34 Mr, Edward B. Poulton [Feb. 11, 



pupae were darker than wlien they had been more isolated. The 

 colours of each were in fact affected by that part of the surroundings 

 made up by the black skins of its neighbours. 



3. Illumination. — Black surroundings produced rather stronger 

 effects in darkness than in light, but the pupas were dark in both 

 cases. 



4. Time of Susceptibility. — The mature larvae, after ceasing to feed, 

 wander (stage i.) until they find a surface on which to pupate ; they 

 then rest upon it (stage ii.), and finally hang, head downwards, 

 suspended by their last pair of claspers (stage iii.), in which position 

 pupation takes place. Stage i. is variable in length, stage ii. may be 

 estimated at 15 hours (but it is also variable), while stage iii. is 

 fairly constant, and lasts about 18 hours ; while the whole period is 

 commonly about 36 hours in length. The larvae are probably affected 

 by surrounding colours for about 20 hours, before the last 12 hours 

 of the whole period, and in this time the puj^al colours are determined. 

 These facts were discovered by a very large number of experiments, 

 in which larvae were placed in surroundings of one colour, and then 

 after a variable time were transferred to another colour producing an 

 opposite effect. It was thus found that stage ii. is more sensitive 

 than stage iii., although there is some susceptibility during the latter 

 stage. 



6. The Part of the Larvse ivhich is Sensitive to Colour. 



(a) The Ocelli. — The most obvious suggestion was that the larval 

 eyes (or ocelli, six on each side of the head) saw the colours, and 

 being influenced, transmitted an impulse to the nervous centres wliich 

 regulate the formation of the pupal colours. When, however, these 

 organs were covered with black varnish, the pupae resembled sur- 

 rounding surfaces to the same extent as when they were produced from 

 normal larvae. 



(^) The Complex Branching Spines. — It seemed possible that these 

 structures might contain some organ which was influenced by the 

 colour, but after cutting them off, the larvae remained normally 

 sensitive. 



(y) The General Surface of the Shin. — This was tested by conflicting 

 colour experiments. It had been previously shown that the larvae 

 were sensitive during stage iii., and therefore they were covered in 

 this stage with compartmented tubes, so constructed that the head 

 and anterior part of the body hung in the lower chamber of one 

 wlour, while the posterior part of the body was in the upper chamber in 

 another colour. In another method, the larvae were hung upon a vertical 

 surface, while the head and front part of the body passed through a 

 hole in a shelf, the vertical surface above the shelf, and the upper 

 side of the shelf itself being one colour, while the vertical surface 

 below the slielf and the lower side of the shelf were of the colour 

 tending to produce the most opposite effects. The result of all 

 these experiments was to show that the colour influence does act on 

 some element of the larval skin, and tjjat the larger the area of skin 



