128 THE PHYSIOLOGY OF INSECT SENSES 



clearly than one. 'Let us assume that the visual acuities (or other 

 thresholds) of both eyes fluctuate independently of one another . . . 

 and that the instantaneous thresholds for monocular vision have the 

 same distribution in both eyes, . . . then if the one eye alone has the 

 chance of a of seeing the symbol, both eyes together have the chance 

 2a— a^. As a is smaller than 1, this expression will always be greater 



7.0 



(f) 

 h- 



d) 

 O 



Q. 

 O 



o 



< 



UJ 



o 

 cr 



UJ 



6.0 



5.0 



• I LEG-RANDOM 

 O 2 LEGS-RANDOM 



•3.0 



2.0 



1.0 



0.0 



LOG MOLAR CONC. SUCROSE 



Fig. 76. Comparison of the distribution of acceptance thresholds for 

 sucrose, as a function of concentration, for flies stimulated unilaterally 

 or bilaterally. The broken line represents the theoretical distribution 

 of bilateral thresholds (two-legged flies) calculated from the expression 

 \—q^, where q equals the fraction of the population of one-legged flies 

 not responding. (From Dethier, 1953.) 



than a - that is to say, two eyes will be able to see better than one solely 

 as a result of random combination' (Barany, 1946 b, p. 127). And as 

 Smith and Licklider (1949) went on to state, the same source of bias is 

 inherent in the procedure as appHed to the determination of thresholds 

 in other sense modalities. 



The idea can be clarified still further by quoting from the study of 

 hearing by these authors (p. 279). *In order to estimate the magnitude 

 of the bias, it is necessary to define the null condition under which we 



