612 F. W. C4AMBLE AND F. W. KEEBLE. 



can be seen all over the body of the " chromatophore/' but 

 particularly clearly at the margins of the " blue spots." It 

 is the close association of the yellow and blue network — 

 perhaps in some regions the two pigments actually form parts 

 of one reticulum — that causes the green colour. 



We saw, in dealiug with the brown Hippolyte, that it 

 was the network alone which gave rise to the colour of the 

 animal, the greater mass of the pigment of "chromatophores" 

 playing no part in the naked-eye colour, being, indeed, con- 

 cealed by the surface colour. Here, in the green variety, the 

 same is largely the case, the finer branches and the super- 

 ficial branches of the colour elements alone give rise to the 

 gross colour. 



In the brown colour form, red is fully expanded, yellow 

 slightly, and blue least of all ; in the green variety, yellow 

 and blue are fully extruded, while the red is so densely 

 retracted as to have the appearance of spherical black dots. 

 To complete the comparison, take the case of a pink specimen, 

 the nearest approach to a red variety which the neighbour- 

 hood of Piel or of St. Vaast permits. Here the " chromato- 

 phores " are large, moderately expanded, that is to say, 

 the red pigment exists in a coarse network (PI. 35, fig. 25), 

 while, except for the " blue spots," the blue and yellow are 

 absent or invisible. 



The general conclusion, which must be of great use in future 

 examination, may be noted : that the colour of the network 

 defines the gross colour of the animal ; and that, in any form, 

 great stores of pigment, more or less concealed by the net- 

 work, exist in the intermuscular " chromatophores." 



Section IV. 



Colour Mutability of Hippolyte varians. 



We have seen that the young specimens of Hippolyte 

 varians often exhibit colorations distinct from that of adults, 

 and we have shown that the patterns of the variegated forms 

 and the tints of the uniformly coloured specimens are deter- 



