PHOTORECEPTION 195 



analysis. They suggested : (1) that the rhabdomeres consists of a doubly 

 refracting central column and a surrounding cylinder of different 

 refractive index, and (2) that the optical axis of each of the retinal cells 

 is different. They reported that there was birefringence in the rhab- 

 doms and that between crossed polarizers each rhabdomere ex- 

 tinguishes about four times during one complete rotation. Many ob- 

 jections to this hypothesis were raised by de Vries (1956), who also 

 pointed out that the report of birefringence was based upon examina- 

 tions of fixed material and that fresh material showed at best too httle 

 birefringence (less than 1 per cent) to meet the requirements of the 

 hypothesis (de Vries, 1956; de Vries, Spoor, and Jielof, 1953; de 

 Vries and Kuiper, 1958). 



An alternative hypothesis to the effect that analysis is accomplished 

 by means of dichroic molecules was advanced by de Vries (1956). 

 Organic molecules tend to absorb light polarized in the direction of the 

 long axis of the molecules. If all of the photosensitive molecules of a 

 sense cell are arranged in parallel and different sense cells are oriented 

 differently the visual organ possesses the necessary mechanism for 

 analysing polarized light. Indeed, this mechanism works to a minor 

 extent in the human eye, which can detect the polarization of Hght by 

 Haidinger brushes. These are seen as a yellow shape (figure of eight) 

 and are due to orientation (about 50 per cent) of the molecules of the 

 yellow macular pigment. Nobody, however, has been able to demon- 

 strate dichroism in the insect eye, even though evidence from electron- 

 microscopy proves that the microvilli of each pair of rhabdomeres are 

 oriented in a different horizontal direction and the visual pigment is 

 presumed to be in the microvilH. De Vries (1956) concluded that if 

 dichroism is present it must be less than 10 per cent; nevertheless, he 

 did not consider that this negative finding disproved his hypothesis. 

 Later, however, he concluded that the oriented structure of the 

 rhabdomeres has nothing to do with the detection of polarized fight 

 and that insects do not *see' polarization (de Vries and Kuiper, 1958). 

 He and his co-worker were influenced in drawing these conclusions by 

 failure to detect analysers and by the experiments of Baylor and Smith 

 (1953, 1957). Stockhammer (1959), in discussing the dichroism 

 hypothesis, advanced a number of arguments from structure which 

 support the idea. 



Baylor and Smith (1953, 1957) proposed that the responses of 

 animals to the plane of polarization of light is in fact a response to a 

 non-uniform distribution of intensities in the pattern of fight reflected 

 from a substratum (in the case of terrestrial animals) or scattered by 



