2i6 COLOUR IN NATURE chap. 



amount of pigment developed increased with the 

 amount of exposure, and that although pigment was 

 developed more rapidly when the larvae were very- 

 young, yet the power of developing pigment was 

 not confined to any particular age. This is interest- 

 ing as showing that the abnormal pigmentation was 

 not a purely larval phenomenon as, for example, was 

 the peculiar colour change shown by Fischel's sala- 

 manders (see Amphibians, p. 233). The illumination 

 of the lower surface during the process of metamor- 

 phosis did not in any way interfere with the ordinary 

 course of development, but throughout the whole of 

 the experiments Mr. Cunningham noticed that the 

 larvae all had what he calls the " objectionable 

 habit " of clinging to the sides of the tank so as to 

 avoid as far as possible the illumination of the lower 

 surface. This suggests that the light had some 

 powerful and unpleasant influence upon the nervous 

 system. The pigmentation of the lower surface was 

 apparently in no case so marked as that of the 

 normal upper surface, but was always the same in 

 kind, and showed the same variations of colour due 

 to contraction or expansion of the chromatophores 

 as the upper surface. Artificially produced pigment 

 always appeared first in the middle region of the 

 body on each side of the lateral line, and last in the 

 head and tail regions. 



Besides these artificially produced abnormalities 

 in flat-fish, many cases have been described in 

 which the lower surface is more or less coloured 

 under natural conditions. In some cases the colour 

 is due to mere ill-defined patches of pigment, but in 

 others the lower or blind surface shows in whole or 



