84 VISION 



increases. A distinct "kink" is seen at about 4 log units; this is associated 

 with the switchover to the cone system. Animals with pure rod or pure cone 

 retinas show a simple or smooth curve. It was this kind of indirect behavioral 

 evidence for cone function that Gruber was seeking. These studies have 

 recently been reviewed, and the reader is referred to Gruber (1975) for 

 details. 



Gruber (1966, 1967) first investigated dark adaptation of the lemon 

 shark. Subjects were intensely light-adapted before being placed in darkness. 

 Determinations of the minimum light stimulus detectable by the shark (i.e., 

 threshold of vision) were made as dark adaptation proceeded. Results 

 demonstrated that the lemon shark is capable of extensive slow dark adapta- 

 tion. Independent electrophysiological investigation (Hamasaki and Bridges 

 1965) confirmed the results at the retinal level. Infrared photography of 

 pupillary changes during dark adaptation indicated that iris dilation (results 

 discussed in the section Iris) accounted for only tenfold (1 log unit) of the 

 more than 1-millionfold (6 log unit) increase in visual sensitivity accom- 

 panying dark adaptation. The final sensitivity of the shark exceeded that 

 of two human subjects measured on the same adaptometer. However, the 

 curve of dark adaptation in the lemon shark was apparently smooth and 

 thus did not provide evidence for a photopic mechanism. 



Gruber (1969, 1975) and Gruber and Hamasaki (in preparation) then 

 turned to the parameter of critical flicker fusion. Figure 21 shows the 

 rationale behind the experiment as well as the result. This time sharks were 

 trained to respond when a steady visual field was made to flicker (flicker is a 

 subjective term). The stimulus beam was increased in intensity, and the 

 number of flashes per second which just elicited a conditioned response at 

 each intensity level was recorded. This test revealed a discontinuity or rod- 

 cone break as seen in Figure 22. Concurrent electrophysiological studies 

 confirmed the presence of a rod-cone break in the cff-vs-log I curve. This was 

 the first behavioral evidence favoring cone function in sharks. 



The next series of investigations involved spectral sensitivity and the so- 

 called Purkinje phenomenon of the lemon shark (Gruber 1969, 1973, 1975; 

 Cohen et al. 1977). One of the most fully supported relationships in visual 

 science is the correspondence between the absorption characteristics of the 

 (rod) visual pigment and the dark-adapted spectral sensitivity of the animal. 

 Spectral sensitivity may be defined as the minimum number of quanta detect- 

 able at each frequency of light within the spectral limits of the particular 

 species. Under dark-adapted conditions the shape of the rhodopsin absorp- 

 tion curve determines the spectral sensitivity of the animal, although ocular 

 features such as a colored lens, oil droplets, or tapetal reflection can affect 

 the curve. As the animal becomes light-adapted a shift in spectral sensitivity 

 may occur, since the rhodopsin (rod) system is suppressed as the cone sys- 

 tem whose spectral characteristics are based on other visual pigments 

 becomes active. This shift is named after its discoverer Purkinje and is 

 indirect evidence for cone function. 



Figure 23 shows the results of a behavioral determination of spectral sensi- 

 tivity in the light- and dark-adapted lemon shark. While the dark-adapted 



