Adjustment of colour in various pwpx, etc. 403 



reflected from the latter is described. The pupa on white 

 muslin (26) was light (3 a), and probably affected by this 

 surface rather than the leaves. 



White jocqycr (27, 28) is of great interest, invariably 

 producing light pupa3 (1 a) ; the influence through glass 

 (28) being the same as that exerted directly (27). We 

 see in this and in the pupae produced by Dutch "gold" 

 (19 — 21) a great advance in the suscei^tibility of these 

 pupse over those of the Vancssidm. Thus the highly 

 sensitive pupa of V. io is influenced in the same direction 

 by bright green (such as those of nature), yellow and 

 orange as it is by golden metallic surfaces and by white. 

 These all alike tend to produce brilliant green pupse 

 with a golden sheen over much of the surface. And 

 yet such pupge would only be concealed on the bright . 

 green backgrounds. The pupa of P. ncqn is similarly 

 influenced by bright green, yellow, and orange, but is quite 

 differently affected by white and gilt. The pupge in the 

 latter case are light hone-coloured (3 a), and certainly much 

 more effectually concealed on a white surface than if they 

 were green. Traces of the same kind of sensitiveness at a 

 much lower level of development are rendered probable in 

 Papilio machaon from the results already recorded, and 

 will be seen to exist in Pieris hrassicm and P. rci'px, in the 

 formation upon a white background of intermediate, grey, 

 and pale forms rather than green ones. The similar effects 

 of bright green, yellow, and orange are certainly to be 

 explained as they have been in many other species, both 

 larvae and pupse, by the fact that all these colours reflect a 

 high proportion of the effective rays, viz. the yellow and 

 orange. This will be proved in the case of P. ncipi by the 

 results of the spectroscopic examination of the back- 

 grounds which were made use of (see Appendix). 



Yellow paper (29 — 31). Produced only the light (3 a) 

 and green pupse (3 h), the latter strongly predominating. 

 The influence through glass (30) and at a distance (31) 

 was less strong in the direction of green, although the 

 number of pupag was too small to carry much weight. 



Orange leno and paper, in some cases combined with 

 yellovj (32 — 38), were even more powerful than the yellow 

 in producing green pupai (3 I) ; in fact if we consider 

 those experiments oidy in which the pupse were directly 

 placed on the backgrounds or were only separated by the 

 thickness of the glass (32, 33, 35, 36, 37), no less than sixty- 



