186 



H. M. JOHNSON 



is generally smaller than one obtained under the same conditions 

 by the method of right and wrong cases, of which the discrimi- 

 nation-method is a special and rough adaptation. The relation 

 between the absolute value of the standard stimulus and a 

 series of thresholds obtained by either method ought to vary in 

 the same way. This comparison between the two sets of data 

 is the only one which I wish to be made directly, although the 

 values obtained for the monkey and for the human subjects 

 indicate that their sensitivities are of the same order. The two 

 human subjects were Dr. A. G. Worthing (W), a member of 

 the research staff of this laboratory, and Mr. B. E. Shackelford 

 (S), of the University of Chicago, late Brush fellow in this labo- 

 ratory. Both observers are physicists possessed of a high degree 

 of skill in optical pyrometry — one of the most difficult types of 

 photometric measurements. Perhaps it should be remarked 

 that both observers considered the visual conditions in the 

 present experiment quite trying. An hour or more was usually 

 required for making ten paired readings, which constituted a 

 single sitting. The results are summarized in table 6. 



TABLE 6 



Width of 



striae on 



standard 



field (mm.) 



0.780{W" 



°- 390 {s Y ' 



0.260|^' 



0.195 



fW. 



(s. 



Upper 



threshold 



in mm. 



0.026 

 0.019 



0.007 

 0.014 



0.009 

 0.006 



0.006 

 0.009 



Lower 



threshold 



in mm. 



0.020 

 0.012 



0.009 

 0.013 



0.008 

 0.008 



0.006 

 0.008 



Mean 

 threshold 

 per cent 

 standard 



3.0 



2.1 



2.1 

 3.5 



3.3 



2.7 



3.1 



4.7 



Mean 

 variation 

 per cent 



mean 

 threshold 



581 

 45/ 



28\ 

 35/ 



40\ 

 27/ 



49 

 34 



Mean 



threshold 



for the two 



observers 



2.6 

 2.8 

 3.0 

 3.9 



The relation between absolute width and magnitude of the 

 threshold for the animals and the two human observers is shown 

 graphically in the accompanying figure. The average threshold 

 for the two observers at each stimulus-value was taken merely 

 for convenience in plotting on the scale selected. 



SUMMARY 



A very large effect of practice was found in the work on the 

 monkey, which indicates the desirability of modifying the dis- 

 crimination-method for use on the higher mammals. 



