layers. However, close inspec- 

 tion reveals that latencies tended to be longer in the 

 superficial layers and to decrease progressively to 

 minimum values in layers III and IV (the locus of 

 the thalamocortical specific afferent plexus). This 

 tendency is shown more clearly in the right half of 

 figure 16, where the mean initial latency for each 

 layer is plotted. The progressively increasing firing 

 latency toward the surface from layers III and IV 

 accords with the temporal and spatial pattern of the 

 spread of electronegativity measured with gross pene- 

 trating electrodes (4; Towe, unpublished observa- 

 tions). 



It has l:)een argued (4) that the rate of progression 

 of maximal negativity is too slow to be accounted for 

 by conduction through apical dendrites at reported 

 characteristic conduction rates (22-24). That the 

 spread to the surface is synaptic is further suggested by 

 the fact that there is an inverse relation between 

 initial latency and maximal repetitive rates which 

 cortical units will follow faithfulK', failure to follow 

 high rates of stimulation being characteristic of cells 

 activated via multisynaptic pathways. 



Activation of the internal lamina is more complex; 

 units in layer V are activated later than those in 

 layers III and IV, suggesting synaptic spread of activ- 

 ity downward; but layer VI contains a population of 

 units (some Betz cells, some not) which fire with brief 

 delays. The differences of mean values for firing 



