192 THE PHYSIOLOGY OF INSECT SENSES 



light patterns and polarized light presented simultaneously by reacting 

 to each as if it were alone. Therefore, polarized light and intensity 

 patterns are distinct visual qualities. 



(6) For Daphnia light contrast reaction deteriorates at low overall 

 levels of illumination, while reactions to the plane of polarization do 

 not. 



(7) Honeybees orient at the same angle with respect to the plane of 

 polarization under a polarizer as to that of corresponding quadrants 

 of the blue sky, despite large differences in intensity in the two situa- 

 tions. 



(8) 'Since bees know the regional distribution of polarized light in 

 the sky (von Frisch, 1948, 1949), they must know the direction from 

 which it is coming. Yet they cannot with ordinary image vision infer 

 the direction of the original source from a hght pattern estabHshed by 

 reflection and refraction.' 



(9) 'Under natural conditions reflection and refraction patterns due 

 to polarized sky light must to a considerable extent cancel each other 

 out because the plane of polarization is different in various parts of the 

 sky. In addition such patterns as do arise will be confused by much 

 more marked intensity patterns due to direct sunlight, which com- 

 prises up to 80 per cent of the total sky light, and to clouds and surface 

 details of the earth. To claim that bees could learn sky polarization 

 patterns from these reflection-refraction cues observed on their field 

 trips seems highly unlikely in view of such conditions.' 



(10) Tn this view, experiments with spiders (Papi, 1955; Corner, 

 1957), ants (Jander, 1957) and bees (Jacobs- Jessen, 1959), have shown 

 in some instances that when the main intensity pattern is dislocated by 

 transposing the sun 180° with a mirror, the animals nevertheless main- 

 tain their orientation direction relative to the blue sky.' 



Additional evidence in support of this hypothesis is derived from 

 the work of von Frisch (1960) and von Frisch, Lindauer, and Daumer 

 (1960). 



If insects do indeed possess a discrete polarized light vision it 

 follows that there must be a mechanism in the eye for analysing the 

 plane of polarization. This could be located either in the dioptric 

 apparatus or in the retinal elements. It could depend on single re- 

 fraction according to the Brewster-Fresnel law or upon double re- 

 fraction. There is no convincing evidence for the existence of an 

 analyser in the dioptric apparatus, and the proposition that analysis 

 depends on simple reflection-refraction phenomenon at the corneal 



