VISION. 229 



dary, locate upon each meridian a point which indi 

 cates the limit of indirect vision in that direction. Join 

 with a continuous line the points located, thus inclos 

 ing an area of indirect vision. 



(2) Test the lower median quadrant in the same way. 

 Is the total area covered by indirect vision in this 

 quadrant greater or less in extent than that in the 

 upper quadrant? 



(3) Test the upper-lateral quadrant and then the 

 lower-lateral quadrant. Are these two quadrants 

 practically equal ? 



Is there any ready explanation why the outer two 

 quadrants should contain such an excess of area over 

 the inner two quadrants ? 



(4) To record the perimeter outline. 



For this purpose one should have printed charts like 

 the one given in Fig. 33. Note that here the circles 

 are equidistant. They represent concentric arcs of a 

 quadrant with 10 of the circle between each two, 

 while the circle upon the blackboard-chart represent 

 a radial projection of these arcs upon a plane tan- 

 gent to the sphere at the point of fixation. 



In transcribing the perimeter upon the record chart 

 one has only to locate the twelve or more points lo- 

 cated upon the observation chart and join these points 

 into a continuous perimeter. 



Point x, Fig. 30 for example, would naturally fall at 

 x' Fig. 31; pointy corresponds to y'; Z to Z' whose 

 reading is : " Upper-lateral quadrant arc 64, 70 

 from vertical. 



(5) In the above experiment we have determined the 

 perimeter for light sensation only; the subject be- 

 ing conscious simply of a light or white spot on a dark 

 ground but not certain whether the spot is circular or 



