550 F. W. (JAMIiLK. 



tion of" many and well-expanded yellow pig-mented chromato- 

 phores with an underlying blue pigment. The light, it is true, 

 is not monochromatic, but in the succeeding section it will be 

 seen that a similar result obtains even when monochromatic 

 light is used. 



Here, then, the result is that in strong red-orange light 

 yellow pigment is well developed, but that red pigment is 

 not. Turning to this green-weed experiment we have the 

 converse conditions and result. In these fish a brown colour 

 and red pigment are strongly developed (Table II). 



Considering that the contrast of green weed to red weed 

 lies in the extension and greater brightness of the green part 

 of the spectrum, the inference is that the development of red 

 pigment is due to the green light and that the strong red 

 light encourages the formation of yellow. The two together 

 give a brown coloration. 



Brown weed in a thin film transmits from red to light blue, 

 but only the red end is of fair intensity. Under a brown 

 screen the fish maintain their green colour and the contracted 

 condition of the red chromatophores. 



(5) Note on Coloration of Larval Gobies. 



Before passing on to these experiments on Hippolyte, I 

 may interpolate a short statement of results obtained by sub- 

 jecting certain larval fish to varying illumination. It is to 

 be expected that tlie discovery of sensitive and responsive 

 species will prove a difficult matter, and these notes may 

 serve to help future workers in their choice of material. 



During the past summer I obtained the eggs of two species 

 of Gobi us (G. paganellus and G. minutus) with a view 

 to determining the rate and direction of pigmentary response 

 to cultural conditions. The larval chromatophoi-es are either 

 black or of that '^yellow" colour wliick is only seen by 

 reflected light. Specimens developed and hatched in dark- 

 ness showed normal pigmentation. On green backgrounds 

 (obtained as explained on pp. 561, 562) the appearance of the 

 larvae after a week's exposure and again after a foitnight 



