290 NATURAL SCIENCE. April, 



Nevertheless, this insect, when placed among oak leaves and twigs, 

 spun a dark cocoon ! Such difference in behaviour in different 

 species should guard observers in future against too hasty generalisa- 

 tion. 



Poulton's recent experiments on butterfly pupae confirm the 

 results given in " The Colours of Animals." The most striking 

 effects were obtained with Vanessa iirtic^r and V . to. These chrysalids 

 vary from dark brown or black to bright golden [uvtico') or metallic 

 green {io). As in the larvae, the latter effects are due to the absence 

 of dark superficial pigment, and are produced by light, or bright 

 metallic surrounding surfaces. As with the larvae, dark surfaces lead 

 to the formation of the dark pigment. The pupal colour is fixed in 

 V . urtico! by the surroundings of the larva when in the early stages 

 preparatory to pupation, wandering in search of a suitable spot for 

 attachment, or resting motionless before attachment. These stages 

 are lengthened by dark surroundings, and also by disturbance ; 

 Poultoc believes that during them the " colourless precursor " of the 

 dark pigment which will appear in the pupa is being formed. The 

 last stage of V. urtico', after the caterpillar has suspended itself, is 

 not sensitive, and little or no effect is produced by changing the sur- 

 roundings then ; but in V. to this stage is longer, and some result 

 can be obtained by transferring the insects to a different environment. 



Caterpillars of V. urticce placed in a gilt cylinder in darkness 

 changed to dark pupae. Dark surroundings, in a strong light, pro- 

 duced slightly greater darkening of the pupae than the same environ- 

 ment in the dark. Some larvae were allowed to pupate in boxes 

 lined with alternate strips of black and gold paper, so that the 

 insect in its motionless stage rested partly on black and partly on 

 gold. These produced pupae of an intermediate tint, but none were 

 parti-coloured. As the position of the head made no difference in 

 the result, it seems certain that the colour seen by the caterpillar has 

 nothing to do with the effect. 



Many experiments besides simple dark and bright surroundings 

 were tried with V . io. The larvae, when exposed to blue light, pro- 

 duced darkish pupae ; when placed on orange and yellow backgrounds, 

 bright green pupae. Chrysalids of intermediate tints were formed 

 amid bright green surroundings. Dark green surroundings and red 

 paper backgrounds led to dark pupae, but light falling through green 

 glass, or through red glass or gelatine, made the pupae light or bright 

 green. When the pupae were exposed to black ventrally, and to white 

 dorsally, or vice versa, the effect of the dark surface was stronger than 

 that of the white. 



The nature of the light-rays reflected from or transmitted through 

 the various substances used, was tested by careful spectral analysis, 

 and it appears that the presence of yellow and orange rays checks the 

 formation of dark pigment, and tends to produce light, green, or bright 

 larvae and pupa^. The difference between the results with green 



