THE VISUAL PIGMENTS 



(1), amphibia (3) and fish (8). The species they used are listed 

 below. 



Species studied by Kottgen and Abelsdorff {IS96) 



Mammals Fishes 



Ape {Kalitrichus sabacus) Perch {Perca fluviatilis) 



Cat Pike-perch (Lucioperca sandra) 



Dog Burbot {Lota vulgaris) 



Rabbit Carp {Cyprinus carpio) 



Birds Tench {Tinea vulgaris) 



Owl {Strixflammia) ^ream {Abramis brama) 



Brown trout {Salmo fario) 



Amphibians Pike {Esox lucius) 



Common frog {Rana temporaria) 

 Tree frog {Hyla arborea) 

 Toad {Bombinator bombinus) 



Reptiles, the fifth class of vertebrates, were omitted from their 

 investigation since, in agreement with the earlier observations of 

 BOLL and kOhne, they found this class to be generally deficient in 

 rods and visual purple, while those members of it which had been 

 reported to possess retinal rods (viz. crocodile, gecko and boa) were 

 not available. 



The visual pigment solutions were prepared in red hght (an 

 improvement on the sodium light used by kOhne) by treating retinae 

 with bile solution. The absorption spectra were measured by a 

 spectrophotometric method. The optical cells containing the retinal 

 extract and the reference solution were situated between the light 

 source and the spectrophotometer (konig's). To avoid undue 

 bleaching during the measurement periods, coloured glass filters 

 were interposed between the light source and the cells. In this way 

 the Hght-sensitive solutions were shielded from all Ught except that 

 in the spectral region required for measurement. 



KOTTGEN and ABELSDORFF measured the optical densities of their 

 solutions within the spectral range 420-700 mfj.. Starting at the red 

 end, measurements were made at 20 m// intervals, 5 determinations 

 being made at each wavelength. Observations were then repeated in 

 the reverse order, i.e. from blue to red. Comparison of the two sets 

 of data showed that the amount of bleaching during the operation 

 (lasting about 45 min.) was hardly detectable in comparison with the 

 rather large errors of measurement. In any case, any distortion of 

 the density spectrum caused by progressive photodecomposition of 

 the visual pigment was more or less compensated by taking the means 



28 



