INTRINSIC RHYTHMS OF THE BRAIN 



293 



the repetition rate of the flash groups is set at about 

 the alpha frequency or a submuhiple of this, many 

 subjects show no signs of the stimulus pattern but only 

 a sharp synchronization of the alpha rhythm. This 

 effect can also exist together with an evoked pattern 

 and the two modes of response may appear in adjacent 

 regions which may exchange modes from time to time. 

 The effects of alpha driving and alpha synchroniza- 

 tion can combine to corrupt an evoked response by 

 interpolation of the ' missing' component in a triplet 

 pattern, or by omission of one of the responses. 



The distinction between rhythm synchronization 

 and true evocation is important in the interpretation 

 of the results of such experiments; it is not always 

 easy to achieve because the identity of a cerebral 

 process can be inferred only indirectly from its elec- 

 trical characters. However, with the toposcopic dis- 

 play .system the peculiar phase or time relations of 

 the alpha rhythms in different parts of the brain can 

 be used to supplement identification by frequency, 

 distribution and responsiveness. Records taken with 

 electrodes on the scalp almost always reveal clear 

 differences in the time of appearance of alpha 

 waves in various regions. In general, the maximum 

 potential change is earlier in the anterior regions 

 than in the occiput, and besides this anteroposterior 

 sweep there is evidence of an even greater discrepancy 

 in phase between the longitudinal and transverse 

 derivations covering the same region. The commonest 

 appearance is for the alpha waves in the transverse 

 derivations to lead those in the longitudinal ones by 

 90°; for example, the peak of the waves seen in the 

 parietotemporal channel occurs at the instant of zero 

 potential in the parieto-occipital one. These phase re- 

 lations are so clear and consistent that they can be 

 used as a diagnostic sign of alpha activity; when the 

 effect of a stimulus is merely to synchronize this ac- 

 tivity, the characteristic phase relations are usually 

 maintained. On the other hand, where there is a 

 true evoked response, this shows the expected latency, 

 which varies slightly from region to region, but the 

 phase relations are not those of the alpha activity. 



Even before these details of alpha activity were 

 known the pos.sibility had Ijeen considered that such 

 rhythms represented a more elaborate process than a 

 simple time cycle of excitability. It may be supposed 

 that the regular rise and fall of threshold in the brain 

 resembles the ebb and flow of a tide round the globe. 

 The time of high tide, so to say, varying from port to 

 port will not merely control the accessibility of the 

 various relav stations but will also act as a clock. 



transforming time into space patterns and contrari- 

 wise. From this conjecture have been derived a num- 

 ber of variations on the theme of scanning, elaborated 

 by McCulloch (41), Wiener (67), Walter (56) and 

 others. The nearest to conclusive evidence of such a 

 process is the phenomenon described by Walter (58) 

 as ' abscission' ; the elements of a visual time pattern 

 are cut off and projected in a spatial pattern in the 

 visual association regions of the brain. The time rela- 

 tions and distribution of this effect suggest that the 

 sweep of alpha waves through the cortex may provide 

 the time-space transformation. Auxiliary subjective 

 evidence is provided by the illusions of mottled mov- 

 ing patterns of colored light seen when gazing at a 

 featureless flickering field. The illusions are powerful 

 enough to produce aberrations of color vision as indi- 

 cated by the Ishihara test when viewed by a flickering 

 light (59) and they are attributed to the same cause 

 as the complex electric patterns evoked by flicker — 

 interaction between rhythmic volleys of impulses in 

 the visual pathways with the intrinsic scanning 

 rhythms. 



Evidence from Intracerebral Electrodes 



The spontaneous brain rhythms as .seen in scalp 

 records seem to have a characteristic geometry as well 

 as a proper frequency and relation to function. Such 

 records are open to the obvious criticism that being 

 derived from electrodes on the surface of the head, 

 they can represent only the average field of vast ag- 

 gregations of neural units, all remote from the elec- 

 trodes in terms of neuronic dimensions. Additional 

 information is now a\ailable from investigations with 

 microelectrodes placed in or near to individual 

 neurons and their processes (38) and also from elec- 

 trodes implanted in the brains of human subjects for 

 clinical study (20, 48, 49). 



These methods are still in the early stages of de- 

 velopment, but they have nevertheless already indi- 

 cated that even in the intimate details of brain 

 mechanisms spontaneous rhythmic activity is a dis- 

 tinct phenomenon; it cannot be considered as an 

 aggregate or envelope of unitary neuronic spike 

 discharges. Nor is there any invariable relation be- 

 tween the spontaneous wave-like potential changes 

 near a neinon and its all-or-none action potentials. 

 When the probability of a cortical unit discharging is 

 low, then its rate of firing may be governed to some 

 extent by the field of the spontaneous rhythms; unit 

 spikes are seen more commonly in the phase when 



