470 H. M. JOHNSON 



expectancy, should react in such case, the fact will be registered 

 by the chronoscope, or by the lamp-buzzer combination, or both, 

 according as SSi, SS 2 , or both are actuated. 



Each daily sitting consists of three series of 100 reactions each, 

 one series being given at each sitting under each of the three 

 experimental conditions under study. The order of arrange- 

 ment of the three series is varied from day to day so as to distrib- 

 ute the effects of fatigue and practice uniformly among them. 



Each series of 100 presentations of the stimulus contains an 

 equal number of right-hand and left-hand presentations. The 

 order is predetermined by shuffling a pack of 100 cards. The 

 stimuli to which reaction is to be inhibited are interpolated 

 irregularly. 



Each series of 100 presentations is subdivided into four groups 

 of 25 presentations each. Each group requires from 4.5 to 6 

 minutes, and a rest of about 2 minutes is allowed between groups. 



For adaptation to the condition in which the interior of C 

 is not illuminated, 15 minutes is allowed; for the other conditions, 

 5 minutes. In the latter cases, the subject keeps the head at 

 the opening in C and looks at any part of the interior, or at the 

 photometric field, as he may elect. Intent fixation during adap- 

 tation is not encouraged or practiced. For adaptation to dark- 

 ness the subject turns away his body so that the photometric 

 field is not in view. The rest of the room is not illuminated. 



During a series of presentations the subject keeps his head 

 in place. This is rendered fairly easy by the use of a mouthpiece 

 containing a wax impression of the teeth. He is instructed to 

 fixate the center of the photometric field as soon as the warning 

 signal is given. 



The three conditions the reactions under which were compared 

 are defined in table 1 . 



All the measurements given are in terms of candles per square 

 meter. 



The brightness of the interior of C under condition Z>i is too 

 low to be measurable with a Beckstein portable photometer, 

 which is of the Lummer Brodhun type. 



The measurements given below are those made near the end 



