THE RELATION BETWEEN LIGHT AND PIGMENT-FOllMATION. 543 



experiment on both these lines. It is well known that the 

 ballan wrasse varies throusfh a series- of monochrome, barred 

 and spotted types of coloration from deep red to blue^ and it 

 has been ascertained by Holt that a given individual is 

 capable of passing through colour phases from a spotted to a 

 uniform livery with accompanying changes of colour. It is 

 also asserted (Noe and Dissard), that these colour varieties 

 are associated with the substratum over which the fish range. 

 Thus Gourret, in his beautifully illustrated memoir on the 

 wrasses of Marseilles^ describes the varieties of several species, 

 associated respectively with Zostera and with NuUipore- 

 grounds, and the seasonal changes which they undergo. 



The existence of a close relation between the coloration of 

 many aniuials and that of their surroundings is a well- 

 e^>tablished conclusion. In the particular and striking case 

 of Hippolyte varians, the development of this relation has 

 been shown by Professor Keeble and the author to take place 

 rapidly if young transparent animals are placed with algae in 

 a strong light. Thus an experiment conducted in bright 

 sunshine at Tregastel showed that out of fifteen colourless or 

 pale red lined Hippolyte, eleven became red after two days' 

 association with red-brown weed ; and that eight out of 

 twelve became green on green weed in the same time. 

 Analogous but much slower changes have been established 

 by entomologists for sensitive geometrical moths during their 

 early larval stages. Except for this group, however, the 

 amount of experimental evidence on the factors that deter- 

 mine this colour sympathy is very limited. Certain insects 

 excepted, Crustacea are apparently the only class in which 

 the action of the environment has been tested ; and even 

 here the light-factors that determine the development and 

 distribution of one or more pigments so as to produce an 

 effect in harmony with the coloration of the environment, 

 are quite unknown. 



With a view of determining these factors I undertook in 

 1907 a series of experiments with one of the wrasses, the 

 common gold sinny (Crenilabrus melops). 



