392 Mr. VonMon's farther experiments upon 



periments, it still remained possible that the light directly 

 influences the developing pupa beneath the larval cuticle, 

 and thus determines the presence or absence of the 

 colourless precursor of the pigment which subsequently 

 appears. If two colours with opposite influences pro- 

 duced opposite effects on the two parts of the pupa to 

 which they had been respectively applied, the suggestion 

 made above would receive very strong support. If not, 

 if some intermediate tint was common to the whole 

 pupal surface, the above suggestion could only hold if 

 we suppose that the superficial layer in which these 

 changes take place is in a condition of such complete 

 physiological unity that each local influence is just as 

 powerful in another part of the layer where an opposite 

 influence is at work as it is in the area directly exposed 

 to its action. Although such a view is difficult to con- 

 ceive, the tendency of recent research has certainly 

 afforded proof of the organic continuity of tissues which 

 such a hypothesis requires. Dr. Michael Foster tells me 

 that he does not by any means consider this hypothesis 

 to be essentially improbable as an explanation of the 

 adjustment of colour, (c) The influence of light through 

 the nervous si/stem. If the nervous system receives the 

 stimulus, and controls the result, a general effect from a 

 local influence is to be expected. There is no difliculty 

 whatever in the supposition that the impulses from con- 

 flicting stimuli applied to different areas of the body 

 would become neutralized when they meet in some 

 nervous centre or centres, and hence result in efferent 

 impulses which produce a uniform intermediate effect. 

 This conclusion is also supported by the power of 

 adjusting the colours of the cocoon, which can still 

 be maintained to exist in the genus Halias, and which 

 receives its most probable explanation on the supposition 

 that the nervous system is concerned. 



In addition to its direct bearing on these important 

 questions, the experiment also affords interesting in- 

 formation as to the relative strengths of stimuli opposed 

 to each other, and (in the form in which it has been 

 • conducted in 1892) as to the possible exercise of choice 

 by the larva. 



The results obtained in 1886 are well known to be 

 negative — a uniform result following the two opposed 

 local influences. I was anxious to apply the experiment 



