THE RELATION BETWEEN LIGHT AND PIGMENT-FORMATION. 555 



orange. Weak lithium carmine solution in a 2 mm. layer was 

 used in 1909. For green a 60 per cent, solution of copper 

 chloride with a trace {ry\y of the volume employed) of 6 per 

 cent, potassium chromate gave a good result in 1*5 and 2 cm. 

 thickness. For blue, ammoniacal solution of copper sulphate 

 was used, a concentrated solution to which strong ammonia 

 was added until the precipitate was thrown down and could 

 be filtered off. Unfortunately this blue screen, probably 

 owing to the ammonia exhaled, is very toxic. 



The light employed was direct, or direct and diffuse, day- 

 light. In the former case the vessels stood for more than half 

 their depth in a tank placed on the south side of the Plymouth 

 Laboratory. In the latter the vessels were placed about 

 10 ft. from the south window on the slate base of the table 

 tanks. In 1909 the vessels stood opposite a north window 

 on a glass shelf, and were illuminated from below by a mirror 

 as well as from above and laterally. The temperature main- 

 tained by a flow of water around the outer vessel was 16'5° to 

 17"o° 0. even in direct light; that of the inner vessel with a 

 continuous water-current was 16° to 17° C. Other experi- 

 mental batches were maintained in clear glass, and under 

 white or black background influence as well as in darkness. 



(2) Variability of the Larval Pigment. 



The chromatophores of Hippo lyte varians at the time 

 of hatching usually contain a single granular pigment of a 

 red (scarlet) colour. No true yellow pigment is present, but 

 there is a substance in the chromatophores that is yellow by 

 reflected light and brownish by transmitted light. This is 

 very constant in all broods. A variable amount of diffuse 

 blue pigment is associated with the red. 



Previous investigations (Gamble and Keeble) have shown 

 that " the progeny of females (in Hippo lyte varians) with 

 much red pigment have more of this substance in each 

 chromatophore than have those derived fi'om green parents 

 in which red pigment is less abundant. 



The question is of some importance since the initial amount 



