250 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 195 



out only one or two of its manj components. An instrument lias been 

 designed, and is now in use in many laboratories, which automatically 

 breaks down the complex oscillations into their various constituents, 

 rather as a spectroscope resolves white light into its component colors. 

 This wave analyzer writes out on the ordinary record every 10 seconds 

 a frequency histogram corresponding to the Fourier transformation 

 of the primary changes. A mathematician would take several days 

 to perform the same task. Another device, still under development, 

 displays the electrical activity on a battery of 24 cathode-ray tubes, 

 each one corresponding to a small area of the brain. In this "topo- 

 scope," voltage is transformed into the brilliance of the cathode-ray 

 screen, and an indication of frequency is given by the way in which 

 the patch of light seems to rotate on each tube. 



Figure 6. — Electroencepbalogram taken during a minor epileptic seizure sliow- 

 ing the prominent .^'wave and spike" activity diagnostic of this condition. 

 Note that the time scale is longer than in the other records and that the 

 amplification has been greatly reduced. 



Ten years ago these devices would have seemed absurdly elaborate : 

 in another 10 years' time they will probably seem childishly simple 

 compared with the intricacy of brain function as represented by the 

 E. E. G. Even now, we can probably understand less than 1 percent 

 of the total information contained in a record such as those in the 

 figures. We are rather in the position of a visitor from Mars who is 

 deaf and dumb and has no conception of the nature of sound, but is 

 trying to build up a knowledge of languages by looking at the grooves 

 on a phonograph record. 



In dealing with a complex signal, there is always the problem of de- 

 ciding how the significant parts of the message can be made most clear 

 and memorable and the insignificant ones least obtrusive. Success in 

 this discrimination depends upon choosing the right criteria for sig- 

 nificance, and this can often be done only empirically until the cipher 

 has been broken, so to speak. 



The second sign of progress is that a more active approach is being 

 made to the brain by providing controlled stimuli instead of merely 



