INTRINSIC RHYTHMS OF THE BRAIN 



287 



this was the schema tentatively suggested for the 

 origin of the alpha rhythms in the human brain. 



Individuality of Alpha Rliytlirns and their I 'ariation 



During the last 10 years several experiments have 

 expanded and modified the methods introduced by 

 Adrian & Matthews and have also widened their 

 survey to the study of a large number of subjects. 

 These obser\ations have shown clearly that, unlike 

 most physiological phenomena, the alpha rhythms 

 must be considered in relation to each individual of a 

 given species and not merely as a specific or generic 

 character. In other words the alpha rhythm patterns, 

 in terms both of spatial distribution, frequency and 

 relation to function, are highly characteristic of every 

 indi\idual. The variation is so wide that classification 

 of alpha type must include a class of normal person 

 in whom no alpha activity whatever is visible (fig. 5), 

 even in those conditions which are most favorable to 

 the appearance of these rhythms in other people. At 

 the other extreme there are people in whom alpha 

 rhythms persist even in circumstances which are most 

 inclined in other subjects to interrupt or suppress this 

 activity. A distribution of this sort is extremely difficult 

 to reconcile with any simple theory of spontaneous 

 activity, and any general scheme to account for 

 these phenomena must include considerations of 

 mental and even social character — aspects of human 

 existence which are in general far removed from the 

 domain of neurophysiology. 



Accepting the need for considerable reserve, it is 

 generally true that the amplitude of spontaneous 

 rhythms in the alpha category is inversely correlated 

 with visual attention. Empirically the frequency 

 range acceptable for alpha rhythms is from 8 to 13 

 cycles per sec, and the distribution of frequency in a 

 large population follows a more or less normal Gaus- 

 sian curve, the mode falling at about 10 cycles per 

 .sec. The tails of the curve should extend to about 6 

 and 15 cycles per .sec; about one in 5,000 individuals 

 does in fact show rhythms at these limits which com- 

 ply with the arbitrary definition of alpha rhythms. 



Complexity oj Alpha Rhythms 



The appearance of alpha rhythms in normal people 

 is usually suggestive of intrinsic complexity and var- 

 ious methods of analysis have jjeen applied to the 

 study of this possibility. Whatever method is used the 

 great majority of alpha rhythms are compound in the 

 sense that there are usually several components within 

 the alpha band (65). The superposition of these vari- 



ous components in a record produces an appearance 

 of continuous but irregular modulation; sometimes 

 the amplitudes and frequencies are so constant over a 

 period of time that a regular pattern of 'beats' is pro- 

 duced. The identification of the various alpha com- 

 ponents can also be accomplished geometrically by 

 recording from electrode patterns in which derivations 

 may be made from orthogonal electrode chains. With 

 this arrangement one component may be found to ije 

 more prominent in records from anteroposterior 

 electrodes and another from traverse ones. Further- 

 more, the components may be distinguished by their 

 functional activity or responsiveness. For example, in 

 those people with persistent alpha rhythms, one com- 

 ponent may continue when the eyes are open while 

 another is more prominent when the eyes are shut 

 during mental activity. The geometrical and func- 

 tional separation of alpha components is perhaps 

 more convincing than their display by instrumental 

 analysis, but the three methods can be employed to- 

 gether to construct a dynamic picture of spontaneous 

 activity in relation to function and behavior. Other 

 methods of analysis of alpha rhythms have been pro- 

 posed by Sato (47), Krakau (35), Burch (15), Koz- 

 hevnikov (34) and Bekkering (8). These have not so 

 far been applied to a very wide range of subjects and 

 situations but their trials have been promising (15, 16, 

 24, 25, 36, 66). 



In experiments designed to exploit the three prin- 

 cipal methods of analysis, simultaneous observations 

 of changes in the autonomic nervous system and the 

 behavior of the subject are of considerable value. 

 These techniques have not yet been fully developed 

 so the information is still inadequate to give a clear 

 indication of alpha significance in terms of somatic 

 change. A further difficulty is that such experiments 

 involve some selection of ' suitable' subjects and this 

 has caused considerable difficulty in comparing re- 

 sults from diiTerent laboratories. For example, there is 

 a natural inclination to choose for study people with 

 large regular alpha rhythms since they seem likely to 

 provide records which are easier to measure and inter- 

 pret. Selection of people in this group inevitably 

 limits the scope of investigation, and there is even 

 some evidence that subjects with extremely prominent 

 and persistent rhythms may display mental characters 

 bordering on the pathological. 



Identification of Alpha Components 



When allowance is made for these limitations, there 

 remain certain general features which seem incon- 

 testable. First, the attenuation and constriction of 



