266 BRADLEY M. PATTEN 



in which the sign of the reaction to hght was reversed by lowering 

 the hght intensity, it would introduce serious comphcations. 

 Since such reversals are not uncommon (Frandsen, '01, p. 214; 

 Hadley, '08, p. 180; Parker, '02, p. 119, etc.), the necessity of 

 obtaining data covering this point is apparent. 



Since the variations in effective intensity in the present ex- 

 periments would all be decreases, animals were tested from the 

 highest intensity used (120 candle meters), through a series of 



70- 



60 - 



50 - 



40 



30 



20 



lOh 



o^ ° ^- ° 



/o 



/o 



0/ U-. oo UB , 2 3 7.4 120 



Fig. 4 Graph showing the increase in deflection exhibited by animal 18 

 under various intensities of lateral illunaination. Each point represents the 

 average of ten reactions. The smoothed curve has been added to bring out 

 more clearly the general nature of the rise in reaction amplitude. The breaks 

 in the graph were made to reduce the extent of the horizontal axis. There is no 

 change in scale in the vertical axis. 



diminishing intensities. For these experiments the movable 

 Ught of the apparatus (C, fig. 2) was used. The curve plotted 

 for animal number 18 (fig. 4) shows the nature of the results 

 obtained. Each point was located by averaging ten reactions. 

 Deflections induced by lateral illumination were plotted along 

 the vertical axes, and along the horizontal axes were plotted the 

 intensities of the illumination in candle meters. A smoothed 

 curve was drawn in to show more clearly the general nature of 

 the changes in reaction amplitude. 



