294 W. J. CROZIER AND LESLIE B. AREY 



shading response is therefore to be sharply distinguished from 

 gill-crown retraction ; the former is a local matter, involving the 

 plumes individually, through their local and probably non-syn- 

 aptie nerve nets (since strychnine has no effect whatever in 

 increasing the shading response); the latter is a reflex effect 

 (compare strychnine experiments cited in section I), so far as con- 

 traction under shading is concerned, and when not secondarily 

 involved, owing to the gill plume reactions, is governed solely by 

 the constant intensity of the light (granted optimal conditions 

 of oxygen supply). 



The extent of the gill-plume reaction to shading is very vari- 

 able in different individuals, and it has not been possible to control 

 this variability. Sometimes every animal in a dish was found 

 markedly sensitive, in other cases only one or two gave detectable 

 reactions. Subsequent investigation showed that strands of 

 shme connecting one individual with another occasionally caused 

 one sensitive animal to stimulate others confined with it (Parker, 

 '08, for a not dissimilar instance in the behavior of Amphioxus). 

 Even when studied in individual aquaria, however, great vari- 

 ability was found. Sensitivity to shading was not enhanced by 

 confinement in the dark or in the light even for lengthy periods. 



Successive shadings at 30-second intervals elicit responses 

 growing rapidly more feeble, commonly failing after the third 

 or fourth. Here again, however, variability is very great; one 

 individual gave twenty such successive reactions. 



Sunlight, diffuse dayhght, light from an incandescent bulb, 

 were each efficient for the production of shading reactions. The 

 visible region of wave lengths is concerned, since responses are 

 obtained on shading through several thicknesses of glass. Tests 

 with ray filters showed that the cutting off of light passing through 

 a blue filter (X 523-450) produced good reactions, whereas that 

 through a red filter (X 690-634) most often failed. Experi- 

 ments with green and yellow filters gave no clear result. This 

 agrees with results obtained in similar experiments, employing 

 the same ray filters, with a barnacle (Crozier, '15 a, p. 273) and 

 with Chiton (Arey and Crozier, '19). 



