PHYSIOLOGY AND BEHAVIOR OF THE ARTHROPODA 



339 



Figure 17.14. Design for studying discrimination in bees. All 16 cards had dishes, 

 but only a few dishes had syrup. By varying background (as on one card above), the 

 ability of the bees to discriminate can be observed. This design was also used to study 

 color vision. (After von Frisch.) 



they easily distinguished one bar from two that occupied less space than 

 the square. Hence, the observed failures cannot be attributed to poor 

 acuity. Apparently shape as such is not recognized by bees when feed- 

 ing, but discontinuity is. All the pairs of objects that the bees can tell 

 apart differ in discontinuity. As the bee flies over the targets, with its 

 compound eyes fixed rigidly on its body, a solid square, circle or bar 

 produces a single wave of darkening across the ommatidia, whereas an 

 open figure or two lines or an X produce two waves, at least in some 

 regions of the eye. The bees appear to be counting interruptions not ob- 

 serving form. 



These results indicate the risk involved in drawing negative con- 

 clusions from experiments with behavior. If only solid figures had been 

 used von Frisch might well have concluded that bees discriminate very 

 poorly if at all. Actually, however, the choices presented to the bees 

 simply would not have provided stimuli appropriate for the response 

 being studied. 



A more dramatic case of this kind occurred earlier in experiments 

 with color vision. Men have long wondered whether other animals per- 

 ceive color, and many early experiments were negative. Again, the 



■ ▲/ 



XDAli 



Figure 17.15. In feeding experiments, bees did not distinguish among the figures 

 of the top row, or among those of the bottom row. They did distinguish between the 

 members of any pair including one upper and one lower figure. (After von Frisch.) 



