COMPLEX NEURAL ARCS. 81 



plate and plate-holder is taken up, and friction rendered relatively 

 constant by a spring guide on one side of the frame. 



HEAD- REST. 



For measurements of the time of the eye-movements, a relatively 

 simple head-rest is sufficient. A forehead-rest above, and a mouth-rest, 

 or rest for the upper teeth, make an adequate support in the simplest 

 possible manner. 



RECORDING LIGHT. 



The best sort of light for recording the eye-movements is an arc light, 

 stopped down by blue glass. In the following experiments we used three 

 thicknesses of blue glass. The consequent light is a soft blue for vision, 

 but is highly actinic. Even direct fixation of this light for several seconds 

 produces an after-image only slightly disturbing to vision. Figure 1 

 (page 31) shows the general orientation of the source of light and the 

 mirrors which reflected it to the eye of the subject. 



EXPOSURE APPARATUS AND STIMULUS. 



The exposure of the peripheral stimulus and the beginning of the 

 record were synchronized mechanically. Both were produced by the 

 movement of a single shutter, shown in figure 1 to the subject's right 

 of the enlarging-camera. 



The nature of the peripheral stimulus, its position, and the instruc- 

 tions to the subject are not unimportant in the eye-reaction experiments. 

 In experiments with the insane, Diefendorf and Dodge 1 used digits for 

 the peripheral objects. The subjects were requested to read these as 

 quickly as possible after exposure. The reaction of the eyes was thus 

 made an incident in the process of reading. The task was a familiar 

 one and the results were fairly consistent for the same individual. A 

 similar arrangement was adopted in the present series of experiments. 

 Letters were used instead of digits, however. They were typewritten 

 on small, uniform strips of paper, which could be inserted in the object- 

 holder of the exposure apparatus. This object-holder was movable 

 in a horizontal plane, and was placed at a different one of six possible 

 positions before each experiment. The order of these positions varied 

 from record to record and from series to series. But it was the same on 

 all days, alcohol and normal alike. 



Some variation of this sort about the prestimulating fixation mark is 

 necessary to prevent anticipatory reactions, which no subject can 

 prevent if he knows exactly where the object is to appear. It is doubt- 

 ful, however, if our arrangement was the best possible. The six posi- 

 tions are probably too few. There is some apparent tendency for the 

 subject to guess where the next exposure will be. Such guesses are usu- 

 ally clear enough on the records, since there is only one chance in six 

 that the subject will guess correctly, and an incorrect guess will result 



Diefendorf and Dodge, Brain, 1908, 31, p. 451. 



